methods, and materials used have been the same as those previously described.

AMINO ACIDS IN THE NUTRITION OF EXCISED TOMATO ROOTS PHILIP R. WHITE (WITH FIVE FIGURES) Introduction A preliminary study of the growth-promoting materials obtainable from yeast and essential for the nutrition of excised tomato roots (10) has shown that all of these materials are soluble in 85 per cent. ethyl alcohol, insoluble in ether, and stable in boiling trichloroacetic acid. They can be further divided into two fractions by extracting with absolute alcohol. Both fractions are essential for satisfactory growth. The present paper deals with results of an analysis of that fraction which is soluble in 85 per cent. alcohol and insoluble in 100 per cent. alcohol and ether. The basic nutrients, cultural methods, and materials used have been the same as those previously described. As was stated in the earlier report, the material soluble in 85 per cent. alcohol contains, in addition to many unknown substances, a very considerable amount of free amino acid. OSBORNE and WAKEMAN (7) give the total N content of a similarly prepared fraction, soluble in 80 per cent. alcohol, as 6.5 per cent. of the total yeast N or 12 per cent. of the weight of the fraction. They do not report the amino nitrogen content. About 3.5 per cent. of the total weight of this fraction, or 29 per cent. of its total N, has been found by the present writer to be amino nitrogen (van Slyke method). Between 20 per cent. and 25 per cent. of the fraction is, therefore, amino acid. As was suggested in the earlier paper, this high amino acid content obviously suggests that this constituent may play an important part in bringing about the growth-promoting activity of the fraction. As a check on this possibility an aqueous extract of yeast was shaken for one-half hour in an atmosphere of NO2 at ph 3.0, adjusted to ph 5.6, and its growth-promoting properties tested. The effectiveness was found to be completely destroyed. One of the primary effects of this treatment is deaminiation, although such treatment is too drastic to furnish an unequivocal diagnosis. This observation supports the presumption that the amino acids might be important factors in producing the effectiveness of the fraction. It seemed desirable, therefore, to determine in some detail the effects of commonly occurring amino acids on growth of these roots. Experimentation EFFECTS OF KNOWN AMINO ACIDS ON GROWTH In a preliminary survey, 19 nutrients were made up, each one containing ing the usual salts and carbohydrate but with a single amino acid taking the 793

794 PLANT PHYSIOLOGY place of accessory organic material. Amino acids were obtained from two sources, as follows: From Dr. C. S. MARVEL, University of Illinois, 1-cystine, 1-tyrosine, dl-valine, dl-isoleucine, dl-norleucine, dl-methionine, dl-phenylalanine, dl-lysine-2hcl. From Pfanstiehl Chemical Co., Waukegan, Illinois, glycine, d-arginine, I-leucine, 1-tryptophane, I-proline, 1-hydroxyproline, l-histidine-hcl, 1-aspartic acid, d-glutamic acid, dl-serine, dl-alanine. Hydroxy-glutamic acid, citrulline, and ROSE's a-amino-f3-hydroxy-n-butyric acid (1) were not available for study. A concentration of 5 mg. of amino acid per liter of nutrient (5 p.p.m.) was first tested. This was slightly less than the total weight of amino acids (8 mg.) shown by MEISENHEIMER to be present (2) in what has been determined to be the optimal concentration of yeast (10). The results are shown in figure 1. At a concentration of 5 mg. per liter, 7 of the 19 amino acids 70- Stimulating 60_ 50-6S0 Effect doubtful Less than 30% more or less F. than the control. 40- bi cc U J30.34 ouc 4c < 20- >?. Depressant 10-4)4) )4QC,C FIG. 1. Relative effectiveness of single amino-acids when tested at a concentration of 5 ng. per liter, in the absence of all other accessory materials. Those marked with an asterisk are considered by Roeg (9) to be essential for growth of rates. studied proved to be depressant, and only 3-glycine, lysine, and glutamic -acid-were definitely stimulating. Under these conditions glycine gave results almost equal to the control. Optimal concentrations were then determined individually for each acid. For this purpose, each was- tested at concentrations of 50, 15, 5, 1.5, 0.5, 0.15, and 0 mg. per liter of nutrient. Isoleucine, norleucine, methionine, serine, and arginine, which gave regularly increasing growth rates with decreasing, concentration throughout this range, with the optimal growth better than the controls, were tested at additional concentrations of 0.05 and 0.015 mg. per

WHITE: AMINO ACIDS IN NUTRITION OF EXCISED ROOTS TABLE I GROWTH INDICES OF ISOLATED TOMATO ROOT-TIPS AT VARIOUS CONCENTRATIONS OF SINGLE AMINO ACIDS. THE GROWTH RATE IN THE ABSENCE OF ALL ACCESSORY MATERIAL IS TAKEN AS 100 795 AMINO ACID CONCENTRATION-MG. PER LITER 50 15 5 1.5 0.5 0.15 0.05 0.015 0 Glutamic acid 35 133 97 102 97 98... 100 Aspartic acid 15 124 109 98 106 86 100 Glycine 38 84 306 146 97 104 100 Lysine.31 157 216 194 200 171 100 Phenyl-alanine 52 86 118 98 100 95 100 Proline 30 41 56 134 93 74 100 Cystine 70 70 100 128 104 100 100 Hydroxyproline 7 9 18 71 127 99 100 Leucine 20 50 110 114 118 103 100 Histidine 16 51 62 92 109 100 100 Valine 5 22 44 104 108 97 100 Methionine 34 42 48 67 100 13.9 100 100 Serine.11 33 66 82 107 111 93 98 100 Norleucine 10 15 22 32 56 128 124 103 100 Arginine 11 26 74 110 110 127 100 100 Tyrosine... 19 42 76 99 84 103 100 Isoleucine 21 31 51 81 124 138 174 99 100 Tryptophane 13 20 36 78 85 86 100 Alanine 6 10 14 24 90 76 100 liter. This series included about 3000 individual cultures so that it was necessary to run them in 10 consecutive sets, each with a separate set of controls. In order to facilitate comparison of the sets within this series with one another, the results have been expressed as percentages of the controls. The results are shown in table I. All amino acids studied proved to be depressant at 50 mg. per liter (50 p.p.m.). Alanine, valine, hydroxyproline, and norleucine were quite definitely so. Tryptophane and alanine were depressant at all concentrations studied. Tyrosine was apparently not beneficial at any concentration tested. Valine, histidine, serine, phenyl-alanine, methionine, cystine, leucine, norleucine, arginine, aspartic acid, glutamic acid, proline, and hydroxyproline showed slight stimulation at some point in the concentration range investigated. Glycine, lysine, and isoleucine were markedly stimulating at favorable concentrations. The optimal concentrations varied from 15 p.p.m. in the case of glutamic and aspartic acids down to 0.05 p.p.m. in the case of isoleucine. A typical series of results is shown in figure 2 (glycine). From these data an amino acid mixture having the provisional constitution shown in the first column of table II was chosen for study. MEISEN-

796( PLANT PHYSIOLOG'Y (, IJvrln 40V-I 35-30- ~25-20- i.15-0 5-0- 50 15 5 1.5 0.5 0.15 0 MG. PER LITER FIG. 2. Histogram showing total increments obtained in a week 's time in nutrients containing various concentrations of glycine, without other accessory material. HEIMER has studied in detail the nitrogenous constituents of hydrolyed yeast (2, 3). Although his figures are much higher than would be the case for an alcohol extract, it is clear that such an extract could never contain more of any amino acid than he has indicated. His figures have, therefore, been recalculated to give comparable units and are presented for comparison. The last column of the table is of especial interest. In most cases the experimentally determined optimum concentration for nutrition of isolated tomato root-tips did not differ widely from the amounts actually found in a hydrolyed yeast extract. But the optimal amounts found for glycine, cystine, glutamic acid, and aspartic acid were so much higher than can possibly be present in yeast (MEISENHEIMER, see table II) that some explanation would seem necessary. It is to be noted that three of these-glycine, glutamic acid, and aspartic acid-are not among the nine amino acids recognied by ROSE (8, 9) as essential for animals. It seemed possible that in a complete mixture these three might prove to be partially or wholly interchangeable with some of the other amino acids. However, it is to be remembered that MUELLER (5) found diphtheria bacilli to require one of these, glutamic acid, in unexpectedly large amounts. On the basis of the information contained in the data presented above, an amino acid mixture was prepared containing 15 mg. each of glutamic acid

WHITE: AMINO ACIDS IN NUTRITION OF EXCISED ROOTS TABLE II CONCENTRATIONS OF SINGLE AMINO ACIDS OPTIMAL FOR GROWTH OF ISOLATED TOMATO ROOT-TIPS 797 TOTAL CONCEN- TRATION IN OPTIMUM AMINO IACET ACID CONCENTRATION OPTIMAL YEAST RATIO (a) CONTENT (FROM a/b MEISENHEIMER) (b) rig./liter Glycine... 5.000 0.03 166.7 Cystine... 1.500 0.04 37.5 Lysine... 5.000 0.52 9.6 Glutamic acid... 15.000 0.62 24.1 Isoleucine... 0.050 0.23 0.2 Norleucine... 0.150...... Methionine... 0.150 0.14* 1.1 Serine... 0.150 1.14 0.1 Proline... 1.500 0.16 0.9 Hydroxyproline... 0.500 0.38 1.3 Phenyl-alanine... 5.000 1.13 4.4 Leucine 0.500 0.70 0.7 Valine... 0.500 0.84 0.6 Arginine... 0.150 0.31 0.5 Aspartic acid... 15.000 0.38 39.3 Histidine... 0.500 0.24 2.1 Total... 50.650 6.86 7.4 * MEISENHEIMER (2, 3) does not mention methionine by name but speaks of "sulfur containing amino acids other than cystine " as being present in this amount. It is assumed that the " amino acid" to which he refers was probably methionine. and aspartic acid, 5 mg. of glycine, lysine, and phenyl-alanine, 1.5 mg. of cystine and proline, 0.5 mg. of hydroxyproline, leucine, histidine, and valine, 0.15 mg. of methionine, serine, norleucine, and arginine, and 0.05 mg. of isoleucine, totaling 50.65 mg. of amino acid in each liter of solution. Since this was about 7 times the total amount of amino acid contained in the optimal concentration of yeast (MEISENHEIMER, 2) and since concentrations, both relative and absolute, have repeatedly proved of great importance, the mixture was tested at 1, 0.3, 0.1, 0.03, and 0.01 times this concentration. The results are shown in figure 3. As was the case with single amino acids, a concentration of 50 mg. per liter was quite depressant, the optimum lying at about 15 mg. per liter, one-third the concentration originally chosen. As has already been suggested, the importance of many of the amino acids used above, especially in the presence of the other amino acids, was subject to considerable doubt, since it is known that in animal nutrition cer-

798 PLANT PHYSIOLOGY Lii I- 'IL -J 0 I-- w MG. PER LITER FIG. 3. Effects on growth of various concentrations of a mixture containing 16 amino acids (see text). tain acids may substitute for one another (ROSE, et al.). It thus seemed probable that some of these amino acids, which when taken singly proved to be beneficial, might nevertheless be unnecessary for optimal growth. To test this, a series of cultures was made, in each of which one amino acid was employed at 10, 1, 0.1, and 0 times the concentration (one-third that shown in table II) found to be optimal in the last experiment, while all other amino acids were present at the optimal concentrations. This test involved 64 combinations, 1280 cultures. In each set of cultures a solution containing all amino acids except the one under examination was used as control. The results, given as percentages of these controls are presented in table III. It will be seen from the table that of the 16 amino acids which have a stimulating effect at some concentration when taken alone, 7 fail to exert such an effect when in the presence of all the others. These 7 are glycine, aspartic acid, hydroxyproline, cystine, methionine, arginine, and norleucine. Reference to table II will show that of these, 3-glycine, cystine and aspartic acid-were among those which, when taken singly, gave optimum results at concentrations greater than those existing in the optimal concentration of yeast. It seems probable that the originally observed effects of these 3 (table

WHITE: AMINO ACIDS IN NUTRITION OF EXCISED ROOTS TABLE III GROWTH INDICES OF ISOLATED TOMATO ROOT-TIPS GROWN IN A COMPLETE AMINO ACID MIX- TURE (TABLE II) IN WHICH THE CONCENTRATIONS OF SINGLE AMINO ACIDS WERE VARIED. THE GROWTH RATE IN A MIXTURE, COMPLETE EXCEPT FOR THE AMINO ACID UNDER EXAMINATION, IS TAKEN AS 100 RELATIVE CONCENTRATION OF THE AMINO ACID WHOSE CONCEN- TRATION IS BEING VARIED. ONE-THIRD OF THE VALUE INDICATED IN TABLE II AS INDIVIDUALLY OPTIMAL AMINO ACID IS TAKEN AS 1 (COLUMN 2) (SEE TEXT) 10-1.0-0.1 0 Histidine... 136 115 108 100 Proline... 85 120 102 100 Serine... 88 122 104 100 Valine... 82 124 114 100 Phenyl-alanine... 48 124 149 100 Glutamic acid... 35 144 138 100 Lysine... 56 162 159 100 Leucine... 80 96 133 100 Isoleucine... 80 99 118 100 Arginine... 100 104 102 100 Methionine... 68 98 96 100 Cystine... 60 89 90 100 Aspartic acid... 14 72 97 100 Glycine... 53 67 89 100 Hydroxyproline... 63 80 92 100 Norleucine... 61 82 77 100 II) may have been due to traces of other essential amino acids present as impurities in the samples studied1, or else that they may function only as building blocks for other essentieal acids present in the complete mixture. A fourth amino acid, methionine, discarded by this process, is known to be a common contaminant in commercial preparations of leucine (MUELLER, 6) and was possibly supplied in the requisite amount in this material. Of the 4 acids giving aberrant results (table II), only glutamic acid remains as apparently essential sui generis. Using diphtheria bacilli, MUELLER likewise found the optimal concentration of glutamic acid to be abnormally high. If the acids found to be essential for rats (ROSE, 9), for diphtheria bacilli (MUELLER, 5), and for tomato roots (this work) are tabulated, the very close correspondence between the three sets is evident. Only two amino acids-proline and serine-present in the list developed for tomato root-tips are not present in either RoSE'S or MUELLER'S lists. Four acids-cystine, methionine, glycine, and tryptophane-are to be found in one or both of these lists but not in the present one. Tryptophane alone, of those found essential by both RoSE and MUELLER, has under no circumstances appeared 799

800 PLANT PHYSIOLOGY beneficial for tomato root-tips. In view of its close chemical relationship to heteroauxin, which is known to have a depressant effect on root growth, tryptophane may have a place which is unique. Further study will be required before its true position can be considered established. TABLE IV AMINO ACIDS ESSENTIAL FOR OR BENEFICIAL TO ORGANISMS OF THREE WIDELY SEPARATED GROUPS RATS TOMATO ROOTS DIPHTHERIA BACILLI (ROSE) (WHITE) (MUELLER) Histidine Histidine Histidine Phenyl-alanine Phenyl-alanine Phenyl-alanine Lysine Lysine... Leucine Leucine... Isoleucine Isoleucine...... Cystine Cystine Tryptophane Tryptophane Threonine* (Not tested) (Not tested)... Valine Valine Glutamic acid..glutamic acid...... -Proline..........,...3Serine *= a-amino-,3-hydroxy-n-butyric acid (4).... Glycine -Methionine A provisional amino acid mixture which appeared to be complete and possibly optimal then contained: -5 mg. of glutamic acid, 1.5 mg. each of phenyl-alanine, lysine, and histidine, 0.5 mg. of proline, 0.15 mg. of valine, 0.05 mg. of serine, 0.015 mg. of leucine, and 0.0015 mg. of isoleucine per liter of nutrient. The total amino acid content of this mixture was 10.2165 mg. per liter. Since such an amino acid mixture was developed in the hope of replacing therewith the fraction of yeast soluble in 85 per cent. ethyl alcohol but insoluble in 100 per cent. alcohol, a series of cultures was carried out using this mixture together with a 100 per cent. alcohol extract of yeast in amount equivalent to the standard amount of whole yeast regularly used as control. The results are shown in figures 4 and 5. Such a mixture of amino acids, when added to the yeast material soluble in absolute alcohol, is adequate for the maintenance of continuous growth for at least three passages, at a level equal to or nearly equal to that obtained with whole yeast. The entire effect of the yeast material soluble in 85 per cent. alcohol but insoluble in absolute alcohol, insofar as it can be observed in this way, is then to be attributed to its amino a;cid content.

WHITE: AMINO ACIDS IN NUTRITION OF EXCISED ROOTS 801 o0-4 4e 2-~ ~ ~ ' I I o s 4~~~~t - 15 S220 DAYS FIG. 4. Graph showing the effects during three passages' of a mixture of 16 amino acids (see text) when taken alone and in the presence of a 100 per cent. alcohol extract of yeast. A; 81 C I) E'J (Photograph by J. A. CARLILE) FIG. 5. Reading from left to right: A, control with yeast extract. B, control without accessory material. C, amino acid mixture alone. D, yeast material soluble in 100 per cent. alcohol, insoluble in ether. E, same as D but with amino acid mixture (C) added. Photograph taken at the end of the third passage. x 1. There remains to be identified only the yeast material soluble in absolute alcohol, representing 0.0005 per cent. of the total nutrient. Summary and conclusions In an earlier paper it was shown that yeast extract contains certain materials essential for growth of isolated tomato roots. These are all soluble 1 "Passage 1,, in this experiment is passage 196, counting from the time of isolation, the roots having been maintained through 195 passages under standard conditions before being used in this experiment.

802 PLANT PHYSIOLOGY in 85 per cent. ethyl alcohol, and insoluble in ether. Further extraction with 100 per cent. alcohol segregates two fractions, both of which are necessary for satisfactory growth. Data have been presented in this paper which show that the fraction which is soluble in 85 per cent. alcohol, but insoluble in 100 per cent. alcohol and ether, may be replaced by a mixture of 9 amino acids without appreciably reducing the growth rate. The amino acid content of this fraction appears to be entirely responsible for its growthpromoting effect, at least so far as the requirements of isolated tomato roottips are concerned. Such a mixture contains the following: glutamic acid, lysine, histidine, phenyl-alanine, leucine, isoleucine, valine, serine, and proline. All other amino acids appear to be unessential under the conditions of the experiment reported. The fraction soluble in 100 per cent. alcohol remains to be identified. DEPARTMENT o01 ANIMAL AND PLANT PATHOLOGY THE ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH PRINCETON, NEW JERSEY LITERATURE CITED 1. MCCOY, R. H., MEYER, C. E., and ROSE, W. C. Feeding experiments with mixtures of highly purified amino acids. VIII. Isolation and identification of a new essential amino acid. Jour. Biol. Chem. 112: 283-302. 1936. 2. MEISENHEIMER, J. Die stickstoffhaltigen Bestandteile der Hefe. Zeitschr. physiol. Chem. 104: 229-283. 1919. 3.. Die stickstoffhaltigen Bestandteile der Hefe. II. Die Purinbasen und Diaminosauren. Ergebnisse. Zeitschr. Physiol. Chem. 114: 205-249. 1921. 4. MEYER, C. E., and ROSE, W. C. The spatial configuration of a-amino-j3-hydroxy-n-butyric acid. Jour. Biol. Chem. 115: 721-729. 1936. 5. MUELLER, J. H. Studies on cultural requirements of bacteria. V. The diphtheria bacillus. Jour. Bacteriol. 29: 515-530. 1935. 6.. Methionine as an impurity in natural leucine preparations. Science, n.s. 81: 50-51. 1935. 7. OSBORNE, T. B., and WAKEMAN, A. J. Extraction and concentration of the water-soluble vitamine from brewers' yeast. Jour. Biol. Chem. 40: 383-394. 1919. 8. ROSE, W. C. The amino acids in nutrition. Yale Jour. Biol. & Med. 4: 519-536. 1932. 9.. Significance of amino acids in nutrition. Harvey Lectures, 1935-1936: 49-65. 1936. 10. WHITE, P. R. Separation from yeast of materials essential for growth of excised tomato roots. Plant Physiol. 12: 777-791. 1937.