No. 5109 May 1, 1951 BIOCHEMICAL INSTITUTE STUDIES IV Individual Metabolic Patterns and Human Disease: An Exploratory Study Utilizing Predominantly Paper Chromatographic Methods From the Biochemical Institute and the Department of Chemistry, The University of Texas and the Clayton Foundation for Research, Austin. PUBLISHED BY THE UNIVERSITY OF TEXAS AUSTIN Publications of The University of Texas PUBLICATIONS COMMITTEE H. Y. MCCOWN A. MOFFIT C. H. EADS E. C. MOSSNER C. D. LEAKE F. H. WARDLAW W. P. WEBB Administrative Publications H. Y. MCCOWN F. L. Cox R. C. ANDERSON B. GONZALES, JR. L. L. CLICK B. E. SHORT The University publishes bulletins twice a month, so numbered that the first two digits of the number show the year of issue and the last two the position in the yearly series. (For example, No. 5101 is the first publication of the year 1951.) These bulletins comprise the official publica­tions of the University, publications on humanistic and scientific subjects, and bulletins issued from time to time by various divisions of the University. The following bureaus and divisions distribute publications issued by them; communications concerning publications in these fields should be addressed to The University of Texas, Austin, Texas, care of the bureau or division issuing the publication: Bureau of Business Research, Bureau of Economic Geology, Bureau of Engineering Research, Bu­reau of Industrial Chemistry, Bureau of Public School Service, and Division of Extension. Communications con­cerning all other publications of the University should be addressed to University Publications, The University of Texas, Austin. Additional cop,ies of this publication may be procured from The University of Texas Publications, The University of Texas, Austin 12, Texas Price, $1.00 per copy THE UNIVERSITY OF TEXAS PRINTING DIVISION ~~ No. 5109: May 1, 1951 BIOCHEMICAL INSTITUTE STUDIES IV Individual Metabolic Patterns and Human Disease: An Exploratory Study Utilizing Predominantly Papel," Chromatographic Methods From the Bioch'eniical Institute and the Department of Chemistry, The University of Texas and the Clayton Foundation for Research, Austin• .... \ PUBLISHED BY THE UNIVERSITY TWICE A MONTH. ENTERED AS SECOND• CLASS MATTER ON MARCH 12, 11U3, AT THE POST OFFICE AT AUSTIN. TEXAS, UNDER THE ACT OF AUGUST 24, 1912 The benefits of education · and of useful knowledge, generally difjiised through a community, are essential to the preservation of a free govern­ment. Sam, Houstpn . Cultivated mind is the · guardian genius of Democracy, and while guided and controlled by virtue, the noblest attribute of man. It is the only dictator that freemen acknowledge, and the only security which freemen desire. Mirabeau B. Lamar Contents PAGE I. Introduction, General Discussion and Tentative Conclusions ------------------------------------------------------------------7 Roger J. Williams II. Development of Paper Chromatography for Use in the Study of Metabolic Patterns ____________________________ 22 Helen Kirby Berry, Harry Eldon Sutton, Louise Cain, and Jim S. Berry III. The Influence of Solvent CompositiOn, Temperature and Some Other Factors on the Rf Values of Amino Acids in Paper Chromatography ________________ 56 B. Jirgensons IV. Quantitative Study of Urinary and Salivary Amino Acids Using Paper Chromatography __c___,_______________ 71 Helen Kirby Berry and Louise Cain V. The Quantitative Determination of Histidine Using Paper Chromatography _____ _____________,___._,___,_______________ 77 Louise Cain and Helen Kirby Berry · · VI. The Quantitative Determination of Creatinine in Urine Using Paper Chromatography ---------------------------------80 Helen Kirby Berry and Louise Cain VII. The Qu'antitative Determination of Creatine in Urine by Paper Chromatography ____c ___________________________________ 82 • Louise Cain ; . . · VIII. The Quantitative Estimation of Uric ,Acid by Paper Chromatographic Methods, with Applications to Human Urine and Saliva ------------------------------------------84 Helen Kirby Berry IX. The Quantitative Estimation of Urea by Paper Chro­matagraphic Methods with Applications to Human Urine ----------------------------------------------------------------------------88 Helen Kirby Berry X. A Micro Determination of Sodium Using Paper Chro­ matography ----------------------------------------------------------------93 Harry Eldon Sutton XI. Quantitative Study of Ketosteroids by Paper Chro­matographic Methods with Applications to Human Urine --------· _ ____ _____ _______--------------·------------------__________ 97 Eric Bloch, Ernest Beerstecher, Jr., and Roy C. Thompson XII. The Determination of Ferric Chloride Chromogens in Human Urine by Paper Chromatographic Methods 109 Harry Eldon Sutton and Ernest Beerstecher, Jr. XIII. The Effects of Single Vitamin Deficiencies on the Con­ sumption of Alcohol by White Rats ________ ___ ____________ 115 Ernest Beerstecher, Jr., Janet G. Reed, William Duane Brown, and L. Joe Berry 4 The University of Texas XIV. Individual Excretion Patterns in Laboratory Rats________ 139 Janet G. Reed XV. A Study of the Alcoholic Consumption and Amino Acid Excretion Patterns of Rats of Different Inbred Strains --------------------------------------------------------------------------144 Janet G. Reed XVI. A Study of the Urinary Excretion Patterns of Six Human Individuals ----------------------------------------------------150 Helen Kirby Berry, Louise Cain, and Lorene Lane Rogers XVII. Further Studies on Individual Urinary and Salivary Amino Acid Patterns ------------------------------------------------157 Helen Kirby Berry XVIII. Individual Urinary Excretion Patterns of Young Children --------------------------------------------------------------_______ 165 Helen Kirby Berry and Louise Cain XIX. A Further Study of Urinary Excretion Patterns in Relation to Diet ----------------------------------------------------------173 Harry Eldon Sutton XX. Metabolic Patterns of Underweight and Overweight Individuals ------------------------------------------------------·--__ _____ 181 Jack D. Brown and Ernest Beerstecher, Jr. XXL Metabolic Patterns of Schizophrenic and Control Groups --------------------------------------------------------------------------189 M. Kendall Young, Jr., Helen Kirby Berry, Ernest Beerstecher, Jr., and Jim S. Berry XXII. Exploration of Metabolic Patterns in Mentally Defi­cient Children ------------------------------------------------------------198 Louise Cain Acknowledgement The authors wish to express their grateful thanks to the many individuals and organizations whose cooperation and help have made possible the investigations described in this bulletin. To Drs. Lorene Lane Rogers and Roy C. Thompson and to Mr. Jim S. Berry we acknowledge extensive help in the preparation of a number of the papers in addition to those on which their names appear as co-authors. Without this help the publication of the material would at best have been materially delayed. Our especial thanks go to Dr. A. T. Hanretta and the Staff of the Austin State Hospital for their kind cooperation in connection with the studies of mental patients; to Dr. C. W. Castner and the Staff of the Austin State School for help.ing with the difficult problem in­volving the mentally deficient children; and to members of the Staff of the Austin Public Schools and of St. Mary's Cathedral School for help in connection with the control children. We wish to ex­press our gratitude to Dr. Pearl S. Swanson, Iowa State College, to Dr. W. F. Dunning, Detroit Institute of Cancer Research, and to Dr. W. F. J. Cuthbertson, Glaxo Laboratories, Middlesex, England, who were kind enough to make available to us highly inbred strains of rats. We also express our appreciation to all those who, often with inconvenience to themselves, have served as subjects for the investigations, and to all other members of the Biochemical Institute Staff who have cooperated in numerous ways. For financial assistance on minor aspects of the extended study we gratefully acknowledge help from the Office of Naval Research in Washington, and from Research Corporation and the Nutrition Foundation, both of New York. THE AUTHORS I. Introduction, General Discussion and Tentative Conclusions by ROGER J. WILLIAMS When the writer published in 1946 the volume The Human Frontier (1) , he had become convinced that in many areas of hu­man interest a careful study of individual people without undue haste in trying to make generalizations would contribute enor­mously to human understanding and the ability to deal successfully with members of the species known as homo sapiens. Too many studies, he felt, were founded on the assumption that there is such a creature as the normal man and that every energy must be directed toward ascertaining the characteristics possessed by this creature. He became convinced that the average man is in a sense a chimera and that further progress will come through an understanding of real individuals and developing the ability to classify them in relation to the various problems which they en­counter. Generalizations with respect to human beings are very much to be desired, but unless the generalizations are backed by adequate information, they are premature and superficial and liable to be incapable of substantiation. The writer envisaged and still envisages the time when real people will be studied from any and every angle and when no promising discipline will be ignored in connection with such studies. There are hopeful moves in this direction, the most im­portant of which is embodied in the Report of the Trustees of the Ford Foundation (1950). This report avoids completely the tend­ency to compartmentalize all human knowledge; instead it rec­ognizes the existence of human problems and of basic knowledge and insight needed for their solution-knowledge and insight which must often come through the agency of a number of dif­ferent disciplines. One of the five broad areas which this Founda­tion will be concerned with supporting is the development of an interdisciplinary science of man. Another has to do with educating individuals to their full capacities, which will entail (interdiscip­linary) methods for determining individual potentialities. Before an all-round, well-directed inquiry into the nature of human beings became possible, we decided in our laboratories that The University of Texas within our own area of specialization-biochemistry-much could be done in line with this point of view which would of itself be valuable and at the same time would help pave the way for broader studies involving other disciplines. It was decided that support would most likely be forthcoming and the most would be accomplished if a specific well defined problem were uppermost in our minds. After due consideration the problem decided upon was alcoholism. It appeared highly probable in this case that an exploitation of the biochemistry of in­dividuals, in contrast to the biochemistry of normal man, could yield valuable and useful information. However, we realized that before we could gain any significant information about the distinctive biochemistry of those afflicted with alcoholism (or any other group) we would have to gain a background of biochemical information regarding individuals in general. While this approach was under way we were fortunate in dis­covering a nutritional attack on the problem of alcoholism which was not anticipated at the outset. This has been pursued with highly favorable results, which have been reported elsewhere (2, 3). Attempts to find metabolic traits or trait patterns which would be distinctive for alcoholics have been continued; this has ne­cessitated studying individual metabolic patterns in general and in turn has led to the preliminary study of groups other than alcoholics-namely schizophrenic individuals, feeble minded indi­viduals, obese and lean individuals, young children and, to a very Hmited extent, identical twins. Ceneral Discussion of Restdts The exploratory nature of the studies here reported must be emphasized. We became convinced that an exploration of individ­ual metabolic patterns was an urgent need, and were determined that we would carry out such an exploration with whatever means we could command. If exploration in a refined manner were impossible, then we would adopt cruder methods. If highly ac­curate measurements could not be taken, we would make rougher measurements. If the best results could be obtained only by having all our subjects hospitalized (clearly an impossible undertaking from our standpoint), we would be content with something less than the best. As our work proceeded we could see that by taking short cuts that ordinarily might be frowned upon, we could gather Introduction, General Discussion, Tentative Conclusions significant or at least highly suggestive data, whereas if we me­ticulously and exhaustively developed all the methods underlying each individual item, our exploration would be too limited in extent to be useful or worthwhile. One of the factors which forced us into the use of "rough and ready" methods with respect to some items was the tremendous volume of work which needed to be done. We have estimated that in connection with this investigation approximately 200,000 paper chromatograms have been run. Allowing only one analytical re­sult for each chromatogram (which is highly conservative) this would mean an average of about 700 analyses each working day for a year, by this method alone. To have carried out anything approaching this number of analyses by long drawn out methods would have been absolutely prohibitive from the standpoint of space, money, and manpower. Two extremely favorable aspects of the situation have made it possible to get significant results : one is the very timely devel­opment of paper chromatography; the other is the fact that dif­ferences observable between individuals are often gross and require no hair-splitting refinements to detect. Corresponding values obtained from different individuals often differ not by 10 or 20 per cent but by as much as several hundred per cent. The meaning of the term metabolic pattern needs to be clarified. Insofar as the total picture of the metabolism of an individual (including the chemical processes in each and every organ and tissue and their effects on each other) is distinctive, it constitutes his metabolic pattern. The complete metabolic pattern of any individual would probably be an exceedingly complicated matter, and of course our studies are only exploratory in this field. We have merely dipped here and there into the vast area and have limited ourselves to items that could be measured with relative ease. The items which we include in our fragmentary metabolic patterns are not necessarily the most striking items nor by any means the most important ones; they are simply items which show distinctiveness and which we are able to measure practically by objective means and without inconveniencing the subjects. It would be a most interesting phase of the subject if one could measure quantitatively in individuals the basal output of the vari­ ous hormones coming from the various glands of internal secre­ tion. This information alone would probably be far more revealing than all that we have been able to gather, but of course science has not developed to the point where this can be done. We h::ivP to be content at this time with those items which can be measured by a reasonable expenditure of effort. Particularly for those outside the field of biochemistry it will be desirable to make clear that metabolic patterns have very little relation to so-called basal metabolism or basal metabolic rates. Basal metabolism values are crude summations of the total energy yielding reactions of all the tissues and organs of an individual, and reveal nothing as to the patterri or details of the metabolic mechanisms. Two individuals might have exactly the same total metabolism, measured in calories, and yet have metabolic patterns differing widely from one another.* Conversely, two individuals might have metabolic patterns very much alike and yet possess total metabolisms very different from one another. For reasons of convenience we have studied principally the compositions of urine and saliva samples. We would have preferred complete 24 hour specimens of urine to the morning samples which we have most often used, but we found it impractical except in a limited number of cases to obtain these. Again we would have liked to have repeated samples of blood for analysis, but we could not expect our voluntary subjects to undergo repeated incon­venience and annoyance, and hence our studies of blood composi­tion have been limited. Another series of items which has been studied in some of our subjects is taste thresholds. These have been included only in a limited number of cases where we had complete cooperation of the subjects. How these thresholds are related to internal metab­olism is not well known, but in the case of taste sensitivity to sodium chloride, for example, there is a definite physiological connection. The following charts depict in summary form some of our findings. In Figure I is charted in polar co-ordinates for a hypo- FIGliRE 1 *To attempt to get information about the metabolic pattern of an individual by studying his basal metabolism might be compared to attempting to de­termine the pattern of a wall paper by measuring the light intensity in the center of a papered room. .. FIGURE 2 FIGURE 3 / .I FIGURE 4 FIGURE 5 FIGURE 6 FIGURE 7 (Fies. 1-13). Taste Sensitivity: 1. Creatinine, 2. Sucrose, 3. KCI, 4. NaCl, 5. HCI. Salivary Constituents: 6. Uric acid, 7. Glucose~ 8. Leucine, 9. Valine, 10. Citrulline, 11. Alanine, 12. Lysine, 13. Taurine, 14. Glycine, 15, Serine, 16. Glutamic acid, 17. Aspartic acid. (contd. p. 6) The Uni-versity of Texas " " " " zo FIGURE 9 FIGURE 8 FIGURE 10 FIGURE 11 FIGURE 12 FIGURE 13 (FIGS. 1-13). (contd.) Urinary Constituents: 18. Citrate, 19. Base Rf.28, 20. Acid Rf.32, 21. Gonadotropin, 22. pH, 23. Pigment/ creatinine, 24. Chloride/ creatinine, 25. Hippuric acid/ creatinine, 26. Creatinine. 27. Taurine, 28. Glycine, 29. Serine, 30. Citrulline, 31. Alanine. Introduction, General Discussion, Tentative Conclusions 13 thetical individual, various numbered items which we are able to measure. The hypothetical individual charted is one all of whose measurements are average, and as will be noted in Figures II-XIII, which represent corresponding charts for twelve actual individuals, the hypothetical individual's chart bears no resemblance to that of any of the actual individuals. It may likewise be noted that there is no such thing as a "normal" pattern. It appears that each individual that we have studied has when­ever tested exhibited a characteristic pattern of measurements which is distinctive for that individual alone. While there are in every case day-to-day variations in saliva and urine composi­tions and in taste thresholds, certain items, at least, stand out as grossly distinctive, and the patterns as a whole remain nearly constant. Several questions naturally arise in connection with these and more extended "patterns." How are the patterns influenced by diet? If diet is an important factor, to what extent is the self selection of diets on the part of individuals important? To what extent do special foods or medication contribute to the production of the patterns? To what degree is the genetic background of the individuals a determininative factor in producing the distinctive patterns? To what extent is the characteristic pattern of urine composition merely a reflection of differences in kidney func­tioning, and to what extent does it indicate more fundamental metabolic differences? None of the above questions can be answered with certainty and definiteness, though our continued investigation has given indi­cations, and in some cases very strong indications, as to what the answers are. In the publication of Kirby and Thompson (4) it was reported that the patterns of urinary amino acid excretion with respect to four amino acids, leucine, lysine, threonine, and arginine, were substantially unchanged when each of the five individuals sub­jected to study was placed for four days on a uniform Army K ration diet. This finding shows that not all the differences, at least, can be ascribed to diet. It does not prove of course that every distinctive item which is discoverable when individuals are on uncontrolled diets, will prove to be distinctive regardless of dietary changes. A further study was made later (p 1'73) which sought to answer the same question. The results are not conclusive with respect to all the items measured, but for most of the items it appears that The University of Texas shifting the·individuals from an uncontrolled diet to one in which they all ate the same selection, did not cause material alteration in their excretion patterns. A very substantial amount of re­search will have to be done before it can be stated with assurance, with respect to all the items, whether or not they are influenced by ordinary changes in the diet. Common medicinal agents such as sa:licylates or sulfa drugs give rise to characteristic spots on chromatograms, but they do not interfere with the determination of amino acids and other metabolites which constitute the more im­portant items in our patterns. We can be reasonably sure that a considerable number of the items we have measured remain con­stant for a · given individual and .that the differences are not due primarily to peculiarities in the diets of the individuals concerned. Evidence of quite a different sort supports the conclusion that metabolic patterns are distinctive for individuals and are not revolutionized by changes in diet: Bloodhounds for example are able to follow the scent of an individual regardless of what his diet may have been. We are unaware of any observations indi­cating that dogs are confused in their identification of human beings (or of other dogs) by changes in diet on the part of the individuals to be identified: It is obvious, however, that certain items which we have meas­ured, such as chloride excretion, must be dependent on the diet. A high percentage of chloride excretion takes place through the kidneys, and it is reported that human consumption of ·salt may vary from 2-30 grams daily (5). Obviously ariyone who consumes a relatively ·1arge amount' of salt must excrete a relatively large arnourit .ih the urine~ ;;The fact that excretion patterns are in­fluenced by diet, however, does not by any means prove that the patterns fack significanc~ar:,mhough we·:.~ notf'kriow whabthe'·sig­nificance is;.iit seems·most! unlikely::tihat ooe person could~spontane­'Ously eat 30-grains·of salt a day and:ahother .2,. gr.a ins per day, with­·out the fact being physiologically;significarit. · It is conceivable (though ribt ·ln accord with ·our. findings) that in di vi duals could .. possess · significant individUal 'urinary 'patterris, based entirely up6ri their own distinctiveness with respect to food selection and consumption. Notwithstanding'tbe fact that mem­bers of the general population (at least within a cultural level) will often accept and partake of identical meals, the resemblance between what two associated individuals eat is probably more apparent than real, because servings are by no means always eaten in their entirety and even when the main items in the menus Introduction, General Discussion, Tentative Conclusions 15 are identical, there still may be choices with respect to bread and butter, beverage, condiments, etc. Furthermore, individuals who are closely associated may have become so partly because of their similarity in tastes, and even though they may eat some individual meals that are identical, they often have chances to choose dif­ferently-at other meals and between meals. It may be that self selection of food by individuals is more important than is now appreciated, in connection with the whole problem of proper nutri­tion. Obviously as has been pointed out elsewhere (3), however, self selection cannot be relied upon as an adequate guide to good nutrition. A number of our studies and findings are related to the question of the importance of hereditary backgrounds in the determination of metabolic patterns as we have measured them. One of the find­ings is that experimental rats on identical diets often show con­sistent differences in their urinary excretion patterns. These can be ascribed to no cause of which we are aware, other than hered­itary differences. Furthermore our extensive studies of individ­ual rats on different dietary regimes with respect to their alcohol consumption, have resulted in a multitude of observations which can only be explained on the basis of the possession of distinctive patterns which depend on the genetic backgrounds of the indi­vidul;J,I animals. Turning to human experiments, one of the studies bearing on the point of the roles of diet and heredity is that having to do .wi.th young children. The diet of very young children, consisting ,:µi;:tinly of milk, may be regarded as relatively uniform. Yet in young babies as in older children and ~dults w~ observe the exist­ence of distinctive urinary excretion patterns. In view of other evidence this finding .PPint11 strongly to the important.rok which • '· -j:._ • • . •_... .. · •· ' • . . .• . • ... .... . per;edity p~obably plays, :in the deterll),im1ti?I1 Qf these .patterns. Jn .the same study a further ev~dence of the im~prtance of genetic fa_ctors wa&.,,gained by comparing,: )?tatistically -.'.the_excretion pat­ltep1s of th~ee brothers with each other and with non-siblings . .The patterns of the three brothers showed sufficient similarity :to suggest strongly inheritance as the basis. . Another type of study which should throw much light on this problem involves the metabolic patterns of so-called "identical" twins. Our investigation in this area has been limited. However, Figures XII and XIII (p. 12) are charts for two "identical" twins, and mere inspection reveals a likeness which is not observed among the others none of whom are twins. In connection with such twin studies it should be made clear that while the term homozygous as applied to twins has a definite meaning, namely that the twins were derived originally from a single egg cell, one cannot conclude that such twins are completely identical in their inheritance. Partial asymmetry reversals are observed, and perhaps related to these is the fact that in some cases homozygous twins are very much alike in appearance and otherwise, while in other cases they differ substantially. Due to mechanisms not well understood and not based on known mor­phological sequences, homozygous twins differ in the degree to which they are alike. These differences, certainly those involving partial asymmetry reversals, can hardly be attributed to nutri­tional differences during fetal life. The homozygous twins which were available for our limited study differed in appearance slightly, particularly with respect to the curliness of their hair, and hence from the standpoint of appearance they were not as much alike as so-called identical twins sometimes are. It should not be a matter of surprise therefore that their metabolic patterns showed some differences. It should be noted that their diets were not controlled. · Our exploratory studies designed to find out whether individ­uals who have a strong tendency to become alcoholic (actually members of Alcoholics Anonymous, during periods of abstinence) possess metabolic features which are distinctive, resulted in posi­tive findings (6). The evidence gained in this study greatly strengthens the thesis that metabolic patterns of individuals are not only distinctive, but are significant and important in connection with certain human problems. The question may well be asked in connection with this study: Do the individuals have distinctive patterns which cause them to become excessive drinkers, or have their patterns resulted from excessive drinking? While we cannot categorically answer this question on the basis of available information, it seems to us very unlikely that excessive drinking would cause a change in pattern which would remain changed for many months after the drinking ceased. The whole background of information on metabolic pat­terns which we possess fits in well with the hypothesis that the tendency toward excessive alcohol consumption has as a forerunner the possession of a characteristic metabolic pattern. The study of schizophrenic and well individuals, though car­ried through only to a limited extent, has also yielded positive results. There are numerous features which appear to be dis­ tinctive to a greater or lesser degree, for individuals in this group. The question here again may be asked: Are the individuals schizophrenic because of the prior possession of characteristic patterns or do the patterns result from schizophrenia? We are unable to answer this question with definiteness since we have not studied the metabolic patterns of individuals both before and after attacks of schizophrenia. Either interpretation of the find­ ings points unquestionably to the physiological basis of schizo­ phrenia. Our numerous observations lead us to think that the hypothesis of the prior possession of characteristic patterns which predispose toward schizophrenia, is more tenable than the alternative one. Kallman's survey (7) of tbe subject points very strongly, we believe, to the importance of inheritance in the predisposition toward schizophrenia, and all our findings are in accord with the supposition that metabolic patterns are inherited and remain rela­ tively constant. To what extent they may change with age, is a problem largely for the future which will be discussed briefly in later paragraphs. Our limited exploration of the metabolic patterns of feeble­ minded children as compared with other children not so afflicted has also shown highly significant differences. There can be no question with respect to the importance of inheritance in mental deficiency, and in this case it seems meaningless to consider the question whether the feeble-mindedness may have produced the distinctive patterns since presumably the inherent mental de­ ficiency and the patterns existed from the beginning. The study of the metabolic patterns of obese and lean individuals has yielded several items, notably urinary phosphate excretion, which appear to be statistically different for the two groups. It is · our opinion that further exploration of the metabolic patterns of such individuals is likely to yield still other items which are signifi­ cantly different. Tentative Conclusions Growing out of our preliminary but nevertheless extensive ex­plorations have been various conclusions, some more tentative than others, which will prove valuable as a guide for further in­vestigations. The conclusions outlined below are in substantial accord with each other and with the facts as we know them, and present a rather different picture of the world of human beings than is customarily entertained. The implications of our explora­tions are far-reaching, not only into the field of medicine but also into the field of education and into the numerous ramifications of the problem of human behavior. Distinctive metabolic patterns are inherent in each fertilized egg.-Each fertilized egg cell which is on its way toward being a human being has within it the potentialities to produce in a suit­able environment an intricate system of organs and tissues which is distinctive for that particular individual. This distinctiveness is not limited to morphological characters, but the entire metabolic pattern-the ability for carrying out the numerous chemical re­actions and the relative efficiency of the various parts of the process-is probably inherent in the original fertilized egg. This conclusion is strongly supported by _the work of Beadle and his school in the field of biochemical genetics as well as by our ex­plorations. Changes in pattern with age are probably inherent.-There is ample reason for thinking that as an organism matures its metab­ olic pattern changes-perhaps not in all respects-and that the rapidity and to some extent the nature of these successive changes-provided the environment is relatively constant-is de­ termined by the genetic make-up of the original fertilized egg. A striking example of a metabolic change which has its basis in inheritance and yet appears only later in adult life is Huntington's chorea. In this unfortunate disease, the devastating change does not appear until the individual has had time to raise a family and pass on his defect to his progeny, but the potentialities for the change were nonetheless in his make-up continuously throughout life. The relative abilities of various organs, tissues, and bodily mechanisms to function year after year and to resist wear, is probably also inherent in each individual's make-up. (Environ­ mental factors are also important as we will note later.) When important limiting chemical mechanisms are unusually effective, this may make for long continued operation, whereas relatively ineffective mechanisms (which we know from Beadle's work to be inherited) may be conducive to wear and early breakdown. There has been a growing appreciation in recent years of the fact that as people grow old, certain of their internal mechanisms tend to wear out first. In other words various parts and func­ tions of the body do not age at the same rate. A large percentage of people who reach the age of 50 have already lost the ability to produce hydrochloric acid in their gastric juice. An increasing number of people are found who become diabetic in middle life or later. In some cases it is merely that the insulin producing machinery is wearing out. Careful investigations have shown that the tendency toward diabetes is inherited-some inherit the tendency so strongly that they are diabetic while they are still children, others may inherit the tendency so weakly that they do not become diabetic until they reach the age of seventy. The metabolic patterns of individuals are crucial in connection with their susceptibility to disease.-lt is well known-though there has been a tendency in medical circles to ignore the fact­that susceptibility to numerous diseases has a relationship to the genetic background of the individual concerned. In some cases the evidence is clear cut; in others, because of the inherently com­plicated nature of the genetic machinery, the evidence has not been worked out. Physicians have, often for humanitarian reasons, been loath to dwell upon the inescapable facts of heredity. They continually stress what can be done environmentally. This policy is based upon a praiseworthy attitude, except that it tends to overlook the possibility that understanding the herediary mechan­isms may make possible an environmental attack which would not be possible without this understanding. We believe we have discovered a clear cut example of such a situation in the disease alcoholism. We believe this disease in an individual is related to his metabolic pattern, but that recognition of this fact makes possible a successful treatment by nutritional means (5). Differences do not denote defectiveness.-There is an unfor­ tunate and quite unwarranted tendency to link together inheritance and defects. People undoubtedly inherit (in an admittedly com­ plex manner) the color of their eyes. Does this mean that every­ one whose eyes are not of a standardized color are defective? Not at all. Is the person who has small hands or feet a defective in comparison with one who has large hands and feet? Not at all. There are some types of activities for which large hands or feet may be desirable, and other activities for which they had better be small. Were Jack Spratt and his wife both defective because of their preferences for lean and fat meat respectively? Not at all. The ability to utilize different types of foods may well reside in different individuals, and the family is fortunate which can avoid waste. Is the mathematician who may lack certain musical abilities a defective, likewise the musician who may lack mathematical abilities? Not at all. There is room in the world for both mathe­maticians and musicians and the two types of abilities do not need to reside in the same individuals. Biological variability is a most important fact. Life as we know it could not exist without it. Far more is involved than mere in­heritance of "normal" and "defective" make-ups. Each of us possesses an individuality which we should cherish, because it is the real basis for our love of freedom. Because two people are different, we cannot draw the conclusion that one is inferior, the other superior. The biological fact of the differences between sexes does not demonstrate that one sex is superior to the other. The fact that one race is different biologically from another does not mean that one is superior and the other inferior. We cannot say of a musician and mathematician that one is the inferior of the other. We have previously spoken of the probability that every indi­vidual possesses a distinctive metabolic personality (8, 9). Com­paring these metabolic personalities is like comparing an infinite variety of wall paper or rug patterns. No two are the same, there is no "normal" pattern, and wide differences can exist without any of the patterns being defective. Metabolic patterns are probably intimately concerned with our psychological make-ups, our mental abilities and our behavior problems.-One clear example of a metabolic trait associated with mental powers (or lack of them) has been known for a number of years. Certain types of feeble-minded children always excrete in their urine phenyl pyruvic acid, and no person excreting this sub­stance has been found to have ordinary mental powers. The children included in this group constitute a restricted and relatively small proportion of the total number of feeble-minded children. What about the others? Do they have metabolic peculiarities of some other sort? Our investigations answer the question in the affirma­tive and point the way toward finding out more about these metab­olic peculiarities. In view of the findings of Thurstone that "primary mental abilities" are inherited as individual traits (10) we are led to suppose that some day correlations will be found between metab­olic patterns and mental traits. Not only is it to be expected that mental traits will be found related to metabolic patterns, but that all sorts of behavioral problems (strikingly those involving sex, for example) will be found closely tied in with the funda­mental and widely inclusive metabolic patterns of individuals­patterns of which up to now we have had only glimpses. Introduction, General Discussion, Tentative Conclusions 21 Environmental attacks on specific problems may become pos­sible only when the genetic backgrounds of these problems are rec­ognized.-An extremely encouraging conclusion has been reached in our study of the problem of alcoholism-namely that in spite of its hereditary roots, a nutritional attack may be successful. As indi­cated elsewhere (3) we are of the opinion that a nutritional at­tack offers promise not only in alcoholism but also in the treatment of numerous diseases of obscure origin, even including the devas­tating mental diseases. Diseases which have a hereditary origin but are capable of being attacked nutritionally we have designated as genetotrophic in nature (11, 12). There is nothing in our findings which precludes the use of psychological means for the treatment of disease. Certainly the maintenance of suitable mental hygiene has possibilities in alle­viating difficulties arising because of peculiarities in metabolic patterns. Of course, the whole problem of how mental attitudes affect the details of metabolic processes remains largely to be explored. That such changes are brought about by psychological means cannot be denied. It should be obvious from the foregoing discussion that our study is, without question, a preliminary and exploratory one and that further study of an intensive and extensive nature must follow before many of the questions which have arisen can be answered with definiteness. Our explorations have convinced us that there is a whole new area of insight into human problems, which lies ahead. BIBLIOGRAPHY 1. Williams, R. J., The Human Frontier, Harcourt, Brace and Co., New York, N. Y., 1946. 2. Williams, R. J., Berry, L. Joe, Beerstecher, Ernest, Jr., Arch. Biochem., 23, 275 (1949). 3. Williams, R. J., Nutrition and Alcoholism, University of Oklahoma Press. (In press). 4. Thompson, Roy C., and Kirby, Helen M., Arch. Biochem., 21, 210 (1949). 5. Borden's Rev. Nutritional Research, 8, 6 (1947). 6. Beerstecher, Ernest, Jr., Sutton, H. Eldon, Berry, Helen Kirby, Brown, William Duane, Reed, Janet, Rich, Gene B., Berry, L. Joe, and Williams, Roger J., Arch. Biochem., 29, 27 (1950). 7. Kallman, F. J., Am. J. Psychiat., 103, 309 (1946). 8. Williams, R. J., Chem. Eng. News, 25, 1112 (1947) . 9. Williams, R. J., Berry, L. Joe, and Beerstecher, Ernest, Jr., Proc. Natl. Acad. Sci., 35, 265 ( 1949). 10. Thurstone, L. L., "Primary Mental Abilities", Centennial Pub. Am. Assoc. Advancement Sci., 61 (1950). 11. Williams, R. J., Berry, L. Joe, and Beerstecher, Ernest, Jr., Lancet, 258, 287 (1950). . 12. Williams, R. J ., Nutrition Rev., 8, 257 (1950) . IL Development of Paper Chromatography for Use in the Study of Metabolic Patterns by HELEN KIRBY BERRY, HARRY ELDON SUTTON, LOUISE CAIN, AND JIM S. BERRY Before paper chromatography had reached its present state of development an ascending technique was devised in this laboratory (1). This involved the use of the simplest vessels, but nevertheless made possible quantitative work. This procedure has been devel­oped and used continuously in our laboratory since it was devised for the specific purpose of studying metabolic patterns. Our experience with this simple procedure has been extensive, and while it may not be superior to the use of the descending method utilizing the most modern apparatus, it is an effective technique and one which can be used in practically any laboratory without the purchase of expensive equipment. Our discussion of procedures is limited to those employed in our laboratory. Reviews dealing with the general field of paper chromatography will be found elsewhere (2, 3, 4, 5). Our pro­cedures will be discussed under the following headings: general techniques, solvents, color developing reagents, compounds studied, complications arising in the analysis of biological fluids and quan­titative aspects. The application of these general procedures to the determination of specific substances will be considered in more detail in the papers which follow. General Techniques The simplicity of the apparatus and the ease with which large numbers of analyses can be performed are the outstanding fea­tures of this method. The solutions to be chromatographed are placed in measured small quantities on filter paper sheets which are stapled in the form of cylinders, care being taken that the edges of the sheet do not touch each other. These paper cylinders are placed upright in Pyrex dishes (10 inch diameter) containing the appropriate solvent for resolution of the chromatogram. Where a number of chromatograms are to be run simultaneously, cylinders of different diameters may be arranged concentrically, care being taken to prevent their touching at any point. ~he sol­vent dish rests on the bottom of a six gallon earthenware crock which is covered with a square of plate glass. By grinding the areas of contact between glass and jar and greasing with desic­cator lubricant, an airtight seal can be obtained. However, abso­lute airtightness appears to be seldom essential, and the greasing of the seal is usually omitted. Whatman No. 1 filter paper has been employed in all the studies described in this bulletin. This grade of paper has proved to be most satisfactory, particularly in regard to its low cost and to its convenience in permitting the overnight resolution of chroma­tograms. Whatman No. 4 has been used to some extent, and it possesses the advantage of rapid solvent ascension. However, its greater density not only causes considerably greater diffusion and a generalized increase in Rf values, but also renders the quantitative color comparison method (p. 50) ineffective. Sheets measuring 18 x 111,4 inches have been used most fre­quently. Solutions to be chromatographed are placed 11,4 inch from the long edge of the sheet, leaving 10 inches for maximum ascent of the solvent, which is normally allowed to reach the top of the sheet. Under these circumstances, Rf values are very con­veniently measured using a transparent plastic ruler graduated in tenths of an inch. It is sometimes possible to use shorter sheets with a consequent saving in the time required for solvent ascen­sion. In the quantitative estimation of urea (p. 88), for example, a solvent ascent of about two inches is sufficient to separate the urea spot from interfering substances. It is not necessary to remove the chromatogram from the crock at the precise instant that the solvent front reaches the top of the sheet. When the solvent reaches this point, resolution of the chroma­ togram ceases, and, except for slow diffusion, no further immediate changes occur. Prolonged standing for several hours is, however, to be avoided, since the diffusion may introduce serious errors. In certain acid-alcohol solvents esterification occurs, and secondary solvent fronts appear on prolonged standing. Solutions to be chromatographed are usually placed on spots at one inch intervals along the sheet. A straight-edge with small "V" shaped notches cut into it at one inch intervals has been found convenient for marking these positions. A pencil run lightly along the edge of this ruler marks the chromatogram, and the light pencil marks do not interfere with the subsequent treatment of the sheet. Samples are applied to the paper in five microliter amounts, using a capillary pipette* (obtained from Microchemical Specialties Co., Berkeley, California). If larger quantities must be applied to a single spot, they are added in five microliter incre­ments, the spot being allowed to dry thoroughly between each application. The use of a hot plate heated to about 50-60°C. is helpful in hastening the drying. This low temperature has no dele­terious effect on the sample itself. By this procedure uniform spots of approximately one centimeter diameter are obtained. Where quantitative deductions of any sort are to be made from the developed chromatogram, it is important that the starting spots be of uniform small diameter. When large numbers of chromatograms are run daily, the most tedious and time-consuming procedure is the application of the sample to the paper. The efficiency of this operation has been considerably increased by the use of a device for stacking the sheets in such a manner that a technician may remain seated in one position while pipetting samples to a number of sheets. This device is shown in Fig. 1'. Each sheet is placed so that the line FIG-. 1 *A convenient means of cleansing these capillary-bore pipettes is to allow first w~ter and the~ acetone to be drawn through them by the aspirator. :rrolongmg. the suct10n 5-10 seconds after the acetone rinse is completed msures drymg. along which the samples are to be placed overlaps the edge of the board under it and does not touch any surface. The sheet im­mediately above is set back far enough so that it does not inter­fere with pipetting to the sheet below. As many as eight chro­matogram sheets can conveniently be prepared in this manner at one sitting. Solvents The various solvent mixtures that have been employed in this laboratory for the resolution of paper chromatograms are de­scribed below in the approximate order of their general useful­ness. Most of the compounds dealt with have been studied using each of the first five solvent mixtures described. The other sol­vents listed have also been found generally useful, with the ex­ception of those which are specifically indicated for special pur­poses. Most of the solvents listed may be used for several deter­minations, but they should be replaced with a freshly prepared solution after 48 hours. An exception to this rule occurs in solvents in which esterification is likely. In these cases, the solvent mixture should always be freshly prepared immediately before use. Phenol solvent (Phen.) : 100 g. phenol (Mallinckrodt, Analytical Grade) saturated with an aqueous solution containing 6.3% so­dium citrate and 3.7% sodium (or potassium) dihydrogen phos­phate.-About 20 ml. of the aqueous solution are required to saturate the phenol. The salts serve a dual purpose: they inhibit the diffu­sion of the spots on urine chromatograms, and they prevent the mi­gration of a contaminant in the paper which otherwise obscures the lower half of the chromatogram. Butanol-acetic acid solvent (BuAc) : 80 ml. n-butanol, 20 ml. glacial acetic acid, and 20 ml. water. This solvent must be freshly prepared for each determination. Butanol-ethanol solvent (BuEt): 80 ml. n-butanol, 20 ml. 95% ethanol, and 20 ml. water. Isobutyric acid solvent (Isobu) : 80 ml. isobutyric acid and 20 ml. water. Lutidine solvent (Lut) : 65 ml. 2,6-lutidine and 35 ml. water. Lutidine is preferred to collidine because of the greater uniformity of the commercially available product. Collidine-lutidine solvent (ColLut) : 33 ml. of the fraction of mixed collidines boiling between 158 and 165°C., 50 ml. of 2,6-lutidine, and 17 ml. water. Pyridine-butanol solvent (PyBu) : 80 ml. pyridine, 20 ml. n-butanol, and 20 ml. water. Butanol-pyrid1'.ne solvent (BuPy) : 50 ml. n-butanol, 50 ml. pyridine, and 20 ml. water. Butanol-ethanol-ammonia solvent (BuEtAm) : 80 ml. n-butanol, 10 ml. 95% ethanol and 30 ml. concentrated ammonium hydroxide. Ethanol solvent (Et) : 100 ml. 95% ethanol. Methanol solvent (Me): 95 ml. abs. methanol and 5 ml. water. Pyridine solvent (Py) : 65 ml. pyridine and 35 ml. water. Ethanol-acetic acid solvent (EtAc): 95 ml. 95% ethanol and 5 ml. glacial acetic acid. This solvent must be freshly prepared before each determination. Ethanol-ammonia solvent (EtArn) : 95 ml. 95% ethanol and 5 ml. concentrated ammonium hydroxide. Butanol-ethylene glycol-hydrochloric acid solvent (BuEGHCl) : 80 ml. n-butanol, 20 ml. ethylene glycol, and 20 ml. 0.1 N hydro­chloric acid. Phenol-acetic acid solvent (PhAc) : Same as phenol solvent ex­cept that 0.5 ml. of glacial acetic acid is added to the bottom of the earthenware crock to maintain an acid atmosphere. This sol­vent is useful in separating acidic substances. Phenol-ammonia solvent (PhAm) : Same as phenol solvent ex­cept that 0.5 ml. of concentrated ammonium hydroxide is added to the bottom of the earthenware crock to maintain an alkaline atmosphere. After 24 hours a pink discoloration is detectable in the solvent mixture, and the solution must be discarded. This solvent is useful in separating basic substances. Butanol-ethanol-hydrochloric llcid solvent (BuEtHCl) : 80 ml. n-butanol, 20 ml. 95 7o ethanol, and 40 ml. 2 N hydrochioric acid. This solvent is especially useful in the determination of cystine and histidine. Ethanol-hydrochloric acid solvent (EtHCl) : 80 ml. 95% ethanol and 20 ml. 0.1 N hydrochloric acid. This solvent has been used exclusively in the determination of sodium and potassium. Isobutyric acid-acetic llcid solvent (IsobuAc) : 80 ml. isobutyric acid, 20 ml. glacial acetic acid, and 20 ml. water. This solvent has been used for the resolution of certain unidentified urinary con­stituents. Color Developing Reagents Listed below are the reagents which have been used at this laboratory for the development of color on paper chromatograms. A few intensely colored compounds may be detected without the Paper Chromatographic Techniques use of a reagent: e.g., picric acid. Other substances may be de­tected by their fluorescence under ultraviolet light (UV). When­ever possible the reagents to be applied are prepared in alcoholic solution. This permits rapid drying and minimizes spreading of the spots which may occur after application of the reagent. When using aqueous solutions of reagents, particular care must be taken that the reagent be sprayed lightly and uniformly. Unless other­wise noted, the reagent is applied to the paper by spraying. Ninhydrin reagent (Nin.) (6) .-This reagent consists of a 0.2 % solution of ninhydrin (triketohydrindene hydrate) in water-sat­urated n-butanol. The reagent, if kept tightly stoppered in a brown bottle, may be used for as long as one month. After having been sprayed, the sheets are heated for five to seven minutes at 90°C. to promote color development. Most amino acids react to give purple spots, although there are notable exceptions. Phenylalanine, tyrosine, and aspartic acid give blue colors; tryptophane, olive­brown; asparagine, brown; proline, yellow. The reagent is not specific for amino acids; certain amines, ethanolamine, peptides, ephedrine, etc., give colors. As little as 0.2 microgram of certain amino acids may be detected. The presence of high salt concen­trations may cause rapid fading of the colors and may even prevent color development entirely. The presence of strong acids in the at­mosphere or on the chromatograms (as with the butanol-ethanol­hydrochloric acid solvent) causes reddening of the usual purple color and promotes rapid fading. Ammoniacal silver nitrate reagent (AmAg) .-Equal volumes of 0.1 N silver nitrate and 5 N ammonium hydroxide are mixed immediately before use. Chromatograms may be sprayed with or dipped into the reagent. It is advisable that the sprayer be washed immediately after use several times with 5 N ammonium hydroxide followed by washing with water to prevent clogging by deposition of silver and insoluble silver salts. Uric acid develops almost immediately as a black spot against a white background. For further development the sheet should be heated for ten minutes at 100°C., thus darkening the background to light brown. Carbohydrates then appear as dark brown spots. Chloride and phosphate produce bleached areas on the paper. The "weak acid" alkaline area (p. 110) produces a spot slightly darker brown than the background, but at high concentrations (50-100 micrograms of sodium acetate) this may also appear as a bleached area. Most of the amino acids cause a slow bleaching of the background color to white or yellow. The bleaching effect of the amino acids appears to be different from the immediate bleaching of inorganic ions. The latter seems to be due to simple interference with the contact of the reagent with the paper. As little as 0.2 microgram of uric acid and 0.5 microgram of glucose can be detected using this re­agent. The amino acids can be seen only at concentrations of 30-50 micrograms; chloride and phosphate are detectable at 5-10 micrograms. The chromatogram becomes very dark after stand­ing for several days, and the originally darkened areas can no longer be detected. This reagent cannot be used for the color development of chromatograms which have been developed in solvents containing phenol unless the phenol is previously removed by drying the sheet for at least 48 hours at room temperature. Alkaline ferricyanide-nitroprusside reagent (FCNP) (7) .­Equal volumes of 10 % sodium hydroxide, 10% sodium nitroprus­side, and 10 % potassium ferricyanide solutions are mixed. The mixture is diluted with three volumes of water. After standing for about 20 minutes the initial dark color changes to a pale yellow, and the reagent is ready for use. This reagent is quite unstable at room temperature, but may be kept for two to three weeks in the refrig­erator without deterioration. Creatinine, guanidine, glycocyamine and arginine give rise to orange colors against a light yellow back­ground. Creatinine gives a blue color. Certain other nitrogenous compounds also respond to the reagent. The colors are stable over long periods providing precautions are taken to exclude phenol vapors during development and contact with phenolic substances afterward, since these cause marked color changes on the chroma­tograms. Ferric chloride reagent (FeCl3 ) .-A 1% aqueous solution of ferric chloride is employed. The reagent may be kept for 7 to 10 days in a refrigerator. Sheets must be sprayed lightly to avoid streaking. Spots appear most clearly after the sheet has dried, except in the case of tartaric acid, which is visible as a yellow spot only on the wet sheet. Phosphate and sulfate produce white areas against a light yellow background. Amounts of phosphate and sulfate greater than 10-20 micrograms can be detected. Urea appears as a pale area slightly lighter than the background and can be detected only in fairly high concentrations (30 micrograms). Basic areas are light brown, presumably due to precipitation of fer­ric hydroxide. Phenol derivatives produce green to purple spots Aspirin appears in urine chromatograms as an intense purple spot. As little as 0.5 micrograms of salicylic acid or acetyl salicylic acid can be detected. Phenol-hypochlorite reagent (PHC) (8) .-Chromatograms are first sprayed with 5% phenol in 95% ethanol. After having been dried, the sheets are sprayed with 5.25 % sodium hypochlorite (Clorox). Colors develop immediately. The background darkens on standing. This reagent reveals a number of common urinary constituents. Urea, at concentrations above about 5.0 micrograms, appears as a yellow-green spot. Chloride, at concentrations as low as 1.0 microgram, produces a blue spot. Basic cations produce yel­low areas. Most amino acids at high concentration (25 micro­grams) produce blue or green colors. Arginine, however, yields orange, cystine and cysteine, brown, and tryptophane, pink. If urea alone is to be measured, preliminary spraying of the resolved chromatograms with 1 N sodium hydroxide solution prevents the color development of other urinary constituents. For chromatograms resolved in a phenol solvent the preliminary spraying with phenol may be dispensed with. In this case the amount of phenol remaining on the chromatogram must be con­trolled by rather careful partial drying for eight minutes (±30 seconds) at 90°C. The sheets must be sprayed immediately after this drying. a-Naphthol-hypochlorite reagent (NHC) (9) .-Chromatograms are first sprayed with a 0.1 % solution of a-naphthol in 1 N sodium hydroxide. The a-naphthol reagent may be used for as long as one to two months. After having been dried, the sheets are sprayed with a solution consisting of one volume of "Clorox" diluted with one volume of water or ethanol. Urea appears as a blue-green spot. Arginine appears as a red spot against a white background but fades rapidly. Arginine is detected only at levels above 10 micrograms. Bromcresol green indicator reagent (BCG) .-This reagent is a typical acid-base indicator, and as such can be used to detect acidic or basic substances on the chromatogram. A 0.04% solution in 95 % ethanol is employed. Before using the solution, the color should be adjusted to blue-green with dilute sodium hydroxide solution. The indicator solution is stable indefinitely. Chromato­grams developed in acidic or basic solvents must, of course, be thoroughly dried before spraying. Acids give yellow spots against a blue-green background, while bases show intensification of the blue basic color. Acidic areas on urine chromatograms correspond to chloride, sulfate, phosphate, citrate, lactate, and hippuric acid. Urea appears as a slightly basic spot, while the weak acid alkali (p. 111) gives a strong basic spot. Some metal ions, e.g., lead 30 The University of Texas and copper, give a bright pink color with this indicator. The colors developed are affected by acidic or basic substances in the atmos­phere, but are otherwise quite stable. Diazotized sulfanilic acid reagent (DSA) .-4.5 grams of sul­fanilic acid are dissolved in 45 ml. of 12 N hydrochloric acid with warming, and the solution is diluted to 500 ml. with water. A 100 ml. aliquot of this dilute solution is chilled in an ice bath and 100 ml. of a 4.5 % solution of sodium nitrite is added. This mix­ture is allowed to stand for fifteen minutes in the ice bath. This reagent may be kept in the refrigerator without deterioration for one to three days. Immediately before use an aliquot of this solution is mixed with an equal volume of a 10 % solution of sodium carbonate. This reagent is particularly useful in the de­termination of histidine and a number of unidentified substances present in urine. Pink, red, orange, and purple spots are developed against a light yellow background. The colors are stable, but con­tact with vapors of phenol, collidine, or lutidine causes darkening of the background. The minimum concentration of histidine de­tectable is about 0.2 microgram. p-Dimethylaminobenzaldehyde reagent (PDAB) .-2.0 grams of p-dimethylaminobenzaldehyde are dissolved in 100 ml. of 1.2 N hydrochloric acid. Storage in a refrigerator will keep this reagent three to five days without deterioration. Urea and allantoin give bright yellow spots against a white background. The colors do not fade on drying or standing. The minimum concentration of urea which is detectable is 5 to 10 micrograms. Bromine reagent (Br2 ) (10) .-0.5 ml. of liquid bromine is added to 50 ml. of glacial acetic acid and 50 ml. of water. This reagent in stable for two or three weeks if kept in the refrigerator. After the chromatograms are sprayed with this mixture they are heated in an oven at 90°C. for three to five minutes. Histidine may be detected by this procedure in amounts of 25 micrograms or more. On exposing the damp sheet to ammonia vapor, the brown histidine spot becomes purple and the sensitivity is in­ creased so that 10 micrograms become visible. The histidine spot does not fade. In addition to histidine, a number of unidentified urinary constituents show up with this reagent as pink, brown and green spots. Some of these spots show a bright yellow flu­ orescence under ultraviolet light as well. Their ability to fluoresce decreases after several days. Azide-iodine reagent (Azl) (11).-50 ml. of a 0.1 N iodine so­ lution (aqueous solution prepared using potassium iodide to effect solution) is added to 50 ml. of 95 % ethanol. 1.5 g. sodium azide is dissolved in this solution. If kept tightly stoppered in brown bottles, this reagent is stable indefinitely. This reagent is par­ticularly useful for the detection of sulfur-containing amino acids. Methionine, cystine, and cysteine decolorize the reagent imme­diately, appearing as white areas against a brown background. After an hour these spots show considerable fluorescence under ultraviolet light. These amino acids may be detected in concentra­tions of about five micrograms or more. Spots must be marked soon after development, because within a few hours the back­ground fades to give a uniformly colored sheet. Rose and yellow spcts due to unidentified constituents then appear on urine chromatograms. Mercuric nitrate-ammonium sulfide . reagent (HgS) (12).­Chromatograms are dipped into a 0.25 M solution of mercuric nitrate in 0.5 N nitric acid and then heated for ten minutes at 80°C. They are next washed by dipping in 0.5 N nitric acid and then in water. After having dried at room temperature the sheets are dipped in a solution of one part ammonium sulfide (reagent strength) and four parts water. Purines and pyrimidines give black spots against a white to gray background. The m1mmum amount detectable is from 5 to 10 micrograms. The spots are permanent. Alizarin-ammonia reagent (AlAm) .-Chromatograms are sprayed with a 0.1 % solution of alizarin in 95 o/o ethanol and then exposed to ammonia vapor. The reagent, tightly stoppered, is stable indefinitely. Calcium and sodium salts of weak acids appear as blue to purple spots against a lavender background. Acid areas appear as yellow immediately; the background fades to yellow on standing. All these substances can be detected in amounts as low as 5 micrograms. Picric acid reagent (Pie) .-Chroma'ograms are sprayed with 0.5 N sulfuric acid and heated for one hour at 100°C. They are then sprayed with a 1.3 o/o solution of picric acid in 95 o/o ethanol which is combined immediately before use with one-fifth its volume of 10% sodium hydroxide. Creatine and creatinine appear as orange spots against a yellow background. Allantoin gives a faint orange spot. A number of other urinary constituents cause a slight fading of the yellow background color. The colors de­veloped for creatinine and creatine are permanent. As little as 0.2 microgram is detectable. If creatinine alone is to be deter­mined, the preliminary hydrolytic treatment with sulfuric acid may be omitted. Hydrolytic reagent (Hyd) .-Following the procedure outlined above but omitting the picric acid spray, a pink color is developed with indole and a gray color with tryptamine. In addition, sev­eral unknown urinary constituents give pink, blue, and green colors. 2,6-Dichlorophenolindophenol reagent (DCPI) .-A 0.4 % solu­tion of 2,6-dichlorophenolindophenol in ethanol is sprayed on the chromatograms. The reagent is stable indefinitely. This reagent acts as an acid-base indicator as well as an oxidation-reduction indicator. Initially the background is blue with acidic areas ap­pearing as pink spots. The most conspicuous spot in urine is chloride (pink). The weak acid alkali spot (p. 111) appears as a darker blue area. Although amino acids do not appear as acidic or basic substances, they become apparent as completely bleached areas in from one hour to several days, depending on the con­centration present. Ascorbic acid causes almost immediate bleach­ing. Creatinine and creatine appear as bleached areas in about 45 minutes. Lactic and hippuric acids, as well as chloride at fairly high concentrations, cause slow bleaching. Urea may appear as a blue area at high concentrations. The bleached areas remain unchanged indefinitely, although the background changes from blue to pink in forty-eight to seventy-two hours, necessitating the marking of the acid areas before this change occurs. Most of the substances mentioned above are detectable in amounts above about 5 micrograms. Zinc uranyl acetate reagent (ZnUAc) (13) .-5.0 grams of uranyl acetate are mixed with 3 ml. of warm 30 % acetic acid and the mixture diluted to 25 ml. A second solution is prepared, con­sisting of 15 g. of zinc acetate mixed with 3 ml. of 30 % acetic acid and diluted to 25 ml. The two solutions are mixed and warmed gently. Sodium chloride (0.1 g.) is added to the warm solution and the reagent allowed to stand for 24 hours. It is then filtered and is ready for use. This reagent is stable indefinitely. It is used exclusively for the determination of sodium (p. 93). Lead cobaltinitrite reagent (LCN) (14) .-11.5 g. of lead nitrate and 15 g. of sodium nitrite are dissolved in a small amount of water. When these are completely dissolved, 10 g. of cobalt nitrate are added and the solution diluted to 100 ml. with water. This solution is allowed to stand for at least one hour before filtering to remove the orange-brown precipitate. A 1 :2 dilution of the filtrate is prepared immediately before spraying. The reagent is stable for one to two days at room temperature. Contact with Paper Chromatographic Techniques the vapor of the reagent while spraying should be avoided. Potassium and ammonium ions produce brown spots against a light yellow background. The determination of potassium by means of paper chromatography using this reagent is reported elsewhere by Beerstecher (15). 2,6-Dichlorquinonechloroimide reagent (DCC).-The reagent is prepared as a 1% solution of 2,6-dichloroquinonechloroimide in 95 % ethanol. When stored in the refrigerator the reagent is stable for two to three weeks. On sheets sprayed with this reagent alkaline areas appear as blue spots, uric acid as a yellow spot, and creatinine as a brown spot. On standing several hours certain unidentified substances in urine appear as brown, pink and red spots on the chromatogram. Contact with phenolic substances must be strictly avoided, and sheets which have been sprayed with DCC should be kept apart from those developed with other reagents because of interaction between this reagent and others, especially sulfanilic acid, the ferricyanide-nitroprusside reagent, and p-di­methylaminobenzaldehyde. Aniline acid phthalate reagent (AHP) (16) .-To 100 ml. of water-saturated n-butanol are added 1.66 g. phthalic acid and 0.93 g. aniline. The reagent is stable for two to three months at room temperature. On sheets sprayed with this reagent and heated five minutes at 100°C., aldopentoses appear as red spots and hexoses as green to brown spots. Orcinol reagent (Ore) (17) .-To prepare the reagent 0.10 g. orcinol in 10 ml. of 95 % ethanol is added to 90 ml. of 2 N hydrochloric acid containing 0.01 % ferric chloride. The reagent should be prepared immediately before use. The sheets are sprayed and heated ten minutes at 85-95°C. Pentoses produce green spots. Heating must be carefully controlled to prevent de­struction of the paper by the strong acid. Naphthoresorcinol reagent (NR) (17) .-To 90 ml. of 2% trichloroacetic acid is added 0.100 g. naphthoresorcinol dissolved in 10 ml. 95 % ethanol. The reagent must be prepared immediately before use. The sheets after having been sprayed are heated ten minutes at 85-90°C. Hexoses produce blue and brown colors. Naphthylethylenediamine reagent (NED) .-Chromatograms are sprayed with a mixture of equal volumes of 0.5 % sodium nitrite and 0.5 N sulfuric acid. After drying, the sheets are sprayed with a 0.1 % solution of N-1-naphthylethylenediamine hydrochloride. This reagent is useful in detecting diazotizable amines. No such compounds were observed in a large number of urine chromato­grams. 34 Compounds Studied Table I is a listing of compounds studied in this laboratory on paper chromatograms. Included in the table are the Rf values obtained for the pure compounds in the various solvents employed, the color developing reagents employed, the color developed, and, where such information has been obtained, the minimum amount of the compounds which can be detected on paper chromatograms using a particular color developing reagent. The abbreviations employed in the table for solvent mixtures and color developing reagents have been indicated in the preceding sections. More de­tails concerning the characteristics of the colors developed, back­ground color of the sheet, fading, etc.. can be obtained by referring TABLE I PAPER CHROMATOGRAPHIC BEHAVIOR OF COMPOUNDS IN PURE SOLUTIONS A. Amino Acids and Amino Acid Derivatives Rf Values in Common Solvents Other Color Min. Solvents Rea- Amt. Compound Phen BuAc BuEt lsoBu Lut Rf Solvent gent Color µg. alanine ----------­-.55 .38 .16 .40 .18 .21 Et Nin purple 0.5 .46 Me .60 Py .53 EtAc .52 EtAm .39 .22 PyBu BuPy .43 BuEtHCI ,8-alanine -----­---­.55 .37 .09 .41 .13 .11 Et Nin blue 0.5 .30 Me .41 Py a-aminobutyric acid ----­------­.65 .45 .42 .53 .18 .60 Et Me Nin purple .64 .55 Py EtAc 'Y-aminobutyric acid -----------­.78 .50 .11 .50 .15 .14 .40 Et Me Nin purple .39 .75 Py EtAc .44 EtAm .25 .12 PyBu BuPy a-aminocaprylic acid ----------­-.74 .85 .21 .55 .031 Et.59 .08 Me.75 Nin purple .17 .86 Py TABLE I-Continued A. Amino Acids and Amino Acid Derivatives (Continued) Rf Values in Common Solvents Other Color Min. Solvents Rea- Amt. Compound Phen BuAc BuEt IsoBu Lut Rf Solvent gent Color µg. a-aminoisobu­ tyric acid ____ .68 .48 .44 .54 .25 .36 .70 Et Me Nin purple .71 Py arginine -­-----­ .41 .20 .04 .23 .07 .03 .05 .19 Et Me Py Nin PHC NHC FCNP purple orange pink red 1.0 10.0 asparagine ---­ .29 .19 .06 .27 .09 .03 .09 .35 Et Me Py Nin brown-purple 1.0 aspartic acid ____ .07 .24 .04 .20 .09 .03 .40 Et Py Nin blue 0.5 .28 EtAc .15 EtAm .08 .04 PyBu BuPy citrulline ____ ___ .56 .25 .08 .41 .13 .02 .15 Et Me Nin purple 3.0 .44 .23 Py EtAc .33 EtAm .06 .07 .32 PyBu BuPy BuEtHCl cysteine ____________ .19 .07 .09 .07 .10 Et Nin brown .57 .23 .12 Me Azl decolors 5.0 .08 Py cystine ____________ 0.08 .08 .02 .10 .06 .03 Et Nin brown 1.0 18 .05 .03 Me Py Azl decolors 5.0 .17 BuEtHCl a, 'Y-diaminobu­tyric acid ____ .25 .12 .05 .15 .06 .03 .03 Et Me Nin purple 2.0 dihydroxyphenyl­alanine ________ .30 .24 .15 .25 .25 Nin FeCla purple green diiodoty­rosine .80 .70 .54 .78 .75 Nin DSA Br, purple orange yellow* glutamic acid __ .16 .30 .04 .31 .12 .04 .20 Et Me Nin purple 0.5 .45 Py .32 EtAc .10 EtAm .21 .08 PyBu BuPy .40 BuEtHCl glutathione ____ .25 .05 .10 .08 .03 .15 Et Me Nin purple .49 Py 36 The University of Texas TABLE I-Continued A. Amino Acids and Amino Acid Derivatives (Continued) Rf Values in Common Solvents Other Col-Or Min. Solvents Rea-Amt. Compound Phen BuAc BuEt IsoBu Lut Rf Solvent gent Color µg. glycine __________ 0.30 .26 .10 .29 .14 .08 Et Nin purple 0.25 .24 Me PHC blue 25.0 .44 Py .38 EtAc .38 EtAm .23 PyBu .10 BuPy .28 BuEtHCl glycylglycine __ .39 .22 .06 .27 .49 Nin purple hippuric acid __ .75 .93 .81 .80 PhAc BCG yellow histamine ________ .52 .22 .18 .03 .33 .34 Et Nin brown str. .20 Me DSA rose­.71 Py brown .80 BuEtAm PHC yellow histidine --------.55 .20 .08 .29 .11 .31 Et Nin brown- purple 3.0 .15 Me DSA red 0.2 .47 Py Br2 brown­.12 EtAc purple .36 EtAm .23 PyBu .12 BuPy .18 BuEtHCl .08 BuEtAm homoserine __ _ .47 .30 .40 .20 .13 Et Nin purple.34 Me .60 Py hydroxy­proline ------.59 .30 .09 .37 .22 .10 Et Nin yellow­.30 Me brown 2.0 .60 Py isoleucine ________ .79 .72 .44 .81 .45 .49 Et Nin purple.82 Me kynurenin ________ .43 .76 .62 .26 Nin purple .66 leucine -----------.79 .73 .42 .82 .43 .50 Et Nin purple 2.0 .82 Me .80 Py .82 EtAc .69 EtAm .68 PyBu .51 BuPy lysine ------------.39 .14 .04 .12 .02 .07 Et Nin purple .25 .04 Me .13 Py .17 EtAc .31 EtAm .02 PyBu .02 BuPy .18 BuEtHCl TABLE I-Continued A. Amino Acids and Amino Acid Derivatives (Continued) Rf Values in Common Solvents Other Color Min. Solvents Rea- Amt. Compound Phen BuAc BuEt lsoBu Lut Rf Solvent gent Color µg. methionine ______ .73 .55 .29 .70 .35 .30 Et Nin purple 1.0 .31 Me Azl decolors 5.0 .76 Py .25 EtAc methionine sulfone ________ .53 .28 .13 .39 .24 .08 .26 Et Me Nin purple .69 .27 Py EtAc .44 EtAm .42 .21 .35 PyBu BuPy BuEtHCl methionine sulfoxide ---­.72 .25 .11 .49 .14 .07 .55 Et Me Nin purple .59 Py .29 EtAc .42 EtAm .27 .11 .30 PyBu BuPy BuEtHCI norleucine ______ .84 .74 Nin purple norvaline ________ .73 .65 .42 .72 .31 .41 .64 Et Me Nin purple .74 Py ornithine -------­ .27 .15 .05 .06 Nin purple phenylalanine __ .78 .68 .30 .74 .49 .54 Et Nin blue . 2.5 .53 Me .75 Py proline -----------­.85 .43 .18 .61 .24 .20 .43 Et Me Nin yellow 2.5 .62 Py serine ---­----­ 0.24 .27 .10 .28 .16 .07 .20 Et Me Nin purple 0.5 .58 .28 Py EtAc .42 EtAm .28 .13 .32 PyBu BuPy BuEtHCl taurine ---·-­----­.29 .19 .12 .17 .27 .10 .35 Et Me Nin purple .72 .18 Py EtAc .53 EtAm .49 .22 PyBu BuPy threonine ________ .39 .35 .14 .40 .22 .10 Et Nin purple 1.0 .35 Me .65 Py 38 The University of Texas TABLE I-Continued A. Amino Acids and Amino Acid Derivatives (Continued) Rf Values in Common Solvents Other Color Min. Solvents Rea-Amt. Compound Phen BuAc BuEt lsoBu Lut Rf Solvent gent Color µ.g. trimethylala­nine ( neo-leu­ cine) ________ .81 .66 .50' .78 .78 Py Nin purple tryptamine ----.85 .73 .55 .89 .85 Nin brown FCNP white DCC pink PDAB purple Hyd gray tryptophan ______ .66 .50 .29 .71 .50 .16 Et Nin brown-purple 3.0 .62 Me DSA red-orange .59 Py PHC pink tyrosine ________ .52 .45 .19 .43 .45 .18 Et Nin blue 3.0 .42 Me DSA rose .82 Py FeCI. brown .15 BuEtAm .38 BuEGHCl valine ___ ________ 0.64 .60 .29 .70 .29 .41 Et Nin purple 2.0 .65 Me .74 Py .73 EtAc .62 EtAm .58 PyBu .38 BuPy .76 BuEtHCl B . Sugars and Sugar Derivatives arabinose --------.49 .42 .25 .18 .52 AHP brown Ore blue AmAg brown fructose ----------.53 .39 .24 .17 .47 NR green Ore green AmAg brown galactose --------.44 .16 .18 .12 .43 AHP brown NR blue AmAg brown galacturonic acid --------------.69 .17 .10 .07 .26 BCG yellow PHC brown glucosamine ____ .34 .17 .15 .07 .42 Nin purpleAmAg brown glucose ------------.35 .16 .20 .12 .53 AHP brown AmAg brown 0.5 lactose ------------.37 .35 AHP brown AmAg brown mannose ----------.33 .21 .25 .13 .52 AHP brown NR brown AmAg brown xylose ------------.41 .27 .29 .17 .58 AHP red Ore green AmAg brown TABLE I-Continued c. Vitamins and Vitamin Derivatives Rf Values in Common Solvents Other Color Min. Solvents Rea- Amt. Compound Phen BuAc BuEt lsoBu Lut Rf Solvent gent Color µg, p-aminobenzoic acid ------------­.84 .86 .82 .80 .78 .80 PhAc FeCl3 brown 1.0 BCG yellow PHC brown ascorbic acid __ .20 .45 .37 .53 .38 PhAc BCG AmAg DCPI yellow black pink- bleaches Nt-methylnico­tinamide _____ .84 UV dark nicotinamide _ _ .77 .75 .90 FCNP yellow nicotinic acid __ .95 .79 .94 .80 .93 DCPI decolors pantonine ________ .83 .43 .39 .62 .24 .45 Py Nin purple pantothenic acid -------­---­ .81 .84 PhAc BCG PHC yellow blue pyridoxine ________ .32 .73 .74 .53 .38 BuEtAm FeCl, brown 1.0 DSA purple riooflavin -----­--.83 .26 .15 .53 .73 UV yellow* thiamine ---------­.25 .36 .30 .55 BuEGHCl DSA red- FCNP orange * trigonelline ____ .94 PHC brown D. Alkaloids brucine ____________ .70 .53 BCG blue ephedrine ________ .95 .86 .69 .80 .85 Nin purple hordenin sulfate -------­.95 .76 .46 .85 .96 Nin purple- brown DSA red PHC · brown scopolamine _ .98 .68 .66 .99 BCG blue theobromine ____ .81 PHC green E. Purines and Pyrimidines adenine ___ ________ .94 .70 .70 .76 Py HgS black .67 BuEtAm DSA pink guanine ___________ .60 .35 .23 .52 .78 .12 .20 Py BuEtAm BuEGHCl HgS DSA black orange-purple hypoxanthine _ .86 HgS black .55 2-methyl-5-amino­methyl-6-amino­pyrimidine __ .91 .95 .95 .96 .89 AmAg black uracil -------------­.80 .45 HgS black PHC green 40 The University ·of Texas TABLE I-Continued E. Purines and Pyrimidines (Continued) Rf Values in Common Solvents Other Color Min. Solvents Rea-Amt. Compound Phen BuAc BuEt lsoBu Lut Rf Solvent gent Color µg, uric acid ________ .25 .25 .06 .27 .33 PhAc AmAg black 0.2 DCC yellow PHC orange xanthine --------.09 .01 streak .08 BuEtAm HgS black .13 BuEGHCl DSA orange F. Aliphatic Carboxylic Acids adipic acid ----.89 .91 .83 BCG yellow citric acid __ _____ .22 .40 .45 .35 PhAc BCG yellow FeCla white fumaric acid ____ .87 .49 PhAc BCG yellow a-ketoglutaric acid .38 .38 BCG yellow lactic acid ______ .70 .85 .69 .70 PhAc BCG yellow FeCla white malic acid __ _____ .34 .65 .29 BCG yellow malonic acid ____ .57 .72 .36 BCG yellow oxalic acid ----.26 .30 .14 .35 PhAc BCG yellow pyruvic acid ____ .39 FeCI. white succinic acid --.75 .83 .68 BCG yellow tartaric acid __ .21 .25 .15 .35 PhAc BCG yellow FeCia yellow (when wet) AmAg black G. Miscellaneous Organic Compounds acetamide _______ .30 PHC green allantoin -------­ .45 .27 .23 .34 PRC FCNP green orange PDAB yellow 5.0 5 (4) -amino­4 (5) imidazole carboxamide .93 .55 .63 PHC Nin pink yellow anthranilic acid -------------­.90 .93 .93 .92 .74 UV blue* FCNP PDAB green red DCC brown cadaverine ______ .55 catechol _____ _____ .17 .95 .09 .05 .74 .62 .97 .01 Py BuEtAm Nin DSA purple purple .98 BuEGHCl choline -·-­-------­.97 .45 .27 .23 .97 BCG blue creatine ---------­.91 .36 .10 .36 .27 · Pie FCNP orange red 2.0 FeCla brown TABLE I-Continued G. Miscellaneous Organic Compounds (Continued) Rf Values in Common Solvents Other Color Min. Solvents Rea-Amt. Compound Phen BuAc BuEt lsoBu Lut Rf Solvent gent Color µg . creatinine --------.90 . 55 .30 .56 Pie orange 2.0 DCPI decolors (after 45 mins.) DCC brown FCNP blue FeCl. brown cysteic acid ----.-03 .06,.09 .00 .02 .12,.18 Nin purple epinephrine ----.76 .47 .28 .32 .78 Br, yellow* AmAg black FCNP red DSA pink DCC brown DCPI decolors ethanolamine _ .65 .44 .24 .70 .35 .35 Et Nin purple streak streak .69 Me .74 Py streak gallic acid ______ .07 .65 .63 .26 .05 AmAg black Feel. purple FCNP purple DCC brown glycocyamine __ .35 .08 .25 FCNP pink guanidine ________ .66 .54 .25 .77 FCNP orange FeCla green PHC green hippuric acid __ .75 .93 .81 .80 PhAc ECG yellow hydroquinone __ .95 .67 .97 Py DSA purple­.22 BuEtAm · brown .97 BuEGHCl indole -------------.97 .95 .96 .96 FeCI. red PDAB purple DCC red Hyd pink indole-3­acetic acid ____ .96 .92 .90 .90 .96 Br, red-yellow FeCl, red PDAB purple DCC brown isatin ------------.89 .88 .87 FCNP blue- green leucinol ------------.31 .02 .02 .32 Nin purple FeCla purple neurine hydrate __, _____ .92 .84 .82 .90 FCNP red-purpleAmAg black phenobarbital _ .85 PRC green TABLE I-Continued G. Miscellaneous Organic Compounds (Continued) Rf Values in Common Solvents Other Color Min. Solvents Rea- Amt. Compound Phen BuAc BuEt lsoBu Lut Rf Solvent gent Color µg. phenyl pyruvic acid ------­------­ .86 .56 .89 FeCL blue picric acid ________ .85 .55 .60 .36 .88 none yellow quinine --------­--­ .81 .77 .86 UV blue* salicylic acid __ .85 .95 .55 .75 .72 .88 .84 .28 Py PhAc BuEtAm ECG FeCI, DSA yellow purple* yellow 5.0 .~,3 BuEtHCl UV blue* sulfadiazine ____ .92 .73 .69 .53 .84 PHC brown sulfa­ guandine --­-.80 .42 .43 .78 PHC orange sulfanilamide __ .70 .60 .53 .48 .82 PHC purple sulfathiazole -­ .88 .72 .68 .58 .84 PHC yellow-brown taurine --­--------­.29 .19 .12 .17 .27 .10 Et Nin purple .35 Me .72 Py .18 EtAc .53 EtAm .49 .22 PyBu BuPy thymol ____________ .73 .97 DSA red urea ---------------­ .78 .55 .52 .75 .43 .49 BuEtHCl PHC PDAB FeCI, green yellow white 5.0 5.0 50.0 *Substance appears fluorescent when viewed under ultraviolet light. to the discussion of the particular color developing reagent in the preceding section. Where the spots exhibiting fluorescence under ultraviolet light have been utilized in our studies this is noted in Table I by an asterisk. Undoubtedly many of the spots not so in­dicated in the table also fluoresce. Color developing reagents listed in Table I are only those which have actually been used in this laboratory for the detection of the compound in question. In many cases, other color developing reagents listed in the preceding sec­tion might have been used. Where more than one color reagent is listed for a given compound, the one listed first is the one which has been found most useful in our experience. It should be pointed out in connection with Table I that Rf values under conditions ordinarily employed in the production of paper chromatograms, may vary by a few percent. The variation may be much greater when the same compound is measured in the presence of varying high concentrations of other substances, an effect which will be discussed in detail later. For this reason the absolute Rf values listed in Table I are of less significance than the relative values of different substances with respect to one an­other, which remain the same in a given solvent. In a few cases two Rf values are recorded for a compound resolved in a given solvent. As will be pointed out later, this may be a real effect in the case of certain compounds capable of existing in different ionic states. In the case of certain of the less common compounds listed in Table I the effect may be due to the presence of impurities in the sample chromatographed. Table II lists information similar to that found in Table I ex­cept that the Rf values reported have been measured on chromato­grams prepared from urine samples rather than from solutions of the pure compound. Section A of Table II deals with inorganic ions present in urine. As will be discussed in detail later (p. 49), the paper chromatographic behavior of inorganic ions cannot be defined except in relation to other substances present in the sample chromatographed. Section B lists a number of as yet unidentified substances which appear quite consistently on urine chromato­grams. Complications Arising in the Analysis of Biological Material While the interpretation of paper chromatograms is a fairly simple process when dealing with pure solutions or simple mix­tures, the complexities of most biological fluids serve to introduce many complications. Difficulties arise from the presence, in rela­tively high concentrations, of substances other than those to which attention is being directed. Thus, the larger amounts of urea and inorganic salts in urine complicate the paper chromatographic study of this material, while in blood the high level of sodium chloride and the large amounts of protein constitute major com­plicating factors. Saliva is remarkably free of interfering sub­stances, and its paper chromatographic study is relatively simple and straight-forward, even though rather large volumes (up to 0.25 ml.) must be chromatographed to detect many of the con­stituents. TABLE II A. Inorganic Ions in Urine Rf Values in Common Solvents Other Color Solvents Rea-Ion Phen BuAc BuEt lsoBu Lut Rf Solvent gent Color ammonium ______ .72 EtHCl LCN brown calcium .05 .10 .19 AlAm purple chloride ----------.25 .25 .20 .30 .17 PhAc PHC blue BCG yellow AmAg gray phosphate ________ .05 .17 .10 FeCl. white potassium _ ______ .19 .15 .50 EtHCl LCN brown BCG blue sodium .19 .30 .15 .30 PhAc ZnUAc blue* .68 EtHCl BCG blue sulfate .20 .02 .04 .33 PhAc FeCl, white BCG yellow B. Unidentified Urinary Constituents 1 ________:_______________ .44 .42 .60 .55 .17 .80 .44 .89 IsobuAc BuEtHCl Br, FeCl, green; yellow* blue .82 .37 Py BuEtAm DCC PDAB orange-brown red DCPI decolors DSA yellow (fades with Hyd NED Na,CO,) green blue 2 -----------------------­ .26 DSA pink 3 -----------------------­ .65 .21 .91 BuEGHCl DSA pink 4 ------­--­-------------­ .85 .54 ;95 .23 BuEtHC BuEtAm DSA orange .95 Py 5 ----------­------------­ .90 .71 .98 BuEtHCl DSA red .24 BuEtAm FeCl. brown .95 BuEGHCl PDAB brown .65 Py NED Azl pink decolors 6 ------------------­----­ .21 .30 BuEtHCl DSA white* 7 ---------------­-------­ .06 DCC pink 8 ------------------------­ .95 PDAB purple 9 --------­--------------­ .80 PDAB purple (rat urine) 10 (No. 16 on map of amino acids, (p.-) .68 .30 .60 PhAm Nin purple *!Substance appears fluorescent when viewed under ultraviolet light. Paper Chromatographic Techniques Interfering substances manifest themselves in a variety of ways : simple masking of spots, distortion of Rf values, distortion of spot size, multiple Rf values, and streaking. In addition the behavior of inorganic ions may lead to much confusion. These complications will be discussed in the following paragraphs, with especial reference to their effect on the paper chromatography of urine. Simple interference (masking of spots) .-In a given solvent, the Rf values of two or more interfering substances having simi­lar color reactions may be the same. In such an event these substances will be indistinguishable and a different solvent must be used in which the Rf values of the compounds in question are not identical. When separating complex mixtures of similar sub­stances, such as amino acids in urine, it may often be impossible to find a solvent mixture which will not result in interference between certain components of the mixture. In such a case the two-dimensional chromatogram may often be employed to ad­vantage. The application of such a procedure to urine and saliva samples is described elsewhere in this bulletin (p. 72). When. overlapping spots react differently to the color develop­ing reagent employed, confusing color distortion may occur, or the development of color with certain compounds may be pre­vented entirely. For example, in urine chromatograms prepared using the butanol-acetic acid solvent, the presence of high salt concentrations may completely inhibit the development of the cystine spot with the azide-iodine reagent. When recognized, in­terference of this type may be overcome by the use of a different solvent mixture or a different color developing reagent. Distortion of Rf Values.-The presence of other substances in relatively high concentrations may have a very appreciable effect on the Rf value of a compound. This is illustrated in Table II where the Rf values of various urinary constituents in pure solu­tion and in urine are compared. Urea, creatinine, the acidic and basic ions, as well as uric acid, may cause a generalized lowering of the Rf values from the values of pure compounds in the same solvent. In the case of urea, if the Rf value of a pure compound is either .1 or .2 Rf unit higher or lower than that of urea, the Rf value of the substance in the presence of urea may be considerably lower than for the pure substance. Where the Rf value of a compound very nearly coincides with that of an interfering substance present in high concentration, the spot may actually be split and appear partly above and partly below the position occupied by the interfer­ The University of Texas ing substance. The judicious selection of a different solvent mixture can frequently resolve these difficulties. TABLE III Comparison of Rf Values of Substances in Urine and in Pure Solution Rf Values in Compound Solvent Pure Solution Urine Alanine_________ _ ____________ ___ _ ____ ____ Phen .57 .53 Arginine·------------------------------------Phen .41 .39 Asparagine__ ____________________________ Phen .34 .30 Aspartic acid__________________ ________ _ Phen .07 .05 Chloride_________________ _______ _____ _____ _ Phen .25 * Creatine______________________________________ BuEt .21 .10 Creatinine________________________________ _ BuEt .49 .37 Cystine_ _____ ______ ______________ _______ Phen .12 .01 Glutamic Acid____________ __ ___________ Phen .21 .23 Glycine________________________________________ Histamine_______ __ _______________________ _ Phen BuAc .33 .22 .28 .10 Histidine ----------------------------------­ Phen .53 .47 Leucine_ ______ ______ ____ _ ___________ _ Lysine______________________________________ _ Methionine______________________________ _ Norleucine_________ __ _ _______________ ___ _ Norvaline________________________________ _ Ornithine_________________________________ __ BuAc Phen Phen Phen Phen Phen Phen .20 .84 .36 .73 .84 73 .27 .05 and .15 .81 .30 .70 .84 .73 .22 Phenylalanine_ ________________________ _ Phen .78 .74Phosphate_________________________ __ ___ _ _ Phen .02 * Proline______ _ ___________ _ _________________ _ Phen .85 .82 Serine______________________ __ ______________ _ Phen .27 .22 Sodium-potassium________ ____ __ __ _ Phen .19 * Sulfate_______________________________________ _ Phen * .20 Tartaric acid__________________ _ __ __ _ BuAc .45 .24 Threonine__________ _ ______ ______ _ ____ _ Phen .43 .39 Urea_ _________________________________________ _ Phen .78 .78 Uric Acid________________ _ ___________ _ Phen .25 .20 Valine_______________ _ __ ______ _ _ _____ _ Phen .77 .73 "See Table IV. The complete elimination of difficulties of this sort in the case of urine can be accomplished by removal of urea by treatment with urease, and by desalting through use of ion-exchange adsorbents. This is a somewhat laborious and time-consuming procedure and renders the urine particularly susceptible to bacterial decomposi­tion. An additional means of minimizing many of the Rf distor­tions which are mentioned above is through the use of the phenol solvent. The variations from the Rf values of pure substances, while still present, are much smaller than for other solvents (see Table Ill) . The most practical procedure, which has been employed routinely in this laboratory for quantitative work involves adding meas­ured small amounts of urine to the standard spots applied to the chromatograms. Thus a fairly uniform distortion of all Rf values is obtained. A more complete explanation of this technique will be found in the section of this paper dealing with quantitative measurements. Spot size distortion.-As will be discussed later in detail, the size of a developed spot on a chromatogram is, under properly controlled conditions, indicative of the amount of substance pres­ent. This relationship between concentration and spot size may be considerably altered by the presence of interfering substances in the sample chromatographed. In general, the effect is to pro­duce an increase in spot size, although in some cases partial inter­ference in color development, or "crowding" by other substances, may result in a marked diminution in the size of the spot when compared to the pure substance at the same concentration. In urine the chief offenders are the inorganic ions, chloride, phosphate, sodium and potassium. The effect is probably due to the hygroscopic nature of these substances on paper, which results in localized "waterlogging." This effect may be directly observed dur­ing the drying of the chromatograms, the areas of high salt concen­tration being the last to dry. With many solvents these salts are not very well localized on the chromatogram but form diffuse streaks, with the result that other constituents of urine with Rf values within the rather wide range of these streaks will form diffuse spots or streaks. This effect is largely eliminated in the case of the phenol solvent by the presence of ammdnia or acetic acid in the developing chamber, which serves to localize the inorganic salt spots. Where the phenol solvent interferes with color development a solvent mixture may be employed in which the Rf values of the com­ pounds of principal interest fall well outside the range of the salt streaks. Multiple Rf Values.-It is ordinarily assumed that the appear­ance of two spots on a chromatogram indicates the presence of two distinct compounds. Such is not always the case. Two discrete spots may be obtained from a single substance, or, more fre­quently, diffuse streaks with what appear to be areas of localized concentration. Aronoff has reported (18) that by varying the pH of aqueous solutions of lysine, a plurality of spots can be produced using a saturated phenol-water system. This effect he ascribes to an association between basic lysine ions and phenol. In this laboratory we have noted a somewhat similar effect in the case of urinary histidine using the butanol-acetic acid solvent. Two separate and distinct spots appear corresponding to histidine in urine, while aqueous histidine solutions result in only a single spot. Addition of histidine to urine previously showing no histi­dine likewise results in two spots. Whatever the precise explana­tion for such phenomena, it would seem to involve the possibility for existence of multiple ionic species. The appearance of two spots as a result of partial interference by a second substance has already been mentioned. Using the butanol-acetic acid solvent, creatinine usually appears comciaent with, or slightly below, urea. When fairly large concentrations of both are present (200 micrograms of urea and 25 micrograms of creatinine) creatinine appears as two spots, above and below urea. If the concentration of creatinine is further increased, it will again appear as a single spot above urea. A somewhat similar effect was encountered in separations of malic acid and potassium ion on chromatograms (five inches in length) using the butanol­acetic acid solvent. Color development with bromcresol green indicator reagent disclosed two acid areas, the lower of which was enclosed in a basic ring. Apparently in this instance the positions of malic acid and potassium partially overlap, and the basic potassium spot entraps part of the malic acid. This inter­ference was prevented by employing smaller samples or by the use of a longer sheet for the chromatogram. If multiple spotting is due to the existence of multiple ionic species, the use of a solvent which incorporates a strong acid or-base serves to concentrate the material in a single distinct spot. The use of hydrochloric acid in the butanol-ethanol-hydro­chloric acid solvent in the analysis of histidine is an excellent example of this technique. Where the multiplicity of spots is due to interference by a second substance, the judicious selection of a different solvent mixture should remedy the situation. Streaking.-Diffuse streaks rather than discrete spots are not infrequently encountered on paper chromatograms. This may be due to any of several causes. As was previously noted, localized "waterlogging" due to the presence of a relatively high concentra­tion of inorganic salts may be one cause. The possession of multiple ionic species as in the case of lysine and arginine may also be responsible. Substances which ordinarily migrate as discrete spots may form streaks when present in higher concentrations. Streak­ing is most frequently observed when using solvents which con­tain a relatively high percentage of water. Except where it is due to high concentrations of the migrating substances, streaking may be prevented by the use of a different solvent. Behavior of inorganic ions.-No thorough study has been made in this laboratory of the apparent erratic behavior of many inor­ganic ions on paper chromatograms. An investigation of the be­havior of these ions is certainly needed and would probably do much to establish a sound theoretical basis for understanding paper chromatography. Such an understanding can scarcely be said to exist at the present time. Our interest has been limited largely to the behavior of these ions and their effects on other substances on urine chromatograms. Unlike most of the organic compounds studied, a particular in­organic ion cannot be expected to migrate consistently to the same position when developed with a given solvent. The position of the anion, to some extent at least, is determined by, as well as determines, the position of the cation. The situation is well illus­trated by the data of Table IV. The relative positions of the acidic and basic spots were determined using bromcresol green indicator reagent, and confirmed by specific reagents for the cation or anion involved: silver nitrate for chloride, ferric chloride for phosphate and sulfate, and zinc uranyl acetate for sodium. The effect of an acid solvent (butanol-acetic acid) in localizing anions may be noted as well as in the production in several instances of multiple sodium spots. No completely satisfactory explanation for the observed separation of anion and cation of a strong electrolyte has as yet been proposed. Westall has postulated (19) the forma­tion of a sodium phenolate ion to account for the separation of sodium and chloride ions when run in a phenol solvent. Since the same phenomenon can be observed in solvents in which there is little probability of combination between the salt and solvent, this explanation would not appear to be of general utility. It seems highly probable that something more than simple partition between solvents is involved and that adsorption forces may play a determining role. For our own purposes the position assumed by these ions on urine chromatograms is of prime importance, and the Rf values for a number of inorganic ions in urine and the color developing reagents employed for the detection have been listed in Table II. Additional Quantitative Aspects A number of methods have been proposed for the quantitative estimation of materials separated on paper chromatograms. Some of these involve rather elaborate techniques of elution and colori­metric measurement (20), photo-densitometer measurements (21) The University of Texas and even electron diffraction (20). Only the simplest procedures have been used at this laboratory-visual comparison of color intensity and measurement of spot area. Compounds which are detectable in very small amounts (0.2 to 2.0 micrograms*) with sensitive reagents, generally show a relation between concentra­tion and color or color intensity, while the spot area remains virtu­ally unchanged over that range. Such compounds may be esti­mated quantitatively by visual comparison of color intensity in contrast to other compounds which become apparent only in con­centrations above three to five micrograms. In the range from five to thirty, or even up to fifty micrograms, the concentration can often be related as a linear function of the logarithm of the spot area. While subject to numerous limitations, these methods are so widely applicable and straight-forward as to be particularly suited to extensive group surveys. TABLE IV Rf Values of Inorganic Ions Derived from Various Compounds in Aqueous Solution Ion Compound Rf Values in Following Solvents Detec.ted Chromatographed Phen BuAc BuEt chloride____________________ HCl .19 .20 .15-.50 str. chloride_ _____ _______________ NH,Cl .27 .17 .16 chloride ----------------------NaCl .18 .13 sulfate_______ ________ ________ H.so, .19-.38 str. .37 .02 -.45 str. sulfate_________________________ (NH,) 2SO, .08-.45 str. .05 .03 sulfate -----------------------Na.SO, .05-.39 str. .03 .11 phosphate___________________ HaPO, .10 .47 .26-.44 str• phosphate__ ______________ NaH,PO, .10 .20 .09phosphate__ ______________ _ Na.HPO, .05 .23 .05 sodium -----------------------NaOH .55 .32 .07 sodium ------------------------NaCl .26 .11 sodium_ ______________________ Na.so, .03 and .23 .11 sodium__________ _ _________ Na.HPO, .03 and .45 .20 .03 and .11 sodium_________________________ NaH2PO, .03 .13 and .23 .03 sodium____________________ ____ NaC2H0, 3.43 .29 .14 Color comparison method.-For a compound to be quantitatively estimated by this procedure, a number of criteria must be met. First, a good color contrast between the spot and background is necessary. Second, there must be a clearly visible gradation in color or color intensity as a function of concentration. Third, the concen­tration range over which the color gradation is apparent should be at least three-to five-fold for practical utility. Fourth, distortion of spot color and/or spot size by interfering substances must be eliminated or reproduced in the standard. By careful selection of *An exception to the range given here is the case of urea, in which the color comparison range is from 8 to 32 micrograms. solvent and color developing reagent, these criteria can be fulfilled for a large number of compounds of biochemical interest. For the comparison of color intensity to be valid, standard and unknown samples must be treated under as nearly identical condi­tions as possible. This can best be achieved by running standard and sample adjacently on the same chromatogram. The principal difficulty encountered lies in the frequent presence of interfering substances in the sample (not duplicated in the standard) which may result in distortion of spot color. When such interference cannot be eliminated by the choice of solvent and color develop­ing reagent (as is most often the case when analyzing so complex a mixture as urine), the practice has been followed of adding to the standard a known amount of the sample to be determined. The standard is then subject to the same distorting influences as the sample. This procedure does not wholly eliminate the diffi­culty, since the concentration of interfering substances in the compared spots will necessarily be different. However, this general method has in many instances proved capable of satisfactory re­sults. A rather standardized form of paper chromatogram has been employed for quantitative estimation by the color comparison methods, and the exact procedure can perhaps best be described by considering a specific example. Table V illustrates the form of a chromatogram for the determination of urea in a urine sample. TABLE V Form of Chromatogram for the Determination of Urea in Urine by the Spot Comparison Method Concentration of Urea in Sample Spot Spot Position Urine Dilution Employed (5 µ[.added) Urea Added (µg) Corresponding to Standard Spot (mg./ml.) 1 (sample) 2 (standard) 3 (sample) 4 (standard) 5 (sample) 6 (standard) 7 (sample) 8 (standard) 9 (standard) 10 (sample) 11 (standard) 12 (sample) 13 (standard)14 (sample) 15 (standard) 16 (sample) 1:2 1:4 1:2 1:4 1:2 1:4 1:2 1:4 1:8 1:4 1:8 1:4 1:8 1:4 1:8 1:4 0.0 10.0 o.o 12.5 0.0 15.0 0.0 20.0 12.5 0.0 15.0 0.0 17.5 0.0 20.0 0.0 8 10 12 16 20 24 28 32 In every case sample spots are placed adjacent to standard spots containing half as much urine as the sample spot. Under these circumstances if a sample spot matches a standard spot, the amount of urea in the urine added to the standard spot will equal the amount of pure urea added to the standard spot. To consider the general case, if x equals the amount of urea present in the urine added to the standard spot, and y equals the amount of urea added to the same standard spot, then the total amount of urea present in this standard spot will be x+y. The amount of urea in the ad­jacent sample spot containing twice as much urine must equal 2x. Now, if the two spots are of equal color intensity, they presumably contain the same amount of urea, or 2x equals x plus y, from which it is obvious that x must equal y. To consider a specific example, suppose that upon development of the chromatogram represented in Table V, spots 5 and 7 were found to match spot 6. The amount of urea present in the urine added to spot 6 must be 15 micrograms. Since 5 microliters of a 1 :4 dilution of urine had been added to spot 6, the concentration of urea in the diluted urine was 15 micro­grams/5 microliters; the concentration of urea in the undiluted urine was 60 micrograms/5 microliters, or 12 milligrams/milliliter. The form of chromatogram illustrated in Table V provides not only that all comparisons will be between adjacent spots, but also that in most cases duplicate sample spots will be available for comparison with each standard spot. Where sample spots do not exactly match any standard spot, interpolation is, of course, pos­sible. The exact form of the chromatogram may be altered to fit circumstances. The only essential feature is the alternate arrange­ment of sample and standard spots with the standard containing a measured fraction of the urine present in the sample spot. The color comparison method of quantitative estimation can be applied to two-dimensional chromatograms. Because separate sheets must be employed for each dilution of standard and sample, the method is much more time-consuming than for the case of one­dimensional chromatograms. The probability of error is also greater, since comparisons must be made between spots on different sheets, which, despite all precautions, will have been subjected to slightly differing conditions during resolution and color develop­ment. The quantitative method using comparison of color intensity has a number of advantages which make it the preferred method in routine analysis, where a choice between the two methods out­lined here is possible. Since the effective range involves very small quantities, dilute solutions of the substance to be analyzed can be measured. (As little as 10 micrograms per milliliter can be conveniently determined.) In addition to its sensitivity, the method has other advantageous features: the volumes of solution necessary for application to the chromatogram are generally small, and the arrangement of standard and unknown substances side by side facilitates matching. A major disadvantage of the method is the relatively high absolute error imposed by the inability to distinguish between increments of less than about 0.2 microgram in a range of 0.2 to 2.0 micrograms. The rather narrow concen­tration range over which the method is usually effective, often necessitates use of several dilutions of the solution to be assayed. This is not only time consuming, but contributes an additional source of error. Details concerning the application of this method to the quanti­tative estimation of individual compounds will be found in later papers in the bulletin. Spot area method.-The measurement of the area of a spot on a chromatogram often serves as a satisfactory method of quanti­tative estimation. Most commonly, the logarithm of the area of the spot is a linear function of the concentration, although this is not always the case. Such a relationship has been noted by others (22, 23). Brimley (23) has derived the relationship mathe­matically, assuming it to be analogous to the heat-flow equation. The criteria necessary for the applications of this method are somewhat more flexible than those for the color comparison method. Color contrast, in this case is not so critical, and varia­tions in color intensity are of no importance. The single impor­tant requirement is that the spot area increase with concentration. Where the substance has migrated above an Rf of .10 on a ten­inch chromatogram, the logarithm of area is generally, but not always, a linear function of concentration; where migration does not occur or is less than an Rf of .10, the area is frequently a direct linear function of concentration. Interfering substances causing spot area distortion must, of course, be absent or their effect reproduced in the standard. In this laboratory the method has, in general, been limited to those cases where interference does not occur or where it may be eliminated by such simple means as the choice of a different solvent. Standards and unknowns must receive, as nearly as possible, identical treatment. It has been the practice at this laboratory to develop the standard and unknown simultaneously in the same crock, although not necessarily side-by-side on the same sheet as in the color comparison method. In this way the factors influenc­ing spot size (diffusion effects, temperature, solvent composition, etc.) can be made uniform for a given determination. Particular care must be taken to apply the color-developing reagent uniformly and lightly, especially where the reagent used has a high water content. Spreading or streaking of the spots may result from too liberal application of the reagent. Subjective errors of judg­ment in the marking of the spot areas may be minimized by having the same person mark (with a sharp pencil) all of the areas in any given determination. Areas must, of course, be marked be­fore the spots have faded, which in the case of many substances means immediately following color development. We have used a polar planimeter (Keuffel and Esser No. 4236) for measurement of the areas. The minimum quantity of a substance which may be quanti­tatively estimated by this procedure is set by the size of the sample spot originally applied to the paper. Variations in the size of the originally applied spot will be reflected in the size of the spot re­sulting upon development of the chromatogram. For these reasons it is of utmost importance that the materials to be chromato­graphed be applied in small, uniform quantities. Applying the samples in five microliter increments (resulting in initial spots of about one centimeter diameter), makes the effective range about five to fifty micrograms for most substances. Standard spots are usually run in quadruplicate at concentra­tions corresponding to 5, 10, 1'5, 20, 25, 30, 40, and 50 micrograms. Unknown spots are applied in such number and at such dilutions that at least four spots will fall in the range of 10 to 40 micro­grams. A standard curve is constructed from the averaged areas of each standard dilution, and values corresponding to the areas of the unknown spots read from this curve. The spot area method is, in large measure, complementary to the color comparison method. It is frequently useful where the latter cannot be used. Its lack of sensitivity to small quantities as compared to the color comparison method can often be put to advantage in avoiding large dilution of samples. Its most evi­dent advantage is that of providing an empirical means for esti­mating the relative amounts of unidentified substances. Although considerably more time consuming than the color comparison method, if carefully performed, it is probably capable of greater accuracy. BIBLIOGRAPHY 1. Williams, R. J., and Kirby, H., Science, 107, 481 (1948). 2. Williams, R. T. and Synge, R. L. M., Biochemical Society Symposium No. 2, "Partition Chromatography," Cambridge University Press, 1950. 3. Clegg, D. B., Anal. Chem., 22, 48 (1950). 4. Consden, R., Nature, 162, 359 (1948). 5. Consden, R., et al., Bice hem J., 38, 223 (1944). 6. Berry, H., and Cain, L., Arch. Biochem., 24, 179 (1949). 7. Levinson, S. A., and MacFate, R. P., Clinical Laboratory Diagnosis, 2d. Ed., Lea and Febiger, Philadelphia, Pa., 1943, p. 294. 8. Berry, H.K., M.A. Thesis, University of Texas (1949). 9. Sakaguchi, S., Biochem. J . (Japan), 5, 25 (1925). 10. Chattaway, F. W., Biochern. J., 41, 226 (1947) . 11. Chargaff, E., et al., J. Biol. Chem., 175, 67 (1948). 12. Vischer, E. and Chargaff, E., J. Biol. Chem., 176, 703 (1948). 13. Mellon, I., Organic Reagents in Inorganic Analysis, The Blakiston Co., Philadelphia, Pa., 1941, p. 560. 14. Frediani, H. A., and Gamble, L., Mikrochemie, 29, 22 (1941). 15. Beerstecher, E., Jr., Anal. Chem., 22, 1200 (1950). 16. Partridge, S. M., Nature, 164, 443 (1949) . 17. Forsyth, W. G., Nature, 161, 239 (1948). 18. Aronoff, S., Science, 110, 590 (1949). 19. Westall, R. G., Biochem. J., 42, 249 (1948). 20. Polson, A., et al., Science, 105, 603 (1947). 21. Bull, H.B., et al., J. Am. Chem. Soc., 71, 550 (1949). 22. Fisher, R. B., Parsons, D. S., and Morrison, G. A., Nature, 161, 764 (1948). 23. Brimley, R. C., Nature, 163, 215 (1949). III. The Influence of Solvent Composition, Temperature and Some Other Factors on the Rf Values of Amino Acids 1n Paper Chromatography by B. JIRGENSONS* Introduction The purpose of this work was to obtain more knowledge about the fundamental dependencies of the Rf values on solvent com­position, temperature, pH and the presence of salts. This knowl­edge might be of certain value for the improving of the results of practical work in the chromatographic determination of amino acids, as well as in obtaining more particular basic information as a background for a satisfactory theory of paper chromatography which might be elaborated in time. From the practical point of view it would be desirable if solvents other than those commonly employed (phenol, lutidine, collidine) were available for use. Systematic research about the water-miscible solvents in the detection of amino acids by paper chromatography was done re­cently by Bentley and Whitehead (1). They used as solvents some aliphatic alcohols, acetone, furfuryl alcohol, tetrahydrofuran, and tetrahydrofurfuryl alcohol mixed with water. Some data about the influence of temperature on the Rf values can be found in the work of Consden, Gordon and Martin (2). According to these data the dependence of the Rf on temperature is compli­cated, especially for the widely used phenol solvent. Some new solvents have been used in the work presented in the following pages. The influence of the mentioned factors was investigated for some of these solvents. The most satisfactory spots and the best resolutions were realized through the use of some ternary solvent mixtures, containing water, propionic acid and such or­ganic solvents as the butanols, ethylene glycol monobutyl ether or diethylene glycol monoethyl ether. Some important facts about the influence of temperature, pH and salt content have been found. *Present address: Texas Lutheran College, Seguin, Texas. Experimental The experiments have been carried out according to the tech­nique as proposed by Williams and Kirby (3). Each spot com­prised 25 micrograms of amino acid; the cylindric stapled sheets were placed in very wide-mouthed brown glass bottles which were then tightly closed with screw caps. Each bottle contained 100 ml. of the solvent. The bottles were put in a thermostatically controlled water bath at definite, constant temperatures; the temperature was kept constant by using a cooling and heating unit with relay in a large bath with three stirrers. The sheets were dried and developed with ninhydrin. The Rf values were then measured in the proper manner by marking the centers of the spots and by measur­ing the distance of the solvent front and the distance Qf the center of the spot from the start line; the Rf is the ratio of the distance of the spot to the distance of the solvent front. The results are presented in the following tables and figures TABLE I ~ The Rf Values for Several Amino Acids Resolved with tert-Butanol-Water Mixtures at 30° C. Vol.-%water 10% 20% 30% 40% 50% 60% Glycine --------­----------­.023 Alanine --------------------.047 Aminobutyric acid .10° Valirie ---------------------­.137 Leucine -------------------­.256 .ll7 .185 .242 .319 .493 .2l3 .300 .36° .444 .604 .467 .536 .535 .639 .767 .58° .636 .68° .no .818 .692 .761 .795 .805 .84° Isoleucine ---­-----------­.258 Norleucine ---­-----­-­.295 Phenylalanine -------­.14° Serine -------------------­--.026 Glutamic acid -------­.00 Lysine-HCl ------------­.00 Methionine ----------­--­.147 .427 .51° .402 .124 .06 .025 .325 .58° .64° .534 .239 .17° .096 .480 .70° .75° .714 .480 .368 .20° .62° .775 .81° .756 .592 .602 .29 .72 .85° .860 .796 .714 .75° TABLE II The Rf Values for Several Amino Acids Resolved with Diethylene Glycol Monoethyl Ether-Water Solvent at 20 ° C. Vol.-o/owater 10 20 30 40 50 .366 .484 .622 .683 Glycine --------------------------------------.256 .400 .530 .637 .735 .776 Alanine -------------------------------------- Aminobutyric acid ------------------.487 .595 .682 .76° .82° .650 .730 .795 .847 Valine ----------------------------------------.556 .632 .688 .742 .815 .85 Leucine ---------------·---------------------­ .293 .410 .538 .648 .71 Serine ---------·-------------.---------------­ .575 .60° .718 .80 Glutamic acid --------------------------.33° .220 .268 .28 Lysine-HCl --------------------------------.098 .273 TABLE III The Rf Values for Several Amino Acids Resolved with Diacetone Alcohol­ Water Mixtures at 30° C. Vol.-o/owater 10 20 30 40 50 Glycine Alanine ---------------------------------------­---------------------------------­----­ .022 .051 .ll2 .190 .270 .37° .432 .515 .58 .66 Aminobutyric acid --------------------Valine --------------­--------­-­---­------------Leucine ----------------------------------------Serine -----------------------------------------­ .085 , 135 .147 .Ol8 .25° .339 .410 .118 .436 .51° .61° .288 .575 .61 6 .664 .45° .70 .73 .77 .60 Glutamic acid ---------------------­-------­Lysine-H Cl --­-----­-­-----------------­----· .022 .00 .128 .05° .340 .128 .48 .20 .64 .23 TABLE IV The Rf Values of Some Amino Acids Resolved with a Ternary Solvent Mixture of 15 vol. Water, 15 vol. Propionic Acid and 70 vol. Diethylene Glycol Monoethyl Ether; T=19 ° C. Glycine ---------------------------------------.360 Aspartic acid ----------------------------.393 Serine -------------------------------------------.378 Glutamic acid ----------------------------.620 Alanine --------------------------------------,520 Arginine-HCl ---------------------------.408 Aminobutyric acid --------------------.597 Histidine-HCl ----------------------------.278 Valine ------------------------------------------.684 Tyrosine (Na-salt) ----------------.640 Leucine ---------------------------------------. 7 40 Tryptophan-HCl ----------------------.726 Lysine-HCl --------------------------------.387 Threonine ----------------------------------.504 TABLE V The Rf Values for Some Amino Acids Resolved with a Ternary Solvent Mixture Containing 20 vol. Water, 20 vol. Propionic Acid and 60 vol. Ethylene Glycol Monobutyl Ether; T=8°, 17° or 28° go 17° 28° Glycine ----------------------------------------------------.l73 .194 .205 Serine -----------------------------------------------------.160 .180 .187 Alanine ----------------------------------------------------.260 .276 .282 Valine ------------------------------------------------------.425 .448 .468 Leucine ----------------------------------------------------.573 .590 .598 Isoleucine ------------------------------------------------.536 .56° .576 Threonine ----------------------------------------------.220 .232 .249 Methionine ----------------------------------------------.461 .474 .506 Aspartic acid ----------------------------------------­.180 Glutamic acid ----------------------------------------.204 .228 .247 Lysine-H Cl ----------------------------------------------.111 .109 .124 Arginine-HCl -----------------------------------------­.147 Histidine-H Cl ---------------------------------------­.135 Tyrosine (Na-salt) -------------------------------­.41° Tryptophan-HCl ------------------------------------.4 78 .52° Phenylalanine ----------------------------------------.58° The most pronounced spots and the best separations were achieved with the solvents containing 30 volumes water and 70 volumes of the organic solvent. The reproducibility was good. Streaking of some amino acids takes place in the solvents with 10% water (e.g. leucine) ; lysine streaks at a higher water con­tent of 40 to 60 vol-%. The differences in capillary ascent of sev­eral amino acids with propionic acid-water solvent mixtures and ethylene glycol monobutyl ether-water mixtures are presented in the Figures 2 and 3. In the case of propionic acid the Rf values differ most when the water content is least . . 9 .8 .7 .6 w :::> _J <{.5 > '+ 0:: .4 .3 .2 .1 NORLEUCINE GLUTAM IC ACID 60 The dependence of Rf values on the water content of tert-butanol-water solvent mixtures. The dependence of the Rf values on temperature was investigated mainly using as solvents the n-propanol-water systems. The tem­perature was kept constant within the limits of approximately one degree; the bottles containing the solvent and paper were tightly closed. The results are presented in the Figures 4, 5 and 6. It was found that excellent spots and a good separation can be obtained by applying as solvents some ternary mixtures containing 15 to 30 volumes water, 15 to 30 volumes propionic acid, and the rest (up to 100) either ethylene glycol monobutyl ether or diethyl­ene glycol monoethyl ether; instead of the latter two solvents, also some esters or alcohols may be used . .... ~----vVALINE .9 ALANINE GLYCINE SERINE .8 w ::> _J <(· 7 > .5 60 10 20 30 PERCENT FIGURE 2 The dependence of the Rf values of several amino acids on the water content in propionic acid. T =17° C. Since it is impossible to dissolve some amino acids to a suffi­cient extent in pure water, making it necessary to add acid or al­kali, it is important to know how the pH of the amino acid solu­tion influences the Rf values. Some of the results are presented in Table VI. It is quite evident that the Rf values depend on the pH of the solutions of the amino acids; especially in the cases of tyrosine and glutamic acid the differences are remarkable. The same was observed if other neutral solvents were used. For instance with a ternary solvent mixture composed of 25 vol. water, 30 vol. n-propanol and 45 vol. n-butanol at pH=0.8 the Rf value of tyrosine was .480 but at pH=l0.7 it was .377; the Rf values for glutamic acid at pH=l.5 and 9.2 were .231 and .085 respectively . . 9 .8 LEUCINE .7 VALINE .6 GLYCINE w ~ _J <(.5 > '+­ 0:: ,4 .3 .2 . 1 60 10 20 30 40 50 PERCENT H20 FIGURE 3 The dependence of the Rf values of several amino acids on the water content in ethylene glycol monobutyl ether. T =18° C. A quite different picture is revealed if the pH of a neutral solvent is changed by adding acid (HCI) or alkali (NaOH). Some of the '!'ABLE VI The Rf Values as Depending on pH of the Solution. Solvent: 32 vol. Water with 8 vol. of tert-Butanol. T =27° C. pH Rf Glycine -------------------------------------------------------------------------2.0 .325 7.1 .293 10.0 .282 Serine_______________________________________________________________________________________ 1. 7 .325 6.5 .289 11.4 .295 Tyrosine -----------------------------------------------------------------------------------0.8 .554 10.7 .446 Tryptophan ---------------------------------------··---------------------------------------1.45 .538 11.8 .492 Glutamic acid ------------------------------------------------------------------------------1.5 .433 3.0 .292 9.2 .178 Lysine --------------------------------------------------------------------------------------------1.4 .15° 6.15 .133 11.0 .158 results are presented in Figure 7. In a strongly acidic solvent of pH one to two, the Rf values are very high; the spots after de­veloping are pink or purple. In a broad range between pH 4 and 11 the Rf values are constant. LEUCINE 5 4 w ~ _J <( .3 > ~ 0::2 .1 GLYCINE 10 20 30 40 0 50 60 TEMPERATURE C. FIGURE 4 The dependence of the Rf values on temperature. Solvent: 10 vol. water + 90 vol. n-propanol. The dependence of the Rf value on the salt concentration of the solvent was investigated mainly with the propanol-water and ethanol-water solvent mixtures. The latter were chosen because the ethanol-water systems, including those containing salts, have been widely used and it was hoped that from these data it would be possible to make some theoretical conclusions with respect to paper chromatography. M,oreover, it might be expected that the salts would make the spots more distinct, perhaps preventing the streaking, or would change the Rf values in such a way as to im­prove the separations. Indeed, there is a definite influence of a neutral salt like sodium chloride on the Rf values as is shown in Table VII and Figure 8. It is noteworthy that sodium chloride completely prevents the streaking of lysine, histidine and arginine .7 -o -0 0-­ .6 ,5 w :) _J <(4 > '+ 0:: .3 .2 ,1 LEUCINE VALINE AMINOBUTYRIC ACID ALANINE GLYCINE • 10 20 30 40 50 TEMPERATURE 0c. FIGURE 5 The dependence of the Rf values for several amino acids on temperature. Solvent: 30 vol. water+ 70 vol. n-propanol. but causes some streaking of aspartic and glutamic acid. The Rf values of most amino acids (most of which are acidic) decrease with .9 .8 .7 .6 w :) ~.5 > \,.,_ a::.4 .3 .2 1 increasing salt concentration, except lysine, arginine and LEUCINE VALINE ALANINE GLYCINE • 10 2 0 3 0 4 0 5 0 60 . TEMPERATURE °C. FIGURE 6 The dependence of the Rf values on temperature. Solvent: 50 vol. water + 50 vol. ~propanol. histidine whose Rf values increase in the presence of salt in the solvent. The spots in the lower part of the sheets, after develop­ment with ninhydrin, appear pink or purple; the tyrosine, tryp­tophan and histidine spots are greyish. TABLE VII The Influence of Sodium Chloride on the Rf Values. Solvent : 66.5 vol. Ethanol+ 33.5 vol. Water Saturated with NaCl. T=l8° . • Rf without NaCl Rf with NaCl Glycine ---------------------------------------------------------------------------.396 .386 Serine -----------------------------------------------------------------------------.412 .429 Alanine ---------------------------------------------------·----------------·------.550 .54° Valine ___ ---·---·-----------------------------------------· -----------------______ .667 .682 Leucine -----------------------------------------------------------------·--------.695 .70° GIutamic acid ---------------------------------___ __ ___________________ ______ .47s .3'66 Lysine-HCl ___________ ------------------·-----------------------------______ .22s .435 Arginine-H Cl ______________ ·------------·----------------------------_____ _ .221 .469 H istidine-HCl ----------------------------------------------------------_____ .239 .408 .7 .6 0 TRYPTOPHAN TYROSINE .5 0 0 GLUTAMIC ACID .2 .1 LYSINE 2 4 6 8 10 12 14 pH FIGURE 7 The dependence of the Rf values on pH of the solvent, composed of 25 vol. dil. HCI or NaOH + 75 vol. n-propanol. T = 28 ° C. Water miscible solvents were used in this study because of the possibility of changing the ratios through wider limits. As Bent­ley and Whitehead recently proved (1) there are some water­miscible organic liquids (some furane derivatives, acetone, pro­panol) which in mixtures with 20 to 40% water give satisfactory spots and good separations. It was found in this work that quite .7 .6 .5 w ::>4 _r <( > '+-·3 0:: .2 .1 L EUC lNE VALINE ALANINE GLYCINE LYSINE 0.05 0.10 0.20 NACL (MOLES PER LITER) FIGURE 8 The dependence of the Rf values on the salt content of the solvent containing32% water and 68% n-propanol. T = 17° C. useful in this respect are also some glycol ethers, especially ethyl­ene glycol monobutyl ether and diethylene glycol monoethyl ether. Again, the best spots and the best separation can be obtained using 20 to 30 % water in the mixtures. Quite satisfactory are also the tert-butanol-water solvent mixtures. In the Figures 1 to 3 the Rf values are plotted as dependent on water content of the solvent, and these graphs show that there are several types of the curves obtained. In Figure 1 (for tert-butanol) it is shown that the Rf values of some amino acids (e.g. glutamic) increase very markedly with increasing water content, whereas the Rf values of leucine and its isomers change much less; consequently we obtain curves bent in opposite directions, i.e. a maximum separa­tion or maximum difference in the Rf values at some intermediate water content (20 to 40 % ) . The same type of curves are exhibited by the propanols, diacetone alcohol and some other solvents. A quite different curve system was obtained in the case of propionic acid (Figure 2) where the differences in Rf values diminish with increasing water content. In such cases the best separation is at the lowest water content used. However, in all the investigated cases, at a low water content of 5 to 10% water some of the amino acids streak. This applies also to the alcohols and glycol ethers. The latter give curve systems which are of somewhat intermediate type (Figure 3). This was one of the reasons which induced the author to try the ternary mixtures of water-propionic acid-glycol ether, and similar systems. Satisfactory spots and good separa­tion were achieved by the use of 15 to 30 % water, 15 to 30 % pro­pionic acid, and the rest ethylene glycol monobutyl ether or di­ethylene glycol monoethyl ether. The same satisfactory results were obtained with ternary solvents containing (a) liquids which are partially miscible with water such as n-butanol or ethyl succi­nate, (b) propionic acid or propanol, and ( c) 15 to 20 % water. The best separation of the difficulty separable glycine and serine was realized with a ternary solvent mixture containing 20 vol. water, 40 vol. propionic acid and 40 vol. ethylene glycol monobutyl ether. The practical conclusion regarding the influence of the tem­perature is that the commonly occurring changes of room tempera-· ture have too little influence on the Rf values to distort the re­sults materially, especially where adequate controls are included. However, for good reproducibility of the chromatograms it is advisable to keep the room temperature as uniform as possible. The reproducibility in work with carefully prepared solvents, uni­form working conditions, and at thermostatically controlled con­stant temperature is good; in repeated experiments the Rf values then vary only within the limits of some two to five per cent. In such cases it is not unreasonable to calculate the values to the third decimal place; however, the third figure is not highly accurate and is so indicated in the Tables. There is a definite illfluence of the pH of the amino acid on the Rf value, especially for tyrosine and glutamic (probably also as­partic) acid, if neutral solvents are used. The pH of the solvent too has a definite influence on the ascent. In the solvents con­taining organic acids the pH influence is present. It is important that the quality of the spots can be improved by changing the pH of the material and of the solvent. Sodium chloride, if added to the solvent, prevents the streaking of the basic amino acids and considerably increases their Rf values; however, the Rf value of glutamic acid is decreased. Thus we have a convenient means of changing the Rf values in certain cases and of improving the definiteness of the spots also in this way. It would be desirable to investigate these problems more exten­sively, using other salts and other solvents. Another broad prob­ -lem is the one involving the salt content of the amino acid solution to be chromatographed. There is comparatively little to be said about the theoretical conclusions. There must be a partition of the amino acids between the stationary phase of the water in the cellulose fibers and the mobile phase of the solvent, and the more water-rich the solvent the higher the amino acids ascend. How­ever, that there is not exclusively a partition merely caused by the differences in solubilities, is now quite generally admitted. If this were the case it is difficult to understand why glycine and serine, whose solubilities are so different, have the same or nearly the same Rf values. According to Dunn and Ross (4) in 74.2% ethanol the solubility of glycine is 0.448 g. per 100 g. solvent, but that of serine is only 0.084 g. Also in pure water glycine is much more soluble than serine. If partition only were involved there should be expected also a great influence of the salts on the partition of all amino acids. Actually, there is little influence of salt on the ascent of leucine. In alcohol-water mixtures the influence of the electrolytes on solu­bility is great; leucine is then dissolved or "salted in" as is also glycine and alanine. Therefore on the partition hypothesis the presence of salt in the alcohol-water solvent mixture should have a marked effect on the Rf values for these amino acids. This is not the case; salt has a great influence only on such amino acids as lysine or glutamic acid. Summary The capillary ascent of amino acids on paper according to the method of Williams and Kirby was investigated and the depend­ence of the Rf values on solvent composition, temperature, pH of the amino acid and of the solvent, and on the salt content of the solvent was studied. Itwas found that: 1. The Rf values increase with increasing water content in a regular manner, and characteristic curves are obtained when the Rf values of the amino acids are plotted against the water content of different solvents. Tertiary butanol-water, propionic acid-water, and ethylene glycol monobutyl ether-water represent three differ­ent types of solvent mixtures. 2. Some ternary solvent mixtures containing 15 to 30% water, 15 to 40 % propionic acid, and the rest ethylene glycol monobutyl ether or diethylene glycol monoethyl ether yield excellent spots and good separation. 3. The Rf values usually increase slightly with increasing tem­perature. For the solvent composed of 30 vol. water and 70 vol. n-propanol, the incre~se within the limits of 6° to 40° C. is linear. 4. The Rf values are influenced by the pH of the amino acid solution applied on the sheet, especially in the cases of tyrosine and glutamic acid. 5. The Rf values depend also on the pH of the solvent. The Rf value increases in strong acid and strong alkaline solutions, but is practically constant within the limits of pH 4-11. 6. The Rf values depend on the salt concentration of the solvent; there is a remarkable decrease of Rf values for glutamic acid with increased salt concentration. On the contrary, the Rf values of lysine, arginine and histidine increase greatly in the presence of salt, and the latter prevents streaking of these amino acids in the solvent tested. Acknowledgment The author is cordially grateful to Dr. Roger J. Williams, Director of the Biochemical Institute of The University of Texas, for the facilities to carry out this research. Many thanks also to Dr. Norman Hackerman for his interest toward this work. BIBLIOGRAPHY 1. Bentley, H. R., and Whitehead, J. K., Biochem. J., 46, 341 (1950); Nature, 164, 182 (1949). 2. Consden, R., Gordon, A. H., and Martin, A. J. P., Biochem. J., 38, 224 (1944). 3. Williams, R. J., and Kirby, H., Science, 107, 481 (1948). 4. Dunn, M. S., and Ross, F. J ., J. Biol. Chem., 125, 309 ( 1938). IV. Quantitative Study of Urinary and Salivary Amino Acids Using Paper Chromatography by HELEN KIRBY BERRY AND LOUISE CAIN Of all the chromatographic procedures which have been de­veloped, perhaps the most extensively used method has been the quantitative determination of the amino acids. At this laboratory the urinary excretion of these substances has been studied for a considerable number of subjects and in some cases for consider­able periods of time. In addition, many saliva specimens have been analyzed. The results of these studies are reported principally in other portions of this bulletin. With a mixture containing as many amino acids as urine fre­quently contains, the use of ordinary one-dimensional chroma­tograms which have been previously described is frequently im­practical. However, after the preliminary use of the two-dimen­sional chromatograms, the one-dimensional method can, with se­lected individuals and for certain amino acids, be effectively em­ployed. One-Dimensional Chromatograms Although the specific details of the one-dimensional chromato­graphic procedure have been previously published (1), certain of the essential features will be repeated here. The method is useful in the determination of urinary glycine, lysine, alanine, leucine, valine, glutamic acid, and aspartic acid. However, when the urine sample in question contains more than very small amounts of asparagine and taurine, glycine cannot be determined using one­dimensional technique. Similarly, where glutamine, ,B-alanine, tyro­sine, or citrulline are present in considerable amounts, alanine cannot be determined. Threonine and arginine likewise may in­terfere with the determination of lysine. Hence, a preliminary test of the sample using two-dimensional chromatograms should always be run to determine the abundance of these substances relative to the amounts of the amino acids to be determined. When the combined interfering substances are more than about 10 o/o of the quantities of glycine, alanine and lysine present respectively, recourse must be made to the two-dimensional technique. The essential features of this procedure are similar to the other color comparison methods which have been previously described. The details concerning ranges and application of standard and sample can be found elsewhere (p. 23). The chromatograms are run in the phenol solvent (p. 25) and dried at room temperature. Color is developed using the ninhydrin reagent described on p. 27 followed by heating for five minutes at 100° C. The matching of spots and the calculation of amounts present are carried out in the same manner as has been previously described. The advantages of the one-dimensional over the two dimen­sional method are twofold : the saving of time in ·· the preparation and development of chromatograms, and the facility and increased accuracy in the matching of spots in close proximity. However, in many individual cases the presence of the interfering substances makes the one-dimensional procedure impractical. Two-Dimensional Chromatograms Chromatographic procedure.-The procedure followed in making two-dimensional chromatograms is essentially that of Polson (2) ; however, there are enough modifications of the method to warrant a more detailed description of the technique. The special problems involved in the preparation of such chromatograms have to do chiefly with the choice and application of the second solvent and with the complications which arise in the quantitative interpretation of the chromatograms. In this laboratory, the phenol solvent (p. 25) has been routinely employed for the initial development, and the 2,6-lutidine solvent for the other dimension.* After the initial development with the phenol solvent, the sheets must be dried thoroughly before using the lutidine solvent. The sheets run in lutidine as a second solvent show a characteristic brown front at an Rf of . 75-.90, apparently due to a residue from the phenol solvent on the paper. The Rf values, based on the total ascension height of the lutidine solvent, are always lower in this case than those obtained from a one-dimensional chromatogram developed with the lutidine solvent alone. If, however, the Rf values are cal­culated on the two-dimensional chromatogram using the brown *Most other workers involved in the analysis of amino acid mixtures em­ploy routinely for the second dimension either a fraction of the mixed collidines or 2,4,6-collidine. We have avoided the mixture because of the variable results it yields and the pure 2,4,6-collidine because of the expense of using it in an extensive series of analyses. front as the effective lutidine solvent height, excellent agree­ment is observed with the Rf values obtained with the one­dimensional chromatograms. This effect can be minimized by dis­carding the phenol solvent at the first appearance of any brown discoloration in the solvent, and in any case within 48 hours after its preparation. The lutidine solvent should also be discarded at the first appearance of its discoloration; otherwise colors and intensities of the amino acid spots after development with ninhy­drin will be distorted and difficult to compare. Since strong acid vapors may cause atypical red colors rather than purple with certain amino acids, contact with such vapors must be avoided during the development of the chromatograms. The chromatograms are prepared using 9xll" sheets. On the individual sheet the single standard (or sample) spot is placed one inch from a corner. The general practice at this laboratory has been to prepare a series of two-dimensional chromatograms containing 0.3, 0.6, 0.9, 1.2, 1.5, 2.0, 2.5, and 3.0 micrograms respectively of each of the following amino acids: aspartic acid, glutamic acid, serine, glycine, lysine, alanine, leucine, valine, taurine, threonine, citrulline*, and methionine sulfoxide. Volumes of urine equivalent to 20 and 40 micrograms of creatinine respec­tively were used to prepare two-dimensional chromatograms for comparison with the standard chromatograms. Not all the amino acids in the standard mixtures appear in each µrine sample, since the pattern of amino acids excreted by each individual is character­istic. The practice of superimposing small amounts of urine on the ~tandards has not been followed here since it is both impractical and unnecessary. . The series of standards and sample chromatograms should be prepared and developed at the same time and in the same manner. It is particularly important that the drying treatment to remove the first (phenol) solvent be identical for standard and sample sheets. Quantitative Estimation.-The quantitative estimation of the amino acids on two-dimensional chromatograms, while the same in principle, is somewhat more tedious and subject to greater errors than is the case with one-dimensional chromatograms. In order to make color comparisons as exact as possible, the sample *Since citrulline, ,8-alanine, and glutamine are not separable using the two solvents employed here, the spot on the sample sheets corresponding to the citrulline spot on the standard sheets may also contain .8-alanine and glutamine. and standard sheets are viewed side by side on a lighted X-ray viewing screen. Inasmuch as the standards and samples are on separate sheets, the chance for errors arising from difference in treatment is somewhat increased, and the ability to make the exact comparisons possible in the alternate spot arrangement of the one-dimensional chromatograms is decreased. In addition, the rather wide intervals between the different standard sheets make the absolute error of a given determination larger. However, for the purpose of comparing individual amino acid excretion patterns the technique gives results of sufficient accuracy. Fig. 1 is a composite diagram of the spots observed on two­dimensional chromatograms of urine and saliva. All of these spots were not observed in any single sample, and some few of them have not, as yet, been identified with any known substance. The identifications given in Fig. 1 were based on comparisons with standard chromatograms of known substances and the addition of "marker" amino acids to the urine samples. Comparisons with amino acid "maps" of other experimenters have also been made (3). Spot 7 is the position taken by citrulline, glutamine and ,8-alanine and we have not found it possible to distinguish between them since each possesses practically identical Rf values in both phenol and lutidine. Spot 16 remains unidentified. Its position corresponds to that of ethanolamine on Dent's (3) map, but is unaffected by addition of ammonia or acid to the atmosphere during phenol development, whereas ethanolamine is shifted markedly under these conditions. Spot 17 has been provisionally identified as methyl histidine. This amino acid was not available to us, but the color and position of the spot corresponds to that of methyl histidine as described by Dent (3). Spots 21 and 22 are unidentified. Spot 27 may correspond to the "nephrosis peptide" of Dent (3). A great deal of effort has been spent in an attempt to identify the substance responsible for Spot 26. This is the principal amino acid in all saliva samples investigated. The paper chromatographic behavior of this substance as compared with known substances of similar behavior is summarized in Table I. The Rf values re­corded were all obtained in the presence of small amounts of saliva to minimize the effect of interfering substances. For a consider­able time it was thought that the substance appearing at Spot 26 might be y-amino butyric acid, but the behavior of the two sub­stances in solvents containing acetic or isobutyric acid is different. 1­ z i.J > ...J 0 II) ...J 0 z i.J :r: a.. Urinary and Salivary Amino Acids 22 ·o 17 21 018 0 0 08 2eoo10 01g 028 ()14 01 Q6 016 240 012 0 5 4 0 03 02 015 Q-i 020 2 16 LUTIOINE SOLVENT FIGURE 1 Composite Diagram of Spots Observed on Two-Dimensional Chromatograms of Urine and Saliva Spot No. Spot No. 16 unidentified (purple) 1 aspartic acid (blue) 17 methyl histidine (?) (blue­ 2 glutamic acid (purple) 3 serine (purple) green) 18 proline (yellow)4 glycine (purple) 19 a-amino butyric acid (purple)5 taurine (purple) 20 cystine (brown)6 alanine (purple) 21 unidentified (purple)7 citrulline and/or glutamine and 22 unidentified (purple)I or fl-alanine (purple) 23 phenyl alanine (blue)8 valine (purple) 24 arginine (purple)9 leucine (purple) 25 tryptophan (brown-purple)10 methionine sulfoxide* 26 unidentified salivary constitu­11 lysine (purple) ent-(position of -r-amino butyric12 threonine (purple) acid) (see text) (purple)13 tyrosine (blue) 27 "nephrosis peptide"? 14 histidine (brown-purple) 28 methionine sulfone 15 cysteic acid (blue) *Since the material for this Bulletin was prepared it has become probable that the substance oTh?inally identified by Dent as methionine sulfoxide is actually {3-aminoisobutyric acid. Nature, 167, 307 (1951). The possibility that this substance might be a peptide was con­sidered and rejected on the basis of its presence in chromatograms run on acid hydrolyzed saliva. A pooled saliva sample was chromatogrammed, (a) without further treatment, (b) after hy­drolysis with hydrochloric acid, (c) after precipitation of pro­teins with ethanol, and (d) after precipitation of proteins, followed by hydrolysis. None of these treatments appreciably affected Spot 26. TABLE I Comparison of Rf Values of Unidentified Substance in Saliva (Spot 26) with Known Amino Acids Suspected of being Identical Rf Values in the Following Solvents* Amino Acid Phen BuAc BuEt But Lut IsoBu Pyt a-amino n-butyric --------------------.76 .41 .37 .19 .21 .56 .37 a-amino isobutyric --------------------. 78 .44 .37 .20 .19 .56 .42 'Y-amino butyric ------------------------.85 .44 .22 .09 .10 .53 .22 leucine ----------------------------------------.89 .67 .63 .49 .40 .76 .57 methionine ----------------------------------.83 .46 .46 .29 .34 .67 .49 methionine sulfone --------------------.66 .16 .24 .06 .21 .38 .33 methionine sulfoxide ----------------.85 .16 .21 .06 .10 .44 .21 pantonine ------------------------------------.81 .42 .39 .21 .24 .60 .44 phenylalanine ----------------------------.89 .55 .59 .42 .40 .79 .51 valine ------------------------------------------.83 .55 .46 .30 .28 .69 .47 "Spot 26" ------------------------------------.81 .54 .22 .07 .10 .70 .23 *A complete description of the solvents employed will be found on p. 25. tThis solvent was composed of 80% pyridine and 20% water. The methods here outlined have found extensive application in our studies and the results of analysis by these methods· will form an important part of the data presented in other parts of this bulletin. BIBLIOGRAPHY 1. Berry, H.K., and Cain, L., Arch. Biochem., 24, 179 (1949) . 2. Polson, A., Biochim. Biophys. Acta., 2, 575 ( 1948). 3. Dent, C. E., Biochem. J., 43, 169, (1948). V. The Quantitative Determination of Urinary Histidine Using Paper Chromatography by LOUISE CAIN AND HELEN KIRBY BERRY Considerable interest has been shown in the study of the urinary excretion of histidine, both in connection with its reported varia­tion during the menstrual cycle and with its attempted use as a i::pecific test for pregmmcy. Boxer and Kapeller-Adler ( 1) demon­strated that the excretion of histidine was greatest about 14 days before the beginning of the menstrual cycle, corresponding to the ovulatory phase. However, Chattaway (2) was unable by testing urine samples at seven-day intervals to demonstrate any appre­ciable difference in histidine excretion during the menstrual period. The marked increase of histidine excretion fairly early in preg­nancy has suggested its use as a diagnostic test (3, 4, 5). The standard colorimetric tests suggested as a basis for the determination of histidine in urine are subject to uncertainty as a result of the lack of specificity of the reagents. The two most common reactions of histidine which are used for its determina­tion are the bromination reaction described by Knoop (6) and subsequently modified by several investigators (7, 8), and the diazotized sulfanilic acid reagent of Roessler and Hanke (9). At­tempts in our laboratory to use the bromination reaction as modi­fied by Chattaway (8) to determine urinary histidine were unsuc­cessful in that no adequate recoveries could be obtained. We have attempted to use both of these color developing reagents for the determination of histidine in our usual paper chromatograms. In most urine samples there were several substances other than his­tidine which gave the same color reaction and caused interference by migrating to the same position. Consequently different solvents were explored. Experimental Solvent Mixture.-Although a number of different solvent mix­tures were used in developing the procedure, all but one had seri­ous disadvantages. In most of the solvents tested the histidine either did not migrate appreciably or it split to form two spots in the presence of urinary salts. However, using a solvent mixture consisting of 80 ml. of n-butanol, 20 ml. of ethanol, and 40 ml. of 2N hydrochloric acid, histidine migrates as a single spot to a posi­tion with an Rf of .20, free from interference by the substances which disturb the colorimetric procedure. Reagent.-Both the bromine reagent and the diazotized sul­fanilic acid reagent were employed in developing the quantitative procedure, but the latter, in addition to being more sensitive, was found to give more consistent and reproducible results. As little as 0.20 micrograms of histidine could be detected using this re­agent, while 5-10 micrograms were required when the bromination reaction was carried out. The details for preparing and using the diazotized sulfanilic acid reagent may be found on p. 30. The ex­clusion of the vapors of phenol, pyridine, lutidines and collidines is essential to prevent interference with development. Procedure.-The chromatograms were prepared using sheets of filter paper 51/2 inches in height. Pure histidine in a standard solution was added in 5 microliter increments to alternate spots % inch apart so as to have the following amounts of histidine present: 0.25, 0.50, 0.75, 1.0, 1.5, and 2.0 micrograms. At this stage a preliminary test was usually carried out by placing 5.0 microliters of the unknown urine on a small strip of filter paper and allowing the solvent to ascend about one inch above the ap­plied spot. The strip was dried and developed with the diazotized sulfanilic acid reagent. If the preliminary test showed a distinct pink to red-orange spot, as the majority of urine samples did, 5 microliters of the urine was placed on the positions alternating with the standard histidine spots. Finally, the urine was diluted 1 :2, and 5 microliters of the dilution was superimposed on the standard histidine spots to correct for the distorting effects of urinary constituents. The effective color range of the test is from a light pink to a conspicuous red-orange, and the unknown and standard were matched as described on page 50. Where the urine was too dilute to show a distinct reaction using 5 microliters, 10 to 30 microliters of urine were added to the sheet containing the standard histidine, and 5 microliters of undiluted urine was superimposed upon the standard. Wherever such relatively large quantities of urine were necessary, chromatograms 11 inches in height were employed to avoid interference and distortion. When the urine was relatively concentrated in histidine, 5 microliters of a 1 :2 dilution was used and a 1 :4 dilution was placed on the standard spots. The sheets were dried for 30 minutes at room temperature before spraying with the diazotized sulfanilic acid. Results.-In Table I are given results of recovery experiments in which histidine was added to urine. TABLE I Recovery from Urine of Added Histidine Histidine Histidine Histidine Added Found Recovered Per Cent (mg./ml.) (mg./ml.) (mg./ ml) Recovery Urine Sample 1 ________________ 0.0 0.30 0 1.00 1.27 .0.97 97 Urine Sample 2 ----------------0.0 0.45 0 0.65 1.16 0.71 109 Urine Sample 3 ----------------0.0 0.46 0 0.50 0.95 0.49 98 Urine Sample 4 ----------------0.0 0.30 0 0.80 1.12 0.82 103 Urine Sample 5 ----------------0.0 0.30 0 0.20 0.52 0.22 110 Summary.-A colorimetric procedure for the quantitative de­termination of histidine in urine employing paper chromatography is described. A special solvent mixture is used and diazotized sulfanilic acid is used for color development. The average re­covery of histidine added to urine was 103% . BIBLIOGRAPHY 1. Boxer, G., and Kapeller-Adler, R., Biochem. Z., 293, 207 (1937). 2. Chattaway, F. W., Biochem. J ., 41, 226 (1947). 3. Kapeller-Adler, R., Klin. Wochschr., 13, 21 (1934). 4. Langley, W. D., J. Biol. Chem., 137, _255, (1941). 5. Dawson, J., B'iochem. J., 38, xvii (1944). 6. Knoop, F., Hofmeisters Beitrage, 11, 396 (1908). 7. Armstrong, A. R., and Walker, E., Biochem. J., 26, 143 (1932). 8. Chattaway, F. W., Biochem. J., 41, 226 (1947). 9. Koessler, K. K., and Hanke, M. T., J. Biol. Chem., 39, 497 (1919). VI. The Quantitative Determination of Creatinine in Urine Using Paper Chromatography by HELEN KIRBY BERRY AND LOUISE CAIN Creatinine, which has been used in all of the subsequent studies on individual metabolic patterns as a basis for correcting for urine concentration (p. 152), has been routinely determined using a colorimetric procedure (1). However, where the urine samples available from young infants and from rats and mice were small (as little as a fraction of a milliliter in some cases), it was neces­sary to develop a technique which would require much smaller quantities than were needed for the colorimetric procedure. Experimental Solvent mixtures.-The solvent mixture used in the determina­tion was composed of 80 ml. of n-butano1, 20 ml. of ethanol, and 20 ml. of distilled water. This solvent may be made up as a stock solution and aliquots used for two to three determinations before replacing with a.fresh aliquot. Reagent.-The reagent used for color development was the alkaline picric acid reagent described on p. 31. Procedure.-Chromatograms for the determination of creatinine were prepared on sheets of filter paper 3 x 11 inches. Experience has shown that a solvent ascension of three inches is ample to remove distorting and interfering substances in this case. Pure creatinine in a standard solution was added to alternate spots 11/2 inches apart so that the spots contained 0.25, 0.50, 0.75, 1.0, 1.25, 1.50, 1.75, and 2.00 micrograms. On the positions alternating with the standard spots was placed 5 microliters of a 1 :2 dilution of the sample of urine. Finally 5 microliters of a 1 :4 dilution of the urine was superimposed upon the standard creatinine spots. After resolution and color development the creatinine content of the urine was calculated as described on p. 52. The Rf value of creatin­ine in the solvent used is .30. The color shade of the spots did not vary over the effective range, but the intensity of the color showed a considerable variation. Occasionally with concentrated Creatinine Determination urines it was necessary to employ a 1 :4 dilution as the unknown and to place a 1 :8 dilution on the standard creatinine spots. Results.-The results obtainable using the paper chromato­graphic procedure are compared with the results obtained from the standard colorimetric technique in Table I. Note that while the absolute errors are approximately the same, the relative error may be rather high when samples are dilute. TABLE I Comp.arison of Creatinine Determination by a Standard Colorimetric Procedure and by Paper Chromatography Creatinine Creatinine by by Paper Percent Sample No. Colorimetric Chromatography Difference (mg./ml.) (mg./ml.) 1.6 -3.7 2 -------------------------------· 2.32 2.4 + 3.4 1 --------------------------------1.66 3 --------------------------------1.56 1.6 + 2.6 4 --------------------------------.48 .50 + .4 5 --------------------------------.32 .25 -22. 6 --------------------------------.37 .40 + 8.1 7 ------------------------------.35 .35 0. Summary.-A paper chromatographic method for the quantita­tive determination of creatinine in very small amounts of urine is described. Alkaline picric acid is used for color development. Comparison of the results obtained by this method to the results obtained using the standard colorimetric technique are presented. BIBLIOGRAPHY 1. "Klett-Summerson Clinical Manual," Klett Mfg. Co., New York. VII. The Quantitative Determination of Creatine 1n Urine Using Paper Chromatography by LOUISE CAIN The determination of creatine by the standard colorimetric method (1) is tedious and time consuming. A relatively simple method for the determination of creatine in urine has been devel­oped using paper chromatography and is here presented. This method is especially useful when the quantity of the urine sample is limited and a large number of samples are to be analyzed. Experimental Screening Procedure.-Since creatine appears only infrequently in the urine, it is desirable to make a rapid screening of a large number of samples to ascertain its presence. For this purpose filter paper sheets eleven inches high are used, and samples of urine containing 50 P-g. of creatinine (previously determined) are placed in the usual manner (p. 23) one inch from the bottom of the sheet and one inch apart. The samples are then resolved by allowing a solvent mixture composed of 80 ml. n-butanol, 20 ml. ethanol, and 20 ml. distilled water to reach the top of the sheet. The sheets are thoroughly dried, sprayed with a warm aqueous solution of 1.3% picric acid, and then heated for one hour at 110° C. in order to convert the creatine to creatinine. Finally the sheet is sprayed with a 1 N sodium hydroxide solution. If creatine was present in the original sample, the characteristic orange color of creatinine on a yellow background appears at an Rf value of .10. Since cre­atinine moves in this solvent to an Rf value of .25, there is no like­lihood that the two substances will be confused. Using this butanol­ethanol-water mixture as a screening solvent has the added advant­age that much valuable information can be gained concerning other substances which react with picrate and which migrate near the solvent front in such a solvent as lutidine. Quantitative Determination.-Creatine, when resolved in the butanol-ethanol-water mixture, is sometimes distorted by an inter­fering substance also appearing in urine. Hence, it is desirable that samples which have been shown by the screening procedure to contain creatine be chromatographed using a mixture of 65 ml. Creatine Determination lutidine and 35 ml. water as the resolving solvent. A sheet of filter paper 51/2 inches high is sufficient for this more precise determina­tion. On alternate positions are placed 5 ,ul. quantities of standard solutions containing 3.0, 5.0, 8.0, 10.0, 12.5, and 15.0 ,ug. of creatine respectively. On the remaining positions are placed 10, 10, 15, 20, and 25 ,ul. quantities of urine. Finally, 5 ,ul. of urine is superim­posed upon each spot of the standard solution. These volumes of urine may be changed depending on the concentration of creatine estimated to be present by the screening procedure. After resolu­tion and color development as above, the concentration of creatine can be determined by matching the appropriate spot from a urine sample with the corresponding standard spot. The Rf value of creatine in this solvent is .27 while that of creatinine is .56. In Table I are summarized the results of some recoveries of creatine as determined by this method. Creatine was added in the amounts listed to samples of urine which originally contained an insufficient amount of creatine to be detected by this method. It is probable, however, that in some cases small amounts were pres­sent. These minute amounts may account for the large error ob­tained when low concentrations of creatine were added. TABLE I Recovery from Urine of Added Creatine Urine Added Creatine Creatine Per Cent Sample (mg./ml.) (mg./ml.) Recovery 1 0.94 1.0 106 2 1.33 1.0 75 3 0.59 0.50 85 4 5 6 7 8 9 1.05 0.91 0.78 0.80 0.19 0.33 1.67 1.60 0.30 0.60 1.22 1.0 86 102 200 96 200 82 BIBLIOGRAPHY 1. Klett-Summerson Clinical Manual, Klett Manufacturing Company, New York. VIII. The Quantitative Estimation of Uric Acid by Paper Chromatographic Methods, with Applications to Human Urine and Saliva by HELEN KIRBY BERRY Uric acid is the principal end-product of purine metabolism in humans, and its concentration in the urine is known to vary widely with the purine content of the diet. Dent has studied the migration of uric acid on paper chromatograms using phenol and collidine as solvents and 1'lo ammoniacal silver nitrate solution as a color developing reagent (1). Experimental Selection of color reagent and solvent.-A solution 0.05 N in AgN03 and 2.5 N in NH40H was employed as the color developing reagent. Chromatograms were dipped in this solution, the uric acid spot developing as a black area within two to three minutes. The following solvent mixtures were investigated: (a) 65 ml. of 2,6-lutidine and 35 ml. of water; (b) 80 ml. of n-butanol, 20 ml. of ethanol, and 20 ml. of water; and (c) 100 g. of phenol and 20 ml. of lO 'lo sodium citrate solution (2). Using 2,6-lutidine, uric acid appears as a well defined spot at Rf .31. With the butanol-ethanol solvent, uric acid forms a streak. In phenol uric acid migrates to an Rf value of .35, but the phenol causes almost immediate reduction of the silver nitrate applied as a color developer. 2,6­Lutidine was selected as the most satisfactory solvent. Both the position of uric acid on the chromatogram and the rapid reduction of ammoniacal silver nitrate are very specific. The uric acid need be allowed to migrate only far enough to pre­vent interference from a substance, probably sodium ion, which does not move in lutidine, but which causes reduction of silver nitrate rather more slowly than does uric acid. Chromatographic assay procedure.-The general techniques em­ployed in preparing and developing the paper -chromatograms have been described elsewhere (2, 3). Because of the specificity of the procedure only a short migration is necessary, and sheets 14 x 2112 inches have been employed. These provide 18 positions for application of sample and standards at 34 inch intervals, 112 inch from the bottom of the sheet. About 20 minutes is required for the solvent to ascend to within 1,4 inch of the top of the sheet. The spot comparison method, described in detail elsewhere in this bulletin (2), was employed for quantitative estimation. Spots developed from five microliters of 1 :2, 1 :4, and 1 :8 dilutions of urine were compared with adjacent spots developed from five microliters of 1 :4, 1 :8, and 1 :16 dilutions of urine plus added uric acid. The nine standard spots on a sheet were made up to correspond to 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.8, 1.0, and 1.2 mg. uric acid per ml. urine. After determining the approximate uric acid content of a urine sample, if greater accuracy is desired, a second sheet can be run with smaller intervals between the amounts in the standard spots. The sheets were dried in air for thirty minutes before treatment with the ammoniacal silver nitrate solution. Spots may be com­pared whiie the sheet is still wet, or it may be allowed to dry. The background of the sheet turns brown after several hours, unless excess silver nitrate is removed by washing. Results Recovery experiments.-The results of recovery experiments on two urine samples are shown in Table I. The recoveries varied from l00-133 7'o with an average error of 17% . Each determination was made using a single chromatogram as described above. More precise values could no doubt have been obtained by employing a second chromatogram with a narrower range of standards. TABLE I Recovery of Added Uric Acid from Urine Uric Ac.id Uric Acid Uric Acid Added Found Recovered Percent (mg./ml.) (mg./ml.) (mg./ml.) Recovery Sample 1 --------------------0.0 0.6 0.0 0.086 0.7 0.1 116 Sample 2 --------------------0.0 0.13 0.0 0.1 0.23 0.1 100 0.2 0.38 0.25 125 0.3 0.53 0.40 133 0.4 0.58 0.45 112 0.5 0.70 0.57 114 Uric acid content of male urines.-Morning urine samples were collected from twelve men for five · days a week over a four week period. The samples from each subject for each week were pooled on the basis of a constant creatinine content. Chromatograms were run in duplicate, and when the variation between duplicates was greater than 20 %, the determination was repeated. The results are shown in Table II. Values for uric acid are recorded as mg. of uric acid per mg. of creatinine, in order that the values for the various subjects may be compared on a basis more nearly inde­pendent of urine volume. Creatinine was determined by a standard colorimetric method (4). Inspection of the data reveals a greater variation in the excretion of a given individual from week to week, than in the average excretion rates of the different indi­viduals. These data are in line with the known variation of uric acid excretion with the diet. TABLE II Urinary Uric Acid/ Creatinine Ratios for Twelve Males First Second Third Fourth Subject Week Week Week Week . Average 1 --------------------.37 .15 .29 .27 .27 2 --------------------.49 .41 .06 .31 .32 .Q7 3 --------------------.66 .35 .13 .31 4 --------------------.35 .41 .03 .26 5 --------------------.30 .23 .47 .33 .33 6 --------------------.24 .37 .23 .04 .22 7 -------------------.52 .22 .21 .11 .27 8 --------------------.39 .20 .08 .40 .27 9 --------------------.73 .55 .27 .13 .42 10 -------------------.55 .24 .31 .24 .33 11 --------------------.42 .34 .25 .10 .28 12 --------------------.32 .38 .21 .30 Uric acid content of saliva.-Daily saliva samples from each of twelve subjects were collected for four five-day periods. Equal volumes of each sample were pooled for a five-day period. For analysis of uric acid in saliva, chromatograms were prepared using 10, 15, 25, and 55 microliters of saliva alternating with standards containing uric acid added to five microliters of saliva. If uric acid could not be detected in 55 microliters of saliva, no attempt was made to determine it by using larger volumes. Table III shows the results of this study. Less than half of the subjects consistently secreted uric acid in excess of the minimum detectable quantity of .0025 mg. per ml. There is at least a 32-fold difference between the six subjects showing no detectable uric acid secretion and subjects 5 and 6 who secreted an average of 0.08 mg. per ml. There is a 10-fold difference between subjects 5 and 6, and subject 7, whose average was 0.008 mg. per ml. Ad­mitting a possible error of 50 % in determinations at these low concentrations, the differences noted are still highly significant. Uric acid determinations have been made on single and dupli­cate saliva samples from a number of individuals in addition to those listed in Table III. In all, saliva samples from 66 individuals have been studied. Of these, 25 secreted less than 0.0025 mg. of uric acid per ml.; 4 secreted in the range from 0.0025 to 0.01 mg. per ml.; 29 secreted in the range from 0.01 to 0.05 mg. per ml.; and 8 secreted in the range from 0.05 to 0.13 mg. per ml. TABLE III Secretion of Uric Acid in Saliva (mg./ml.) First Second Third Fourth Subject Week Week Week Week Average 1 ------------------* * * * * 2 --------* * ----------* * * 3 -----------------­ * * * * * 4 -----------------­ * * * * * 5 ------------------0.037 0.042 0.09 0.15 0.08 6 ------------------0.050 0.052 0.15 0.068 0.08 7 ------------------0.010 * 0.0025 0.020 0.008 8 ------------------* * * * * 9 ------------------0.0075 * * * * 10 ------------------0.037 0.055 0.063 0.050 0.05 11 ------* * ------------* * * 12 ------------------0.05 0.022 0.013 0.03 * Less than 0.0025 mg./ml. Summary.-A method is described for the quantitative deter­mination of uric acid in urine by a paper chromatographic method, employing 2,6-lutidine as a solvent and ammoniacal silver nitrate solution as the color developing reagent. The average error in recovery experiments was 17% . The method has been applied to the determination of uric acid in weekly pooled urine and saliva samples from 12 subjects. The variation in the weekly urinary uric acid excretions of a given individual was greater than the variation of the averages of the different subjects. Variations in diet are doubtless important in this connection. Highly significant variations between individuals were noted in the salivary secre­tion of uric acid. BIBLIOGRAPHY 1. Dent, C. E., Biochem. J., 41, 240 (1947). 2. Paper on general chromatographic methods, this Bulletin. 3. Williams, R. J., and Kirby, H., Science, 107, 481 (1948). 4. "Klett-Summerson Clinical Manual," Klett Mfg. Co., New York. IX. The Quantitative Estimation of Urea by Paper Chromatographic Methods with Application to Human Urine by HELEN KIRBY BERRY As the principal nitrogenous end-product of protein metabolism in mammals, urea is of central importance in any thorough study of urine composition. The migration and identification of urea on paper chromatograms of urine have been previQusly described by Dent (1). Experimental Phenol-hypochlorite color reagent.-On urine chromatograms prepared using phenol as a solvent, a bright green spot with Rf value of .80 appeared after spraying with "Clorox" (commercial grade contains 5.25 '/o sodium hypochlorite). This spot was in the position occupied by urea as developed with mercuric nitrate-sodium hydroxide, the color developing reagent used by Dent (1) . Identi­cal Rf values were obtained with the two color reagents using either pure urea solutions or urine, and regardless of the solvent em­ployed in developing the chromatogram. To confirm the identity of the substance reacting with sodium hypochlorite, 10 ml. of urine to which 0.35 g. of urea had been added was incubated overnight with urease and then chromatographed. No urea could be detected on the chromatograms of the treated urine, using either color reagent. In addition to the green spot at Rf .80, allchromatograms pre­pared using phenol as solvent and developed with sodium hypo­chlorite showed a blue spot at Rf .28, which appeared to correspond to a strong acid anion, chloride or sulfate. The qualitative appear­ance of the urine chromatograms could be duplicated by chromato­graphing synthetic mixtures of urea and ammonium chloride. The phenol-hypochlorite reagent is a color reagent of rather wide application. It has been used in a specific test for glycine (2), and for the detection of small concentrations of ammonium ion (3). A number of compounds occurring in urine, which might possibly interfere with the determination of urea, were tested for reaction with phenol and sodium hypochlorite. Those which developed a color included: allantoin, arginine, uric acid, creatinine, glycine, taurine, p-aminobenzoic acid, tryptophan, histamine, pantothenic acid, lactic acid, tartaric acid, HCl, NaCl, NH4Cl, H2S04, and (NH4 ) 2S04 • The color developed and the Rf values for these substances are recorded elsewhere in this bulletin (4). None of the substances with Rf values similar to urea are present in urine in sufficient concentration to cause substantial interference. The color forming reaction between urea and sodium hypo­chlorite does not occur in the absence of phenol. The color and intensity of the spot is markedly affected by the amount of phenol present on the sheet when sprayed with sodium hypochlorite, too much phenol resulting in very faint spots, and too little resulting in spots which fade rapidly to a yellow-green. With too little phenol present the background of the whole sheet turns brown almost immediately. When phenol is used as the solvent in urea determinations, it is therefore important to control carefully the drying of the sheet prior to color development. Best results have been obtained by drying the sheet for eight minutes at 90° C. When properly developed the green color of the urea spot remains stable for long periods, but the background may slowly turn brown due to the action of sodium hypochlorite on the filter paper. Selection of solvent.-The following solvent mixtures were in­ vestigated: (a) 80 ml. of n-butanol, 20 ml. of ethanol, and 20 ml. of water; (b) 65 ml. of 2,6-lutidine and 35 ml. of water; and (c) 100 g. of phenol and 20 ml. of 10% sodium citrate solution (5). Chromatograms run in the first two solvents were sprayed with a solution of 5% phenol in 95 % ethanol and allowed to dry before spraying with sodium hypochlorite. Using 2,6-lutidine, urea migrates to an Rf value of .43, but the background turns blue very soon after spraying with sodium hypochlorite. With the butanol­ethanol solvent, urea migrates to an Rf value of .50, and the acid anion spot appears at Rf .24. In the phenol-citrate solvent urea appears at Rf .80 and the acid anion moves only slightly to Rf .15. The phenol-citrate solvent was chosen for general use because the wide spread in Rf values between the two substances in urine which develop color makes possible their rapid separation using short ascents. The use of phenol as solvent also eliminates the phenol spraying step which would be required if any other solvent were employed. Chromatographic assay procedure.-'-The general techniques em­ployed in preparing and developing the paper chromatograms have been described elsewhere (5, 6). Because of the total lack of interfering substances of similar Rf value it is convenient to use short paper sheets; an ascent of 1112 to 2 inches by the solvent front is sufficient to separate clearly the urea spot from the acid anion spot. Sheets 13 x 21/2 inches were employed, providing six­teen positions for application of sample and standards at % inch intervals, 112 inch from the bottom of the sheet. Only 20 minutes is required for the solvent to ascend to within %, inch of the top of the sheet. The spot comparison method, described in detail elsewhere in this bulletin (5), was employed for quantitative estimation. Spots developed from five microliters of 1 :2 and 1 :4 dilutions of urine were compared with adjacent spots developed from five microliters of 1 :4 and l' :8 dilutions of urine plus added urea. The eight standard spots on a sheet were made up to correspond to 10, 15, 20, 25, 30, 40, 50, and 60 mg. urea per ml. urine. Having determined the approximate urea content of a given urine sample on such a chromatogram, the concentration can be more accurately de­termined on a second sheet on which the standard spots are present at two mg. per ml. intervals. The necessity for the first "rough" chromatogram is eliminated if the urea content of the sample can be estimated by other means. A sufficiently accurate estimate for this purpose can usually be made, based on the specific gravity of the urine. Table I shows the range of urea concentrations usually encountered for a given specific gravity range. TABLE I Specific Gravity and Urea Content of Urine Specific Gravity Probable Range of Range Urea Cone. (mg./ml.) Below 1.006 --------------------------------------------4-10 1.006-1.012 ----------------------------------------------8-20 1.012-1. 032 ----------------------------------------------16-40 Above 1.032 ------------------------------------------32-80 It is important that the standard spots be developed in the presence of urine. Creatinine present in the urine, although it does not form a colored spot with phenol-hypochlorite, does affect the Rf value of the urea spot slightly, and also causes a somewhat more diffuse spot. Results Recovery experiments.-The results of recovery experiments on three different urine samples are shown in Table II. For each determination a preliminary chromatogram was run to determine the approximate urea content, and then a second chromatogram with smaller intervals between standards was used to obtain the recorded value. The recoveries varied from 87-102 % with an average value of 96%. TABLE II Recovery from Urine of Added Urea Urea Added Urea Found Urea Recovered Percent (mg./ml.) (mg./ml.) (mg./ml.) Recovery Sample 1 --------------------0.0 10 0 16.7 26 16 96 26.7 36 26 97 Sample 2 -------------------0.0 19 0 13.3 32 13 98 33.4 48 29 87 Sample 3 --------------------0.0 9 0 12.1 20.5 11.5 95 28.0 37.5 28.5 102 Urea content of male urines.-As a further test of the method, morning urine samples were collected from twelve men for five Clays a week over a four-week period. The samples from each sub­ject for each week were pooled on the basis of a constant creatinine content. The results of this study are shown in Table III. Values for the urea content are recorded as mg. of urea per mg. of creatinine, in order that the subjects may be compared on a basis more nearly independent of urine volume. Creatinine was de­termined by a colorimetric method (9). The week to week ratios, urea :creatinine, sometimes vary for a given individual by a factor of two. It will be noted that the individual and average values for subjects No. 1 and No. 6 are much lower (average a little over TABLE III Urinary Urea/ Creatinine Ratios for Twelve Males First S econd Third Fourth Subject lVeek lVeek lVeek lVeek A verage 1 2 3 4 5 6 7 8 9 10 11 12 -------------------­-------------------­-------------­--­--­--­----------------­---­---------------­-------------­-----­-------­------­----­------­------------­---------­--------­--------------------­-------------------­---------­-----­---­7 14 19 13 13 8 21 13 17 14 17 9 17 20 15 14 10 13 15 18 12 17 11 12 16 15 16 12 14 21 19 21 18 12 13 9 8 13 23 9 11 12 19 13 24 10 10 14 16 14 16 10 15 15 18 15 19 11 50 %) than those of subjects No. 9 and No. 11. This seems to in­dicate, since the diets were self chosen, that subjects No. 9 and No. 11 consistently tend to eat high protein diets compared with subjects No. 1 and No. 6. Summary.-A method is described for the quantitative deter­mination of urea in urine by a paper chromatographic method, employing phenol as a solvent and phenol-hypochlorite as the color developing reagent. Recoveries of urea added to urine average 96 %. The method has been applied to the determination of urea in weekly pooled urine samples from 12 males. BIBLIOGRAPHY 1. Dent, C. E., Biochem. J., 41, 240 (1947) . 2. Engel, E., Ber., 8, 699 (1875). 3. Snell, E. D., and Snell, C. T., "Colorimetric Methods of Analysis," Vol. 1, Van Nostrand, New York, 1936, p. 658. 4. Summary of compounds studied by paper chromatography, this Bulletin. 5. Paper on general paper chromatographic methods, this Bulletin. 6. Williams, R. J., and Kirby, H., Science, 107, 481 (1948). 7. "Klett-Summerson Clinical Manual," Klett Mfg. Co., New York. X. A Micro Determination of Sodium Using Paper Chromatography by HARRY ELDON SUTTON In working with biological fluids, it is frequently desirable to estimate the sodium concentration when only a small amount of material is available. To accomplish this, the following test has been developed which involves only a few micrograms of sodium and requires only a small amount of inexpensive equipment. The analysis employs paper chromatography and is based on the highly specific and sensitive reaction of sodium with zinc uranyl acetate to form a fluorescent compound. The sodium is quantitated by visual comparison with a standard on the same chromatogram. Experimental Procedure Choice of solvents.-In testing various solvent mixtures for the resolution of sodium, it was found that the sodium tends to mi­grate to the positions occupied by the various anions present. With neutral solvents this makes comparison with a standard composed of one substance impossible because of the multiple spots. To overcome this, strongly acidic solvents were tried so that the sodium would migrate almost exclusively to one position with the one anion. The solvent which was decided upon consists of 80 parts of 95 % ethanol to 20 parts of 0.1 N hydrochloric acid. In this solvent, the sodium migrates to an Rf value of .68. The solvent mixture can be used repeatedly. Preparation of chromatograms.-Since the sodium separates from the interfering substances rather rapidly, it is not necessary to allow the solvent to rise very high. Sheets about 12 cm. high and 26 cm. wide are used with the spots placed 2.5 cm. from the bottom of the chromatogram and 2 cm. apart. Six spots each of standard and unknown are placed on the chromatogram in alter­nate positions, the standard occupying the first spot on the right and decreasing in concentration to the left and the unknown occupying the first spot on the left and decreasing in concentra­tion toward the right. For the tests reported here, urine in 1 :8 dilution was used as the unknown. The quantities used were 4, 5, 6, 7, 8, and 9 microliters, no more than five microliters being applied at a time. For the standard, similar volumes of a sodium chloride solution (LOO mg./ml.) were used. The solvent is allowed to rise to about one cm. from the top of the sheet (60-90 minutes). In selecting the standard the following salts were tested in con­centrations containing equivalent amounts of sodium: NaCl, Na2S04,NaN03, sodium acetate, NaHC03, sodium hydrogen tar­trate. There was no visible difference in the ability of the salts to form the fluorescent compound. Color development of the chromatograms.-After the chromato­grams are dried, they are developed by either of two methods with a solution of zinc uranyl acetate prepared according to directions given on page 32. Care must be taken in preparing this reagent according to directions, otherwise the reagent may not be sufficiently sensitive to the amounts of sodium to be detected. The preferred method of development is to spray the chromatogram lightly with the reagent. The spray should be very fine, and it is only necessary to develop the area to which the sodium is known to migrate (de­tectable by a thin yellow line due to a second acid front). The reagent should be sprayed only if adequate facilities are available to prevent contact of the zinc uranyl acetate with the skin. An :ilternate method of development is to dip the chromatogram into the reagent. This procedure can give equally as good results as spraying, but greater precautions must be taken to prevent con­tamination with sodium from the hands or other objects. The chromatogram should not be allowed to touch the dipping tray as this causes the spots to spread as well as causing contamination. Rubber · gloves should be worn if the chromatograms are to be manipulated by hand. Marking the chromatogram.-After the chromatograms have been developed they are observed under ultraviolet light, the sodium appearing as bright blue fluorescent spots. If the chromatograms were dipped, they should be observed immediately to avoid streaking; if they were sprayed, they are more easily observed in the dry state. The sodium in the unknown is quantitated by se­ lecting the two adjacent spots which appear to be equal in intensity and equating the amount of sodium in them. Thus if a spot of the urine sample is equal in intensity to seven micrograms of standard sodium chloride, then from the known volume and di­ lution of the urine, the original concentration of sodium in the urine can be calculated. The spreading of the sodium during resolution is a function of the concentration, and the area of the spot is considered along with the intensity of fluorescence in matching spots. It is sometimes possible to estimate amounts by interpolation. Expression of results'.-Table I has been used to convert the results of the chromatograms into concentration of sodium in the unknown. The figures given as concentration of sodium should be multiplied by the appropriate concentration or dilution factor and rounded off to the proper number of significant figures to obtain the value of sodium in the original unknown. Assuming no error in matching adjacent spots, the maximum error in the extremes of the table would be 11 % with the accuracy increasing toward the center. In practice, if sufficient precautions are taken, it is possible to stay within this maximum theoretical error. TABLE I Table for Converting Matched Spots on Chromatograms into Concentration of Sodium. Volume Volume Cone.Na Unknown Standard in Unknown (µl) (µl) (mg./ml.) 9 4 0.176 8 4 0.198 8 5 0.246 7 5 0.282 7 6 0.341 6 6 0.397 6 7 0.464 5 7 0.556 5 8 0.635 4 8 0.794 4 9 0.893 Results To determine the accuracy of the method as applied to urine, several samples were analyzed by one experimenter using this technique, and the values recorded. These samples were then diluted 1 :2 and sodium chloride was added by a second experimenter in amounts unknown to the first experimenter but sufficient to restore the concentration to the range normally found in urine. These altered samples were then analyzed as "unknowns," single deter­minations being made. The results are recorded in Table II. The average error for these nine determinations is 9.4 % . The first five samples were sprayed and the last four dipped, but there is no significant difference in accuracy. Since the original value of the samples was also determined by this method, part of the error may undoubtedly be attributed to that source. TABLE II Comparison of Experimentally Determined Sodium Concentrations with Known Values. Sodium Sodium found in diluted in original original plus Sodium as samples added amount determined Per Cent Per Cent Sample (mg./ml.) (mg./ml.) (mg./ml.) recovery error 1 ···················· 4.1 3.2 3.5 109 9 2 4.8 2.7 3.2 118 18 3 2.3 1.7 1.8 106 6 4 3.2 2.6 2.7 104 4 5 3.7 2.6 2.3 88 12 6 4.1 2.7 2.9 107 7 7 4.8 4.4 3.7 84 16 8 3.2 3.8 3.7 97 3 9 3.7 4.1 3.7 90 10 Discussion The methods in use at present for the determination of sodium generally require expensive equipment or involve considerable time. Perhaps the chief advantage of the method presented here is that it requires neither. The results of a given determination can be obtained within two hours, and if many determinations are to be done, one person can average five to ten determinations an hour. The materials required are very simple and easy to obtain. The accuracy of the test compares rather favorably with that of other techniques, especially when the range of concentra­tions is known in advance. Another advantage of this technique is the small amount of unknown material required. For most samples of human urine, only 39 microliters of a 1 :8 dilution are actually used. This appli­cability to very small amounts becomes particularly advantageous, for example, when working with small animals. In addition to use in the analysis of urine, the method should also be adaptable to such materials as saliva, blood, tissue extracts, etc. XI. Quantitative Study of Ketosteroids by Paper Chromatographic Methods with Applications to Human Urine by ERIC BLOCH, ERNEST BEERSTECHER, JR., AND ROY C. THOMPSON Zaffaroni, Burton, and Keutmann have described a method for the qualitath~e separation of ketosteroids by paper partition chro­matography (1). The critical feature of their method lies in the fact that prior to being chromatographed the ketosteroids are con­verted to their hydrazone derivatives through the use of Girard's Reagent T (trimethylaminoacetohydrazine) (2). Water solubility is enhanced by this treatment to a varying degree depending on the number of carbonyl groups present in the original molecule, and consequently the spread of Rf values for the various ketosteroids is increased. The problem of locatixg the ketosteroid spots on the developed chromatogram is also simplified by the presence of the reactive quarternary nitrogen groups which produce colored com­pounds by reaction with suitable alkaloid reagents. In the present study, Zaffaroni's method has been further inves­tigated and adapted to the quantitative estimation of ketosteroids. Experimental The reagents used and their methods of preparation were essen­tially those described by Zaffaroni, et al., (1). Hydrazones were formed by Zaffaroni's alternate "Method B" in which the keto­steroid, Girard's Reagent T, and acidified methanol were incubated at 40° C. for two hours or more. Zaffaroni employed water­saturated n-butanol and a 1 :1 :1 mixture of tert-butanol, n-butanol, and water as developing solvents. In the present investigation a large number of solvent mixtures were studied over a wide range of composition ratios. Best results were obtained using a mixture containing 84 parts n-butanol and 16 parts water (which is essen­tially a water-saturated solution). A tert-butanol-n-butanol-water mixture in a volume ratio of 1 :2 :1 also proved satisfactory. The techniques employed in preparing and developing the paper chro­matograms have been described elsewhere (3, 4). Color was de­veloped with K2Ptl6 as described by Zaffaroni, et al. (1). A general description of the quantitative procedures employed will be found elsewhere in this bulletin ( 4). Quantitative inferences were drawn from the area of the developed spots. Since the color developed with the K2Ptl6 solution fades rather rapidly on drying, the outline of the spot was marked with a pencil while the sheets were still damp. Urine samples for determination of ketosteroids were obtained from 14 subjects, seven male and seven female. One of the subjects (No. 6) was in the eighth month of pregnancy. Approximately 800 ml. of urine were collected from each subject. Samples collected after periods of sleep were not included. A modified form of the Pincus method of hydrolysis and extraction ( 5) was followed in the separation of the ketosteroid fractions from the urine samples. To 750 ml. of urine, 110 ml. of concentrated hydrochloric acid were added and the mixture was refluxed for seven to nine minutes. The solution was then quickly cooled to room temperature under running water. The hydrolyzed urine was continuously extracted with 220 ml. of diethyl ether for six to seven hours, a modified form of the Hershberg-Wolfe cyclic extractor being used ( 6) . The ether extract was washed successively with 50 ml. of distilled water, twice with 25 ml. of a saturated sodium carbonate solution, four times with 2 N sodium hydroxide and three times with 30 ml. of distilled water. The resulting ether extract was evaporated to dryness and the residue dissolved in five milliliters of 10% methanolic-acetic acid. For conversion of the ketosteroids to hydrazones, 20-40 mg. of Girard's Reagent T were added and the solution was incubated at 40°C. for two hours or longer. Although much pigmented material was removed during the v,rashing of the ether extract, the methanolic-acetic acid ketosteroid hydrazone solution was still highly colored. Most of this color was removed by filtering the solution through Darco charcoal. Any D.dsorbed hydrazones were eluted by repeated washings with 0.5 ml. of 10 6.1-cholestenone ________ 1 <:r" Dehydroisoandro­ ~ sterone-acetate ________ 1 0 1 1 21 .63 .84 .49 -;:s .... Pregnane-3,20-~ (1> dione ------------------------1 1 0 0 21 -.91 ( ?) .48 .70 ~ .... .,... Allopregnane-cs::'!3,20-dione _1 1 0 0 21 .61 .86 .48 -0 ...... Isoandrosterone ________ 1 0 1 0 19 .58 .83 .49 ­l""3 (1> Testosterone --------------1 0 1 1 19 .56 .81 .49 .69 ~ Estrone ---------------------1 0 1 3 18 .52 .92 .47 -~ Progesterone ______________ 2 0 0 1 21 .19 .45 .15 .28 (.59) (.84) Desoxycortico­sterone --------------------2 0 1 1 21 .17 (.66) *Ketosteroids were purchased from Bios Laboratories,with the exception of dehydroisoandrosterone-acetate and 6,1-cholestenone, which were supplied through the courtesy of Schering Corp. and Eli Lilly and Co., respectively. Urinary Ketosteroids T (color developed with Tollen's reagent) was noted in the case of one preparation. In Table II are summarized the results of experiments designed to establish a relationship between the quantity of ketosteroid hy­drazone applied and the area of the resultant developed spot. The solvent was 84 :16 n-butanol-water. The areas recorded are the average of four determinations. Where a single value differed from the mean by more than four times the·average deviation of the three remaining values, it was eliminated. The area measurements are quite reproducible except at the five microgram level. The marked effect of temperature on spot area is shown by the two estrone and dehydroisoandrosterone-acetate experiments conducted at 15° and 25° C. Since the temperature in these experiments was controlled only within limits of perhaps +2° C., the observations should be interpreted accordingly. They indicate the importance of careful control of temperature, or the incorporation of standards on the same chromatograph sheet with unknown samples, if the areas measured are to be of quantitative significance. TABLE II Relationship Between Amount of Ketosteroid Hydrazone Chromatographed and Area of the Resulting Developed Spot Steroid hydrazone Micrograms of ketosteroid applied Approx5 10 15 20 25 30 Temp. Testosterone area (cm2 ) 1.08 1.98 3.63 3.60 4.15 5.93 15°C. avg. dev. ( o/o) ----13 9 6 2 1 8 .61 .60 .60 .60 .59 .58 Rf ---------------------------­ Progesterone area (cm2 ) 0.54 1.41 1.91 1.99 2.48 2.58 15°C. avg. dev. ( o/o) ----11 12 8 7 7 12 .21 .24 .25 .21 .21 .21 Rf ---------------------------- Isoandrosterone (cm2 ) __________1.76 3.80 5.13 5.43 6.78 7.47 25°C. area avg. dev. (%) ----22 4 2 3 4 3 .72 .72 .70 .68 .70 .70 Rf ---------------------------- Estrone (cm2 ) 2.40 3.81 4.32 4.94 5.41 15°C. area avg. dev. (%) ----7 2 5 2 .71 .69 .64 .65 .68 .69 Rf ---------------------------- Estrone 1.78 5.98 8.28 9.02 8.95 10.58 25°C area (cm2 ) ---------­ 2 3 6 2 5avg. dev. (%) ---- Dehydroisoandro­ sterone-acetate 2.58 4.04 5.00 6.15 7.46 15°C. area (cm2) ----------0.64 3 3 2 2 1avg. dev. ( o/o ) ----22 Rf ---------------------------­ .66 .69 .65 .66 .67 .67 Dehydroisoandro­ sterone-acetate 5.49 7.17 8.37 11.58 12.39 25°C. area (cm2 ) ----------2.68 avg. dev. ( o/o) ----25 7 5 7 2 1.2 1.0 --=" ~ 0 d0.8 Cll C> 0 ~ c( Q6 w a: c( C> 0 _, 0.4 0.2 (5b) (9b) (7) (5cr) (3) (9cr) (4) 15 20 25 30 jJG. Of" KETOSTEROID FIGURE 1 Log Graph of the Area of Spots Produced by Varying Concentrations of Ketosteroids. Solvent: 84: 16 n-butanol-water mixture. 5b = dehydroisoandrosterone at 25° C. 9b = estrone at 25° C. 7 = isoandrosterone at 25° C. 5a =dehydroisoandrosterone at 15° C. 3 = testosterone at 15° C. 9a = estrone at 15° C. 4 =progesterone at 15° C. Urinary Ketosteroids Also included in Table II are the measured Rf values for the different ketosteroid hydrazones at different concentrations. No ~ignificant change in Rf values with change in concentration is observed. In Figure 1, the logarithms of the areas recorded in Table II are plotted against micrograms of ketosteroid applied. An approx­imately linear relationship is obtained except in the lowest con­centration range. Results on Urine Samples Figure 2 reproduces the patterns obtained by chromatographing 50 p.l. portions of four of the urine samples. The shaded spots were purple in color, the degree of shading indicating the intensity of the color. The unshaded spots were gray-black in color. The purple spot (or spots) of high Rf value is in a position to be expected of keto­steroids with one Girard's Reagent T group bound. The purple spots of low Rf value are in a position to be expected of ketosteroids with two or more Reagent T groups bound. The purple ring at Rf. value .00 was not observed in the chromatography of crystalline keto­steroids. The gray spot of lowest Rf value, labeled G.R.T. in Figure 2, is due to uncombined Reagent T. The gray spots of intermediate Rf value are of unknown nature. They might be due to ether soluble pigments, but they were not affected by charcoal treatment, nor does their position correspond to that of any of the several fluorescent spots which were visible under ultra-violet light. The apparent multiple character of the purple spot of high Rf value led to obvious difficulties in interpretation and in the meas­urement of areas. The line of demarkation between the deep purple spot and the lighter purple areas bordering it on either side (in most of the chromatograms) was quite sharp. The area of the darker spot could therefore be determined quite accurately. Areas of the light purple regions bordering the central spot were deter­mined separately. Despite these difficulties obvious similarities :;,,re apparent in the chromatograms from different urine samples, as are also certain dissimilarities. In Table III are recorded the Rf values and areas of the spots cbtained from the 14 urine samples. Recorded values are the averages of duplicate runs. There was excellent agreement between Rf values in duplicate runs, the difference never exceeding .03. Agreement was not so good in the area measurements. Crystalline ketosteroids chromatographed with the urine samples gave the ( I CJ w CJ 0 CJ C) 0 I ~ I ' ~ . ---~ 8 ' ~ I I \ J ~ e e e N0.12cf N0.6~ NO.Bo' N0.1 7 ----------------.60 .... .,... Male 8 ----------------.61 1.1 .54 4.5 .45 1.6 .30 1.5 .22 1.5 .094 0.54 --~ 9 ----------------.62 1.0 .53 6.0 .41 0.7 --.21 0.4 .096 0.47 --0 10 ----------------.60 1.1 .53 3.6 ----.091 0.34 -­ - 11 ----------------.62 1.1 .53 4.5 .43 0.6 .29 0.5 .22 1.2 .096 0.42 --""3 (1) 12 ---------------.59 1.4 .53 2.5 .47 x .27 x .20 0.6 ----~ 13 ----------------.62 1.2 .51 4.5 .41 0.7 .28 0.5 .22 1.1 .093 0.30 --Cl> ~ 14 ----------------.63 0.7 .55 5.2 .46 0.9 .29 0.5 .22 0.5 .110 0.44 *pregnancy urine. x spot too small for precise quantitative measurement. Not included in the above table at'e gray spots with an Rf value of .38-.39 observed in three urine.samples, nor the spots due to unreacted Reagent T with Rf values of .04-07. Urinary Ketosteroids The comparison of these semi-quantitative data with the pub­lished results of others is complicated by the method of urine collection employed in this study. The averages of 1.8 mg. of 17­ketosteroids per liter of daytime urine for females, and 2.4 mg. per liter of daytime urine for males, compare with the following TABLE IV Quantitative Estimation of 17-Ketosteroids in Urine Urine 17-Ketosteroid Sample (mg/ liter) Female 1 ------------------------------------1.9 2 -----------------------------------1.0 3 ------------------------------------1.2 4 ------------------------------------2.5 5 ------------------------------------1.5 6* ------------------------------------5.8 7 ----------------------------------1.1 Female Avg. (excluding 6) ------------------------------------1.8 Male 8 ------------------------------------1.8 9 -----------------------------------4.2 10 ----------------------------------1.3 11 ----------------------------------1.8 12 ------------------------------------0.8 13 ------------------------------------2.0 14 ------------------------------------2.2 Male Avg ~------------------------------2.4 *pregnancy urine. values obtained for normal young adults by other workers: "androgen" : male: 3.8 mg./ l. ( 9 ) , 9.9 mg./ day (10), 9.05 mg./day (11) female; 4.2 mg./l. . (10), 6.75 mg./ day (11) androsterone : male: 1.6 mg./ l. (12), 1.2 mg./l. (14), 2.9 mg./day (13) female: 1.3 mg./ l. (13) eticholane-3 (a)-ol-17-one : male: 1.4 mg./ l. (12) female: 1.3 mg./ l. (12) dehydroisoandrosterone: male: 0.2 mg./ l. (12), 2.0 mg./day (13) female: 0.2 mg./ l. (12) It would seem reasonable to conclude that, while the quantitative estimates of the present investigation are no doubt low, they are nevertheless of the correct order of magnitude and sufficiently precise to permit of valid comparison between individual keto­ steroid excretion patterns. The urine sample standing in unique contrast to all others tested is the pregnancy urine (No. 6) . The very large "l7-keto­:3teroid spot" is probably due largely to steroids of the pregnane series which are found in large amounts in pregnancy urine and which would be expected to be chromatographically indistinguish­able from the 17-ketosteroids. Summary.-Zaffaroni's method for separating ketosteroids by paper partition chromatography has been adapted for quantitative estimation of such compounds. The logarithm of the area of the spots obtained on the chromatogram was shown to be linearly related to the amount of ketosteroid chromatographed, for amounts in excess of 1'5 micrograms. The areas of the spots were markedly affected by temperature, but were quite reproducible under con­stant environmental conditions for amounts of 10 micro­grams or more. Ketosteroid fractions of the urine of 14 subjects were studied using this method. The chromatograms showed general similarities, but qualitative and quantitative evidences of individuality were apparent. All female urines were characterized by two ketosteroid hydrazone spots of low Rf value, while only one such spot was pre­sent on chromatograms from male samples. The single pregnancy urine sample was marked by an increased content of ketosteroid hydrazones migrating with high Rf values. Quantitative estimation of the ketosteroid content of the urines was hampered by diffuse­ness and/or overlapping of the principal spots, and by uncertainty as to their identification with known ketosteroids. BIBLIOGRAPHY 1. Zaffaroni, A., Burton, R. B., and Keutman, E. H., J. Biol. Chem., 177, 109 (1949). 2. Girard, A., and Sandulesco, G., Helv. chim. acta, 19, 1095 (1936). 3. Williams, R. J., and Kirby, H., Science, 107, 481 (1948). 4. Paper on general chromatographic techniques, this Bulletin. 5. Johnston, C. D., Science, 106, 91 (1947). 6. Hershberg, E. B., and Wolfe, J. K., J . Biol. Chem., 133, 667 (1940). 7. Lieberman, S., Dobriner, K., Hill, B. R., Fieser, L. F., and Rhoads, C. P., J. Biol. Chem., 172, 263 (1948). 8. Lieberman, S., Fukushima, D. K., and Dobriner, K., J. Biol. Chem., 182, 299 (1950). 9. McCullagh, E. P., and Lilga, H. V., Endocrinology, 26, 753 (1940). 10. Dingemanse, E., Borchardt, H., and Laquer, E., Biochem. J., 31, 500 (1937). 11. Callow, N. H., Callow, R. K., Emmens, C. W., and Stroud, S. W., J. Endocrinol., 1, 76 (1939). 12. Callow, N. H., and Callow, R. K., Biochem. J., 34, 276 (1940). 13. Hoskins, W., and Webster, B., Proc. Soc. Exptl. Biol. Med., 43, 604 (1940). 14. Callow, N. H., Biochem. J., 33, 559 (1939). XII. The Determination of Ferric Chloride Chromogens in Human Urine by Paper Chromatographic Methods by HARRY ELDON SUTTON AND ERNEST BEERSTECHER, JR. Ferric chloride has long been used as a qualitative reagent for the identification of certain classes of compounds, such as phenols, oximes, enols, etc., and tables of compounds and their correspond­ing colors with ferric chloride can be found in the literature (1, 2). Ferric chloride would therefore seem to be a useful reagent for the development of color with certain compounds separated on paper chromatograms. The present investigation is an application of this reagent to the qualitative, and, where possible, quantitative determination of certain constituents of human urine on paper chromatograms. Experimental The general chromatographic techniques employed have been previously described (3, 4). The solvent employed for the develop­ment of the chromatograms was a mixture of four parts n-butanol, one part glacial acetic acid, and one part distilled water. Other solvents commonly used in paper chromatography such as phenol, collidine, and pyridine, could not be used since the solvents them­selves react with ferric chloride to form colored products. The paper used was 28 cm. in height. An aliquot of urine con­taining 40 to 60 micrograms of creatinine was used as the test amount. This volume contained enough of the chromogenic material to react fully and not too much to prevent good resolution. Chromatograms were allowed to dry in air. Warm air was some­times passed over the sheets to speed the process, but they were never subjected to the higher temperatures for longer than five minutes. Color was developed by spraying the sheets with a one percent aqueous solution of ferric chloride. This concentration gives maximum color, while keeping the background color at a minimum. With the exception of tartrate, the spots were not clearly visible until the water had been removed from the sheets. Rapid drying in warm air seemed to deepen the colors. As the maximum color intensity was obtained immediately after drying, the chromatograms were marked as soon as possible after color development. Quantitative estimates, where possible, were based on the areas of the developed spots. The techniques involved and a general con­sideration of this procedure will be found elsewhere in this bulletin (3). Results Qualitative identification of ferric chloride chromogens in urine. -The spots obtained by spraying urine chromatograms with ferric chloride solution are summarized in Fig. 1. All of these spots did not appear in the urine of every person, nor in all samples from a given person. Most of them, however, appear to be normal daily constituents of urine. The identification of these spots with known compounds involved the paper chromatographic testing of over 125 different compounds known to be, or possibly, present in urine. These included the naturally occurring amino acids, a large number of sugars, organic acids, vitamins and other metabolites, and many inorganic salts. The results of these tests are recorded elsewhere in this bulletin (5). In addition to this study of pure compounds, a "synthetic urine," with a composition falling within the normal range of variation of the major urine constituents (6), •vas prepared and studied. While several of the amino acids showed chromogenic activity, only faint spots were obtained even at concentrations much higher than would be expected to occur in urine, and none of the Rf values corresponded to those of similarly colored spots on urine chromato­grams. None of the sugars showed color formation at concentra­tions corresponding to or even higher than those occurring in urine. Sodium tartrate gives a bright yellow spot at Rf .25 which disappears on drying the sheet. This Rf value and the unique be­havior on drying correspond to spot E of Fig. 1, which is there­fore ascribed to tartrate. Citric acid corresponds in Rf value and color to spot F of Fig. 1. This spot is usually very faint in urine samples and is difficult to mark accurately. The sodium salts of lactic acid, acetic acid, and a-ketoglutaric acid give brown spots which correspond very closely to spot G. All sodium salts of weak acids seem to give such a brown spot, N a2 C03 and NaHC03 behaving similarly. Presumably these salts of weak acids react with the acetic acid in the solvent to form sodium acetate which migrates corresponding to an Rf value of about .35, where it reacts, upon spraying with ferric chloride solution, to precipitate ferric hydroxide. Other ions such as magnesium, potassium, calcium, and cadmium give the same effect as sodium a.nd may contribute to this "alkaline" spot. Spot G may therefore Rf 1.00­ .90­ .80­ .70­ .60­ .50­ .40­ .30­ .20­ .1 o­ 0.00­ 0 0 0 8 0 B t::) 0 COLOR DESIGNATION PURPLE N BROWN BROWN L BROWN K BLUE-GREENJ WHITE BROWN H BROWN G GRAY F YELLOW E (VISIBLE WHEN WET) YELLOW D WHITE C BROWN B WHITE A FIGURE 1 IDENTIFICATION SALICYLATE PABA PYRIDOXINE UREA CREATININE ~WEAK ACID" ALKALI CITRATE TARTRATE CHLORIDE PHOSPHATE SULFATE Spots Obtained from Urine with Ferric Chloride Chromatograph. be considered a measure of the weak acids present in the urine. Creatine, which migrates to the same position as spot G and which gives a color similar to the alkaline salts, occurs in such small quantities in normal urine that its contribution to spot G may be ignored. Of the vitamins tested, pyridoxine and p-aminobenzoic acid show the greatest chromogenic activity, one microgram of each being capable of producing a clearly defined spot. These spots correspond roughly to spots found in most urineS-pyridoxine to spot K and p-aminobenzoic acid to spot M-but the urinary excretion of these substances is so low that it is difficult to draw definite conclusions. These spots found on urine chromatograms are usually too faint to mark accurately and while they may be presumed to be caused by the two vitamins, they cannot be safely used as a measure of excretion of these vitamins. Urea produces a bleached area of Rf .60. The minimum detect­able quantity is about 30 micrograms or the amount present in about one microliter of urine. Spot 1 of Fig. 1 may therefore be assigned to urea. Creatinine, which corresponds to spot H, may best be demonstrated by use of the picric acid reagent (7) which is specific for creatinine and which forms the basis for the standard colorimetric test for creatinine. By the use of this reagent, it can be shown that the creatinine in urine and spot H always occupy the same positions, are always distorted in the same manner by excessive concentrations of urea, and are always of the same relative concentrations on the basis Qf color intensity. Salicylic acid, which corresponds to spot N, gives an intense purple color with ferric chloride-a fact which is the basis for the clinical test for salicylates, which appear in the urine only after consumption of salicylate drugs (8). The specificity of the ferric chloride color reaction together with the intense fluorescence leave little doubt as to the identity of salicylic acid with spot N. Inorganic constituents of urine account for three more of the urinary chromogens; sulfate corresponds to spot A, phosphate to spot C, and chloride to spot D. Not all of the different salts of the same acid show the same ability to produce spots. As most of the urine acids are present as sodium and potassium salts, these are of particular importance in establishing the identity of the spots. The identity of the phosphate spot may be best established by spraying the chromatogram with the specific ammonium molyb­ date reagent (9). The spot produced by such a procedure coincides with spot C. Pure chlorides give no color reaction with ferric ::hloride. However, repeated tests with synthetic urines and the use of color developing reagents such as bromcresol green indicate that spot D must be due to chloride, perhaps in combination with some other constituent of urine. The mechanisms of the chloride and sulfate reactions are not clear. The remaining spots, B, J, and L, have not been identified with any known substance. Spots Band Lare very faint and frequently cannot be detected. Spot J is a bright bluish green, usually small and clearly defined, but seldom occurring in quantities sufficient to be detected with ferric chloride. The reactions of this chromogen with other reagents are reported elsewhere in this bulletin (10). Quantitative estimation of phosphate and urea.-Most of the ferric chloride-developed spots on urine chromatograms are of such a shade of color as to make color intensity comparisons very difficult. This is particularly true in view of the yellow ferric chloride background. Many of the spots are also i'!O ill-defined as to make area measurements impractical. Only in the case of phos­phate and urea has it been found possible to devise suitable pro­cedures for quantitative estimation. This is not to say, of course, that gross quantitative information of a relative nature cannot be deduced for many of the other constituents. For phosphate, a reasonably linear relationship was observed between the amount of phosphate chromatographed and the area of the resulting spot. Comparison of the areas of urinary phosphate spots with the areas of standard phosphate spots to obtain quantitative values of the phosphate concentration gives results which average forty to fifty percent too high when compared with the values obtained on the same samples by standard colorimetric procedures (9). Nevertheless, the chromatographic detection of phosphate coupled with its approximate determination has proved to be useful in studies involving large numbers of urine samples, particularly since from the same chromatogram other valuable information can also be derived. For the quantitative estimation of urea, the spot comparison technique (3) was employed, spots being matched on the basis of visually estimated area only. Spots developed from five microliters of urine plus varying amounts of urea were compared with adjacent spots from 10 and 15 microliters of urine alone. The results of such determinations on six urine samples are shown in Table I. Values are also given for the urea cont~nt of these samples as determined by the standard urease method (10). The average percent difference between the paper chromatographic determi­nations and the urease determinations was 14%. The error of the paper chromatographic method could no doubt have been somewhat reduced by measuring the areas of the compared spots with a planimeter and interpolating between standard spots. TABLE I Urea Content of Urine (mg./ml.) Paper Chromatographic Urease Percent Sample Values Values Difference 1 ------------------------12 13.5 11 2 18 23.6 24 3 21 19.9 5.5 4 27 23.2 16 5 12 15.3 21 6 7.5 7.9 5.0 Summary.-Spots obtained on paper chromatograms of urine prepared using a n-butanol-acetic acid-water solvent and developed for color with ferric chloride have been identified as sulfate, phosphate, chloride, citrate, "weak acid" alkali, creatinine, urea, pyridoxine, p-aminobenzoic acid, and salicylate. Three additional spots remain unidentified. Methods are presented for the roughly quantitative estimation of phosphate and the more satisfactory determination of urea. BIBLIOGRAPHY 1. Hickinbottom, W. J., Reaction of Organic Compounds, Longmans Green, New York, 1936, p. 118. 2. West, E. F., and Brode, W.R., J. Am. Chem. Soc., 56, 1037 (1934). 3. Paper on general chromatographic techniques, this Bulletin. 4. Williams, R. J., and Kirby, H., Science, 107, 481 (1948). 5. Paper on compounds studied chromatographically, this Bulletin. 6. Williams, R. J., and Beerstecher, E., An Introduction to Biochemistry, Van Nostrand, New York, 1948, p·. 695. 7. Peters, J. P., and Van Slyke, D.D., Quantitative Clinical Chemistry, Vol. 2, Williams and Wilkins, Baltimore, 1932, p. 599. 8. Gross, M., and Greenberg, L. A., The Salicylates, Hillhouse Press, New Haven, 1948. 9. Peters, J. P., and Van Slyke, D. D., Quantitative Clinical Chemistry, Vol. 2, Williams and Wilkens, Baltimore, 1932, p. 878. 10. Unknown compounds, this Bulletin (Chapter II, Table llb, Compound 1). 11. Levinson, S. A., and M.acFate, R. P., Clinical Laboratory Diagnosis, Lea and Febiger, Philadelphia, 1937, p. 453. XIII. The Effects of Single Vitamin Deficiencies on the Consumption of Alcohol by White Rats by ERNEST BEERSTECHER, JR., JANET G. REED, WILLIAM DUANE BROWN, AND L. JOE BERRY I. Introduction Previous nutritional studies with individual rats (1-3) have shown that their ad libitum consumption of 10% ethyl alcohol is highly individualistic and may be profoundly influenced by dietary means. On diets marginal in B vitamins, nearly all of the animals studied consumed large quantities of alcohol, while on the same diets supplemented with a mixture of B vitamins and other growth fac­tors, only a few animals of one strain consumed any alcohol whatso­ever. Based on this evidence as well as evidence for the involvement of a genetic mechanism in the appetite for alcohol, the concept of ge­netotrophic disease was developed ( 4). As a result of this concept, a treatment has been proposed for alcoholism in humans and sub­jected to clinical trial ( 5) . In the course of such studies, however, it has not as yet been possible to determine just which growth fac­tors were most specifically involved in controlling the appetite for alcohol, and it has therefore been impossible to study in detail the exact mechanisms involved. A modification of the previous experi­mental approach was thus expedient in order to obtain further information as to which specific vitamins were involved in the phenomena. The present study involves the assessment of the role of a num­ber of specific vitamins in relation to appetite for alcohol, as de­termined with single vitamin deficiency studies with rats, and the evidence presented shows conclusively that the absence of any one of a number of nutritional factors from the diet may cause the de­velopment of alcoholic consumption in rats. The results are in line with the genetotrophic concept and also with recent findings relat­ing to the distinctive metabolic patterns in human compulsive drinkers ( 6) . II. Experimental The experiments were all performed on the two strains of white rats previously described as "O" and "H" ( 1) . The rats were kept in individual false bottom cages, each equipped with two drinking vessels, one containing lO o/o by volume of alcohol, and the other water. The positions of the two vessels were interchanged from day to day to rule out the element of habit. The animals were placed on the described deficiency diets, and their weights and consump­tion of 10% alcohol and of water were measured at two to three day intervals. When the alcohol consumption had arisen to a suit­able level, generally about 0.5 ml. absolute alcohol/100 gm. rat/ day, the vitamin missing from the diet was administered for a three day period, and the effect on alcohol consumption was noted for a period of two to three weeks before other supplements were tried. When the specifically deficient vitamin failed to abolish the appetite com­pletely, other dietary supplements were tried until final abolition of the appetite was achieved. In some cases, a final cure was achieved only after a period of nearly a year, thus affirmatively answering a question raised earlier (1) of whether a long standing appetite could be cured in older animals. The diets employed were the standard vitamin deficient diets shown in Table I, all however containing Salt Mixture No. 2 (7) as the mineral supplement. In later experiments, several special diets were prepared as follows: a. High protein diets were prepared by reversing the usual per­centages of casein and carbohydrates; i.e., approximately 68 % casein and 18 % sucrose. b. High amino acid diets were prepared by substituting half of the casein (18% ) with an enzymatic yeast-protein digest, containing 57 % of the total nitrogen as amino nitrogen. c. Nucleic acid diets were prepared by adding 1.5 gm. of Fleisch­mann's technical grade nucleic acid to each 100 gm. of the regular deficiency diet or of the other diets as indicated. d. A typical high protein-nucleic acid-amino acid diet was pre­pared by combining 60 % casein, 17 % sucrose, 8 % Crisco, 2% cod liver oil, 9% enzymatic yeast-protein digest, 1.5% nucleic acids, 4 % salt mixture #2, and the usual vitamin supplement. At an early stage in the experiments it was found (evidence to be presented in later paragraphs) that the classical deficiency diets in Table I were in many cases deficient in more than a single vitamin. A later group of test animals was therefore fed the de­ficiency diets fortified (per kg. of diet) as follows: thiamine, 60 mg.; riboflavin, 100 mg.; pantothenic acid, 600 mg.; pyridoxine, Alochol Consumption and Vitamin Deficiencies 117 30 mg.; choline, 2 gm.; inositol, 600 mg.; PABA, 600 mg.; nicotinic acid, 200 mg.; biotin, 1 mg.; and folic acid, 10 mg., except that the deficient vitamin in question was omitted in each case when appro­priate. Strain 0 rats on this supplement are designated hereinafter as "0+9V" and strain Has "H+9V". The nature of other special supplements used is indicated in the discussion of the results. TABLE I Tabulation of Composition of Single Vitamin Deficiency Diets Component Vitamin Deficient from Diet of Vitamin A Thiamine Riboflavin Pantothenic Vitamin Choline Biotin Diet (8) (9) (10) Acid (11) B. (12) (13) (14) Sucrose o/O------------68% 68% 68% 73% Cornstarch %--------65% 63% (Argo) Dextrinized rice starch %------------64% Casein* o/o ________ l8% 13% 18% 18% 18% 17.7%tt 8% Eggwhitet %------------10% Crisco %------------3% 8% 8% Wesson oil %--------5% 5% Lard %------------10% Saltst %--------4% 3% 4% 4% 4% 4% 4% C.L.0.§ %------------1% 2% 2% 2% II Cystine %------------0.3% Yeast1f %--------8% 20%** Thiamine .20 .20 .20 .20 .20 mg./lOOg. diet Riboflavin .20 .30 .20 .20 mg./lOOg. diet Niacin 2.50 2.50 mg./lOOg. diet Choline 200. 200. 100. mg./lOOg. diet Calcium 250. 250. .20 1.00 pantothenate mg./ lOOg. diet Pyridoxine .30 .20 .20 .20 mg./lOOg. diet *Borden's Labco Brand Vitamin Free. tPalpable desiccated powdered egg white. :IGBI Salt Mixture No. 2. §Rexall, 2000 USP units Vitamin A and 200 USP units Vitamin D per gram. llTwo drops Navitol with Viosterol per rat per week (3300 USP units Vitamin A, 660 units Vitamin D) . VAnheuser-Busch, Primary Dried Yeast, Strain G. ••Autoclaved six hours. tt Extracted four times for two hours each time with boiling 95% ethanol. III. Results Because of the complexity of the results obtained and the large number of individual animals used, it is most convenient to present the results for each deficiency study separately, noting at the same time any evidences of efficacy of the vitamin in question in treating animals not responding to other vitamin supplements. Each of the TABLE II Summary of Experimental Data on Alcohol Consumption of Rats Fed Vitamin A Deficient Diet* (Alcohol consumption in ml. alcohol/100 gm. rat/day) Strain -Strain Total Variable "O" "H" All Rats No. of animals in original group____ No. of animals completing expt._____ Initial consumption level __________________ 9 5 0.27 9 3 0.08 18 8 0.18 (0.01-0.42) (0.00-0.33) (0.00-0.42) Level at initiation of treatment________ 0.57 0.74 0.63 (0.50-0.62) (0.59-0.85) (0.50-0.85) Rats responding to deficient vitamin 1/ 5 2/3 3/8 Mean drop in intake due to administering Vitamin A ___ _____ _ 0.37 0.29 0.32 Rats responding to 10 B vitamins____ Mean drop in intake due to 10 B______ Rats responding to 12 vitamins________ Mean drop in intake due to 12 V______ Rats responding to vitamin B,,________ Mean drop in intake due to Bu-------­Rats responding to protein________________ 5/5 0.36 (0.23-0.54) 3/5 0.42 (0.24-0.57) 0/ 5 2/ 2 (0.29-0.43) 3/3 0.28 (0.20-0.41) 1/ 2 0.23 2/ 2 0.10 (0.09-0.13) 1/1 (0.29-0.43) 8/8 0.34 (0.20-0.54) 4/7 0.37 (0.23-0.57) 2/7 0.10 (0.09-0.13) 3/3 Mean drop in intake due to protein Rats responding to nucleic acid__ _____ Mean drop in intake due to nucleic acid ---­---­-------------------------­0.39 (0.32-0.45) 2/2 0.20 0.30 0.36 (0.30-0.45) 2/2 0.20 Rats responding to amino acids________ Mean d~op in _intake due to (0.20-0.20) 1/ 1 1/ 1 (0.20-0.20) 2/2 ammo acids ---------------------------------­ 0.38 0.44 0.41 Rats r~spon?ing t? protein-nucleic acid-ammo acids -------------­-----------­Mean drop in intake due to protein-nucleic acid-amino acids ____________ 3/3 0.29 1/ 2 0.42 (0.38-0.44) 4/5 0.32 (0.13-0.46) (0.13-0.46)Rats responding to continuous 15 vitamins ---------------------------------­4/4 1/2 5/6 Mean drop in intake due to continuous 15 vitamins ______________ 0.26 0.67 0.37 (0.09-0.51) (0.09-0.67)Average total days on test diet__________ 378 349 367 (340-410) (251-412) (251-412)Average final level of alcohol intake 0.29 0.22 0.27 (0.04-0.52) (0.09-0.32) (0.04-0.52) *It should be noted in Tables II to VI that the data further down in each table apply to rats which have not been completely "cured.. by the preceding treatment indicated. Rats on a deficiency diet which were supplied with the missing vitamin were continuously supplied thereafter. Similarly after 10 B vitamins were given, the animals were continually supl)lied with abundant amounts, etc. Rats to which one vitamin was administered may have been affected 100%, yet the alcohol consumption of several may have remained at an appreciable level so that they were suitable objects for further study. Due to the com­plicated nature of the experiment and the fact that revision of plans was required after preliminary results had been obtained, it is impossible in tables to give the exact history of each individual rat. Alcohol Consumption and Vitamin Deficiencies 119 seven deficiency diets was studied separately. Diagrammatic rep­resentation of individual drinking patterns are presented only to illustrate specific effects, since the patterns in general follow closely upon those previously presented (1) . In the diagrams, the re­sponses in terms of alcohol intake per day are averages of five day periods, and the data so obtairied were grouped to provide the tab­ular material. For each variable, the mean response and the ranges seen in the test group are reported, as well as the proportion of the test group giving any response to the indicated supplement. The vast amount of data thus grouped indicates both the nature of the general responses and also the extreme individual variability which was evident. Effects of Vitamin A Deficiency on Alcohol Consumption.-A study of the data in Table II indicates that vitamin A deficiency does not promote high levels of alcohol consumption in rats, and that vitamin A is not strikingly effective in decreasing alcohol in­take levels. The data do indicate that the "vitamin A deficient diet" is inadequate in regard to its content of the B vitamins, and at least suggest the desirability of higher protein-nucleic acid levels in this particular deficiency diet. In no animal throughout the study did there occur a single large drop in alcoholic intake such as occurred when other vitamin deficiencies were corrected in separate studies. Three of eight test animals showed decreased intakes of only 0.29, 0.37, and 0.43 ml. of alcohol per 100 gm. rat per day. The "index of response" (mean response X fraction responding X 10) was thus 11.3. According to this imperfect measure, vitamin A is one third as important as the next most effective vitamin. It is there­fore concluded that vitamin A may be a factor but is not a major one in the etiology of alcohol consumption in rats. Effects of Thiamine Deficiency on Alcohol Consumption.-A study of the data in Table III indicates that a thiamine deficiency produces in most rats a higher level of alcohol intake, and that thiamine administration is able to bring about a decided lowering of this level. An example of the responses obtained is that shown in Figure I. In this animal, a secondary rise occurred after about a month, reaching a level somewhat lower than that of the intake before treatment. This rise was refractory to thiamine and disap­peared only after continued treatment with vitamin B12• It has been observed that in many animals on deficient diets, supplying the de­ficient vitamin diminishes the alcohol intake, but is ineffective in eliminating a subsequent increase in the alcohol consumption. Some entirely unrelated vitamin may be effective in this case, but it in TABLE III Summary of Experimental Data on Alcohol Consumption of Rats Fed Thiamine Deficient Diet* (Alcohol consumption in ml. alcohol/100 gm. rat/day) Variable Strain "O" Strain "0" + 9V Total Strain "0" Strain "H" No. of animals in original group____________ _ No. of animals completing expt._____________ Initial consumption leveL___________________ ____ 9 7 0.31 10 6 0.18 19 13 0.26 9 7 0.13 Level at initiation of treatment_____________ (0.01-0.67) 0.76 (0.09-0.26) 0.93 (0.01-0.67) 0.85 (0.00-0.41) 1.18 Rats responding to thiamine_ ____________ _ __ Mean drop in intake due to thiamine administration___________________ (0.44-1.06) 6/9 0.43 (0.61-1.33) 8/9 0.63 (0.44-1.33) 14/18 0.55 (0.59-2.00) 5/8 0.66 (0.15-0.79)Rats responding to 10 B vitamins_________ 9/ 9 Mean drop in intake due to 10 B_____________ 0.52 (0.32-1.05) -------­ (0.15-1.05) 9/9 0.52 (0.14-1.23) 6/8 0.44 Rats responding to 12 vitamins______________ Mean drop in intake due to 12 y _______ ____ (0.31-1.00) 2/5 0.48 -------­-------­-------­ (0.31-1.00) 2/5 0.48 (0.09-0.85) 4/5 0.43 Rats responding to vitamin B,,_______________ Mean drop in intake due to Bu­---------------· Rats responding to protein______________________ Mean drop in intake due to protein.._____ (0.26-0.70) 3/4 0.17 (0.10-0.21) 1/1 0.25 -------­-------­-------­-------­---­---­-------­ (0.26-0.70) 3/4 0.17 (0.10-0.21) 1/1 0.25 (0.28-0.50) 5/7 0.28 (0.11-0.43) 1/3 0.20 ----­--- Rats responding to nucleic acid__________ Mean drop in intake due to nucleic acid______ ___________________________________ 1/2 0.32 -----­---------­ 1/2 0.32 -------­-------­ -------­ -------­ ------­- Rats responding to amino acids______________ Mean drop in intake due toamino acids____________________________ _____________ 1/1 0.12 -------­---­---­ 1/1 0.12 1/1 0.21 Total All Rats 28 20 0.58 (0.00-0.67) 0.95 (0.44-2.00) 19/26 0.58 (0.14-1.23) 15/17 0.49 (0.09-1.00) 6/10 0.45 (0.26-0.70) 8/11 0.24 (0.10-0.4?) 2/4 0.23 (0.20-0.25) 1/2 0.32 2/2 0.17 (0.12-0.21) ..... N> 0 ~ ;;::.­ ~ ~ ~. ~ ~ ~ .... .,... ~ c ..... '""'3 ~ ~ ~ ~ Rats r~spon~ing t<;> protein-nucleic ac1d-am1no acids__________________ _ ____________ 0/3 0/3 4/5 4/8Mean drop in intake due to protein-nucleic acid-amino acids______ 0.25 0.25 (0.23-0.30) (0.23-0.30) Rats responding to continuous 15 vitamins_________________________________________ _ ;i:.. 7/7 6/6 13/13 6/7 19/20 C)" Mean drop in intake due to c continuous 15 vitamins_____________________ 0.57 0.32 0.45 0.42 0.44 03" c (0.20-1.34) (0.20-0.63) (0.20-1.34) (0.10-0.64) (0.10-1.34) ...... Average total days on test diet______________ 375 292 334 396 354 (320-410) (276-302) (276-410) (386-410) (276-410) ~ Average final level of alcohol intake_______ 0.22 0.35 0.28 0.35 0.31 i ~ (0.05-0.49) (0.12-0.82) (0.05-0.82) (0.06-0.67) (0.05-0.82) •see footnote p. ll8. .,..,.. .... c <2 [ ~ .... .,..,.. ~ ~· t::l ~ ~ .... ~ <2 ~. ~ Cl> ...... N> ...... 1.2 30 MG. THIAMINE 1.0 >­ < ~0.8 ._ < a: . C> 0 0.6 0 - ' _J ~ I 0.4 _J 0 :c 0 u _J < 0.2 10 20 30 40 DAYS ON DEFICIENT DIET FIGURE I Alcohol Consumption of a Thiamine Deficient Rat and the Effect of Thiamine Administration. Alcohol Consumption and Vitamin Deficiencies 123 turn may be ineffective in controlling still subsequent increase, etc. It is therefore concluded that in these cases, a number of different deficiencies are developing at different rates, depending upon the individual vitamin requirements of the animal in question. If for example, an animal on a single deficiency diet has a high require­ment for some other vitamin two deficiencies may occur at the same time, and supplying the single missing vitamin may be ineffective, unless the secondary deficiency is also corrected. In nearly every case where no response was obtained from thiamine administration to thiamine deficient rats a directly subsequent administration of nine other B vitamins along with the thiamine was effective. If secondary and tertiary deficiencies develop at slower rates, they are manifest in the data as succeeding rises in the alcohol consump­tion which are refractory to the vitamin previously administered. Nineteen of twenty-six experimental animals studied responded to thiamine administration with average drop in alcohol intake of 0.58 ml. The "index of response" as calculated above was thus 42.3, second only to that of riboflavin (below). As with vitamin A, there appeared to be little significant difference between the two strains of rats. The response of strain H on the deficient diet was apparently greater than that of strain 0 to the protein-amino acid-nucleic acid supplement. From these data it seems inescapable that thiamine is involved in some critical fashion with the develop­ment of the appetite for alcohol in rats. Effects of Riboflavin Deficiency on Alcohol Consumption>-A study of the data in Table IV shows that a riboflavin deficiency is highly effective in most rats in bringing about a considerable increase in the appetite for alcohol, and that riboflavin adminis­tration is effective in bringing about a prompt drop in the alcohol intake. Eighteen of a total of twenty-three rats responded to the riboflavin, with an average drop in alcohol intake of 0.90 ml! The "index of response" was thus 70.2, the highest obtained for any of the vitamins studied. This may be attributed in part to the degree of deficiency of the riboflavin deficient diet. Figure II illus­trates a drop typical of those obtained from the administration of riboflavin. This animal was one from "0" strain in which the de­ficient diet was supplemented by large quantities of the other B vita­mins as previously indicated. The curve illustrates the development of successive deficiencies on a diet supposedly deficient in only one respect. In this case there is an excellent indication that the specific deficiencies which appeared to develop successively on this diet were those of riboflavin, thiamine, and vitamin B12• It should be noted that each point on the graph represents a five day average. TABLE IV Summary of E xperimental Data on Alcohol Consumption of Rats Fed Ribofiavin Deficient Diet* (Alcohol consumption in ml. alcohol/100 gm. rat/day) Strain Strain Total Strain Variable "O" "0" +9V Strain "0" "H" No. of animals in original group__________ __ 9 10 19 15 No. of animals completing expt._____________ 4 7 11 8 Initial consumption leveL________________________ 0.33 0.23 0.28 0.20 (0.07-0.81) (0.14-0.42) (0.07-0.81) (0.00-0.62) Level at initiation of treatment______________ 1.02 1.47 1.32 1.01 (0.55-1.80) (0.34-2.80) (0.34-2.80) (0.56-1.54) Rats responding to riboflavin -----------------3/4 7/8 10/12 8/11Mean drop in intake due to riboflavin administration______ ____________ 0.86 1.36 1.21 0.51 (0.45-1.57) (0.64-2.51) (0.45-2.51) (0.17-0.90)Rats responding to 10 B vitamins_________ 2/2 (6/ 6t) 2/ 2 9/ 12 Mean drop in intake due to 10 B___________ 0.34 1.10 [ (0.64t) ] 1.10 (0.70-1.50) (0.25-1.55) (0.70-1.50) (0.29-0.56) Rats responding to 12 vitamins_ ____________ 3/3 3/3 --------7/ 11 Mean drop in intake due to 12 V_____________ 0.40 --------0.40 0.24 (0.22-0.65) (0.22-0.65) (0.12-0.35)Rats responding to vitamin B,,_ _____________ 3/3 2/2 5/5 9/11 Mean drop in intake due to ~,._________________ 0.46 0.33 0.40 0.49 (0.17-0.78) (0.20-0.46) (0.17-0.78) (0.16-1.07)Rats responding to protein__________________,___ 0/ 1 --------0/1 0/ 2 Mean drop in intake due to protein_______ Rats responding to nucleic acid______________ 0/1 0/1 --------0/2 Mean drop in intake due to nucleic acid__________________________________________ Rats responding to amino acids______________ 2/ 2 2/2 5/5 Total All Rats 34 19 0.24 (0.00-0.81) 1.17 (0.34-2.80) 18/ 23 0.90 (0.17-2.51) 11/14 0.49 (0.29-1.50) 10/14 0.28 (0.12-0.65) 14/16 0.47 (0.16-1.07) 0/3 0/3 7/7 .... ~ ~ ~ ~ ~ ~· ~ ~ .... C'io ~ c ....... ~ ~ ~ ~ Cl) / Mean drop in intake due to amino acids------------------­·---------­--------------------­---· Rats r~spon~ing t? protein-nucleicac1d-am1no acids_______ _________________________ Mean drop in intake due to protein-nucleic acid-amino acids_____ _ Rats responding to continuous 15 vitamins________________ ______________________ _ Mean drop in intake due to continuous 15 vitamins._ _ _. Average total days on test diet ·--­Average final level of alcohol intake *See footnote p. 118. tThiamine only. 0.37 (0.18-0.56) 3/3 0.83 (0.56-1.20) 4/ 4 0.47 (0.25-0.74) 387 (368-410) .09 (0.02-0.16) 297 (294-300) 0.28 (0.02-0.68) 0.37 (0.18-0.56) 3/ 3 0.83 (0.56-1.20) 4/4 0.47 (0.25-0.74) 330 (294-410) 0.21 (0.02-0.68) 0.46 (0.36-0.67) 7/10 0.59 (0.31-1.21) 1/ 2 0.55 -------­222 (171-392) 0.20 (0.03-0.39) 0.43 (0.18-0.67) 10/ 13 0.66 (0.31-1.21) 5/ 6 0.49 (0.25-0.74) 284 (171-410) 0.21 (0.02-0.68) ~ ...... Q c ~ ~ C':l c ~ ~ ~ '1::l <-+­""·c ~ ~ ~ ~ ""·.,.... ~ ~­ t::i Cl) ';:::t, Q""·Cl) ~ Q""·Cl) Cl> .... Nl Ol •RIBOf LAVIN 10MG. ~ 0 !;(2.5 ' a: ~ g <;: _j2.0 ~ z 0 t= o.1.5 ~ ::> en THIAMINE t MG. z 0 0 1.0 .J 0 :c 0 0 .J <0.5 ~ N> O') 1--3 ~ (1:> ~ ~· (1:> ~ ~ ~ c ........ 1--3 (1:> !.'l ~ <:!> In general the strain differences noted in the responses were ap­preciable but not large. It seemed that the animals on diets supple­mented with large amounts of other vitamins tended to become more deficient in the one missing vitamin (as manifest by increased alcohol intake) than did animals which were on diets not so supple­mented. This trend was observed in most cases throughout this study. The response to riboflavin was apparently considerably less dramatic in the strain H animals than in strain 0. The data leave little question as to the importance of riboflavin in the de­velopment of the alcoholic appetite in rats. Effects of Pantothenic Acid Deficiency on Alcohol Consumption. -An examination of Table V indicates clearly that pantothenic acid is involved in the mechanism of the development of an appetite for alcohol in rats. Fifty-two out of seventy deficient animals re­sponded to pantothenic acid with a lowered intake of alcohol, giving an "index of response" of 38.5. A considerable strain difference was apparent. The "H" strain animals appeared not to be as de­ficient on the supplemented diet as the "O" strain animals, and did not therefore respond as markedly. A considerable mortality was encountered in the "O" strain before treatment was instituted, and additional data not in the table indicated that on the diet supple­mented with other B vitamins, this was particularly true. A con­sideration of the data thus leads us to the conclusion that strain "0" has a considerably higher pantothenic acid requirement than does strain "H". A typical response of a strain "H" animal to pantothenic acid is shown in Figure III. This animal did not de­velop the appetite for alcohol for some time (about 100 days on the diet'), differing in this manner from litter mates that showed a rise in intake in about half this period. A prompt response was obtained with a massive dose of pantothenic acid, and the intake remained at essentially zero for several months until the animal was killed. Doses of the vitamin used for most other animals were from one tenth to one third of this amount, and appeared to be equally effec­tive. The fact that 28 out of 30 "O" strain rats responded to panto­thenic acid administration shows clearly that this vitamin is in­timately concerned with the control of alcohol appetite. The fact that riboflavin caused on the average larger decreases in alcohol consumption might be interpreted as meaning merely that the deficiency diet in this case was more deficient. The relative im­portance of the magnitude of the consumption decrease, as com­pared with the proportion of the animals responding, is difficult to evaluate. ,..... Summary of Experimental Data on Alcohol Consumption pf Rats N) Fed Pantothenic Acid Deficient Diet* (Alcohol consumption in ml. alcohol/100 gm. rat/day) Strain Strain Total Strain Strain Total Total Variable "0" "O" + 9V Strain "0" ''H" "H"+9 Strain "H" All Rats No. of animals in original group__ __ __ __ No. of animals completing expt.________ __ 9 4 0.31 20 9 0.41 29 13 0.38 15 12 0.22 10 10 0.27 25 22 0.24 54 35 0.31 Initial consumption level________ _ __ __ ____ (0.01-0.92) 1.02 Level at initiation of treatment___________ (0.76-1.66) Rats responding to pantothenic acid____ 5/ 6 Mean drop in intake due to 0.59 pantothenic acid_______________________________ (0.13-1.64) (0.10-0.92) 0.82 (0.60-1.29) 9/ 9 0.50 (0.25-1.09) (0.01-0.92) 0.91 (0.60-1.66) 14/ 15 0.53 (0.13-1.64) (0.00-0.48) 0.92 (0.30-1.42) 7/10 0.71 (0.40-0.91) (0.12-0.60) 0.51 (0.20-1.63) 5/10 0.24 (0.12-0.36) (0.00-0.60) 0.74 (0.20-1.63) 12/20 0.51 (0.12-0.91) (0.00-0.92) 0.81 (0.20-1.66) 26/ 35 0.52 (0.12-1.64) 1-3 ~ (1) c:l Rats responding to 10 B vitamins_______ 4/ 4 0.45 Mean drop in intake due to 10 B____ ____ (0.35-0.55)Rats responding to 12 vitamins____ _______ 1/1 0.13 --··­---­----··-­--------­-------­ 4/ 4 0.45 (0.35-0.55) 1/1 0.13 10/11 0.53 (0.23-1.03) 7/8 0.30 -------­-------­-----· -­-------­-------­ 10/11 0.53 (0.23-1.03) 7/8 0.30 14/15 0.51 (0.23-1.03) 8/9 0.28 ~ ~· (1) ~ CZ>........ ~ Mean drop in intake due to 12 V___________ Rats responding to vitamin B,,____________ Mean drop in intake due to B,,___________ ___ Rats responding to protein_ ___ __ __ ___ ____ ____ --­----­-------­-------­-­-----­-------­ ----­----------­---­--­- -------­-------­---·---­--­----­------­- (0.15-0.58) 8/8 0.29 (0.10-0.85) ------­- -------­-------­-------­-------­-------­ (0.15-0.58) 8/8 0.29 (0.10-0.85) ------­ (0.13-0.58) 8/8 0.29 (0.10-0.85) -------­ 0 '-+. 1-'.3 (1) H ------­- -------­ -------­ ---­---­ --­---­- ------­ ~ Mean drop in intake due to protein___ __ Rats responding to protein-nucleicacid-amino acids______________________ ___ __ 2/ 2 Mean drop in intake due to 0.39 protein-nucleic acid-amino acids __ (0.32-0.45) Rats responding to continuous 15 vitamins______________ ______________ ___ __ _ _ 2/2 Mean drop in intake due to 0.22 continuous 15 vitamins_______ ______________ (0.14-0.30) 334 ' -------­---­-­·-·­--·--·· -­-------­181 2/ 2 0.39 (0.32-0.45) 2/2 0.22 (0.14-0.30) 224 7/ 9 0.37 (0.16-0.84) 6/7 0.26 320 -------­-------­-------­--­----­-------­-------­88 7/9 0.37 (0.16-0.84) 6/70.26 -------­215 9/11 0.37 (0.16-0.84) 8/9 0.27 (0.14-0.30) 218 Average total days on test diet___________ (135-410) 0.08 (85-300) 0.25 (85-410) 0.19 (170-405) 0.30 (85-100) 0.40 (85-405) 0.35 (85-410) 0.29 Average final level of alcohol intake____ (0.02-0.14) (0.08-0.51) (0.02--0.51) (0.07-0.52) (0.08-1.46) (0.07-1.46) (0.02-1.46) 10 20 30 40 50 60 70 80 90 . 100 110 120 130 140 DAYS ON DE.flCIEN.T DIET FIGURE III Effect of Pantothenic Acid Deficiency and Supplementation on Alcohlol Consumption by a Rat. Effect of Vitamin B6 Deficiency on Alcohol Consumption.-The data in Table VI leave little doubt that vitamin B0 is also involved in a definite manner in the etiology of alcoholism in rats. Twenty out of a total of thirty-three animals responded to give a mean drop of 0.56 ml. and an "index of response" of 33.6. Significant strain differences were not apparent. Figure IV illustrates a typical re­sponse to vitamin B0 administration. 1.6 1.5 1.4 1.3 ~ ~ 1.2 ~ a: 1.1 ci 0 0 1.0 ~ ...J ~0.9 z 0 i= 0.8 Q. ~0.7 en z 30.6 ...J 00.5 :c 0 ~0.4 < 0.3 0.2 0.1 ACID l1G.l TABLE VI Summary of Experimental Data on Alcohol Consumption of Rats Fed Pyridoxine Deficient Diet* (Alcohol consumption in ml. alcohol/100 gm. rat/day) Strain Strain Total Strain Variable "O" "O" + 9V Strain "0" "H" No. of animals in original group __ ______ __ No. of animals completing expt._____________ Initial consumption level.____________ _________ __ _ 9 4 0.35 10 9 0.21 19 13 0.28 15 7 0.55 Level at initiation of treatment -----··---·-· (0.15-0.70) 0.86 (0.06--0.60) 0.69 (0.06-0.70) 0.77 (0.07-2.30) 1.09 Rats responding to pyridoxine_____________ ___ (0.56--1.63) 6/ 9 (0.45-0.98) 8/ 10 (0.45-1.63) 14/ 19 (0.53-2.43) 6/ 14 Mean drop in intake due to pyridoxine administration________ ___ __ __ 0.58 0.43 0.50 0.69 (0.38-0.91)Rats responding to 10 B vitamins _ __ __ 4/5 Mean drop in intake due to 10 B ------···­0.51 (0.24-0.81)Rats responding to 12 vitamins ___ _________ 1/4Mean drop in intake due to 12 V__ _ ____ _ 0.27 (0.08-0.88) (0.08-0.91) (4/4)t 4/5 (0.45)t 0.51 (0.12-0.74)t (0.24-0.81) -------­1/ 4 --­·-­-­0.27 (0.35-1.56) 8/ 14 0.67 (0.41-1.20) 6/ 13 0.28 Rats responding to vitamin B,, __ __ _ _ Mean drop in intake due to B,,_____ ____ _·-·­ 1/1 0.10 -----·· -­---·--··­ ..... 1/1 0.10 (0.12-0.43) 8/9 0.39 Rats responding to protein______ ___ _ __ __ __ __ Mean drop in intake due to protein_______ ----­--­-----­---------­ -----··-­-------­-------­ -------­-------­---­---­ (0.18-0.90) 1/2 0.20 -------­ ----­--­ -------­ Rats responding to nucleic acid_______________ Mean drop in intake due to nucleic acid_ _____ _____________________ ___ ____ _ __ -------­---­---­ ----- --­---­---­ 3/4 0.30 Rats responding to amino acids____________ __ ------­--------­ -------­___ __.. __ ------··­---­---­ (0.18-0.47) 0/1 Total All Rats 34 20 0.40 (0.06-2.30) 0.91 (0.45-2.43) 20/33 0.56 (0.08-1.56) 12/ 19 0.62 (0.24-1.20) 7/ 17 0.28 (0.12-0.43) 9/ 10 0.36 (0.10-0.90) 1/2 0.20 3/4 0.30 (0.18-0.47) 0/1 ~ Cl:> 0 "-3 ~ ~ c::: ~ ~ ""· ~ ~ '""· <::: -""" 0 "-3 ~ H ~ Cl,) Mean d~op in .intake due to amino acids___ __ ___ _______ _ ________________ _ _ Rats r~spon1ing t<_> protein-nucleic ac1d-am1no acids________________ _______________ Mean drop in intake due to protein-nucleic acid-amino acids______ Rats responding to continuous 15 vitamins___ ______________________________________ Mean drop in intake due to continuous 15 vitamins_______ ________ ____ Average total days on test diet______________ Average final level of alcohol intake_______ •see footnote p. 118. t'I'hiamine only. 2/4 0.28 (0.21-0.36) 4/4 0.40 (0.15-0.77) 406 (400-412) .10 (0.05-0.15) 6/9 0.43 (0.13-0.87) 290 (286-294) 0.30 (0.05-0.59) 2/4 .0.28 (0.21-0.36) 10/13 0.42 (0.13-0.87) 326 (286-412) 0.24 (0.05-0.59) 0/11 3/6 0.23 (0.10-0.40) 409 (392-431) 0.36 (0.12-0.67) 2/15 0.28 (0.21-0.36) 13/ 19 0.37 (0.10-0.87) 355 (286-431) 0.28 (0.05-0.67) ~ ..... (') 0 ~ 0 .... CJ 0 <:2 ~ ~ .,... ""· 0 <:2 [ -0:::: "".,...· ~ ~· t1 ~ ~ (') ""· ~ <:2 ~­ Cf> ~ ~ ~ In many of the deficiencies studied most of the animals tested responded by a diminished alcohol intake to an increase in the protein or amino acid content of the diet. Pyridoxine deficient rats in general did not respond to high protein diets; in fact there was often a distinct rise in alcohol consumption. This finding is in accord with the known fact that pyridoxine is required for the catabolic use of amino acids for energy purposes. A summary of the results obtained from the five vitamins dis­cussed is presented in Table VII in order that the responses of the various groups studied may be more readily compared. 1.2 >­ c( 0 l-' e( a: 1.0 cj 0 0 ... _j0.8 ' ~ z Q 1­~0.6 ::::> U) z 0 0 ...J0.4 0 I: 0 0 ...J <0.2 10 20 30 40 50 60 DAYS ON DEF"ICIENT DIET FIGURE IV Effect on Vitamin B, on Alcohol Consumption by a Rat. TABLE VII Summary of Effect of Treatment of Single Vitamin Deficiency With Various Vitamins ~ ..... on Alcohol Consumption of Rats <:"':> 0 (Alcohol consumption in ml. alcohol/100 gm. rat/day) 0 ~ ..... Pantothenic (") Strain Variable Vitamin A Thiamine Riboflavin Acid Vitamin B. 0 ~ 0 Rats responding 1/5 6/9 3/4 5/6 6/9 Mean drop in intake 0.37 0.43 0.86 0.59 0.58 ~ ~ Range --------(0.15-0.79) (0.45-1.57) (0.13-1.64) (0.38-0.91) '13 <""+­ 0+9 Rats responding --------8/9 7/8 9/9 8/10 .... 0 Mean drop in intake 0.63 1.36 0.50 0.43 ~ Range ( 0.32-1.05) (0.64-2.51) (0.25-1.09) (0.08-0.88) Total 0 Rats responding 115 14/18 10/12 14/ 15 14/19 §Mean drop in intake 0.37 0.55 1.21 0.53 0.50 ~ Range ( 0.15-1.05) ( 0.45-2.51) (0.13-1.64) (0.08-0.91) -.:::: H Rats responding 2/3 5/8 8/11 7/10 6/14 .... Mean drop in intake 0.29 0.66 0.51 0.71 0.69 Range (0.29-0.43) (0.14-1.23) (0.17-0.90) (0.40-0.91) ( 0.35-1.56) @ H+9 Rats responding ------------------------5/10 -------­ ~­ Mean drop in intake 0.24 -------­ Range ----------------(0.12-0.36) t::i Total H Rats responding 2/3 5/8 8/11 12/20 6/14 ~ Mean drop in intake 0.29 0.66 0.51 0.51 0.69 .... <:"':> Range (0.29-0.43) (0.14-1.23) (0.17-0.90) (0.12-0.91) (0.35-1.56) (1:> ;:S Total Rats responding 3/8 19/26 18/23 26/35 20/33 <:"':> .... All Rats Mean drop in intake 0.32 0.58 0.90 0.52 0.56 Cl> Range (0.29-0.43) (0.14-1.23) (0.17-2.51) (0.12-1.64) (0.08-1.56) ~ Index of Response 11.3 42.3 70.2 38.5 33.6 ...... i:.o i:.o t 1.7 1.6 1.5 1.4 ,,,.,,,--­ 1.3 ,,_,, t .-"' I I I I I 1.2 z I 1.1 0 I I ->­ I-<10 n.o. :::i;, ,, : I ..., I ' I ,' ' ·-__....., ,.,.-J Za: ~~ -9 i""..,. "' o· I u~.8 I t I ..Jo I I Q~.7 J: ~ I 0...J I u:::i; .6 ,, I I'~ ...J < .5 I -------',,/ I .4 ,,,--­ f / .3 I .2 I I I I t'f'ft'f I I .11 I --•... I '\_..........-..._____, 220 210 ....... 200 o:> ii:-. 190 180 170 160 150j ~ C> (I) 140Z c::! ...: 130~ ~. <:;) (I) ~ 120 ~ 110 c ...... ~ 100 (I) ~ 90 80 70 60 20 40 60 80 100 120 140 160 180 200 220 DAYS ON DIET FIGURE V. Effects of Multiple Vitamin Supplementation on Growth and Alcohol Consumption of a Rat. Solid line represents alcohol consumption; broken line represents weight of the animal. A represents the administration 0Ł desoxycorticosterone,. B the administration of cortisone. and the arrows represent the ad-ministration of the 15 vitaniin diet plus 20 unit liver. Miscellaneous Effects.-Whereas there was a tendency toward increased alcohol intake in rats on choline and biotin deficient diets, addition of these vitamins to the diet and parenteral biotin therapy had little or no effect. The data do not merit presentation in detail, and suggest that in most rats neither biotin nor choline is inti­mately associated with the alcoholic appetite. A group of fourteen rats were placed on Diet A, (3) and when their alcohol intake had arisen to an appropriate level, individual rats were treated parenterally with desoxycorticosterone, cortisone, gonadotropic hormone (Gonadophysin Searle), and 0.90 % saline. In no case was a significant response elicited as indicated by al­cohol consumption. Supplementation of the diet of these animals for two days with the ten vitamin supplement previously described brought about a prompt drop in alcohol intake. The intake later rose again, and could be cured again with the supplement. Figure 5 illustrates the successive "cure" of an animal three different times. The weight of the rat, which shows a very close correlation with the nutrition as judged by alcohol intake, is also included in this graph. There is therefore a strong probability that under proper conditions the alcohol consumption of rats can be used effectively to study vitamin deficiencies and in developing vitamin assay methods. IV. Discussion A large amount of experimental data has been presented to show that riboflavin, thiamine, pantothenic acid, and pyridoxine de­ficiencies are effective in increasing the appetite of rats for 10% alcohol, and that the addition of these vitamins to the diet of the ap­propriately deficient rats causes a remission of the high levels of alcohol consumption. The data bear out strongly the concept of genetotrophic disease previously presented (6). The evidence for the development of different secondary deficiencies in various in­dividual rats emphasizes the degree to which individual vitamin requirements may vary. Although it has been inexpedient to in­corporate the evidence into this paper, it would be safe to say that no two animals in this work have reacted uniformly throughout the course of the treatment. There seems little doubt that, if the requirements for each of the vitamins by a hypothetical average rat were set arbitrarily along a base line, the real requirements of individual animals would fall well above and below this line. Those animals whose requirements fall appreciably above the line suffer mild deficiencies on so-called normal diet and exhibit these defici­encies by increased alcohol consumption. The vitamins which have been shown to be most effective in the abolishment of alcohol appetite in rats are those which function in the early stages of carbohydrate metabolism (riboflavin, thiamine and pantothenic acid) and in the utilization of amino acids as an energy source ( pyridoxine) . It seems pertinent to the subject to point out that most strong appetites about which we have some physiological understanding arise as the result of a deficiency of the element for which the ap­petite exists. It seems possible that the appetite for alcohol may arise as the result of its sparing action on some materials which are lacking in the body in adequate amounts. Further study of the exact modes of the metabolism of alcohol may show that as an energy source, alcohol spares the metabolic mechanisms needing the above vitamins in a manner analogous to that in which it is known to spare thiamine containing systems. It should be pointed out, however, that in early experiments in which animals were on diets deficient in B vitamins generally (1) the administration of the four vitamins above did not "cure" the rats whereas a more complete supplement of B vitamins did. It seems that to be free from alcohol appetite, rats must have their entire metabolic machinery functioning satisfactorily but that cer­tain deficiencies are more effective than others in promoting the appetite. Regarding the development of the deficiency, it is of some in­terest. to note that the four vitamins under discussion all bear hydroxyl groups which are phosphorylated in the functional form of the vitamin. Dr. Thomas Bardos of the Biochemical Institute has pointed out that Verzar and coworkers have previously shown that the phosphorylation of riboflavin is under adrenocortical con­trol, and that a similar situation may well hold for the conversion of the other vitamins in this group to the active co,enzymes. Such a consideration may help explain the data obtained by other work­ers indicating an interrelationship of the adrenal cortex with alcoholism. Summary 1. Evidence is presented to show that deficiencies of thiamine, riboflavin, pantothenic acid, or pyridoxine are relatively effec­tive in increasing the ad libitum consumption of 10% alcohol by 2. 3. 1. 2. 3. 4. 5. Alcohol Consumption and Vitamin Deficiencies 137 rats, but that other vitamins may also function to a lesser de­ gree. Data are presented to show that on many standard deficiency diets, some rats develop secondary and tertiary deficiencies due to individual differences in their requirements ·for specific vitamins. The findings are interpreted in terms of the genetotrophic concept. 10 v Code for Vitamin Supplements Per Day Thiamine ·--------·-· ----·------·-------------0.6 mg.Pyridoxine --------------------------------------0.3 mg.Calcium Pantothenate ________________ 6.0 mg. Riboflavin --------------------------------------1.0 mg. Choline -------------------------------------------20.0 mg. Inositol --------------------------------------------6.0 mg. P ABA ----------------------------------------------6.0 mg. Nicotinic Acid ·-----------------------------2.0 mg.Biotin ------------------·---------------------------0.02 mg. Folic Acid --------------------------------------0.2 mg. 12 v Same as 10 V with the addition of Per Day Vitamin A -------·---·--------------------------800 units Vitamin D --------------------------------------170 units 15 v Same as 12 V, with the addition of Per Day Vitamin C ----------------------------------------12.0 mg. Vitamin K --------------·------------------------0.4 mg. Vitamin E ----------------------------------------1.0 mg. BIBLIOGRAPHY Williams, R. J., Berry, L. J ., and Beerstecher, E., Jr., Arch. Biochem., 23, 275 (1949) . Williams, R. J., Berry, L. J., and Beerstecher, E., Jr., Proc. Natl. Acad. Sci. U.S., 35, 265 (1949). Williams, R. J., Berry, L. J., and Beerstecher, E., Jr., Texas Repts. Biol. Med., 8, 238 (1950). Williams, R. J., Beerstecher, E., Jr., and Berry, L. J., Lancet, 258, 287 (1950). Williams, R. J ., Nutrition and Alcoholism, University of Oklahoma Press. (In Press.) 6. Beerstecher, E., Jr., Sutton, H. E., Berry, H. K., Brown, W. D., Reed, J., Rich, G. B., Berry, L. J., and Williams, R. J., Arch. Biochem., 29, 27 (1950). 7. The Pharmacopoeia of the United States of America, XIII, Mack Print­ing Co., Easton, Pa., 1947, p. 721 8. The Pharmacopoeia of the United States of America, XIII, Mack Print­ing Co., Easton, Pa., 1947, p. 720. 9. Coward, K. H., The Biological Standardization of the Vitamins, 2nd. ed., The Williams .and Wilkins Company, Baltimore, 1947, p. 70. 10. Street, H. R., J. Nutrition, 22, 399 (1941); also, Wagner, J. R., Axelrod, A. E., Lipton, M. A., and Elvehjem, C. A., J. Biol. Chem., 136, 357 (1940). 11. Gyorgy, P., and Poli'ng, C. E., Science, 92, 202 (1940). 12. Gyorgy, P., and Eckhardt, R. E., Biochem. J., 34, 1143 (1940) ; also, Conger, T. W., and E'lvehjem, C. A., J. Biol. Chem., 138, 555 (1941). 13. Engel, R. W., and Salmon, W. D., J. Nutrition, 22, 109 (1941). 14. Nielsen, E., and Elvehjem, C. A., J. Biol. Chem., 144, 405 (1942). XIV. Individual Excretion Patterns in Laboratory Rats by JANET G. REED In order to test the validity of the tentative conclusion that in­dividual patterns in humans exist independent of diets and en­vironmental conditions, we have explored the possible existence of such patterns in laboratory rats when these variables are con­trolled. Preliminary tests were carried out with about 20 rats in order to select those which seemed to show distinct differences. It appeared from inspection of the results obtained from this larger group that each of the individual rats had a somewhat consistent excretion pattern and that some were more distinctive than others when considered from the standpoint of a limited number of measurements made. The rats used in this study were selected for further experimentation because their excretion patterns appeared to be relatively distinctive. · Observed differences in the alcoholic consumption of individual rats as well as their responses to vitamin supplements (1) have al­ready indicated differences in metabolic patterns. A further study by this approach was underway when this investigation was started and is reported elsewhere (p. -). The present study is an out­growth and extension of these earlier studies. Experimental Rats numbers 3, 5, 12 and 13 are from a highly inbred strain (brother-sister mated for 101 generations) of Wistar rats, from the Nutrition Laboratories at Iowa State College, Ames, Iowa. Rat No. 16 (47th generation of brother-sister mating) is from the Wistar W-1-B strain obtained from the Glaxo Laboratories in Middlesex, England. Rats "H" and "J" are from other colonies not highly inbred, but originally from the Wistar strain. The rats were placed in individual false-bottom cages and given water and food (Purina dog chow) ad lib. Each day for a period of 7 to 11 days the rats were placed in individual metabolism cages for a few hours until an adequate volume of urine was collected. Dur­ing the collection period food and water were withheld to prevent contamination of the urine sample. Immediately after collection the urine samples were frozen until analyzed. The urine samples were analyzed for creatinine by the standard picrate colorimetric method (2), for phosphorus by the method of Fiske and Subbarow (3), and for amino acids by two-dimensional paper chromatography using the method of Berry and Cain (4). To facilitate expressing the amino acid values on a comparable basis aliquots of urine containing 50 p.g. of creatinine were used as the test volume for chromatographing. The results of these determinations are summarized in Table I which shows the average value for each rat and the standard de­viations. Creatinine is expressed as mg./ml. and the values for the amino acids and phosphate are expressed as mg./ mg. of crea­tinine in order to eliminate the effects of variability in the concen­tration of the urine samples. Each average is derived from at least seven samples, more often from 9 to 11. On the amino acid chromatograms there appeared several ninhy­drin positive spots the identities of which have not been estab- TABLE I Urinary Patterns in Laboratory Rats No.3 No. 5 No.12 No.13 No.16 "H" "J" Creatinine Aver. (mg.I ml.) __________ 2.04 1.76 1.87 1.62 2.25 .96 1.66 ­ ;30.8 .... ' < a: ._;0.8 0 S! ' 50 z < :z: ::;0.2 .J :i; ' 6 14 16 9 16 23 24 • 7 RAT NUMBER 1.0 ~ ~0.8 0-•.14 ~ FISCHER 344 a: cio.6 0 0 :::: cjo z < :z: t;o.2 ...i ::i; 1 "' 7 6 9 11 12 16 20 23 RAT NUMBER FIGURE I. Average Daily Alcohol Consumption by Individual Rats of Six Different Strains. The unshaded bars represent average daily alcohol consumption by all rats within each strain. IOWA 25 -32 (j • •15 .> 7 • 9 3 10 11 12 13 RAT NUMBER 1.0 >­ < ~08 .... < a: go6 0 .J 004 ' z < :z: .... "'0.2 .J ::i; 1.0 IOWA 65 -76 (j • .52 RAT NUMBER WI STAR W -2 -A 0-• .10 I 2 7 • 10 11 12 13 14 • RAT NUMBER PIEBALD 0-• .04 26 27 6 7 6 9 10 H 12 13 RAT NUMBER Inspection of the individual averages of alcohol consumption for the 30 day period (Figure 1) shows a wide variation within a strain. The Piebald strain, which is the most homogeneous in re­gard to alcohol consumption (o-=.04) is found to be significantly different (using the customary 5% level) from the Iowa 65-76 substrain and the Wistar W-2-A substrain, but is not significantly different from the others. In contrast Iowa 65-76 which is the most heterogeneous with respect to alcohol consumption (o-=.52) is sig­nificantly different from every other strain with the exception of Wistar W-1-B. It is interesting to note that the most homogeneous group (alcohol consumption) has been brother-sister mated for only 9 generations, while the most heterogeneous group on the same basis (Iowa 65-76) has been brother-sister mated for 101 gen­erations. Although there is a great deal of individual variation in urinary excretory products within each strain, some significant differences between the strains can be shown. The phosphate excretion (Figure 2) of the two Iowa substrains is not significantly different, nor is · the phosphate excretion of the two Wistar substrains significantly different; however, the differences between the Iowa substrains and the Wistar substrains are highly significant. The Iowa substrains are also different from the Fischer and Piebald strains, although these strains are not significantly different when compared with each other or with the Wistar substrains. It should be noted that although the rats are on identical diets there is a maximum of a six fold variation between urinary phos­phate excretion of one strain compared with another. Between individual rats on identical diets there is nearly a 500 fold varia­tion! It seems inescapable that the fecal phosphate excretion must be correspondingly high when the urinary phosphate is low but this has not been investigated. FIGURE II. Excretion Patterns of Rats of Different Strains. Verticle bars represent the ratios of average exc.retions of urinary con­stituents (expressed as mg. constituent per mg. creatinine) by the rats in each strain divided by the average excretion for the rats in all strains. Broken line indicates average excretion for all strains. Cr = creatinine; P = phos­phorus; As= aspartic acid; Gl = glutamic acid; T = taurine; Al= alanine; Ci "7 citrulline; MS = methionine sulfoxide*; V = valine; Ly= lysine; Le = leucme. Ale = average alcohol consumption by all rats in each strain expressed as ml. ethanol/100 g. rat/ day. *See footnote p. 75. Excretion Patterns of Rats of Different Strains 147 2.0 2.0 IOWA 25-32 1.5 1.5 1.0 1.0 0.5 0.5 CR p As C.L T AL C1 MS v Lv LEALC CR P As GL T At C 1 MS V Lv LE ALC 2. 2 .0 PIE.BALO 1.5 1.5 1.0 1. 1.0 0.5 05 CR P AsGL T ALCt MS V LE LvALc Differences in lysine excretion are apparent when Wistar W-1-B is compared with each of the other groups with the exception of its sister substrain W-2-A. In the excretion of lysine the two Iowa substrains are significantly different from each other. Large variations in the amount of taurine excreted by the various strains are also apparent from Figure 2. The Fischer 344 strain is outstanding for the large volume of taurine excreted, while the Iowa strains excrete very small amounts. Both the Fischer 344 and Piebald strains are significantly different from the other strains in their taurine excretion. The Wistar substrain W-2-A is significantly different from the other strains with the exception of its sister substrain W-1-B in the excretion of methionine sulfoxide. * For the excretion of prac­tically each amino acid some statistically significant difference can be found to exist between at least two of the strains. However, the greatest strain differences appear to be concerned with the products previously discussed, _namely: phosphorus, lysine, taurine, and methionine sulfoxide. * Several spots whose identities have not been definitely established occasionally appeared on the amino acid chromatograms. They are designated as: 1. purple spot believed to be cystine, Rf .17; 2. purple spot before methionine sulfoxide, Rf .68; 3. Blue spot above valine, Rf .71; 4. orange spot below serine, Rf .15; and 5. yellow spot below serine Rf .12. The frequency of occurrence of these spots is shown in Table I. TABLE I Per Cent Occurrence of Unidentified Spots Strain: Spot No. 1 Spot No. 2 Spot No. 3 Spot No. 4 Spot No.5 Iowa 25-32 60 80 30 0 30 Iowa65-76 0 50 0 10 60 Wistar W-1-B 0 0 0 10 80 Wistar W-2-A 10 0 0 50 90 Fischer 344 50 10 0 0 0 Piebald 80 80 30 0 30 Any attempt to correlate the metabolic pattern of a strain or sub­strain with the alcohol consumption of that strain or substrain is made extremely difficult because of the great individual variation found in each group. Actually in the study of the metabolic pat­terns of alcoholics, no amino acid excretion figures were found to be significantly different from the controls. It was not possible to measure in rats a number of the items which were included in the •see footnote p. 76. study of the alcoholics because of the unavailability of rat saliva and of sufficiently large samples of rat urine. The findings here recorded reflect the existence of a complicated genetic situation and lead to the conclusion that many genes (and other inheritance factors?) are involved in the production of dis­tinctive urinary patterns in rats and in the determination of the appetite of the animals for alcohol. It should be theoretically pos­sible to develop strains of animals which would have any combina­tion of metabolic traits, but of course if they are not inbred with respect to the specific traits in question, there is no assurance that they will show uniformity with respect to these traits within the strain. It would appear that our initial observation in which we found closely inbred mice to show uniformity in their alcohol con­sumption was to a degree fortuitous. Among the six strains and substrains of rats used in this study only one, the Piebald, showed a high degree of uniformity in alcohol consumption. Summary A study of the alcohol consumption and the excretion patterns of 60 rats belonging to four strains and two substrains of highly in­bred rats on identical diets indicated the existence of several strik­ing strain differences and also highly individual differences in the animals within a strain. Strain differences involved different levels of alcohol consumption, and of phosphate, lysine, taurine and methionine sulfoxide excretion. Individual differences occurred with respect to all the items measured. No conclusions with respect to relationship between specific excretions and the appetite for al­cohol could be drawn. BIBLIOGRAPHY 1. Berry, H. K., and Cain, L., Arch. Biochem., 24, 178-89 (1949). XVI. A Study of the Urinary Excretion Patterns of Six Human Individuals by HELEN KIRBY BERRY, LOUISE CAIN, AND LORENE LANE ROGERS As a preliminary to further investigation, the respective urinary excretion patterns of six human subjects have been explored over a period of several weeks. A limited investigation of this nature had been made previously in this laboratory (1) through the use of microbiological assay methods. With the development of paper chromatographic techniques, it was possible to analyze a given sample for a number of components in a single determination (2), and a much more extended study of urinary excretion patterns be­came feasible. In the present investigation each urine sample was analyzed for creatinine and glucose as well as for the amino acids that could be determined by one-dimensional paper chromatography. Experimental Subjects.-The individuals used in this study included two young adult females (A and B), two young adult males ( C and D), and two adolescent "identical" male twins (E and F). All subjects were in good health and were on self selected diets throughout the period of investigation. Urine samples.-Vrine samples were collected daily from the two female subjects over a period of two months, from the two male adults over a period of one month, and from the twins over a period of two weeks. Because of the difficulty of collecting 24-hour samples during an extended period of time, only morning samples were used in this study. Of course there was the problem of the com­parison of results due to the variability in urine concentration ob­served in different individuals as well as in the same individual at different times. In order to minimize these concentration effects, all results have been expressed as milligram of constituent per milli­gram of creatinine as is discussed below. Determination of urinary constituents.-Creatinine was deter­mined by a standard colorimetric method ( 3) . A measure of total carbohydrate (which we have termed glucose) was obtained using the colorimetric procedure described by Morris (4). Amino acids were separated on one dimensional paper chromatograms and quantitatively estimated by the spot comparison method (5). Excretion Patterns of Six Individuals 151 Results and Discussion The statistically summarized results of assays are shown in Table I. It will be noted that certain of the averages and standard deviations are reported as ranges. This means of designation is used in recognition of the fact that concentrations too small to be detected by the methods employed were not necessarily zero. In cases where as many as 10% of the determinations of a given constituent for a given individual resulted in values of zero (i.e. less than the minimum detectable quantity), averages and standard deviations have been calculated on two bases: first, assuming the zero values to be actually zero, and second, assuming them to TABLE I Individual Urinary Excretion Patterns Female A Female B Male C MaleD Twin E TwinF Creatinine No. samples_____________ ___ 54 58 24 25 12 14 Avg. (mg./ml.) ___ ____ _ 0.82 1.60 1.58 1.57 1.23 0.98 Std. Dev._____ ________________ 0.47 0.78 0.52 0.42 0.53 0.45 Range_ ______ ____________ _ 0.26-3.24 0.16-3.23 0.47-2.45 0.78-2.47 0.50-2.20 0.36-2.00 Glucose No. samples___ ________ __ 46 46 24 25 8 7 Avg. (mg./ mg.Cr.) ____ 0.53 0.41 0.41 0.48 0.57 0.46 Std. Dev.______________________ 0.25 0.11 0.14 0.17 0.19 0.14 Range________ _ _ ___________ 0.21-1.32 0.26-0.61 0.20-0.87 0.28-0.99 0.35-0.92 0.34-0.68 Alanine No. samples________________ 54 58 23 25 12 14 Avg. (mg./ mg.Cr.) __ _ 0.084 0.13 0.065 0.054 0.064 0.041 Std. Dev.__________________ _ _ 0.044 0.085 0.031 0.031 0.034 0.022 Range_ __________ ____ _________ 0.01-0.13 0.03-0.16 0.01-0.10 0.03-0.21 0.02-0.50 0.00-0.16 Glycine No. samples_________________ 54 58 24 25 12 14 Avg. (mg./ mg.Cr.) ____ 0.18 0.21 0.045 .027-.030 0.052 0.041 Std. Dev.____ _ __ ____________ _ 0.094 0.14 0.022 .021-.024 0.036 0.018 Range_________ _______ _ ______ 0.00-0.48 0.04-0.68 0.00-0.16 0.00-0.08 0.00-.14 0.01-0.08 Serine No. samples_ __ __________ 54 58 23 25 12 14 Avg. (mg./ mg.Cr.) ____ .071-.074 0.14 0.045 .025-.028 0.056 0.034-0.035 Std. Dev.______ _ ______________ .051-.055 0.092 0.024 .019-.022 0.019 0.019-0.022 Range_ ______ _ __ __________ 0.00-0.28 0.00-0.50 0.00-0.10 0.00-0.08 0.02-0.08 0.00-0.08 Glutamic Acid No. sampJes _ _____________ _ 42 54 24 25 12 14 Avg. (mg./mg.Cr.) ____ .034-.035 .013-.014 .012-.019 .026-.030 0.017-0.021 0.012-0.019 Std. Dev._________________ ____ .050-.051 .027-.028 .025-.027 .018-.022 0.011-0.016 0.017-0.021 Range_ __ __ __ __ __ ___ ____ 0-.282 0-.159 0.00-0.12 0.00-0.06 0.00-0.04 0.00-0.06 Lysine No. samples ___________ _____ 54 58 24 25 12 13 Avg. (mg./mg.Cr.)__ _ .132-.134 .007-.008 .035-.042 0-.0005 .014-.015 .012-.013 Std. Dev.__ ___________ _ _______ .009-.OlQ .085-.089 .017-.018 .028-.032 0 .011-.012 Range_ __________ _____ __ ________ 0.00-0.29 0-.073 0.00-0.12 0-.0005 0-.032 0-.032 pH No. samples________________ 52 50Avg. __ __ _______ _ ____ ___ ___ 6.85 5.69 Std. Dev._______ _ ___________ 0.78 0.57 Range_ _______ ________________ _ 5.3-8.8 5.2-7.7 be equal to the minimum detectable quantity. The results of these two methods of calculation represent the range reported in Table I. Constituents other than creatinine are, as previously noted, expressed as milligrams of constituent per milligram of creatinine. This procedure was adopted as the best substitute for 24-hour samples in minimizing the effect of the rather wide variations in urine concentration. The choice of creatinine as a reference sub­stance was based in part on the fact that, in our experience, it has shown less variation between individuals than any other sub­stance studied. This is in line with general experience. Of the four adult subjects studied here, the average creatinine excretion for three of them, on a volume basis, is practically identical. Further evidence that data from morning samples, expressed as creatinine ratios, may be considered comparable to the results of analyses on 24-hour samples is presented in Table II. The average creatinine ratios for four amino acids determined by analysis of approximately 50 morning samples are compared with the average amino acid/ creatinine ratios from seven 24-hour samples obtained from the same subject during the same period. An examination shows that the two sets of data are substantially alike. TABLE II Urinary Excretion Pattern of Subject A Based on Based on Constituent Approx. 50 Morning Samples Seven 24-hour Samples Creatinine (mg./ml.) 0.82 0.95 Alanine (mg./ mg. Cr.) 0.084 0.053 Glycine (mg./mg. Cr.) Serine 0.18 0.14 (mg./mg. Cr.) Lysine (mg./mg. Cr.) 0.072 0.13 0.072 0.11 The urinary excretion patterns of the individuals studied are summarized in graphic form in Fig. 1. The broken line on each profile represents the group average excretion for each substance in question, and the vertical blocks plot on the same scale the ratio of the individual averages to that of the group average for each of the individuals studied. Even a cursory glance at Fig. 1 shows that the excretion profiles of the identical twins, E and F, are much more nearly alike than those of any two other individuals studied. According to current ideas in the field of biochemical genetics simi­lar excretion patterns would be expected from identical twins. 2 2 TWIN f TWINE Gu C GA S A Gv L Gu C GA S A Gv L Gu C GA S A G v L 2 2 MALE C MALE 0 Gu C GA S A G v L Gu C GA S A G v L Gu C GA S A G v L FIGURE I. Urinary Excretion Patterns of Six Individuals. G = glucose; C = creatine; GA = glutamic acid; S = serine: A = alanine; Gy =glycine; L =lysine. Broken line represents group average excretion (mg. constituent per mg. creatinine), and vertical bars represent ratio of individual average to group average. The differences in the excretion patterns of various individuals other than the twins are quite striking. One of the most interesting of these differences is that of lysine excretion. Subject A consis­tently excreted a relatively high concentration of lysine while subject D never excreted any detectable amount throughout the period of investigation. All other individuals studied excreted a much lower average amount of lysine than did subject A. Comparison of the amino acid excretion rates of the male sub­jects with those of the female subjects reveals a consistently higher rate of excretion by the females. This sex difference has been previously observed in total a-amino nitrogen excretion as de­termined on single 24-hour samples ( 6), and is supported by a great bulk of additional data which has been gathered in our labora­tories in connection with other studies. There was some indication of cyclic variations in the female ex­cretion patterns coincident with the menstrual cycle. The most obvious example was in the pH values which exhibited pronounced maxima during the menstrual period. Subject B exhibited minima in creatinine and glucose excretion at the onset of menstruation, with corresponding maxima in the amino acid/ creatinine ratios. The effect was of only one or two days duration and was not ob­served at all in the case of subject A. The differences in the average excretion values reported for the various individuals in Table I have been treated statistically, using the "t" test for significance (7). The results of these calcu­lations are recorded in Table III. The numbers recorded represent minimum values for the percentage probability that the difference observed is real and not due to random fluctuations. Thus, a value of 99 recorded in Table III represents a greater than 99 % prob­ability that the difference in question is significant; a value of 98 represents a greater than 98 % (but less than 99 % ) probability that the difference is significant, etc. All values below 90 are simply recorded as such. In cases where the excretion averages are re­ported in Table I as ranges, the "t" test was applied to the dif­ference in the maximum values as well as to the difference in the minimum values, and the results are reported in Table III as a range of probabilities. That there are significant differences in the individual excretion patterns is evident from Table III. The question arises, of course, as to the extent to which the use of creatinine as a reference sub­stance contributes to these observed differences; i.e., to what extent are the differences in creatinine ratios attributable to variations in creatinine excretion. A careful consideration of the data has clearly indicated that the differences are attributable primarily to the urinary constituents other than creatinine. Comparisons between individuals of the same sex do not show one subject consistently excreting more of all of the amino acids than the other subject, as would be the case if differences in creatinine excretions were pri­marily responsible for the differences in the creatinine ratios. TABLE III Sta,tistical Significance of Differences in Urinary Excretion Patterns Per Cent Probability of Significance as Measured by "t" test (7) Comparison Comparison Intercomparison Comparisonof Females of Males of Females and Males of Twins Avs.B Cvs. D Avs. C Avs.D B vs. C B vs. D E vs. F Creatinine 99 < 90 99 99 < 90 < 90 < 90(mg./ml.) Glucose 99 < 90 95 < 90 < 90 90 < 90(mg./mg.Cr.) Alanine 99 < 90 90 99 99 99 90 (mg./ mg.Cr.) Glycine < 90 95-98 99 99 99 99 < 90(mg./mg.Cr.) Serine 99 98-99 95-98 99 99 99 98--99 (mg./mg.Cr.) Glutamic Acid 99 < 90-> 90 < 90-> 90 < 90 < 90 95-98 < 90 (mg./ mg.Cr.) Lysine 99 99 99 99 99 90-95 < 90 (mg./mg.Cr.) pH 99 It is of course impossible to present in a short space a complete picture of the extensive data collected, and average excretion values covering an extended period do not reveal all that is significant. For example the average alanine excretions (per mg. of creatinine) of the two adult females were not very far apart, but the day to day graphs showed a marked dissimilarity. For one individual the the excretion was relatively uniform from day to day; for the other there were tremendous fluctuations which appeared con­sistently for weeks. Differences such as these (and the case cited was by no means unique) appear to be distinctive and a part of the "pattern" of the individual concerned, but do not appear in the tabulations. A separate study of these fluctuations is being made. It may also be noted that during the course of this study new solvents and developing agents were tried and chromatographic spots corresponding to unknown substances were discovered. Many additional observations indicative of individuality of excretion pat­terns were made which are not recorded specifically here. The extent to which differences in diet may have influenced the results of the present investigation cannot be assessed with cer­tainty. It seems likely, however, that in experiments of such long duration many dietary effects would be eliminated. It would ap­pear that the pronounced and consistent sex differences observed cannot be attributed to differences in diet. Other studies bearing on the effects of diet are reported elsewhere in this Bulletin. Summary Creatinine, glucose, alanine, glycine, serine, glutamic acid, lysine, and pH have been determined on consecutive morning urine samples from six subjects over a period of from two weeks to two months. The subjects all possess excretion patterns which are sig­nificantly distinct from each other. Amino acid excretion is consis­tently higher in the female subjects than in the male. The urinary excretion patterns of a pair of identical twins show greater simi­larities than do the patterns of any of the other subjects investi­gated. BIBLIOGRAPHY 1. Thompson, R. C., and Kirby, H. M., Arch. Biochem., 21, 210 (1949). 2. This Bulletin, p. 22. 3. "Klett-Summerson Clinical Manual," Klett Mfg. Co., New York. 4. Morris, D. L., Science, 107, 254 (1948). 5. Berry, H.K., and Cain, L., Arch. Biochem., 24, 179 (1949). 6. Thompson, 1R. C., and Abdulnabi, M., J. Biol. Chem., 185, 625 (1950). 7. Snedecor, G. W., "Statistical Methods", Ames, Fourth Ed., 1946, p. 82. XVII. Further Studies on Individual Urinary and Salivary Amino Acid Patterns by HELEN KIRBY BERRY Irt a series of exploratory studies such as are described in this bulletin, it has not been feasible to do detailed advance planning because, since the work was pioneering in nature, it was impos­sible to tell in advance of the actual work what particular lines of study would develop the greatest interest. In the present article there are set forth results from various studies which have a bearing on individuality in urinary excretion patterns and in saliva composition. Two dimensional urine chromatograms.-In Figures 1-10 are presented diagrams depicting comparatively the urinary excretion patterns of 10 young healthy individuals, 2 boys, 2 girls, 5 adult males, 1 adult female, picked from -a large group to show differ­ences and similarities. These individuals were not under dietary control. The figures are drawn to scale representing the size of the amino acid spots obtained from samples of urine containing 40 micrograms of creatinine (20-75 microliters). Each figure OMETHIONINE SULFOXIDE CITRULLINE 0 0-ALANINE GLUTAMINE p-ALANINE­ OGLYCINE 0-TAURINE ~ .J 0 z ., :i: ... l 0 -2,6-LUTIOINE (2) 0 OMETHYL HISTIDINE LEUCINE-0 OYALINE METHIONINE SULFOXIOE · -\) Q-a-AMINO BUTYRIC HISTIDINE CITRULLINE 00-ALANINE GLUTAMINE jl-ALANINE- GLYCINE-a p-TAURINE :::: QSERIN J Q_CLU MIC ACID 0 z "'J: .. l. o-2 ,6-LUTIDINE (2) FIGURE 2 FIGURE 1 represents the chromatogram from a single particular urine sample from a given individual. In each case, however, several chromato­grams of samples taken from each subject at different times were available for comparison, and the patterns as pictured are charac­teristic respectively of the 10 individuals. CITRULLINE-OG GLUTAMINE ALANINE p-ALANINE 0- METHIONINE SULFOlCIDE CITRULLINE 0­ GLUTAMINE_J) ALANINE j!l-ALANINE 0GLYCINE 0 ..J: .. .J z ~GLUTAMIC ACID 0 -?,6 -LUTIOINE (2) GLYCINE-a ..z .J 0 ..J: o-2,0-LUTIOINE (2) F IGURE 3 F IGURE 4 0-METHYL HISTIDINE UNKNOWN AMINO ~t~~~J~t'ts5 IN CITRULLINE-00 OFROM THIS GLUTAMINE INDIVIDUAL) j!l-ALANINE ALANINE QCLYCINO-E TAURINE ~GLUTAMICO ~ ACID .J 0-sERINE 0 z "'J: .. t 0 -2,e -LUTIDINE (2) METHIONINEO SULFOXIOE HISTIDINED n,_-AMINOBUTYRIG vv·ACID :ITRULLI NE GLUTAMINE 0 0 TYROSINEO j!l-ALANINE ALANINE OOYSINE ()-THREONINE GLYCINE 0-TAURINE .J 0 z "'J: .. f o--2,&-LUTIDINE (2) FIGURE 5 FIGURE 6 The chromatograms were prepared using phenol as the fir::it solvent and 2,6-lutidine as the second solvent. Color development was based upon the use of 0.2% ninhydrin in water-saturated butanol. This reagent was sprayed on the sheets and they were then heated seven minutes at 90°C. 0METHYL HISTIDINE QMETHIONINE SULFOXIDE OiUNKNOWN AMINO ACID (APPEARS IN ALL SAMPLES FROM THIS INDIVIDUAL) &HISTIDINE %1~uR~k~I~~ UNKNOWN MAINO DALA NINE !'-ALANINE Q 0 ACID (SEE FIG. 5) ALANINE 0-LYSINE 0-THREONINE a.LYCINE 0­OGLYCINEO-TAURINE ~ TAURINE .::. SERINE .J 0 .::. z (}GLUTAMIC ACID .., ..J 0 :z: z CL .., .. :J: (}ASPARTIC AC ID f o-2,8 -LUTIDINE (2) t o-2,6-LUTIDINE (2) FIGURE 7 F IGURE 8 OMETHIONINE SULFOXIDE (\.GLYCINE U Q-TAURINE o.GLUTAMIC ACID .J z .., .. :J: t o-2,6 -LUTIDINE (2) .J 0 z .., ..:z: o- ALANINE-a 0-LYSINE GLYCINED 2,&-LUT IDINE (2) {):UNKNOWN AMINO ACID (SEE FIG. 5 AND 7) 0-TAURINE FIGURE 9 FIGURE 10 It may be noted that on Figures 5, 7 and 10 a spot appears which corresponds to an unidentified amino acid. This substance appears consistently in the urine of the three individuals indicated and also in samples from other individuals not included in this particu­lar study. This unknown substance has an Rf value of .55 in phenol and .40 in 2,6-lutidine. It will also be noted that one spot is designated as "citrulline­glutamine-,8-alanine." All three of these substances have been detected in urine samples. They are not separated using these two solvent mixtures and hence are not differentiated. Urinary amino acid data.-In Table I are presented data on the urinary amino acid excretions of 52 individuals derived from chromatographing a total of 178 urine samples-a minimum of 3 samples from each individual. Included in the group are 35 males and 17 females; 12 mental patients and 40 well individuals; 7 chil­dren and 45 adults. TABLE I Urinary Amino Acid Excretion Patterns (amino acid concentrations expressed as µg./mg. creatinine) Subject lM* 2 3M 4M 5 6 7 8 Sex M F M M F F F F No.I\ Amino Acid No. Samples 3 3 3 3 3 3 3 3 glycine__________________________ _________________________ 4 +:i: 37 + + 32 43 42 54 6 4 20 25 26 21 alanine__ _________________________________________________ + + + citrulline_____ __________________________________________ 7 8 20 16 18 22 + + + serine______________________________________________________ 3 19 26 24 25 19 + + + taurine_______ __________________________________________ 5 15 18 24 18 19 + + + 10 methionine sulfoxide**---------------------20 50 10 10 50 2 10 valine_____ _______________________________________________ 8 10 10 10 8 5 5 10 2 glutamic acid__ ____________________________________ tyrosine_________________________________________________ + + + 13 15 10 20 15 30 5 5 5 leucine_________________________________________ _________ 9 10 5 15 10 10 5 5 5 lysine________________________ ____________________________ 11 50 5 5 10 5 cystine____________________________________________________ 20 20 15 20 16 unidentified_____________________ _____________________ 10 5 10 20 17 methyl histidine________ _______________________ 5 15 5 -§ 5 10 10 5 threonine_ ____________________________________________ 12 5 15 15 5 5 5 prorine___________________________________________________ 18 19 a-amino butyric acid __ ________ 28 methionine sulfone___________10 10 1 aspartic acid________15 *M denotes mental patient. tC denotes child. . ;This designation (+>.is used to indicate that the amino acid was found but that quantitative informati< is not recorded here. With regard to the mental patients and the corresponding controls quantitative da are reported in another portion of this bulletin. , §This designation (-) indicates that the amino acid was not found by the procedure used. llFor the significance of this see Figure 1, p. 75. ••see footnote p. 75. 161 TABLE !-(Continued) Urinary Amino Acid Excretion Patterns SpotNo. 4 6 7 3 5 21 8 2 13 9 11 20 16 17 12 18 19 1 9M lOM M M 3 3 + + + + + + + + + +50 20 4 20 +5 5 2 8 5 6 10 20 15 llM M 3 + + + + +5 20 +15 30 185 10 12Ct 13 M M 3 4 54 +20 +44 +29 +26 +80 15 5 7 10 10 15 5 20 10 10 5 5 14 F 3 49 14 17 17 21 10 5 +5 5 10 15M 16M 17M M M M 3 6 3 + + ++ + ++ + ++ + + + + +20 5 10 4 16 5 +15 10 10 8 7 5 50 70 5 10 5 18 M 3 + + + + +10 +10 8 10 5 19 M 4 + + + + +15 5 +10 5 5 20 M 4 + + + + + 10 10 +15 10 10 21C F 3 410 110 23 58 73 40 50 69 15 20 15 22 F 3 75 29 29 38 48 10 +5 5 10 23 F 3 95 29 49 45 28 5 5 5 50 10 5 TABLE !-(Continued) Urinary Amino Acid Excretion Patterns Spot No. 24 M 7 25 M 4 26 F 3 27 F 3 28M M 3 29M M 3 30 M 4 31 M 4 32C M 4 33M M 3 34 M 6 35 M 4 36 F 3 37 M 3 38 M 3 4 6 7 3 5 21 8 2 13 9 11 20 16 17 12 18 19 1 + + + + +5 5 8 10 15 + + + + +15 5 + 5 30 42 12 15 21 7 5 5 5 5 91 26 29 32 33 5 5 5 5 + + + + +30 25 + 10 + + + + +10 12 + 20 + + + + + 10 + 3 10 + + + + 5 +15 10 10 110 50 130 70 30 130 20 50 30 + + + + + + 20 10 + + + + 2 2 4 3 + + + 5 + 8 20 5 31 10 10 20 26 5 15 + + + + + 2 10 + + + + + +50 TABLE 1-(Continued) Urinary Amino Acid Excretion Patterns % of 81 jects E 39C 40 41 42 43 44 45C 46 47 48 49C 50 51C 52 creti Spot F F F M M M F M M M F M M M Ami No. 3 3 3 3 3 3 3 5 4 4 3 4 3 3 Aci 4 130 56 47 500 260 56 100 + + + + + + + + 6 77 19 8 96 16 98 + + + + + + + + 7 99 19 9 88 + + + + + + + 3 47 23 20 81 + + + + 5 15 22 81 + + + + + 21 60 8 5 60 8 20 5 69 8 15 15 15 62 2 68 120 7 56 + + + 13 5 50 9 46 11 10 30 40 20 5 29 16 10 10 10 29 17 29 12 27 18 30 6 19 1 A major part of the information was derived from anaylses of a preliminary nature performed before the methods were refined to the extent that they were for succeeding studies. The prelimi­nary studies of the mental patients and of the children were sub­sequently expanded into more thorough and extensive investiga­tions which are :reported elsewhere in this bulletin. It will be noted that the amino acids, etc., in Table I are listed in the order of the frequency of their urinary occurrence. They range from glycine and alanine which are excreted by all and all but one of the individuals respectively, down to proline, a-amino butyric acid, and methionine sulfone which were found to be ex­creted by three individuals and to aspartic acid which was found by the procedure employed to be excreted by only one. The range of concentration for glycine in the recorded value is from 31 to 500 p.g. per mg. of creatine. For alanine it was 4 to 110 p.g. per mg. of creatinine. For citrulline on the same basis (excluding, of course, those individuals who were not found to excrete measurable amounts) the range was from 8 to 99. For serine the corresponding range was not so large-17 to 58. For taurine it was 7 to 130. The wide variations which exist emphasize the fact that methods of de­termination need not be more than crudely quantitative in order to establish the existence of distinctive patterns of excretion. Salivary amino acid secretion patterns.-Table II gives data re­garding salivary amino acid secretion patterns for a number of indi­viduals. The saliva samples were collected immediately after the subjects arose, and no stimulus was used to induce saliva flow. The saliva was chromatographed (one dimension only) in amounts of 25 to 200 microliters using the phenol-sodium citrate-sodium phosphate solution previously described (p. 25). The table is con­structed in a manner similar to that of Table I, except that the quantitative values are expressed as micrograms per milliliter of saliva. TABLE II Salivary Secretion of Amino Acids Unknown Amino Aspartic Glutamic Acid Subject Sex No. Acid Samples mg./ml. Acid mg./ml. Serine mg./ ml. Glycine mg./ ml. Alanine mg./ml. Lysine Rf-.65 in mg./ml. phenol A (f) 21 .0017 .013 .012 .014 .013 .015 .017 B (f) 24 .0013 .013 .005 .010 .005 .0015 .017 c (m) 2 .010 .014 .012 .035 D (f) 1 .050 E (m) 2 .020 .036 .029 .045 F (m) 1 .005 .005 .015 .005 G (m) 1 .0075 .005 .012 .015 H I (m) (m) 3 1 .0033 .007 .005 .0033 .005 .008 .005 .007 .oio .010 .005 .016 .020 Average .0015 .012 .0066 .014 .012 .0077 .024 The existence of consistent individual patterns of secretion is quite apparent. One individual secreted no detectable amount of any of the known amino acids, four individuals secreted only 3 of the 6 known amino acids listed, one secreted 5, while the remaining three secreted all of them, but in varying amounts. It is of espe­cial interest that the individual A whose saliva contains ten times as much lysine as individual Bis the same individual whose urinary lysine is unusually high (p. 154). This indicated that the high uri­nary lysine is not due merely to a kidney difference, but has a broader metabolic significance. Sumrnary Data are presented showing the existence of highly individual amino acid excretion patterns by 10 individuals using two dimen­sional chromatography. Tabulated data are also given for the urinary excretion of amino acids and related compounds by 52 individuals (at least 3 samples from each). The amino acid content of salivas from 9 different people are also tabulated. One amino acid of unknown nature is consistently present in relatively high concentration. One individual who excretes relatively large amounts of urinary lysine was found to have a high content in the saliva also. XVIII. Individual Urinary Excretion Patterns of Young Children by HELEN KIRBY BERRY AND LOUISE CAIN Urine samples from young children were analyzed following pro­cedures developed for studying metabolic excretion patterns, in order to gain background information as to the existence of pat­terns in early life and how they compare with those of adults. Dur­ing the early months of life children receive a relatively standard diet in which milk is the main item. If differences in diet are in a considerable measure responsible for differences in patterns, then one would not expect to find differences in patterns to be prominent in early ages. Experimental Subjects.-Ten male and 11 female children ranging in age from 6 weeks to 7 years were employed in this study. The ages of the various subjects are shown in Table I. The boys, H2, H3, and H4, and the girl Hl were siblings; of the girls, Bl, B2, and B3 were sisters; the other subjects were unrelated. The diets of the children were not controlled or modifi.ed in any way. Urine samples.-Three urine samples (morning samples wher­ever feasible) were collected on consecutive days. The effect of varying urine concentration was minimized by expressing all re­sults using creatinine as a base (p. 152). Analytical methods.-Creatinine was determined by a standard colorimetric procedure (1). All other constituents were determined by paper chromatographic methods. Amino acids, except histidine, were determined on one-dimensional paper chromatograms using phenol as solvent and ninhydrin as the color developing reagent (p. 71). In some instances, where there were interferences on the one-dimensional chromatograms, two-dimensional chroma­tograms (p. 72) were employed. Urea was determined on paper chromatograms resolved with phenol and developed for color with sodium hypochlorite (p. 29). Phosphate was determined on chro­matograms resolved in butanol-acetic acid solvent, and developed for color with ferric chloride solution (p. 28). Another chroma­togram resolved with the butanol-acetic acid solvent was sprayed with bromcresol green reagent (p. 29) for the location of acidic areas due to such constituents as citric acid, lactic acid and hip­puric acid; and of a basic area which can be interpreted as a meas­ure of total "weak acids" present (p. 30). Still another chromato­gram resolved in the butanol-acetic acid solvent was developed for color with the p-dimethylaminobenzaldehyde reagent (p. 30). A number of characteristic spots whose identities have not been de­termined are developed by this reagent. Histidine and an additional group of unidentified constituents were determined on chromato­grams resolved with the butanol-ethanol-hydrochloric acid solvent (p. 77) and the butanol-acetic acid solvent (p. 25) respectively, and were developed for color with the diazotized sulfanilic acid reagent (p. 30). Another unknown urinary constituent of common occurrence was determined on a chromatogram resolved in the isobutyric acid-acetic acid solvent (p. 26) and developed for color with the bromine reagent (p. 30) . Residts The averages of determinations on three urine samples from each of the subjects are given in Table I. A blank in the table indicates that the constituent in question was not determined. A dash in the table indicates that determinations were performed but nothing was detected. Values for creatinine are expressed in mg./ml. of urine. Values for urea and the amino acids are ex­pressed in mg. of constituent per mg. of creatinine. The other constituents are either unidentified, or some doubt exists as to their identity, and the quantitative values for these substances are expressed in terms of the area of the spot developed from a quan­tity of urine equivalent to 40 micrograms of creatinine; however, in the case of those substances developed for color with diazotized sulfanilic acid, the areas reported were developed from a quantity of urine equivalent to 100 micrograms of creatinine. Most of the data, it will be seen, are expressed relative to creatinine. A justifi­cation for this procedure will be found elsewhere in this bulletin (p. 152), though actually the justification has not been shown to exist in the case of very young children. Since previous workers (2,3) have indicated that the excretion of creatinine is a direct func­tion of body weight, or, more particularly of muscle tissue weight, the results expressed in terms of creatinine provide a rough means of comparing the excretion of different children in terms of mg. of constituent per unit body weight. It should be borne in mind that the areas recorded as quantitative measures of certain of the con­ TABLE I Urinary Excretion Patterns in Young Children PO, Agein Subjects months Females Creat­inine mg./ ml. Urea mg./ml. Area (in.2 ) corres. 40 µg. Cr. Hist­idine mg./ ml. Glutamic mg.I m!. Serine mg./ml. A 5 0.09 41 0.31 .13 Bl 11 0.22 36 0.31 0.22 .060 c 18 0.52 37 0.29 0.15 .12 .11 D 19 0.44 52 0.26 0.29 .010 .015 E 29 0.45 40 0.45 0.17 .068 .047 F 32 0.70 45 0.22 0.30 .037 .065 G 36 0.62 31 0.29 0.36 .069 .058 B2 36 0.36 30 0.22 0.39 .023 Hl 48 0.78 17 0.09 0.08 .044 .017 I 48 0.72 50 0.17 0.41 B3 54 0.30 50 0.11 0.18 .059 .071 Males J 11h 0.13 :33 0.26 .13 K 11 0.34 66 G.42 0.07 .030 .021 L 19 0.99 50 0.45 0.19 .007 .029 M 19 0.39 30 0.29 0.36 .22 .13 H2 20 1.27 16 0.23 0.25 .044 .Q70 N 21 0.54 45 0.26 0.06 .021 0 37 0.91 33 0.20 0.09 .007 .007 H3 42 0.92 17 0.08 0.23 .017 .11 H4 69 0.88 32 0.22 0.13 .020 .050 p 84 1.28 23 0.15 0.35 Average: 10 boys 11 girls 0.76 0.51 35 39 0.26 0.24 0.19 0.26 0.055 0.054 0.046 0.042 Methio­ nine Subject Glycine mg./ml. Taurine mg./ml. Alanine mg./ml. Citrulline Sulfoxide* mg./ mg.Cr. mg./mg.Cr. Females A .047 .11 Bl .62 .074 .013 .18 c D .50 .26 .006 .096 .016 .031 .046 .Q18 E .13 .077 .099 .061 F .091 .11 .043 .027 .052 G .41 .073 .11 .023 .041 B2 .069 .047 .023 .019 Hl ;072 .027 .004 .013 I B3 ;18 .024 .039 Males J K L .11 .023 .054 .13 .026 .050 .020 .044 .071 .032 .079 M ~22 .10 .025 .065 H2 .13 .026 .050 .063 .016 N 0 .005 .056 .038 .006 .011 .013 .004 H3 .17 .067 .067 H4 .10 .012 .046 .044 p Average: 10 boys 11 girls 0.096 0.26 0.026 • 0.027 0.038 0.052 0.028 0.025 0.031 0.052 •see footnote p. 76. stituents listed in Table I are probably not linear functions of con­centration (p. 50), and while the existence of differences may be validly inferred, the magnitude of these differences is not accu­rately indicated. Discussion The data of Table I strongly indicate the existence of character­istic excretion patterns for the children studied. The spread be­tween the highest and lowest individual values recorded for the eleven identified constituents, beginning with histidine, averages about 26 fold and in the case of no item is the spread less than 7 to 1. This statement tends to minimize the differences, because for 8 of the 11 items, several individuals showed no detectable quan­tity. Furthermore the differences with respect to the unidentified constituents appear to be greater than for the identified ones. TABLE I-(Continued) Color developed with bromcresol green reagent Color developed "basic with (citric) (lactic) (hippuric) cations" Bromine yellow yellow yellow blue , Reagent Rf .24 Rf .79 Rf .33 Rf .33 Rf .40 Subject (Area (in.2 ) corresponding to 40 'Y creatinine) Females A .03 Bl .31 .95 c .50 .32 .26 D .07 .16 .02 .17 .10 E .52 .38 F .63 .16 .07 .18 G .50 .17 .15 .10 B2 .16 .06 .27 .11 Hl .25 .14 .07 .17 .16 I .13 .08 .06 .10 B3 .06 .07 .08 Males J .65 .13 1.20 .06 K .15 .82 L .55 .25 M .30 .95 .13 H2 .36 .26 .28 .13 N .09 .44 .18 0 .26 .10 H3 .33 .23 .21 .13 H4 .45 .16 .lQ .11 p .08 .07 .09 Average: 10 boys 0.22 0.15 0.13 0.35 0.12 11 girls 0.28 0.05 0.11 0.24 0.10 TABLE I-(Continued) Color developed with diazotized sulfanilic acid, pink pink br. pk. orange orange yellow pink orange pink Subject Rf .20 Rf .23 Rf .26 Rf .33 Rf .52 Rf .59 Rf .60 Rf .80 Rf .92 (Area (in.2) corresponding to 100 y creatinine) Females A Bl .007 .17 .18 c .19 .14 .17 .05 D .13 .03 .18 .08 E .05 .28 .22 .18 .20 F .17 .06 .11 G .14 .08 .06 .11 B2 .13 .10 .04 H1 .08 .13 .04 I .35 .05 .05 B3 .03 Males J .30 .13 .04 .09 K .31 .08 .09 .23 .12 L .29 .25 .23 .18 M .18 .13 .13 1.02 .04 .40 .20 H2 .13 .10 .14 .14 N .02 .08 0 .20 .12 .17 .11 H3 .08 .08 .24 .09 H4 .30 .27 .30 .17 .09 p .16 .12 Avera~e: 10 oys 0.19 0.07 0.16 0.12 11 girls 0.13 0.05 0.09 0.09 TABLE I-(Continued) Color developed with p-dimethylamino benzaldehyde reagent (allan­toin) yellow purple brown grey pink purple yellow Subject Rf .33 Rf .35 Rf .40 Rf .77 Rf .88 Rf .92 Rf .20 (Area (in.2) corresponding to 40 y creatinine) Females A ,33 Bl .24 c .16 .24 D .18 .18 E .21 .10 .26 F .15 .18 .17 G B2 Hl .16 .22 I B3 Males .18 .14 .35 J .14 .19 K L .22 M H2 .17 .14 N 0 .24 H3 .19 .30 H4 .15 .31 p Inspection and statistical analysis of the results indicate the presence of some sex differences. For example, the average excre­tion of glycine for the girls is nearly three times that for the boys. A difference in the same direction has also been observed in the case of adults (p. 154). In general, the average values of the constit­uents appear remarkably uniform for the sexes. There also appear to be a few age differences. Only the two youngest children were found, for example, to excrete detectable amounts of allantoin. Creatinine excretion, as might be expected, is low in the youngest children. Since initial inspection of the data appeared to indicate that the three brothers included in the study (H2, H3, and H4, ages 51/2, 31/2 and 1112 respectively) showed patterns which resembled each other strikingly, we applied statistical analysis to reveal objectively the extent of similarity of all the siblings involved in the study. The "coefficients of similarity"* between each of the tabulated values for each of these brothers and the corresponding values for the other brothers and for five other boys in about the same age group were calculated. The same type comparison was made for the three sisters (Bl, B2, and B3, ages l, 3 and 4112 respec­tively) and five girls of about the same age group. The girl, Hl, age 4, was compared to her three brothers and to the three sisters from another family Bl, B2, and B3. For this purpose the same 11 consecutive items beginning with histidine in Table I were used. Where a substance was not detected in a sample, it was taken, for purposes of calculation, to be present in an amount equal to one-half the lowest amount found in any of the samples. The results of these calculations are shown in Table II. These calculations definitely confirm the observation as to the pro­nounced similarities of the brothers' patterns and strongly suggest that these patterns of excretion are inherited and are relatively stable, since the three brothers exhibited similarity in patterns regardless of their ages. On the other hand, the three sisters showed no greater average similarity to each other than to the five unrelated girls with whom they were compared. The three brothers' sister, Hl, showed only a slightly greater average sim­ilarity .to her brothers than to the three sisters; this slight differ­ence is probably not significant. Since Hl and her three brothers presumably have access to the sa~e food the comparative dissimi­larity between the girl and her brothers is in line with the idea * " ffi . t f . .1 . ,, lower value coe c1en o s1m1 anty = X 100 average va1ue TABLE II Average Coefficient of Average of Average Similarity for 11 Items Coeff. of Similarity 1. a. Between brothers H4 and H3_____________________________________ _ H4 and H2______________________________ __ __ H3 and H2___________ __ ___ _ _______ __________ _ b. Between male non-siblings H4 and K_______________ _______________________ _ H4 and L __ __ __________________________________ H4 and M __ __________ _ __ _ __________ ____ __ _ H4 and N ______________________ ___ ____ _ ___ ___ _ H4 and O________________________________ ____ _ _ H3 and K____________________ __ ________________ _ H3 and L ______ _ _ __ _ _____________ _ H3 and M________________________________ __ ____ _ H3 and N ________________ __ ___________ _ ______ _ H3 and o_ ________ _ _______________________ _ H2 and K _____________________ _ ________________ _ H2 and L _________________________ __ ___ __ _____ _ H2 and M________________ _ ____ __ _____ _ ______ H2 and N _____________________________________ _ H2 and o_____ _ ____ _ ______ __ _ __________ _ 2. a. Between sisters Bl and B2____________________________________ _ Bl and B3__________ __ ________ _ ______________ _ B2 and B3__________________________ __________ _ b. Between female non-siblings Bl and C_______________________ _ _____ _________ Bl and D_______________________________________ _ Bl and F ________________________________ ______ Bl and G_____ _ ______________________________ _ Bl and Hl_____________________________________ _ B2 and c__ ________________________________ __ _ B2 and D__________________________________ ____ _ B2 and F ____________________________ _ ________ B2 and G---------------------------------------­ B2 and HL__ _________________________________ _ B3 and c_________________________ _ __ ____ _ B3 and D_______________________________________ _ B3 and F ______________________________ ____ _ _ B3 and G---------------------------------------­ B3 and HL______________________________ __ __ _ 3. a. Between the sister and her three brothers Hl and H3______________________________ _ ___ _ Hl and H2 ____________________________________ _ Hl and H4_______________ _ __ ____________ _ _ b. Between the same girl and three sisters not her siblings Hl and BL_________________________ __ _ _____ _ Hl and B2__ __________________________________ _ Hl and B3_______________________ __ ____________ 80 73 76 46 65 46 40 53 44 54 58 28 50 41 64 53 38 39 62 56 52 59 49 44 59 43 56 60 63 64 58 53 54 57 61 59 53 57 66 43 58 59 76 48 57 57 59 53 that the differences are of an internal nature. The existence of strong similarity between certain siblings and lack of similarity between others, as in the sisters above, is in line with what would be expected on a hereditary basis. It may be .noted that of the non-sibling boys, L's pattern is most like that of the three brothers and N's pattern is least like them; the average coefficient of similarity for Lis 61, while for N the cor­responding value is 35. Whether these similarities and differences in patterns denote similarities and differences in patterns of mental abilities or personality traits is unknown and can only be deter­mined by extended interdisciplinary research. BIBLIOGRAPHY 1. "Klett-Summerson Clinical Manual," Klett Mfg. Co., New York. 2. Talbot, N. B., Am. J. Diseases -Of Children, 55, 42 (1938). 3. McClugage, H. B., Booth, G., and Evans, F. A., Am. J. Med. Sci., 181, 349 (1931). XIX. A Further Study of Urinary Excretion Patterns in Relation to Diet by HARRY ELDON SUTTON All of the studies presented in this bulletin both on animals and human beings indicate that excretion patterns are determined in large part by factors other than diet. It is well known, however, that unusual urinary constituents do appear as the result of eating particular foods (of which asparagus is an outstanding example) and the question of diet seemed worth investigating again from the standpoint of individual excretion patterns. For one thing it was hoped that additional light might be thrown on the problem of whether the observed individual patterns are caused in part by differences in the self selection of foods by the individuals con­cerned. In an earlier experiment reported from this laboratory (1) several individuals were placed on identical diets but only four urinary amino acids were determined (microbiologically) and found to be essentially unchanged by changes in diet. The present experiment was similar in principle to the former one; many more items were included but fewer individuals participated. Experimental Preliminary studies failed to show that the use of special foods, including asparagus, caused the appearance of any unusual urinary constituent as determined by paper chromatography. It is known that asparagus consumption causes the excretion of methyl mer­captan which has a characteristic odor (2). Due presumably to the minute amounts present and to its volatility this substance does not show up on chromatograms. No quantitative information is available as to the amounts of methyl mercaptan which may be present in urine, but because the sense of smell for compounds of this type is extremely sensitive, it is probable that the amounts are extremely minute (3). For a more systematic study three young healthy males were chosen because in preliminary experiments they apparently ex­hibited noticeable differences in their urinary excretion patterns. During the experimental period of five consecutive days, these in­dividuals ate together at all times, choosing a diet that did not differ greatly from the normal regimen of any of them. No attempt was made to eat the same foods on each day of the five day period; rather the diet was varied as much as possible, keeping within the realm of "common" foods. Urine samples were collected upon rising for three mornings prior to the experimental period, for the five days of the experimental period, and for the three mornings following. The urine samples were frozen until ready for analysis. No attempt was made to control the amount of exercise of each person, but as the subjects are in similar occupations and undergo approximately the same amount of exercise, no error should be an­ticipated from this source. None of the subjects engaged in an unusual amount of exercise during the experimental period. The analytical procedures used were similar to those used in previous studies of this nature performed in this laboratory. Crea­tinine was determined by the picrate method ( 4) and was used as a basis for comparing the results of other analyses in order to over­come the dilution factor. The amino acids were determined on two-dimensional paper chromatograms according to the techniques reported by Berry and Cain (5) with the exception of histidine which was done by the method reported elsewhere in this bulletin (p. 77). The compounds which form colored products with diazo­tized sulfanilic acid were detected by chromatographing on paper an aliquot of urine containing 100 micrograms of creatinine, re­solving it in the butanol-acetic acid-water mixture (p. 25), and developing the resulting sheet by spraying with diazotized sulfanilic acid ( p. 30) . The substances reacting with bromcresol green (p. 29) were chromatographed by resolving in the butanol-acetic acid-water mixture an aliquot of urine containing 60 micrograms of creatinine. Ferric chloride chromogens ( p. 28) were likewise detected using an aliquot of urine containing 40 micrograms of creatinine. Chromatograms were also prepared by resolving aliquots of urine containing 40 micrograms of creatinine in the butanol-acetic acid­water mixture and developing respectively with ammoniacal silver nitrate (p. 27), 2,6-dichloroquinonechloroimide (p. 33), p-dimethyl­aminobenzaldehyde (p. 30), or bromine reagent (p. 30). The substances developed by these last reagents are for the most part either the same as detected by other methods of analysis or of unknown composition with no accurate means of quantitation. Visual comparison of the size and intensity of the urea-and the uric acid spots as well as those of other chromogens indicated that there was no significant difference between samples when considered from the point of view of either the individual or the diet; hence, no further attempts were made to determine these substances quantita­ tively. Phosphate was determined by the colorimetric method of Fiske and Subbarow (6). Results The results of these determinations are shown in Table I, most of the data being expressed using creatinine as a basis. Where no absolute standard was used for reference and the spots on the chromatograms were of sufficient resolution for the area to be ac­curately measured with a polar planimeter, the results are ex­pressed in square inches. When the spots could only be compared visually for intensity, the results are expressed as -, +, + +, etc. TABLE I Creatinine (mg./ml.) Histidine hydrochloride (mg./mg. Cr.) Day Subj. A Subj. B Subj. C Subj. A Subj. B Subj. C Cl 0.96 2.70 2.14 0.10 0.07 0.05 C2 1.68 1.74 1.92 0.06 0.17 0.16 C3 1.02 1.12 2.64 0.10 0.20 0.15 El 1.h. 1.06 1.80 0.09 0.14 0.17 E2 1.26 3.46 2.46 0.10 0.10 0.16 E3 1.04 1.38 1.76 0.10 0.22 0.17 E4 1.08 1.94 1.44 0.09 0.15 0.10 E5 1.14 1.86 1.42 O.D7 0.19 0.21 C4 0.80 1.32 2.62 0.11 0.15 0.17 C5 1.52 0.92 2.98 0.09 0.19 0.13 C6 0.82 2.20 2.58 0.10 0.14 0.14 Glycine (mg./mg . Cr.) Alanine (mg./m g. Cr.) Day Subj. A Subj. B Subj. C Subj. A Subj. B Subj. C Cl 0.07 O.D7 0.03 0.02 0.01 0.02 C2 0.02 0.05 0.04 0.02 0.01 0.02 C3 0.03 0.12 0.04 0.02 0.04 0.01 El 0.04 0.03 0.06 0.01 0.03 0.02 E2 0.02 0.03 0.04 0.01 0.01 0.01 E3 0.015 0.06 0.07 0.01 0.02 0.02 E4 0.04 0.03 0.04 0.02 0.03 0.02 E5 0.04 0.05 0.07 0.005 0.02 0.02 C4 0.05 0.06 0.06 0.02 0.02 0.02 C5 C6 0.05 0.05 0.04 0.05 0.04 0.03 0.01 0.02 0.04 0.01 0.01 0.01 Serin e (mg./ mg. Cr.) Citrull ine (mg./mg. Cr.) Day Subj. A Subj. B Subj. C Subj. A Subj. B Subj. C Cl C2 C3 El E2 E3 E4 E5 C4 C5 C6 0.03 0.02 0.03 0.03 0.02 0.01 0.02 0.03 0.05 0.03 0.05 0.02 0.03 0.04 0.03 0.02 0.04 0.03 0.02 0.02 0.02 0.03 0.03 0.04 0.03 0.03 0.02 0.02 0.03 0.05 0.03 0.02 0.03 0.06 0.06 0.04 0.03 0.06 0.04 0.07 0.06 0.12 0.05 O.D7 0.09 0.06 0.12 0.03 0.06 0.07 0.07 0.11 0.05 0.08 0.06 0.07 0.10 0.05 0.11 0.10 0.09 0.05 0.05 0.05 0.07 0.06 TABLE !-(Continued) Taurine (mg./mg. Cr.) Methionine Sulfoxide* (mg./mg. Cr.) Day Subj.A Subj. B Subj. C Subj. A Subj. B Subj. C Cl 0.00 O.Q7 0.07 0.07 0.06 0.03 C2 0.10 0.11 0.06 0.04 0.04 0.04 C3 0.00 0.11 0.06 0.12 0.11 0.01 El 0.09 0.14 0.11 0.06 0.09 0.03 E2 0.08 0.07 0.10 0.06 0.06 0.10 E3 0.02 0.11 0.11 0.02 0.07 0.06 E4 0.11 0.08 0.10 0.10 0.05 0.04 E5 0.09 0.13 0.25 0.05 0.06 0.04 C4 0.19 0.11 0.08 0.05 0.08 0.02 C5 0.05 0.08 0.07 0.08 0.11 0.02 C6 O.Q7 0.05 0.10 0.08 0.03 0.10 Valine (mg./ml.) Leucine (mg./ml.) Day Subj. A Subj. B Subj. C Subj. A Subj. B Subj. C Cl 0.03 0.02 0.09 C2 0.02 0.03 0.007 0.03 0.03 C3 0.01 El 0.002 0.04 E2 0.02 0.05 E3 0.007 0.03 E4 0.01 0.03 E5 0.007 C4 0.01 C5 0.01 0.01 0.03 0.03 C6 0.01 0.02 0.03 "Under Citrulline" "Over Threonine" Rating/ 20 µ.I. Urine Rating/ 20 µ.l. Urine Day Subj. A Subj. B Subj. C Subj. A Subj. B Subj. C Cl ++ ++ ++ + C2 ++ + +++ C3 + +++ ++ El + ++ + + E2 ++ + + + E3 ++ + + E4 + + + E5 + C4 + + + + C5 + + + + C6 + + ++ + + ++ + + Sulfanilic Acid Orange Spot (Rf .85) Lysine (mg. / mg. Cr.) Rating/100 µg. Cr. Day Subj. A Subj. B Subj. C Subj. A Subj. B Subj. C Cl 0.02 0.015 +++++ ++++ +C2 0.015 0.015 ++++ + + C3 +++ +++++ +++ El ++++ +++ +++ E2 0.01 +++ +++++ ++++ E3 ++++ +++++ +++++ E4 0.02 0.015 ++++ ++++ ++++ E5 ++ + ++ C4 0.01 ++ + + C5 0.04 0.015 ++ + + C6 O.Q15 ++ ++++ + •see footnote p . 75. TABLE !-(Continued) Sulfanilic Acid Purple Spot (Rf .90) Hippuric Acid Area (in.2/lOOµg. Cr.) Area (in.2/60µg. Cr.) Day Subj. A Subj. B Subj. C Subj. A Subj. B Subj. C Cl 0.06 obscured 0.11 0.41 0.45 0.25 C2 0.00 O.Q7 0.15 0.63 0.33 0.35 C3 0.08 0.33 0.17 0.35 0.34 0.27 l<.;l 0.25 0.29 0.20 0.40 0.34 0.43 E2 0.05 0.13 0.11 0.30 0.43 0.26 E3 0.12 0.15 0.22 0.32 0.44 0.40 E4 0.21 0.18 0.23 0.51 0.43 0.37 E5 0.26 0.31 0.28 0.31 0.37 0.37 C4 0.00 0.15 0.16 0.60 0.39 0.23 C5 0.16 0.46 0.15 0.25 0.29 0.26 C6 0.00 0.26 0.05 0.17 0.31 0.37 Bromcresol Green Acid Spot (Rf .28) Bromcresol Green Acid Spot (Rf .30)Area (in.2/ 60µg. Cr.) Area (in.2/60 µg. Cr.) Day Subj. A Subj. B Subj. C Subj. A Subj. B Subj. C Cl 0.15 0.13 0.17 0.00 O.Q7 0.00 C2 0.10 0.28 0.18 0.00 0.14 0.00 C3 0.23 0.13 0.21 0.06 0.00 0.00 El 0.35 0.16 0.22 0.00 0.10 0.11 E2 0.33 0.10 0.19 0.00 O.Q7 0.04 E3 0.30 0.18 0.21 0.07 0.10 0.07 E4 0.07 0.14 0.17 0.00 0.06 0.05 E5 0.28 0.14 0.18 0.05 0.05 0.00 C4 0.19 0.27 0.25 0.00 0.10 O.Q7C5 0.23 0.20 0.27 0.08 0.21 0.07 C6 0.13 0.20 0.29 0.00 0.07 0.09 Phosphate (mg./ mg. Cr.) Day Subject A Subject B Subject C Cl 0.51 0.78 0.90 C2 0.60 0.28 0.89 C3 0.68 0.56 0.68 El 0.56 0.98 0.77 E2 0.56 0.72 0.72 E3 0.53 0.61 0.73 E4 0.56 0.72 0.66 E5 0.69 0.98 0.89 C4 0.43 0.73 0.86 C5 0.44 0.68 0.89 C6 0.55 0.25 0.79 Two unknown substances are reported which were detected on the ninhydrin-developed amino acid chromatograms. They are "under citrulline," (ascending method) a blue spot which migrates in phenol to a position just below that of citrulline and in lutidine to the same position as citrulline, and "over threonine," a purple spot which migrates in lutidine to a position just above threonine and in phenol to the . same position as threonine. In Table I the control days are indicated by the letter C, the experimental days by E. Creatinine is abbreviated Cr. A study of the results presented in Table I does not yield as much conclusive information as might be desired, because for many of the items determined the individuals did not show, during the short period, significant differences. A number of interesting facts, however, may be noted. First, the changes in the diet which were instituted did not appear to modify the excretion of most of the substances to any material degree. Secondly, certain substances appeared at characteristic levels for each individual and did not change when the diet was modified. Histidine is a clear cut example of this, and the substances designated "under citrulline," "over threonine" and "Bromcresol Green (Rf .30)" appear to fall in the same category. Thirdly, the excretion of certain substances varies with the diet. The clearest example of this is "Sulfanilic Acid Purple (Rf .90) ." When the values for this substance are plotted for the days on which collections were made, a marked parallelism is noted between the curves of the individuals during the experimental period (Fig. 1). Spot "Sulfanilic Acid Orange (Rf .85)" appears to be in the same class but the data are not so clear cut. Not included in the tabulated material are three additional un­known substances: A sulfanilic acid red streak, Rf .22-.36, occurred in much higher quantities in the urine of subject B than in that of subjects A and C and did not seem to vary with the diet. A ferric chloride alkaline spot (Rf .34), which reflects weak acid excretion (7), consistently occurred strongly, regardless of diet, in samples from subject A, less strongly from subject B, and hardly at all from subject C. A sulfanilic acid orange spot Rf .60 occurred in all samples but seemed to vary with the diet as did the other sulfanilic acid orange spot Rf .85. Phosphate and the bromcresol green acid spot (Rf .28) are excreted at characteristic levels, but the daily parallelism in two of the subjects indicates that the excretion of these substances may fluctuate to some extent with the diet. However, the excretion of these substances by the third subject seems not to be related to that of the others. The majority of cases seem to fall between the two extremes. Most of the substances are excreted at levels apparently charac­teristic of the individuals, but the rather large daily variation of many of these substances makes it difficult to evaluate the im­portance of diet in their excretion. This is true of the amino acids other than histidine; in these, the accuracy of the method of de­termination will not allow detection of possible small changes 0 .48 Cl) 0.36 LL.I J: 0 z -0.30 LL.I a: < ::::> 0.24 CJ Cl) z - < 0.1 8 LL.I a: < 0.12 0.06 E3 E4 E5 C4 CS . C6 FIGURE I. Area of Sulfanilic Acid Purple Spot (Rf. 90) vs. Day of Excretion. caused by diet. The methods are suffic_ient to indicate, however, that there are no major changes due to dietary differences and that when treated statisically, the levels of excretion as a whole are characteristic of an individual. Summary Experiments are described in which the relationship of diet to urinary excretion patterns was studied in three individuals who were on an identical diet for a five day period. The results indicate that of the many substances which were studied, only two showed an obvious dependence upon diet. These were the sulfanilic acid orange spot (Rf .85) and the sulfanilic acid purple spot (Rf .90). Of the remainder of the substances, some may vary slightly with diet, but in no case does the diet appear to be as important in determining the excretion patterns as do other factors seemingly characteristic of the individual. Several substances, notably histi­dine, appeared to be excreted at levels which were quite character­istic of each of the three individuals. BIBLIOGRAPHY 1. Thompson, R. C., and Kirby, H. M., Arch. Biochem., 21, 210 (1949). 2. Gautier, CI., Compt. rend. Soc. biol., 89, 239 (1923). 3. Moncrieff, R. W., The Chemical Senses, John Wiley and Sons, Inc., New York, N. Y., 1946, p. 79. 4. Bonsnes, R. W., and Taussky, H. H., J. Biol. Chem., 158, 581 (1945). 5. Berry, H. K., and Cain, L., Arch. Biochem., 24, 179 (1949). 6. Fiske, C.H., and Subbarow, Y., J. Biol. Chem., 66, 375 (1925). 7. Sutton, H. E., M.A. Thesis, The University of Texas (1949). XX. Metabolic Patterns of Underweight and Overweight Individuals by JACK D. BROWN AND ERNEST BEERSTECHER, JR. Recent studies by Williams, et al. (1-5), of individual metabolic differences have proved to be very suggestive in connection with such problems as alcoholism, mental disease, etc. With this in mind it was decided that a similar biochemical approach might be help­ful in elucidating the causes for the differences between persons who are decidedly overweight and those who are underweight but who do not exhibit any other striking abnormalities. A survey of the literature reveals a considerable divergence of opinion as to the types and possible causes of obesity and leanness. The accent in the past has been on the study of obese persons as compared with so-called normals. The available information suggests the ad­visability of avoiding subjects who are prodigiously fat or who are of the walking-skeleton variety because in these, pathological con­ditions may exist which are not to be found in those who deviate less from the average. Consequently, the subjects in the present study have been selected for their tendency toward being over-or under­weight. Extreme cases were not selected. Likely male subjects, ranging in age from 20 to 31 years, were obtained through the co-operation of the University Health Service. One group of seven, henceforth designated as the overweight group, had for no apparent reason, a decided tendency to be over­weight and experienced great difficulty in controlling their weight downward toward the average. Members of the underweight group, comprised of ten individuals, were definitely on the lean side and, for no apparent reason, encountered considerable difficulty in gain­ing weight. Prior to final selection of the above groups, prospec­tive subjects were screened to exclude those with any previous his­tory of diseases, such as diabetes, which might introduce factors of error into the study. Personal data of the underweight and over­weight subjects studied are presented in Table I. Urine samples were selected for use in this study because it has been shown (1-5) that a study of urinary constituents is very en­ lightening in observing individual metabolic differences. Urine re­ flects in a measure a true image of the metabolic processes which take place in the body. No attempt was made to control the diets of the individuals being studied. As is demonstrated elsewhere in this bulletin (6), a number of urinary constituents remain relatively constant for an individual independent of his diet. That self-selection of food by the subjects should not be eliminated from consideration is further supported by other investigations (4). TABLE 1 Personal Data of Overweight and Underweight Subjects Difficulty in Subject Number Age (years) Height Weight (pounds) controlling wt. toward "normal" Overweight:10 25 5' 8" 215 Great 11 20 6' 4" 262 Great 12 29 6' 0" 255 Very great 13 28 5' 10" 230 Very great 14 31 6' l" 230 Great 15 29 5' ll1h" 270 Great 16 26 5' 9" 205 Very great Underweight:20 25 6' 0" 135 Great 21 25 6' 0" 140 Great 22 23 21 25 6' 5' 0" 8" 130 122 Very great Great 24 25 26 27 22 29 28 27 6' 41h" 5' 4" 5' 111h" 5' 111h" 162 98 135 141 Very great Very great Very great Great 28 29 26 25 5' 10" 5' 111h" 115 128 Very great Very great Experimental Three complete morning samples of urine, collected at intervals of three to five days, were obtained from each subject. The volume, pH, and specific gravity were measured for each sample, and the samples were then refrigerated under toluene. Methods Spectrochemical analyses of urine for phosphorus, potassium, sodium, calcium, copper, boron, magnesium, and iron were per­formed by the Pacific Spectro-Chemical Laboratory of Los Angeles, California, by standard techniques. Duplicate determinations re­vealed that the average deviation for all elements except iron and copper was 2.2 ro' indicating a probable error of about 3.3% of the true amount present. The probable error for iron and copper was calculated to be about 15 % . In addition to the elements reported, strontium and silicon were detected in all samples. Strontium con­centration appeared to vary directly with the calcium content of the urine. Metabolic Patterns of Obese and Lean The method of Bonsnes and Taussky (7) for the colorimetric determination of creatinine was used. Since twenty-four hour urine samples were not available, Shaffer's creatinine coefficient (8) was modified and defined as the concentration (milligrams/ milliliter) of creatinine per kilogram of body weight. Pigment was measured as the per cent of absorption of 420 mu light by the clear urine sample as determined by a Coleman spectrophotometer. Phosphate was determined colorimetrically by the method of Fiske and Subbarow (9). Chloride was determined by the Mohr method of titration (10). Total carbohydrates were determined using Dreywood's reagent (11) and reported in terms of glucose. Ketonic substances were determined colorimetrically by a method originally described by Behre and Benedict (12) and modified for this study. Choline was determined by microbiological assay through use of the method described by Horowitz and Beadle (13). Results and Discussion Those items which exhibited substantial group differences were selected from the larger mass of data (14), and are presented in Table II. Data on pH, specific gravities, and volumes are also in­cluded in Table II, although these items showed no marked group differences. Standard statistical methods (number of standard errors as calculated using the standard deviation from the arith­metical mean) were employed (15) to determine the relative sig­nificance of each item examined. A difference between the over­weight and the underweight groups which gave two or more standard errors was considered statistically significant, since at two standard errors the odds against the difference of the two means being due merely to chance is slightly greater than 20 to 1. In biological work in which small numbers of samples are involved it is customary to accept the 5o/o level (20 to 1) as being strongly indicative of statistical significance. The items studied are listed in descending order of significance in Table III. It is interesting to note that three of the items (pH, volume, specific gravity) did not show any marked group differences, thus lending greater credence to those differences observed. When significant results are obtained in a study of the present kind, the problem remains to determine whether the observed differences are causes or results. However, regardless of which they are, some idea of the metabolic processes involved can be obtained. Hence it is desirable to examine the experimental data for any metabolic aberrations which may be involved in obesity. ~ TABLE II 00 tl'>­ Summary of Analytical Data Including the Items Showing Substantial Differences Sample Number Volume ml. (Specific gravity -1) x 103 pH Between Overweight and Underweight Groups Modified Pigment : Phos-Creatinine Creati­phate Calcium Phosphorus Creatinine Coefficient nine mg./ ml. mg./ ml. mg./ ml. mg. / ml. Chloride mg./ml. Boron µg./ml. Choline µg./ml. Copper µg/ml. Under­ weight: 20a 220 31 5.2 18.1 1.50 0.153 1.20 2.60 9.00 2.35 14.5 0.13 b i!OO ~;:s 5.2 3.48 19.7 1.65 0.203 1.73 1.88 5.30 2.52 12.5 0.20 c 21a 305 170 32 31 5.6 5.4 16.9 14.8 1.40 2.45 0.074 0.245 1.15 2.15 1.72 3.12 9.85 7.15 2.45 1.81 8.0 12.0 0.07 0.13 1--3 ~ b 235 28 5.6 4.46 15.6 1.65 0.422 1.42 2.56 7.25 1.92 5.2 0.15 {1' 22a 175 24 5.6 14..7 2.35 0.210 2.55 3.00 3.75 7.00 2.0 0.21 C::l b c 145 155 26 27 5.6 5.6 4.48 14.8 14.4 2.35 2.50 0.050 0.143 1.32 2.38 2.84 2.84 5.10 4.05 4.00 8.00 0.4 3.8 0.12 0.18 ~ ..... ~ 23a b 215 190 25 30 5.3 5.4 4.84 12.1 12.0 1.90 1.65 0.170 0.164 1.87 1.68 2.40 3.00 5.95 6.30 2.60 2.72 16.8 14.5 0.14 0.14 {1' '"'! Cr.>..... c 24a b 190 305 440 28 26 24 5.4 5.4 5.4 2.57 10.1 14.1 19.1 2.30 2.70 1.60 0.275 0.194 0.139 2.08 2.77 1.59 2.68 2.56 1.88 5.45 4.25 5.80 2.40 2.60 2.30 0.9 6.9 4.7 0.20 0.17 0.16 No <;::: 0 ....... c 25a 430 205 16 24 5.2 5.7 22.5 17.6 0.85 1.50 0.046 0.104 1.65 1.50 1.20 1.88 1.55 5.15 1.50 8.40 3.5 7.5 0.21 0.14 1--3 {1' b 26a 160 275 32 20 5.5 5.8 4.67 13.4 23.7 1.65 1.65 0.245 0.155 1.32 1.58 2.24 1.60 5.75 3.60 4.50 3.75 7.1 4.9 0.11 0.10 R p Cr.> b 405 18 6.0 2.63 19.2 0.90 0.155 0.96 1.20 4.05 3.30 0.2 0.10 c 210 23 5.3 18.0 1.60 0.129 1.68 2.00 4.40 3.00 3.6 0.18 27a 185 21 5.3 16.2 1.65 0.063 1.62 2.04 5.30 3.90 10.2 0.12 b 255 20 5.4 2.80 14.1 1.25 0.042 1.62 1.84 3.10 4.10 8.5 0.30 c 265 15 5.5 17.5 1.05 0.030 1.56 1.48 2.40 2.10 9.6 0.20 28a 480 17 6.6 20.3 0.90 0.097 0.89 1.28 4.35 1.20 2.3 0.10 b 120 21 5.6 2.70 20.6 0.85 -------­ -­---­ 1.60 8.05 -----­ 4.3 c 190 18 6.3 19.9 0.65 ·­-·---­ -----­ 1.36 6.60 -----­ 2.5 29a 230 30 5.7 22.3 1.75 0.174 1.50 2.24 7.55 2.75 4.0 0.12 b 235 30 5.6 3.49 20.5 1.40 0.210 1.19 2.00 8.50 3.30 2.0 0.12 c 215 27 6.0 18.5 1.20 0.275 1.03 1.84 10.25 2.00 2.0 0.12 TABLE Il-(Continued) Summary of Analytical Data Including the Items Showing Substantial Differences Between Overweight and Underweight Groups Sample Number Volume ml. (Specificgravity -1) x 10• pH Modified Creatinine Coefficient Pigment: Creati-nine Phos­phate mg./ml. Calcium mg./ml. Phosphorus mg.jml. Creatinine mg./ml. Chloride mg./ml. Boron µg./ml. Choline µ&" ./ml. Copper µg/ml. Over­ ~ weight: lOa b c lla b c 12a b c 13a b c 14a b c 15a b c 16a b c 240 205 270 105 115 97 325 310 213 475 430 490 195 375 160 245 170 220 350 475 365 33 31 34 32 25 28 27 23 30 16 31 28 33 25 35 25 26 34 18 10 24 5.6 5.5 5.6 5.8 5.7 5.5 5.4 5.4 5.3 5.4 5.3 5.2 5.6 5.6 5.3 4.9 5.0 4.9 5.5 5.9 5.8 3.24 4.12 2.14 1.56 2.52 2.03 1.26 13.0 10.9 12.0 12.5 8.7 9.4 15.0 12.9 11.3 22.8 16.5 18.4 21.5 22.2 13.0 12.1 13.2 11.5 17.0 12.5 14.6 3.55 3.30 3.65 5.90 4.25 3.85 2.45 1.70 2.00 0.5·5 2.05 1.05 1.80 1.05 2.20 1.80 3.65 2.25 0.70 0.80 1.15 0.166 0.111 0.180 0.093 0.147 0.2 17 0.418 0.323 0.069 0.222 0.267 0.142 0.080 0.215 0.205 0.348 0.320 0.094 0.039 -------­ 3.18 2.79 2.40 3.80 3.62 3.36 2.38 1.74 2.22 1.00 1.49 1.17 1.77 1.17 1.85 2.00 1.95 2.25 0.97 0.80 -----­ 3.48 3.04 3.00 3.84 5.32 5.52 2.40 2.48 2.56 1.14 2.40 1.36 2.60 1.48 3.84 2.24 2.12 3.12 1.24 0.64 1.64 5.20 4.75 5.85 2.90 3.00 3.40 5.40 4.45 6.40 2.85 6.55 6.60 6.50 4.85 8.75 4.30 5.00 3.40 4.55 2.65 6.10 3.50 2.52 3.00 3.00 3.55 5.20 1.81 1.88 1.85 2.90 3.90 2.00 2.48 2.20 2.40 3.70 1.80 2.50 2.90 0.72 -----­ 18.3 10.5 14.5 22.5 19.5 1.3 20.0 1.0 16.5 3.3 6.3 4.9 8.3 9.3 13.0 5.0 3.0 5.1 1.9 0.7 0.8 0.22 0.21 0.21 0.24 0.20 0.31 0.17 0.15 0.15 0.15 0.11 0.11 0.17 0.14 0.11 0.19 0.15 0.18 0.12 0.13 -----­ ~.,... ~ O' 0......... (":. '"tl ~ .,....,.... ~ ~ ~ Cf,) 0 ....... 0 O" ~ Cf,) ~ § R. t-< ~ ~ ~ Overweight Mean 277 27.0 5.45 2.41 14.3 2.37 0.230 2.10 2.64 4.93 2.69 8.9 0.17 Underweight Mean 248 24.5 5.60 3.65 17.2 1.60 0.160 1.62 2.10 5.71 3.32 6.2 0.15 ..... 00 01 There are some nineteen items for consideration in this regard, (excluding duplication due to the use of two methods for phosphate determination), four of which are statistically significant. TABLE 3 Summary of Results of Urinary Studies of Overweight and Underweight Groups in Order of Decreasing Significance Factor was higher in Over~ Under­ No. of Factor weight Group weight Group Standard Errors Odds• Modified Creatinine Coefficient x 2.76 173:1 Pigment/ Creatinine ---------------­--­----­Phosphate ------------------­-----------------­----­Calcium -------­----­---------------------­---­---­--­ X X x 2.71 2.49 2.34 148:1 78:1 52:1 Phosphorus --------­------­-------­--­-----­------Creatinine ---­-------­----------­-----­---------­--­Boron -------------------------­-------------------­--­Choline ---------------­----------­----­--­---------­--­ X X X x 2.23 1.86 1.51 1.50 38 :1 15:1 7:1 6.5:1 Chloride -----------------­-----------­--------------­Copper ---------­-----­----------------­---­--------­-X Ketonic Substances ------­------------------­X x 1.50 1.38 0.95 6.5:1 5:1 2:1 Volume -------------­----­---------------------------­ X 0.84 1.5 :1 Iron -----------­----­-------­--­---­--­---------­--­---­Specific Gravity ----------­---­---­----------­Potassium ---------------------­-------------------­Magnesium ---------------------------------------­pH -----­------­----------------­-----------------­---­--­Sodium ___ -------­----­--------­----­----------------­Total Carbohydrates ----­---­--------------­Pigment -----------­---­-----------­-------------­--- X X X X X X x 0.73 0.71 0.67 0.60 0.55 0.43 0.12 0 < < < < 1:1 1:1 1:1 1:1 1:1 1:1 1:1 *Odds against the difference between the two means being due merely to chance. Since the excretion of such substances as sodium, potassium, and magnesium ions were approximately the same in the two groups, it would appear that the members of the two groups had a similar mineral intake. It is very unlikely that anyone in this age group would store large amounts of calcium and phosphorus. Inasmuch as the underweight persons studied here excreted less of these two elements in the urine, it follows that more must have been ex­creted through the feces. Other work which is reported in this bulletin (16) indicates that different strains of rats show marked differences in urinary phosphate excretion. Genetic factors may similarly be operative in human phosphate metabolism. Speculation with respect to this situation leads to the idea that obese people, who have a large amount of fat which is in constant dynamic equilibrium, might require larger amounts of phospholi­pides for fat transport. The presence of greater amounts of phospholipides would mean that normal catabolic processes would lead to greater phospholipide breakdown, with correspondingly higher amounts of breakdown products in the blood and finally in the urine. Higher phosphate values in the urine of overweight per­sons are in line with this idea as are also the higher choline values. The elevated urinary excretion of calcium in the overweight group may be the result of the fact that calcium and phosphorus are intimately related in metabolic processes. Another possible ex­planation for the increased calcium may be in the fact th~.t calcium is involved in some obscure manner in fat metabolism, since in arteriosclerosis the deposition of calcium is preceded by a laying down of a layer of lipide material. The relative values for the modified creatinine coefficient merely support well-established findings and probably indicate that the selection of the groups was performed in a valid manner. Drabkin (17,18) observed that the output of urinary pigment is independent of the diet, but bears a relation to the level of basal oxygen con­sumption and is, therefore, a product of endogenous metabolism. The ratio of pigment to creatinine as an index of basal metabolic rate was suggested by Ostow and Philo ( 19) and studied further by Vorzimer, et al. (20). The ratio of pigment to creatinine ob­tained in the present study indicates that the overweight group tends to have a lower basal metabolic rate than the underweight group. The fact that the ketonic substances in urine were approximately the same in the two groups indicates strongly that the obese persons do not have a specific impairment in their ability to oxidize com­pletely the fat acids. Slightly elevated excretion of copper in the overweight group is of some interest in that copper deficiencies result in a deteriora­tion of the lipoidal membranes of nerve cells (21), and it is possible that the catalytic effect of some material such as copper may eventually be shown to be intimately associated with conditions of obesity. Further work is necessary in order to elucidate and estab­lish such relationships as actually exist. Aside from their bearing upon the problem of obesity, various of the data clearly demonstrate how many commonplace urinary constituents are highly individualistic. While it is not practical to consider this aspect of the experimental work at great length in this paper, it is desirable to call attention to the magnitude of the many consistent individual differences that manifest themselves. The data obtained are not sufficient to permit definite conclusions as to the etiology of obesity. However, further investigation of cer­tain aspects, particularly of phospholipide metabolism, should bear fruitful results. Summary 1. The following fourteen metabolic factors were measured and studied in the urine of a group of seven overweight and ten under­weight individuals: potassium, sodium, boron, magnesium, iron, calcium, phosphorus, copper, chloride, creatinine, pigment, total carbohydrates, ketonic substances, and choline. 2. The following items showed sufficient differences between the two groups to be suggestive of the involvement of metabolic factors in obesity: modified creatinine coefficient, pigment to creatinine ratio, urinary phosphate excretion, and urinary calcium excretion. 3. The data suggest that obese people have characteristic metab­olic differences involving calcium and phosphorus but that their ability to catabolize fats may be normal. 4. Highly characteristic individual metabolic differences were observed in the subjects studied. BIBLIOGRAPHY 1. Williams, R. J., Quart. J. Stud. Alcohol, 7, 567 (1947). 2. Berry, H.K., and Cain, L., Arch. Biochem., 24, 179 (1949). 3. Williams, R. J., Berry, L. J., and Beerstecher, E., Jr., Arch. Biochem., 23, 275 (1949). 4. Beerstecher, E., Jr., Sutton, H. E., Berry, H. K., Brown, W. D., Reed, J., Rich, G. B., Berry, L. J., and Williams, R. J., Arch. Biochem., 29, 27 (1950). 5. Wiliams, R. J., Beerstecher, E., Jr., and Berry, L. J., Lancet, 258, 287 (1950). 6. This Bulletin, p. 173. 7. Bonsnes, R. W., and Taussky, H. H., J. Biol. Chem., 158, 581 (1945). 8. Shaffer, T., Am. J. Physiol., 23, 1 (1908). 9. Fiske, C.H., and Subbarow, Y., J. Biol. Chem., 66, 375 (1925). 10. Scott, W. W., Standard Methods of Chemical Analysis, 5th ed., D. Van Nostrand Co., Inc., New York, N. Y., 1939. 272 pp. 11. Dreywood, Rowan, Ind. Eng. Chem. (anal. ed.), 18, 499 (1946). 12. Behre, J., and Benedict, D., J. Biol. Chem., 70, 487 (1926). 13. Horowitz, N. H., and Beadle, G. W., J. Biol. Chem., 150, 325 (1943). 14. Brown, Jack D., M.A. Thesis, Univ. of Texas, 1950. 15. Arkin, H., and Colton, R. R., An Outline of Statistical Methods, 4th ed., Barnes and Noble, Inc., New York, N. Y., 1943. 16. This Bulletin, p. 146. 17. Drabkin, D., J. Biol. Chem., 75, 443 (1927). 18. Drabkin, D., J . Biol. Chem., 75, 481 (1927). 19. Ostow, M., and Philo, S., Am. J . Med. Sci., 207, 507 (1944). 20. Vorzimer, J. J., Cohen, I. B., and Joskow, J. J. Lab. Clin. Med., 34, 482 (1949). 21. Campbell, A. M. G., Daniel, T., Porter, R. J., Russell, W.R., Smith, H. V., Innes, J. R. M., Brain, 70, 50 (1947). XXL Metabolic Patterns of Schizophrenic and Control Groups by M. KENDALL YOUNG, JR., HELEN KIRBY BERRY, ERNEST BEERSTECHER, JR., AND JIM S. BERRY Test Group and Controls: The eighteen male mental patients used in this study were free from any known "organic" disease, and all but one were diagnosed as suffering from dementia praecox. They ranged in age from 20 to 45 years, with the mean age being 3r years. Patients numbers 1 through 10 were described as hebe­phrenic; 11 through 14 were classified as simple type; 15 and 16 were paranoid; 17, in addition to being schizophrenic, suffered from epilepsy. Patient No. 18 was classified as manic depressive, de­pressed, and was included in this study because a number of work­ers consider this type of mental disorder biologically the same as dementia praecox. The control groups were also free from any known organic dis­ease and were from an age group similar to that of the patients. Urine and saliva samples were collected on three successive days from both groups and refrigerated until the analyses could be made. Samples obtained immediately after the subjects arose in the morn­ing were collected in order to have fairly representative samples. Analyses: Creatinine was determined colorimetrically by the standard method of using alkaline picric acid (1). The carbohydrates in both urine and saliva were determined spectrophotometrically using Dreywood's Reagent. The procedure described by Morris (2) was employed. Magnesium in urine was determined colorimetrically by the for­mation of the titan yellow-magnesium hydroxide lake (3). The method reported by Gillam (4) was employed, using 0.2 ml. of the titan yellow solution rather than the ten drops which Gillam used. The amino acids with the exception of histidine were determined on two-dimensional paper chromatograms by the method described by Polson (5). Several of these values were checked using the one-dimensional chromatographic method reported by Berry and Cain (6). Urea and uric acid were determined on one-dimensional paper chromatographs as described in other portions of this Bulletin (7). Histidine and diazonium-coupling compounds: Histidine and a number of unidentified substances present in urine were determined using one-dimensional paper chromatograms. Volumes of urine containing approximately 100 micrograms of creatinine were placed 11,4 inches apart, 1 inch from the bottom of an 11x18 inch sheet of Whatman No. 1 filter paper. The sheets were then developed in a solvent mixture consisting of 80 ml. of n-butanol, 20 ml. of glacial acetic acid, and 20 ml. of water using the ascending method de­scribed by Williams and Kirby (8). The sheets were removed from the solvent when it reached the top edge and were allowed to dry thoroughly at room temperature. Color development was achieved by the following treatment: The dried sheets were sprayed with 0.4 % diazotized sulfanilic acid solution, allowed to dry, and then sprayed with a 10 7o solution of sodium carbonate to develop color. Figure 1 represents a composite chromatogram of the substances revealed by the diazo reagent, their relative size, and their posi­tions. The concentrations, reported as spot area in square inches, may be found for histidine in Table I and for the unknown dia­zoni um-coupling compounds in Table II. Results and calculations: The data for the patients and the con­trol groups are reported as the average of three or more samples per individual in Tables I and II. All the data, with the exception of salivary glucose and creatinine which are reported on a milli­gram per milliliter basis, are based on creatinine excretion. Crea­tinine was chosen as a basis because for healthy subjects it is the most constant excretory product for a given individual from day to day (9) and its use as a basis compensates for concentration differ­ences in the urine samples. The validity of using this substance as a basis for calculation in the schizophrenic group is discussed below. The standard deviation, ...J < 0.5 > ... a: 0.4 0.3 0.2 0.1 • -• • • • • • • • • • • • • • • • • • • • • I ••• • •• • .. • • " • • • • • • • • .. - 6 2 3 4 5 6 7 8 9 10 11 i2. 13 14 15 16 2 3 4 5 6 PATIENT NO. .CONTROL NO. FIGURE I. Diagram of Chromatogram of Urine of Mental Patients and Control Group. ~ Cl> No ~ O" 0 - <:'"> ""· "ti ~ No No Cl> ?J "' 0 ........ VJ <:'"> ;s- N""· 0 ~ ;s­ -;: Cl> ;::S ""· <:'"> "' ...... ~ ...... (Solvent: Butanol-acetic acid-water. Developing reagent: Diazotized sulfanilic acid). TABLE 1 " ~ . ~ ~ 0 Po ti) 8 ..: 0 ., bi .s~ ~~ 88 ..: 0 "tO.s 8... ," . " bOE-< s ..: 0 ., tO .s 8 "' .!!!bi <8 ..: .,o ~~ "'­.... a~ ..: 0 tO., 8 ~~ >S ., ..: .s .,I:.? " ... bO.$? ';( 8 ..co--.. ~:a ~ :il .. 8 0 bi .s; ... 0 ·~ +> -.._.... ·~":i:: .s 8..:,,o ·~ tOi!l 5"--.. ~ tO :il 8 ~v.....: ; i!l 8 &g~ ~68 1 2.52 1.026 15 0.39 0.28 0.002 0.030 0.040 0.016 0.016 0.050 0.010 vis 0.43 0.065 2.70 2 2.66 1.027 14 0.37 0.40 0.000 0.034 0.034 0.015 0.019 0.049 0.004 0.000 0.40 0.072 0.34 3 0.59 1.010 12 0.37 0.000 0.050 0.050 0.000 0.020 0.100 0.000 0.000 0.00 0.140 0.27 4 1.62 1.018 15 0.28 0.28 0.000 0.050 0.060 0.020 0.030 0.060 0.000 0.000 0.34 0.084 0.00 5 3.10 1.037 16 0.52 0.42 0.005 0.026 0.023 0.006 0.012 0.020 0.013 vis 0.19 0.077 1.65 6 3.96 1.035 13 0.35 0.44 0.003 0.020 0.030 0.008 0.070 0.023 0.018 vis 0.28 0.053 0.76 7 1.89 1.026 14 0.16 0.33 0.002 0.030 0.040 0.013 0.020 0.025 0.004 0.002 0.38 0.066 1.62 8 1.93 1.021 19 0.36 0.39 0.005 0.026 0.036 0.000 0.030 0.040 0.000 vis 0.15 0.075 0.54 9 0.72 1.007 10 0.28 0.50 0.000 0.006 0.018 0.000 0.000 0.000 0.000 0.000 0.21 0.120 0.93 10 1.16 1.012 20 0.23 0.33 0.006 0.018 0.051 0.045 0.010 0.018 0.026 0.030 0.30 0.107 0.25 11 0.62 1.009 16 0.19 0.84 0.012 0.036 0.048 0.000 0.012 0.000 0.000 0.000 0.08 0.166 0.52 12 1.10 1.015 13 0.18 0.29 0.011 0.018 0.075 0.036 0.008 0.019 0.012 0.011 0.38 0.118 0.25 13 3.29 1.028 12 0.15 0.22 0.005 0.020 0.028 0.012 0.017 0.049 0.010 vis 0.44 0.070 0.32 14 3.84 1.031 13 0.26 0.32 0.021 0.019 0.045 0.022 0.019 0.033 0.008 0.012 0.28 0.041 15 2.56 1.026 11 0.39 0.52 0.021 0.050 0.050 0.012 0.006 0.000 0.008 0.012 0.23 0.094 Q.36 16 1.64 1.026 22 0.50 0.55 0.008 0.060 0.060 0.012 0.000 0.000 0.000 0.015 0.22 0.092 0.00 17 1.35 1.009 13 0.15 0.59 0.015 0.014 0.032 0.035 0.009 0.022 0.000 0.000 18 1.31 1.017 23 0.21 0.56 0.013 0.024 0.054 0.037 0.021 0.054 0.020 0.000 Control No. 1 1.72 1.011 0.34 0.003 0.008 0.018 0.011 0.006 0.015 0.002 0.001 0.36 0.037 0.14 2 2.98 1.026 0.23 0.000 0.018 0.027 0.011 0.006 0.015 0.000 0.005 0.35 0.039 0.21 3 1.94 1.025 0.40 0.000 0.014 0.014 0.000 0.009 0.029 0.000 0.008 0.49 0.083 0.18 4 1.66 1.008 0.12 0.000 0.000 o.U08 0.000 0.007 0.010 0.000 0.000 0.32 0.025 0.22 5 3.78 0.20 0.000 0.034 0.49 6 1.47 1.015 0.33 0.003 0.000 0.005 0.000 0,013 0.034 0.000 0.000 0.53 0.066 0.33 7 3.57 0.20 0.000 0.012 0.022 0.032 0.014 0.035 0.000 0.000 0.031 0.42 8 1.79 0.43 0.000 0.005 0.015 0.006 0.006 0.019 0.000 0.008 0.095 0.33 9 2.26 0.35 0.000 0.000 0.006 0.000 0.005 0.013 0.000 0.000 0.050 10 2.12 0.32 0.004 O.o18 0.018 0.000 0.014 0.040 0.000 0.000 0.078 11 2.56 0.44 0.044 Aver. Patients 1.99 1.016 15 0.29 0.42 0.007 0.030 0.043 0.016 O.o18 0.031 0.007 0.004 0.27 0.090 0.70 Controls 2.35 ----­--- ---­ -----­ 0.30 0.001 0.008 0.015 0.007 0.009 0.023 0.000 0.002 0.41 0.053 0.29 Stand. Dev. Patients 1.07 -------­ ---­ -----­ 0.15 0.007 0.016 0.015 0.014 0.016 0.026 -­---­-­ 0.008 0.13 0.033 0.75 Controls 0.78 ---­ -­---­ 0.10 0.002 0.008 0.007 0.011 0.004 0.011 -­-----­ 0.005 0.09 0.024 0:12 uPatients/uControls 2.2 -­-----­ ---­ ---­-- 1.5 3.5 2.0 2.1 1.3 4.0 2.4 -­-----­ 1.6 1.4 1.4 6.3 P= 0.66 -­-----­ ---­ -----­ 0.34 0.99 0.99 0.99 0.91 0.90 0.70 -----­-­ 0.96 0.96 0.99 0.85 Discussion The data which appear in the tables, when analyzed statistically, show some interesting and significant differences between the two groups studied. If the arbitrary assumption is made that a P value equal to or greater than 0.90 is statistically significant, it is notable that the excretion values for magnesium, a number of the diazo­nium-coupling compounds and all the amino acids except citrulline, possess significance. Several other constituents approach this value of 0.90 and probably are also significant. TABLE II Imidazole Derivatives and Related Compounds Area (in2 ) /100 g. Creatinine Patient No. Rf= .25 .32 .60 .65 .85 .90 .92 1 0.00 0.00 0.09 0.00 0.38 0.16 0.00 2 0.00 0.20 0.12 0.00 0.35 0.15 0.00 3 0.00 0.04 0.00 0.00 0.00 0.00 0.00 4 0.00 0.00 0.19 0.00 0.11 0.10 0.00 5 0.12 0.15 0.08 0.12 0.32 0.14 0.00 6 0.00 0.00 0.12 0.13 0.51 0.14 0.00 7 0.00 0.04 0.07 0.13 0.18 0.16 0.00 8 0.34 0.00 0.00 0.12 0.25 0.11 0.00 9 0.00 0.00 0.26 0.11 0.21 0.00 0.26 10 0.16 0.12 0.26 0.23 0.83 0.32 0.00 11 0.00 0.00 0.25 0.03 0.44 0.32 0.00 12 0.15 0.00 0.00 0.07 0.59 0.00 0.25 13 0.15 0.22 0.16 0.16 0.41 0.30 0.00 14 0.17 0.00 0.15 0.13 0.11 0.10 0.00 15 0.00 0.00 0.18 0.00 0.44 0.21 0.00 16 0.00 0.00 0.28 0.05 0.20 0.17 0.00 17 18 Control No. 1 0.15 0.00 0.00 0.13 0.18 0.00 0.00 2 0.00 0.00 0.00 0.08 0.00 0.34 0.00 3 0.14 0.00 0.00 0.00 0.37 0.10 0.00 4 0.23 0.00 0.00 0.00 0.00 0.00 0.00 5 0.20 0.00 0.00 0.00 0.00 0.00 0.00 6 0.24 0.00 0.00 0.00 0.00 0.00 0.00 Aver. Patients 0.07 0.05 0.14 0.08 0.33 0.15 0.0-0 Controls P = 0.16 0.93 0.00 0.00 0.03 0.76 0.09 0.98 0.07 0.82 0.00 The lowered creatinine average for mental patients and the P value of 0.66 raise the question of the validity of using creatinine as a basis for correcting for urine concentration. An inspection of the standard deviation values demonstrates that for creatinine ex­cretion, just as for the excretion of the other substances measured, there is a much greater day to day variation in the schizophrenic group than in the controls. However, if the difference in the two groups is compensated by multiplying the individual control values by the ratio 2.35/ 1.99 and recalculating the P values, it is found that the new values are not markedly different from those given in the table; so the apparent invalidity of the creatinine basis can be discounted. A survey of the literature pertaining to the biochemical aspects of the schizophrenic syndrome discloses a confusing and contra­dictory picture. Bellak (11), in his exhaustive survey of the sub­ject, has even stated that "its only consistent difference from the normal lies in the greater variability of values for almost any factor investigated." The increased variability is amply confirmed in the data presented in Table I ; the ratios of standard deviations of mental patients to controls range from 1.3 to 6.3 with not a single value less than unity. The study of urinary magnesium excretion among mental pa­tients has not been previously reported, but the P value and the comparative group averages show that excretion of this ion is almost certainly considerably higher among schizophrenics than in the normal population. This argues for altered magnesium utiliza­tion, and this theory offers several attractive speculative approaches to the problem of its relation to mental disease. The ratio of cal­cium to magnesium ions has long been known to have an intimate connection with tissue irritability; an increase in this ratio is associated with a rise in irritability. In addition, the deprivation of magnesium gives rise to vasodilation and hyperexcitability of the nervous system in rats, similar in symptomology to tetany in cal­cium deficient animals (12). Recent work has indicated that mag­nesium is required as an activator for enolase (13) and carboxyl­ase (14) and is thus implicated in the metabolism of carbohydrates. Considerable attention has been devoted to anomalies in carbohy­drate metabolism in dementia praecox (15), the principal inade­quacy noted being delayed metabolism of blood sugar following intravenous glucose injection (16). The decreased action of the enzymes of the carbohydrate cycle due to magnesium depletion could well contribute to this effect. The increased excretion of this important constituent suggests further studies, especially with re­gard to the relation of blood levels to urinary levels, to determine its possible connection to the etiology of dementia praecox. With regard to the possible relation of the urinary amino acids to schizophrenia, the foundation is more tenuous. With the excep­tion of histidine, all are excreted in markedly higher amounts. Un­published studies made at this laboratory have indicated that high levels of urinary excretion of these amino acids are found in normal women. This finding, coupled with the report by Hoskins and Pincus ( 17) that the androgen to estrogen ratios in male schizo­phrenics are very markedly depressed from normal values, argues that lowered androgen production associated with testicular dys­function (18) in dementia praecox might be correlated with the observed increased amino acid excretion. Histidine excretion, on the other hand, is somewhat lower for the group of mental patients. This could be associated with the increased excretion of unusual substances such as some of the diazonium-coupling compounds or others not detected. Alternatively, it could be related to the ex­istence of the higher histamine blood levels reported by Strengers and Gooszen ( 19) . The nature and significance of the diazonium-coupling compounds are totally unknown. However, the very fact that some of these substances were found solely in the samples from the mental patients indicates that they are potentially important and certainly interesting in the study of dementia praecox. Roessler and Hanke (20, 21) indicate that among the urinary constituents which react with diazotized sulfanilic acid are histidine, imidazole propionic acid, imidazole lactic acid, and imidazole acetic acid. These authors mention a study of the diazo reaction in the urine of normal and pathological cases in which they found an increased excretion of imidazole derivatives in pathological states. The need for further study is most apparent here; until the identity of these compounds can be determined, no speculation is feasible concerning their re­lation to the schizophrenic state, despite the striking contrast ex­hibited by the two groups. The remaining substances reported in the table show only slight differences between the two groups, and the P values are low. Sali­vary gucose alone seems to have some significance, the nature of which is not known. It might be expected, in view of the disturb­ances in carbohydrate metabolism which have been reported, that the urinary excretion might reflect these differences, but aside from a slightly higher excretion, which is not significant, and a greater variability among the patients, the difference is slight. The data presented support the thesis that dementia praecox is essentially a disease of physiological deficiency and demonstrate the existence of a biochemical pattern strikingly different from that of normal controls. More than all, this work has shown the need for additional concentrated and systematized study, on a sound scientific basis, of the problem of mental disease. The aid of power­ful, facile tools has made this study practicable; the magnitude of the problem, both from a scientific and from a sociological view­point, makes it necessary. BIBLIOGRAPHY 1. Klett-Summerson Clinical Manual, Klett-Summerson Mfg. Co., New York, N. Y., 2. Morris, D. L., Science, 107, 254 (1948). 3. Ginsberg, H. Z., Electrochem., 45, 829 (1940). 4. Gillam, W. S., Ind. Eng. Chem. (anal. ed.), 13, 499 (1941). 5. Polson, A., Biochem. et Biophys. Acta, 2, 575 (1948). 6. Berry, H.K., and Cain, L., Arch. Biochem., 24, 179 (1949). 7. This Bulletin, pp. 88 and 84. 8. Williams, R. J., and Kirby, H. M., Science, 107, 481 (1948) . 9. Personal Communication, Dr. John W. Tukey. 10. Snedecor, G. W., Statistical Methods, Collegiate Press Inc., Ames, Iowa, 1937, p. 83. 11. Bellak, L., Dementia Praecox, Grune and Stratton, New York, 1948, p. 27. 12. Tufts, E. V., and Greenberg, D. M., J. Biol. Chem., 122, 693 (1937). 13. Warburg, 0., and Christian, W., Biochem. Z., 310, 384 (1942) . 14. Green, D. E., Westerfeld, W. W., Vennesland, B., and Knox, W. E., J. Biol. Chem., 140, 683 (1941). 15. Bellak, Dementi.a Praecox, Gruna and Stratton, New York, 1948, p. 102. H~. Freeman, ·H., Rodnick, E. H., Shakow, D., and Lebeaux, T., Psychosomat. Med., 6, 311 (1944). 17. Hoskins, R. G., and Pincus, G., Psychosomat. Med., 11, 102 (1949). 18. Hemphill, R. E., Reiss, M., and Taylor, A. L., J. Mental Sci., 90, 681 (1944). 19. Strengers, T., and Gooszen, J. A. H., Ence'Phale, 36, 283 (1946-7). 20. Koessler, K. K., and Hanke, M. T., J. Biol. Chem., 39, 497 (1919). 21. Koessler, K. K., and Hanke, M. T., J. Biol. Chem. 59, 803 (1924) . XXII. Exploration of Metabolic Patterns in Mentally Deficient Children by LOUISE CAIN Significant differences in metabolic patterns have been found in this laboratory to exist when alcoholics were compared with non­alcoholics (1), when obese individuals were compared with lean individuals (2) and when mental patients were compared with well individuals (3). The study here presented was made to ex­plore the possible significant differences between the patterns of mentally deficient children and those with normal faculties. This particular investigation is limited, however, to a study of the uri­nary excretion patterns. Thirteen mentally deficient children, 7 to 11 years of age, from the Austin State School were chosen as subjects. These were se­lected by the medical staff to be as free as possible from physical defects. For comparison, ten control children, ages 5 to 11, were studied in a similar manner. Three morning urine samples were obtained from each individual child. The mentally deficient children were placed in the hospital ward overnight to insure proper urine collection. Amino acid determinations were made according to the methods of Berry and Cain ( 4). Three two-dimensional sheets were run on each individual urine sample, using aliquots of urine of 20 p.l., 40 p.l., and the amount containing 50 p.g. of creatinine. When the amino acid concentrations were low enough so that they were not detected on the sheets involving the smaller aliquots, the values obtained on the sheets involving larger aliquots were used for computation. Otherwise the values obtained from the three sheets were averaged. Most of the amino acids were determined by direct comparison with standards employing known amounts of the amino acids in question. Phenylalanine, however, was determined on the basis of its detectability at 5 p.g. This amount is necessary to make a clearly defined spot when, as under ordinary conditions, leucine is also present and migrates to the same position. Proline and methionine also were not included in the standards. Proline was evaluated on the basis of its detectability at 7 p.g. and methionine by comparison with alanine standards. An unidentified ninhydrin­positive blue spot (Rf value in phenol .21, in lutidine .78) was determined only on a relative basis, but values were assigned to it on the arbitrary assumption that it is detectable at 1 µ,g. Histidine was determined by the method previously described in this bulletin (p. 77). Creatinine was determined colorimetrically in all samples by the standard picrate method (5), phosphorus by the colorimetric method of Fiske and Subbarow (6) and carbohydrates by the color­imetric method employing Dreywood's anthrone reagent (7). Creatine, uric acid and urea were determined by paper chroma­tography in accordance with methods described in this bulletin (8,9,10) except that in the case of urea a modification was made. Instead of using sodium hypochlorite as the color developing agent, the sheet after resolution with phenol solvent was thoroughly dried and then developed with 1'fo p-dimethylaminobenzaldehyde in 10% HCI-alcohol (p. 30). TABLE I Additional Urine Chromatograms Aliquot of urine Color developing Substances detected 1. = 100 µ,g. creatinine agent diazotized sulfanilic acid Unidentified spots: Rf .63 ­yellow Rf .65-pink 2. = 3. = 60 µ,g. creatinine 40 µg. creatinine ferric chloride p-dimethylaminobenzaldehyde Rf .80 ­orange-yellow Rf .89 ­purple Phosphate Unidentified spots: Rf .30 ­blue-gray Rf .62-blue-gray Unidentified spots: Rf .36 ­purple Rf .51-pink 4.= 40 µ,g. creatinine 5.= 40 µ,g. creatinine 6._ 40 µ,g. creatinine ferricyanide-nitroprusside reagent aniline acid phthalate 2,6-dichloroquinone chloroimide Rf .75 ­purple No arginine detectable No spots other than those already detected No spots corresponding to sugars found urea uric acid creatinine 7.= 40 µ,g. creatinine bromcresol green lactic acid* hippuric acid* tartaric acid* 8._ 40 µg. creatinine bromine Unidentified spots: Rf .52 ­fluorescent 9.:= .50 µ,g. creatinine . picric acid Rf .10 ­Rf .30 ­ creatine creatinine Un.dentified spots: Rf .48 ­Rf .35 ­ gray-greent yellow *These were determined by measuring the area of the spots using a polar planimeter. tThis unidentified spot was evaluated on the basis of color intensity and was rated from O to 4+. !\:) 0 TABLE II Data with Respect to Items which Appeared to Show Differences c Łi~ .B·u~ ~c;:a ;:ii Ao .,.. < K., tll : a r ., ..; "~o ..... Cl>.""" ..• S·;; 8 ~.Ł'-:.,_ .. :a~ 8 ..; 0., ."..·:;; s o -...... . >. .. E-<8 ..; 0.,.­"8 ~~ P..8 >. ;; k oS .... 0 ~ ~~ ~ .se ~~.~~ Ci ~.e 8 ";? k ~..... o 0 ... "'""'""'» E w~--;. ~ .... bO ~Ea ..; 0., ."..;; 8 ~~ :r: s ..; 0 ., .. ·= s.. ....." ..,tio E-< E ..; 0 "'~E ·­'­ 7 8 9 10 7 10 11 11 F F F M 70 62 42 22 0.043 0.028 0.136 0.070 0.021 0.007 0.014 0.047 0.061 0.149 0.126 0.058 0.33 0.33 1.00 1.33 0.007 0.000 0.012 0.008 0.121 0.155 0.048 0.126 0.036 0.056 0.022 0.070 0.007 0.010 0.019 0.033 0.031 0.060 0.086 0.040 c::::;::s.... c::: (':> 11 9 M 52 0.023 0.000 0.058 2.00 0.007 0.060 0.000 0.020 0.102 ~ .... 12 13 Ranges Controls 10 8 M M 68 23 0.058 0.101 0.023­0.226 0.047 0.064 0.000­0.083 0.000 0.120 0.000­0.253 1.33 3.33 0.33­3.33 0.016 0.035 0.000­0.035 0.184 0.167 0.06­0.355 0.023 0.000 0.000­0.077 0.029 0.068 0.007­0.14 0.072 0.173 0.031­0.173 <:-;.. «::'! c-"'-3 (':> 14 15 9 9 F F 125 119 0.055 0.024 0.000 0.025 0.000 0.052 0.00 0.33 0.012 0.007 0.023 0.153 0.037 0.060 0.033 0.014 0.000 0.099 ~ ~ <:I> 16 9 F 84 0.053 0.019 0.187 0.00 0.007 0.181 0.060 0.056 0.086 17 18 8 5 F F 108• 0.053 0.040 0.000 0.000 0.000 0.073 1.00 0.66 0.000 0.008 0.152 0.057 0.038 0.006 0.047 0.000 0.142 0.000 19 7 M 93 0.046 0.036 0.147 0.00 0.023 0.155 0.040 0.000 0.114 20 10 M 99 0.030 0.039 0.000 1.33 0.000 0.050 0.030 0.030 0.035 21 7 M 107 0.032 0.022 0.000 0.66 0.000 0.115 0.052 0.030 0.000 22 9 M 93 0.052 0.014 0.000 0.66 0.010 0.161 0.038 0.010 0.079 23 11 M 103 0.042 0.016 0.000 0.00 0.007 0.087 0.065 0.041 0.036 Ranges 0.024­0.055 0.000­0.039 0.000­0.187 0.00­1.33 0.000­0.023 0.023­0.181 0.006­0.065 0.000­0.056 0.000­0.142 •Preschool age--not determined. tSee footnote p. 7 5. Seven additional one dimensional chromatograms (nos. 1-7, Table I) were run on each sample using butanol-acetic acid solvent; one additional chromatogram (No. 8, Table I) was made on each sample using isobutyric acid and one (No. 9, Table I) using butanol­ethanol. The aliquots of urine used, the color developing agents employed and the substances determined in these nine sets of chromatograms are listed in Table I. TABLE III Statistical Analysis of Differences between Control and Mentally Deficient Groups Average No.of Excretion Standard Standard Standard Odds* mg./mg. Cr. deviation error errors Methionine sulfoxide Controls ______________ 0.042 0.024 Mentallydeficient __ _______ 0.087 0.068 0.011 3.9 10,390 to 1 TyrosineControls _____________ 0.017 0.018 Mentallydeficient __________ 0.035 0.029 0.005 3.6 3,142 to 1 Proline Controls ______________ 0.040 0.077 Mentally 0.023 3.1 515.7 to 1 deficient __ __ _____ 0.113 0.112 "Green-Gray Spot" Rf .48 Controls --------------0.45 0.80 Mentally 0.23 2.6 106.3to1 deficient __________ 1.15 1.10 "Blue Spot" (ninhydrin) Controls --------------0.008 0.011 Mentally 0.003 1.9 16.4 to 1 deficient _ ________ 0.014 0.016 Histidine Controls ______________ 0.112 0.084 Mentally 0.021 1.9 16.4 to 1 deficient __________ 0.152 0.092 Taurine Controls -------------0.043 0.030 Mentally 0.007 1.37 5.19 to 1 deficient _________ 0.033 0.033 Lysine Controls _____________ 0.027 0.026 Mentally 0.008 1.4 5.19 to 1 deficient __________ 0.033 0.039 Phenyl alanine Controls _____________ 0.059 0.096 Mentally 0.021 .89 1.12 to 1 deficient ----------0.078 0,075 •Odds against the differences between the two averages being due merely to chance, In Table II are summarized the data with respect to all the items on all the chromatograms for which by inspection it appeared that there might be significant differences between the mentally deficient and the control groups. It may be noted that glycine, serine, alanine, citrulline, valine, leucine, aspartic acid, glutamic acid, threonine, methionine, cysteic acid, lactic acid, hippuric acid, tartaric acid, phosphate, carbohydates, uric acid, urea, creatinine and creatine, all of which were determined, are not listed in Table II, because in these cases the differences, if present, were small. The same statement holds for most of the unidentified spots. In Table III is summarized the statistical analysis of the items included in Table II. It will be noted that for four items there appear to be statistically valid differences between the controls and the mentally deficient children according to conventional standards for biological work. For the two other items the differences are such as to indicate that they may be significant. TABLE IV Statistical Analysis of Differences Between Morons and Imbeciles Average No.of Excretion Standard Standard Standard Odds* mg./mg. Cr. deviation error errors Methionine sulfoxide Imbeciles --------0.115 0.068 0.016 5.2 1,744,000 to 1Morons ___________ 0.033 0.028 Carbohydrates Imbeciles __ _____ 2.20 0.889 Morons _____ ______ 1.10 0.328 0.212 5.2 1,744,000to1 Lysine Imbeciles _____ __ 0.052 0.029 0.008 3.8 Morons ____________ 0.020 0.024 6,915 to 1 Creatine Imbeciles --------0.37 0.235 Morons _______ ____ 0.18 0.149 0.062 3.1 515 to 1 Tyrosine Imbeciles _______ 0.046 0.040 0.011 2.5 79.5 to 1Morons------------0.020 0.023 Threonine Imbeciles --------0.034 0.031 Morons _________ __ 0.022 0.028 0.011 1.8 12.9 to 1 Blue Spot Imbeciles ________ 0.017 0.016 Morons ____________ 0.009 0.014 0.005 1.7 10.2 to 1 Proline Imbeciles ___ _ 0.131 0.115 0.033 1.38 5.2 to 1Morons ------------0.085 0.093 Phenyl alanine Imbeciles ________ 0.087 0.076 Morons ___ __ ____ 0.062 0.070 0.024 1.1 2.7 to 1 Glutamic acid Imbeciles --------0.006 0.007 0.003 .so 1.4 to 1Morons ------------0.004 0.008 •Odds· a&'a.inst the'difference··between the two averages being due merely t<> ebance; Metabolic Patterns of Mentally Deficient Children 203 In the course of studying the data it appeared that the patterns of the morons and imbeciles differed significantly from each other. Iri Table IV is summarized the statistical analysis of the data with respect to ten items for which differences appeared probable. It will be noted that there are five significantly different items; in the majority of these the differences appear highly significant. In the data assembled in both Tables III and IV the interesting fact should be noted that in general the greater excretion is asso­ciated with a higher degree of mental deficiency. In Table III eight of the nine items (taurine is the exception) are excreted in larger amounts by the mentally deficient than by the control children. For all of the significantly different items, the excretion by the mentally deficient children averages more than double that of the control children. In Table IV every one of the ten items is excreted in larger amount by the imbeciles than by the morons. An interesting observation has to do with the carbohydrate ex­cretion by controls, by morons and by imbeciles. Curiously carbohy­drate excretion by the control group gives an average intermediate between that of the imbecile and moron groups. The probability of a significant difference between the imbeciles and morons is nearly 2 million to 1. Because the morons have lower excretion than the controls, however, the probability of a significant difference be­tween the controls and the mentally deficient group as a whole is only 10 to 1. The findings with respect to carbohydrate excretion are summarized in Table V. TABLE V Statistical Analysis of Carbohydrate Excretion Average Excretion mg./mg. Cr. Standard deviation Standard error No. of Standard errors Odds* Imbeciles ____________ Morons -----­---------­Controls ----­---­----­Imbeciles -----------­Controls --­----------­Morons ---------------­ 2.20 1.10 1.46 2.20 1.46 1.10 0.889 0.328 0.664 0.889 0.664 0.328 0.212 0.218 0.164 5.2 3.4 2.2 1,744,000 to 1 1,483 to 1 35 to 1 Controls ______________ 1.46 0.664 Mentally 0.173 1.7 10.2 to 11.77 0.775 deficient ---------­ *Odds against the difference between the two averages being due merely to chance. A suggestive fact is that from inspecting the data with respect to the excretive patterns of the control individuals it was noted the pattern of one individual, No. 16, was different from the others in the control group and in some respects resembled the patterns of the individuals in the mentally deficient group. This observation was made before the "I.Q." values for the control children had been collected ; in fact this observation led us to collect these additional data which are included in Table II. It was found that this child not only had the lowest "I.Q." of any in the control group but was considered to be a special problem. In spite of relatively affluent surroundings, she had demonstrated antisocial behavior with re­spect to property rights. In Table VI are presented, partly because of this special case and partly because of other data, a summary of the excretions for twenty-nine substances for which quantitative data are available. It may be noted that the imbecile group excreted more than the modified control group in 24 cases and less in 5. The moron group TABLE VI Summary of Excretion Averages Control Av. Imbecile Av. Moron Av. No. 16 Av. (Omitting 16) "l.Q." 95-'-125 "1.Q." 22-42 "l.Q." 51-70 "l.Q." 84 mg./mg. Cr. mg./mg. Cr. mg./mg. Cr. mg./mg. Cr. Alanine ___________________ 0.040 0.048 0.056 0.048 Aspartic acid _______ __ 0.005 0.006 0.006 0.015 Blue spot(Ninhydrin) ________ 0.008 0.017 0.009 0.007 Citrulline -----­---­--------Cysteic acid ______________ 0.046 0.008 0.054 0.006 0.048 0.006 0.049 0.027 Glycine ----------------------Glutamic acid __ ___ __ 0.074 0.003 0.082 0.006 0.080 0.004 O.o76 0.000 Histidine __________________ 0.101 0.147 0.141 0.181 Leucine ________________ ____ 0.017 0.024 0.023 0.040 Lysine .---------------------0.024 0.052 0.020 0.056 Methionine -------------0.013 0.006 0.007 0.008 Methionine sulfoxide ___________ 0.040 0.115 0.033 0.053 Phenyl.alanine ________ 0.056 0.087 0.062 0.086 Proline ----------------------0.021 0.131 0.085 0.187 Serine ------------------------0.067 0.070 0.057 0.076 Taurine --------------------0.041 0.035 0.030 0.060 Threonine ________________ 0.031 0.034 0.022 0.060 Tyrosine ------------------0.017 0.046 0.020 0.019 Valine --------------------0.018 0.023 0.023 0.040 Carbohydrates ________ 1.300 2.200 1.100 2.600 Creatine __________________ 0.31 0.37 0.18 0.26 Creatinine* --------------1.07 1.12 1.03 1.27 Green spott ____________ 0.52 1.20 1.27 0.000 Hippuric acidtt ______ 0.15 0.28 0.19 0.31 Lactic acidtt ----------0.17 0.24 0.18 0.27 Phosphorus ------------0.777 0.800 0.740 1.033 Tartaric --------------.-----0.25 0.31 0.22 0.32 Urea _________________________15.9 15.1 15.9 14.1 Uric acid _________________ 0.214 0.210 0.213 0.179 --;c;;atinine values represent milligrams per milliliter. tGreen spot values determined by color intensity. ttHippuric acid and lactic acid values represent spot areas measured in square iilches. averaged more in 16 cases, less in 12 and the same in 1. Individual No. 16 averaged more in 23 cases and less in 6. These data are highly suggestive of the desirability of collecting this type of in­formation with respect to human individuals, and studying its re­lationship to problems of human mentality and behavior. It should be obvious that such studies will probably not yield results that will be readily interpretable in a simple manner. Summary An exploratory study of the urinary excretion patterns of 13 mentally deficient and 10 control children have yielded results of a highly suggestive nature. It appears on the basis of the investiga­tion up to this point that moron, imbecile, and control groups each exhibit excretion patterns which are distinctive in several respects. The findings also suggest the desirability of further investigation of these individual patterns in the case of so-called normal in­dividuals, because of the probability that the findings may throw light on mental and behavior problems. BIBLIOGRAPHY 1. Beerstecher, E., Jr., !Sutton, H. E., Berry, H. K., Brown, W. D. Reed,, J., Rich, G. B., Berry, L. J., and Williams, R. J., Arch. Biochem., 29, 27 (1950). 2. This Bulletin, p. 181. 3. This Bulletin, p. 189. 4. Herry, H.K., and Cain, L., Arch, Biochem., 38, 179 (1949). 5. Folin, 0., and Wu, H., J. Biol., 38, 81 (1919). 6. Fiske, C. H., and Subbarow, Y., J. Biol. Cham., 66, 375 (1925) .. 7. Dreywood, R., Ind. Eng. Chem. (anal. ed.), 18, 499 (1946). 8. This Bulletin, p. 82. 9. This Bulletin, p. 84. 10. This Bulletin, p. 88.