THE ROLE OF ANTHRANILIC ACID IN THE NUTRITION OFLACTOBACILLUS ARABINOSUS

L. E. RHULAND' AND R. C. BARDDepartment of Bacteriology, Indiana University, Bloomington, Indiana

Received for publication July 26, 1951

Tryptophan has long been known as an essential amino acid for many formsof life. However, nuimerous microorganisms do not require nitrogen in the formof tryptophan, and several nitrogenous compounds are known which replace thenutritional requirement for tryptophan. The demonstration that anthranilic acid(Snell, 1943) or indole (Schweigert et al., 1946) supports growth of Lactobacillusarabinosus in the absence of tryptophan suggests the existence of an anthranilicacid -* indole -÷ tryptophan pathway of synthesis. Such a metabolic pathwayhas been described in Neurospora (Tatum and Bonner, 1943; Tatum, Bonner,and Beadle, 1943), and cell-free extracts of Neurospora sitophila form tryptophanin the presence of indole, serine, and pyridoxal phosphate (Umbreit, Wood, andGunsalus, 1946). Since serine stimulates the growth of L. arabinoms in the pres-ence of indole (Schweigert, 1947), it has been concluded (Rydon, 1948; Work andWork, 1948) that, as in Neurospora, indole is condensed with serine to yieldtryptophan by this organism and, moreover, that anthranilic acid is probablythe precursor of indole. Thus, essentially by analogy, the apparent pathway oftryptophan synthesis in Neurospora has been claimed to exist in bacteria.

This study was initiated to clarify the biochemical steps leading from anthra-nilic acid to tryptophan in L. arabinosus. The growth supporting capacity of aseries of potential intermediates along the pathway of anthranilic acid bioconver-sion was tested, and conditions for the utilization of anthranilic acid and indolewere studied. The tryptophan content of cells grown in the presence of anthranilicacid, indole, or tryptophan was also determined. The results appear to cast doubtupon the simple metabolic pathway of anthranilic acid conversion to tryptophanpreviously described.

MATERIALS AND METHODS

Lactobacillus arabinosus, strain 17-5, from the Cornell University AgriculturalExperimental Station, and Leuconostoc mesenteroides, strain P-60, from Merckand Company (through the courtesy of Dr. J. L. Stokes), were maintained asstab cultures at 4 C in the following stock agar: tryptone, 10 g; yeast extract,10 g; K2HPO4, 5 g; CaCO3, 3 g; glucose, 1 g; liver extract, 20 per cent (by volume);distilled water, 800 ml. The cultures were passed semimonthly for 12 to 15 hoursin medium A consisting of tryptone, 10 g; yeast extract, 10 g; glucose, 10 g;K2HPO4, 5 g; distilled water, 1 liter; 20 ml salts solution (see hereafter) and thenreturned to stock agar. All cultures were incubated in a 37 C water bath.

Present address: Research Division, The Upjohn Company, Kalamazoo, Michigan.133

on March 1, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

L. E. RHULAND AND R. C. BARD

The components of the semisynthetic medium used to determine the effect oftryptophan substitutes upon growth and to obtain metabolically active cells arelisted in table 1. The inoculum for the growth studies was prepared from a 12hours' old medium A culture. The 10 ml culture was washed twice with 10 mldistilled water, resuspended in the same volume, and one drop inoculated intothe growth tubes containing 10 ml of the semisynthetic medium. To obtainmetabolically active cells, the following inoculating procedure was developed;strict adherence to this procedure was required to obtain cells in the same meta-bolic state. A transfer from stock culture to medium A was allowed to incubatefor 15 to 17 hours. A second transfer to another medium A tube (0.05 per centinoculum) was incubated 8.5 to 9 hours and then transferred (0.15 per cent) tomedium A for an additional 5 to 5.5 hours. Finally, the semisynthetic medium

TABLE 1Complete medium for cultivation of LactobaciUlus arabinosus

Casein hydrolysate (acid hydro- Calcium pantothenate........... 1 mglyzed)................... 100 ml Nicotinic acid... 1 mg

Phosphate buffer, M pH 7.5....... 100 ml Riboflavin ....................... 1 mgGlucose.................... 10 g Thiamine hydrochloride... 2 mgSodium acetate...........1 g para-Aminobenzoic acid......... 0.5 mgDL-Serine............... 125 mg Pyridoxine hydrochloride. 2.5 mgHistidine............... 100 mg Folic acid.5 mgCystine hydrochloride............ 200 mg Biotin.0.5 jigAdenine sulfate............... 10 mg Anthranilic acid, pH 7.0, or ... 25 mgGuanine............... 10 mg Indole,* or....................... 6.3 mgUracil............... 10 mg Tryptophan..................... 2 mgXanthine............... 10 mg Salts solution*t ................. 20 ml

pH adjusted to 7.1; distilled water to 1,000 ml. Sterilized at 120 C for 20 min.* Sterilized by filtration.t Salts solution: MgSO4*7H20, 10 g; NaCl, 0.5 g; FeSO4 7H20, 0.5 g; MnSO4 H2O, 0.4 g;

water to 250 ml.

was inoculated (1 per cent) from the last transfer in medium A. Cells were har-vested after 14 to 16 hours' incubation (pH 4.9 to 5.0), washed in one-fifth thegrowth volume with 0.85 per cent NaCl, and resuspended in the desired volumeof distilled water.When indole was substituted for anthranilic acid, the foregoing procedure was

used except that cells were resuspended in 10 ml of distilled water before beingused as an inoculum. This procedure was necessary to cultivate cells capable ofutilizing indole in vitro; use of an unwashed inoculum resulted in cells with poorindole activity.Measurement of bacterial growth was made with an Evelyn colorimeter, using

a 660 m,u filter and recording turbidity as optical density. Dry weight estimationsof cell suspensions were made from a previously calibrated curve.

Anthranilic acid was determined by the sulfanilamide method of Bratton andMarshall (1939) modified as follows: after addition of 2 ml 0.5 per cent solution

134 [VOL. 63

on March 1, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

ANTHRANILIC ACID IN L. ARABINOSUS

of N-(l-naphthyl)ethylenediamine dihydrochloride, the color was allowed todevelop for 10 min at 37 C and for another 10 min at room temperature. Indolewas determined by the method of Umbreit, Wood, and Gunsalus (1946) andtryptophan by the method of Nason, Kaplan, and Colowick (1951). N-acetylan-thranilic acid was assayed as anthranilic acid after hydrolysis for 11 min in NH2SO4 at 100 C. Phenylglycine-ortho-carboxylic acid was determined by measur-ing the optical density at 330 mit, the characteristic absorption maximum of thiscompound.The commercially available dehydrated medium (Difco tryptophan assay me-

dium) of Green and Black (1944) was used for all microbiological assays fortryptophan; L. mesenteroides was employed as the assay organism. All growthresponses were calculated on the basis of standard curves obtained with L- andDL-tryptophan (Prescott et al., 1949). Cells to be assayed microbiologically fortryptophan were obtained from 12 to 15 hours' old liter cultures, removed fromthe medium by centrifugation, washed twice in 200 ml distilled water, and driedin vacuo. The dried cells were subjected to barium hydroxide hydrolysis (Shawand McFarland, 1938) for 12 hours at 120 C and the Ba+ removed with 2 NH2SO4. The hydrolysates were extracted 4 times with a twofold volume of ether,neutralized to pH 6.8, and diluted to the desired volume.

RESULTS

Growth. Green and Black (1943) demonstrated that several indole derivatives(skatole, indole-3-acetic acid, beta-(indole-3)-propionic acid, and (indole-3)-n-butyric acid) do not substitute for tryptophan in the growth of L. arabinosus;and similarly, Fildes (1940) reported negative results with Salmonella typhosafor other indole-like compounds (indoleacetic acid, indoleacrylic acid, indolepro-pionic acid, indolecarboxylic acid, indolepyruvic acid, indoleethylamine, andindolealdehyde). In this study attention was focused upon compounds whichmight play roles as intermediates in the conversion of anthranilic acid to indole,and these experiments were performed by adding the compound to be tested tothe semisynthetic medium (table 1) in place of tryptophan. The amount ofgrowth obtained with medium containing tryptophan was equated as 100 percent activity. The activities found with the following compounds were: indole,89; anthranilic, 72; phenylglycine-ortho-carboxylic acid, 11 (with an equimolarquantity) and 34 (with four times the molarity of anthranilic acid); less than 10per cent activity: acetanilide, acetophenone oxime, N-acetylanthranilic acid,ortho-aminobenzaldehyde, ortho-aminobenzyl alcohol, DL-aminophenylacetic acid,ethylaniline, ortho-ethylaniline, ethylanthranilate, ortho-hydroxyethylaniline,beta-hydroxyethylaniline, isatin, oxindole, phenylacetamide, phenylacetic acid,N-phenylglucine, ortho-formotoluidine, and ortho-toluidine. Thus, using growthas a criterion, none of the compounds tested may be considered as intermediatesin the biosynthesis of indole or tryptophan from anthranilic acid.

Anthranilic acid metabolism. To determine the conditions under which anthra-nilic acid utilization occurs and to measure the formation of indole or tryptophan,resting cells were employed. As shown in table 2, glucose is required for maximum

19521 135

on March 1, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

L. E. RHULAND AND R. C. BARD

anthranilic acid utilization by resting cells; the glucose requirement, presumablyas a source of adenosine triphosphate, is absolute. A typical substrate saturationcurve is obtained with varying amounts of glucose, 0.03 M being the optimumconcentration; with 0.04 M glucose, 15 per cent inhibition is obtained, probablydue to the accumulation of lactic acid. Determination of the effect of pH uponanthranilic acid utilization over the pH range 6.2 to 8.0 revealed a rapid decreaseof activity on either side of the optimum pH of 7.5 to 7.6; only half maximumactivity was obtained at pH 6.8.

In addition to the glucose requirement, acetate and coenzyme A also stimulateanthranilic acid utilization. These data suggest a coenzyme A-acetate systemsimilar to that described in pigeon liver for the acetylation of aromatic amines byLipmann (1945) or the formation of citrate from oxalacetate and acetate in

TABLE 2Requirements for anthranilic acid utilization by Lactobacillus arabinosus

ADDITIONS ANTHRANILIC ACID UTILIZED

JugAll ...................... ............................ 37All, less glucose............................................... 0All, less acetate............................................... 31All, less coenzyme A........................................... 22All, less coenzyme A + Ca-panthothenate..................... 25

Protocol: anthranilic acid (200 pg per ml), 0.2 ml; M phosphate buffer, pH 7.8, 0.2 ml;0.1 M glucose, 0.3 ml; ms potassium acetate, 0.1 ml; Clostridium butylicum (30 mg dry wt perml), as coenzyme A source, 0.2 ml, or Ca-pantothenate (100 ug per ml), 0.2 ml; cells (24mg dry wt per ml), 0.5 ml; distilled water to 2.0 ml. The reaction was run 30 min at 37 Cand terminated with one ml 15 per cent trichloracetic acid.

Escherichia coli (Stern and Ochoa, 1949; Novelli and Lipmann, 1950). However,neither N-acetylanthranilic nor phenylglycine-ortho-carboxylic acid is utilizedby resting cells under the same conditions required for anthranilic acid metabo-lism, and as shown before, neither compound supports appreciable growth of L.arabinosus.

In all experiments involving the uptake of anthranilic acid, repeated tests forindole were made, and in no case was a positive result obtained. Similarly, addi-tion of serine and pyridoxal phosphate to the reactants necessary for anthranilicacid utilization failed to yield tryptophan. Moreover, attempts to obtain conver-sion of anthranilic acid to indole or tryptophan by the addition of vitamins(thiamin, riboflavin, pyridoxine, pyridoxal, B12, niacin, biotin, pantothenic acid),other adjuncts (yeast extract, tryptone, boiled yeast juice, boiled L. arabinosuscells, pyruvate oxidase factor of O'Kane and Gunsalus (1948), ascorbic acid,cysteine, thioglycolate, glutathione, glycine, alanine, alpha-ketoglutarate, serine,indole), and inorganic salts (Fe++, Mn, Mg++, Co++, Zn++) all yielded negativetests for indole and tryptophan. These negative data indicate that either a labileintermediate(s) is formed in the reaction or that anthranilic acid is not convertedto indole or tryptophan.

136 [VOL. 63

on March 1, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

ANTHRANILIC ACID IN L. ARABINOSUS

Indole metaboli8m. Study of indole utilization by resting cells of L. arabinosusrevealed (table 3) that indole activity is completely lacking in the absence ofglucose; serine and pyridoxine stimulate indole utilization only moderately. Asin the case of anthranilic acid metabolism, the glucose requirement is verymarked: 0.03 m glucose supports maxmum indole activity while 0.04 M glucoseis inhibitory. With dried cells (table 3) adenosine triphosphate is required as wellas serine and pyridoxal phosphate for indole utilization. The pyridoxal phosphaterequirement is not replaced by adenosine triphosphate and pyridoxal, indicatinganother role for adenosine triphosphate, possibly as a phosphorylating agent foreither the indole or the serine. While establishing the conditions for maximumindole utilization, it was observed consistently that the expected mole for molerelationship of indole utilization and tryptophan formation does not occur, a

TABLE 3Indole utilization by Lactobacillus arabino8us

DOLE UTILIZEDADDMONS

Resting cels Dried cells

All.................................. 40 10All, less glucose............................. 1All, less adenosine triphosphate 2All, less pyridoxine HCl .................... 32All, less pyridoxal phosphate 0All, less serine.............................. 32 0All, less serine and pyridoxine-HCIO.l........ 34All, plus yeast extract (20 pg) 10

Protocol: 40,ug indole; phosphate buffer, pH 7.8, 0.1 M final concentration; 1 mg serine;for resting cells: also 20o M glucose, 20 pg pyridoxine HCl, 4 mg dry wt cells, 40 min at 37 C;for dried cells: also 20 pM adenosine triphosphate, 200 pg pyridoxal phosphate, 20 mg dry wtcells, 120 min at 37 C; total volume of all reaction mixtures was 2.0 ml.

finding not in agreement with the quantitative conversion of indole to trypto-phan in Neurofpora (Umbreit et al., 1946; Nason, 1950; Nason, Kaplan, andColowick, 1951). Study of indole utilization and tryptophan formation by restingcells, using both chemical and microbiological methods for the determination oftryptophan, revealed (figure 1) that although marked indole utilization occurs,tryptophan is not the major product of indole metabolism.

Since the Neuroepora type of indole conversion to tryptophan appears to beabsent in L. arabinosu, the possibility arises that anthranilic acid is also notconverted to tryptophan via a mechanism similar to that described for Neuro-spora. This possibility made it necessary to ascertain the actual formation oftryptophan in L. arabinosus grown with anthranilic acid, for if anthranilic acidis the precursor of tryptophan, the tryptophan content of cells grown with eithersubstrate should be similar.CeU hydrolysates. Cells grown in medium containing tryptophan, indole, or

anthranilic acid were dried and hydrolyzed as described in Methods. The trypto-

19521 137

on March 1, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

L. E. RHULAND AND R. C. BARD

TIME (MIN)Figure 1. Indole utilization and tryptophan synthesis by resting cells of Lactobacillus

arabinosus.Protocol: 0.1 ml indole (430 ug per ml)

0.2 ml M phosphate buffer, pH 7.80.3 ml 0.1 M glucose0.2 ml DL-serine (5 mg per ml)0.2 ml pyridoxine hydrochloride (200 Ag per ml)0.5 ml cell suspension (24 mg dry wt per ml)0.5 ml distilled water.

Tryptophan assayed chemically (curve 1) and microbiologically (curve 2).

TABLE 4Tryptophan content of Lactobacillus arabinosus grown under various conditions

PER CENT TRYPTOPHAN

GROWTH MEDIUM Microbiological Chemical

L DL D and L

Medium A (tryptophan) ........ ......... 0.17 0.18 0.4Complete (anthranilic acid)*.0.07 0.08 0.3Complete (anthranilic acid)t.0.03 0.03Complete (indole) ....................... 0.06 0.06 0.2

Complete medium contained 8 mg anthranilic acid or 6.3 mg indole per liter.* Using 11-hr inoculum.t Using 5-hr inoculum; 11-hr inoculum contained about 4 times as many cells as the

5-hr inoculum.

phan content of L. arabinosus cells was calculated on a dry weight basis fromstandard curves using both L- and DL-tryptophan. As shown in table 4, thetryptophan content of cells grown with anthranilic acid is only 17 to 46 per cent

138 [VOL. 63

on March 1, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

ANTHRANILIC ACID IN L. ARABINOSUS

of the value found in those grown with tryptophan. Similarly, the tryptophancontent of cells grown with indole is only 34 per cent the value found in thosegrown with tryptophan, a finding paralleling the data with anthranilic acid-growncells.

Microbiological analysis of hydrolysates prepared from L. arabinosus cellsgrown in the presence of anthranilic acid was also performed using L. arabinosus,

Q7Q,q I'

0.6

0.5

I-

z 0.4w

-j Q3

O 0.2

0.1

WAVELENGTH: m.LFigure 2. Spectrophotometric analysis of products of anthranilic acid metabolism by

resting cells of Lactobacillus arabinosus.Protocol: 0.1 ml anthranilic acid (400,ug per ml)

0.2 ml M phosphate buffer, pH 7.80.3 ml 0.1 M glucose0.5 ml cell suspension (26.5 mg dry wt per ml)0.9 ml distilled water.

Incubated 30 min at 37 C; terminated with 1 ml N H2SO4. Curve A: zero time; curve B:30 minutes.

an organism which does not demonstrate the tryptophan nutritional specificityof L. mesenteroides. Such analysis revealed 0.13 per cent L-tryptophan, beforeand after ether extraction of the hydrolysates, indicating that, in addition tothe small amount of tryptophan found in the cell hydrolysates, another substancehaving growth supporting capacity for L. arabinosus is formed. Although con-clusive proof is not available at this time, it is possible that the unknown product

1952] 139

on March 1, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

L. E. RHULAND AND R. C. BARD

formed is the terminal product of anthranilic acid metabolism and that thisproduct is amino acid-like in nature.The demonstration of the low tryptophan content of cells grown with indole

or anthranilic acid, coupled with the in vitro studies demonstrating that indoleis not condensed with serine to yield tryptophan, strongly suggests that anotherpathway of anthranilic acid or iindole metabolism exists which either does notresult in tryptophan formation or which is operative via a pathway yielding alow level of cellular tryptophan.

Further aspects of anthranilic acid metabolism. In the attempt to idenitify a com-ponent of in vitro anthranilic acid metabolism, the reaction mixture (less coen-zyme A source) described in table 2 was allowed to incubate 30 min at 37 C. Itwas then acidified to pH 1, extracted with diethyl ether, and the ether-solublefraction placed in M K2HPO4 buffer at pH 8.8; the ether was removed in vacuo.The absorption spectrum of the resulting liquid was obtained using a Beckmanspectrophotometer and the data plotted as shown in figure 2; another curve pre-pared with a reaction mixture similarly treated at zero time is included forcomparison. The results indicate an absorption peak at 270 m,u which is absentin the zero time control. The zero time plot exhibits an absorption maximum at310 m,u, representing anthranilic acid which disappeared after 30 min incubation,with the subsequent formation of the absorption peak at 270 m,u. The absorptionmaximum at 270 m,u is probably due to a product of anthranilic acid metabolism,and from its properties of ether solubility and absorption maximum at 270 m,u, thecompound appears to be indole-like. However, chemical tests for indole werenegative indicating that the postulated intermediate is not indole as such. Fur-ther efforts to characterize this intermediate by microbiological assay with L.arabinosus proved to be inconclusive. Although some evidence was availablethat the ether-soluble fraction containing the presumed intermediate does sup-port partial growth of L. arabinosus, the results were irregular and no de finiteconclusion could be drawn. It is possible that the intermediate concerned is labile(e.g., phosphorylated), explaining the limited success in the latter experiments.

DISCUSSION

The data presented previously do not support the generally accepted conceptof tryptophan synthesis in bacteria (Rydon, 1948; Work and Work, 1948), ametabolic pathway assumed to exist in bacteria and derived, by analogy, fromthe data obtained with the mold Neurospora. Actually, other reports offer similarindications.

Using a strain of Salmonella typhosa able to utilize ammonium salts as a solesource of nitrogen for growth, Fildes (1945) found no in vitro tryptophan synthe-sis wsith indole as the substrate, even ini the presence of seriine and glucose. Apyridoxal phosphate deficiency probably did not underlie this failure silnce un-washed cells were used by Fildes, and as shown ini table 3, the stimulation bypyridoxal phosphate of indole uitilizationi by w-ashed restinig cells of L. arabinosusis Inot very marked.

Kikkawva (1950) founld that injectioni of tryptophan inlto silkwvorm pupa yields

140 [VOL. 63

on March 1, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

ANTHRANILIC ACID IN L. ARABINOSUS

a pigmented product, referred to as plus chromogen and measured as kyneurine.Injection of indole does not lead to plus chromogen but is excreted unchanged.Anthranilic acid injection results in the formation of plus chromogen which,however, the author states is not kyneurine. From these data Kikkawa concludesthat the pathway of anthranilic acid metabolism in insects is different from thepathway described for the mold Neurospora.

Recently Davis, Henderson, and Powell (1951) reported that Xanthomonaspruni requires tryptophan or nicotinic acid, in addition to an organic nitrogensource, for growth. Tryptophan, indole, or anthranilic acid substitutes for nico-tinic acid, but anthranilic acid stimulates growth in lower molar concentrationsand exhibits a shorter growth lag response than tryptophan and indole. From theseand related results, it was suggested that anthranilic acid may not be the pre-cursor of indole in X. pruni in the manner proposed for Neurospora and L. ara-binosus.

These data, although isolated, represent the only attempts to study the meta-bolic fate of anthranilic acid in organisms other than Neurospora, and even in thelatter case, no published evidence exists to explain precisely the mechanism ofanthranilic acid utilization. From the results presented before dealing with L.arabinosus, it appears that indole, as such, is not an end product of anthranilicacid metabolism. Also, indole is not converted to tryptophan by the one-stepcondensation reaction reported in Neurospora. The low tryptophan content ofcells grown with anthranilic acid or indole supports this conclusion. However,these cells contain a substance, presumably tryptophan-like, which supportsgrowth of L. arabinosus above and beyond the growth attributable to the trypto-phan content of these cells. It appears, therefore, that the additional growth isdue to the formation from anthranilic acid of a substance substituting for tryp-tophan in the cell material and that this substance is the actual end product ofanthranilic acid metabolism. This substance is nonether soluble and stable toBa(OH)2 hydrolysis, as is tryptophan; it is not, however, able to support thegrowth of L. mesenteroides, an organism specifically requiring tryptophan forgrowth. It is possible, of course, that the postulated substance is a tryptophanprecursor; yet it is difficult to visualize the accumulation of an amino acid pre-cursor to an extent equal to the amino acid itself.

ACKNOWLEDGMENTS

The authors are indebted to Dr. L. S. McClung for his interest during thecourse of this investigation and for his valuable criticisms in the preparation ofthe manuscript. Some of the growth intermediates were kindly supplied by Drs.J. H. Billman, E. E. Campaigne, C. E. Kaslow, and R. L. Shriner of the Depart-ment of Chemistry, Indiana University. This study was aided by a researchgrant from the Graduate School, Indiana University.

SUMMARY

Of nineteen possible intermediates in the pathway of indole synthesis fromanthranilic acid, none supports growth of Lactobacillus arabinosus, strain 17-5.

1952] 141

on March 1, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

L. E. RHULAND AND R. C. BARD

Glucose, acetate, and coenzyme A play significant roles in the utilization ofanthranilic acid by resting cells of this organism; indole is not formed under theseconditions. Indole utilization also requires glucose, and in addition, serine andpyridoxal phosphate; tryptophan, however, is not formed.Growth of this organism in medium containing anthranilic acid or indole in

place of tryptophan yields cells with significantly low tryptophan levels in cellu-lar material, suggesting that these nutritional substituents are not converted totryptophan but to another substance replacing tryptophan in the cell.These data indicate that the pathway of tryptophan synithesis described in

the mold Neurospora is not present in L. arabinosuts.

REFERENCES

BRATTON, C., AND MARSHALL, 1E K., JR. 1939 A new coupling component for sulfanil-amide determination. J. Biol. Chem., 128, 537-550.

DAVIS, D., HENDERSON, L. MI., AND POWAELL, D. 1951 The niacin-tryptophan relationshipin the metabolism of Xanthomonas pr2uni. J. Biol. Chem., 189, 543-549.

FILDES, P. 1940 Indole as a precursor in the synthesis of tryptophan by bacteria. Brit.J. Exptl. Path., 21, 315-319.

FILDES, P. 1945 The biosynthesis of tryptophan by- Bacteriuml tlyphos unw. Brit. J. Exptl.Path. 26, 416-428.

GREEN, R. D., AND BLACK, A. 1943 A microbiological method for determination of trvpto-phane. Proe. Soc. Exptl. Biol. Med., 54, 322-324.

GREEN, R. D., AND BLACK, A. 1944 The microbiological assay of tryptophane in proteinsand foods. J. Biol. Chem., 155, 1-8.

KIKKAWA, H. 1950 Tryptophane synthesis in insects. Science 111, 495-496.LIPMANN, F. 1945 Acetylation of sulfanilamide by liver homogenates and extracts. J.

Biol. Chem., 160, 173-189.NASON, A. 1950 Effect of zinc deficiency on the synthesis of tryptophan by Neurospola

extracts. Science, 112, 111-112.NASON, A., KAPLAN, N., AND COLOWICK, S. 1951 Changes in enzymatic constitution in

zinc deficienit Neurospora. J. Biol. Chem., 188, 397-406.NOVELLI, D., AND LIPMANN, F. 1950 The catalytic function of coenizyme A in citric acid

sy-nthesis. J. Biol. Chem., 182, 213-228.O'KANE, D. J., AND GUNSALUS, I. C. 1948 Pyruvic acid metabolism: a factor required

for oxidation by Streptococcius faecalis. J. Bact., 56, 499-506.PRESCOTT, J. M.. SCHWEIGERT, B. S., LYMAN, C. M. AND KUIKEN, K. A. 1949 The effect

of D-tryptophan on the utilization of the L isomer by some lactic acid bacteria. J.Biol. Chem., 178, 727-732.

RYDON, H. N. 1948 Anthranilic acid as an intermediate in the biosynthesis of tryptophanby Bacterium7sl typhosuin. Brit. J. Exptl. Path., 29, 48-57.

SCHWEIGERT, B. S. 1947 The role of vitamin 136 in the synthesis of tryptophane fromindole and anthanilic acid by Lactobacillus arabinosus. J. Biol. Chem., 168, 283-288.

SCHWEIGERT, B. S., SAUBERLICH, H. E., BAUMIANN, C. A., AND ELVEHJEM, C. 1946 Thetryptophanie activity of various compounds for Lactobacillus arabinosus and their in-fluence oni the determination of trvptophane in natural materials. Arch. Biochem.,10, 1-8.

SHAW, J. D. L., AND MCFARLAND, W. D. 1938 The determination of tryptophane by amodified glyoxylic acid method employing l)hotoelectric colorimetry. Can. J. Re-search., Sec. B., 16, 361-368.

SNELL, E. E. 1943 Growth promotion on tryj)tophane-deficient miiedia by o-aminobenzoicacid an(l its attempted reversal with orthoarnilamide. Arch. B3iochem., 2, 389-394.

142 [VOL. 63

on March 1, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

1952] ANTHRANILIC ACID IN L. ARABINOSUS 143

STERN, J., AND OCHOA, S. 1949 Enzymatic synthesis of citric acid by condensation ofacetate and oxalacetate. J. Biol. Chem., 179, 491-492.

TATUM, E. L., AND BONNER, D. M. 1943 Synthesis of tryptophane from indole and serineby Neurospora. J. Biol. Chem., 151, 349.

TATUM, E. L., BONNER, D. M., AND BEADLE, G. W. 1943 Anthranilic acid and biosyn-thesis of indole and tryptophane by Neurospora. Arch. Biochem., 3, 477-478.

UMBREIT, W., WOOD, W. A., AND GUNSALUS, I. C. 1946 Activity of pyridoxal phosphatein tryptophane formation by cell-free enzyme preparations. J. Biol. Chem., 165, 731-732.

WORK, T. S., AND WORK, E. 1948 The Basis of Chemotherapy. Interscience Publishers,Inc., New York.

on March 1, 2020 by guest

http://jb.asm.org/

Dow

nloaded from