MODE OF ACTION OF CHLORAMPHENICOLIV. FAILURE OF SELECTED NATURAL METABOLITES TO REVERSE ANTIBIOTIC ACTION

H. E. HOPPS, C. L. WISSEMAN, JR., F. E. HAHN, J. E. SMADEL, AND R. HODepartment of Virus and Rickettsial Diseases, Walter Reed Army Institute of Research, Washington, 12,

D. C., and Department of Microbiology, School of Medicine, University of Maryland, Baltimore

Received for publication April 18, 1956

Chloramphenicol has been shown to exert astrong inhibitory action on microbial proteinsynthesis (Gale and Folkes, 1953; Maxwell andNickel, 1954; Wisseman et al., 1954). Hence, thehypotheses that the antibiotic may act simply asthe antagonist of one amino acid or another areespecially attractive (Mentzer et al., 1950; Wool-ley, 1950; Bergmann and Sicher, 1952; Molho andMolho-Lacroix, 1952). Nevertheless, examina-tion of the evidence in support of this proposedmechanism of action reveals that it is derivedfrom experiments in which reversal of growthinhibition by the respective substances wasgenerally minimal and was demonstrable onlyunder special circumstances. Moreover, results ofthis kind are not limited to selected amino acids;indeed, a wide variety of other substances hasalso been found to yield comparable degrees ofreversal of growth inhibition (Smith, 1952;Swenseid et al., 1952; Foster and Pittillo, 1953;Roblin, 1954; Hitchings and Elion, 1955). Al-though it is conceivable that a complex relation-ship of antagonists exists here, as in the caseof the sulfonamides (Work and Work, 1948),it is also possible that these minimal "reversaleffects" may be due to indirect factors whichinfluence bacterial growth in some other mannerand whose relation to chloramphenicol inhibitionis more non-specific in nature than is that usuallyincluded in the concept of antimetabolites asdiscussed by Work and Work (1948), Martin(1951) and Woolley (1952).The structural relation of chloramphenicol to

the aromatic amino acids phenylalanine, tryosineand tryptophan in particular, bears reconsidera-tion in discussing this problem. When the formulais written in the conventional two-dimensionalprojection (see A, figure 1), there is a suggestionof a resemblance to the aromatic amino acidsor to D-serine. However, the x-ray diffractionstudies of Dunitz (1952) on crystalline chloram-phenicol and bromamphenicol indicate that in

this physical state, a second six membered ringmay be formed by a hydrogen bond in thepropanediol moiety, giving the molecule a muchdifferent appearance (see B, figure 1). It is, ofcourse, possible that this ring does not actuallyexist in solution or under the conditions of anti-biotic action, but compounds in which a similarbut more stable ring structure has been sub-stituted actually have been synthesized andpossess antibiotic activity (U. S. Patent, 1952;Taylor, 1953). Formula C in figure 1 illustratesone of these compounds. An extensive discussionof the structural and configurational factorswhich contribute to the antibiotic activity ofchloramphenicol is published elsewhere (Hahnet al., 1956). Furthermore, the theoretical rela-tion of phenylalanine to chloramphenicol requiresthe assumption of four distinct substitutions(Woolley, 1950). This multiplicity of alterationsmight change the properties of the resultingcompound to such an extent that it could nolonger function as an antimetabolite of phenyl-alanine. In any case, the structural basis forrelating chloramphenicol to the aromatic aminoacids in an antimetabolite relationship wouldappear less well-founded than formerly supposed.The present report is an outgrowth of studies

which have been continued intermittently sinceour first reported failure to detect reversal ofgrowth inhibition by a variety of natural sub-stances (Wisseman et al., 1954). The results arepresented at this time because of their bearingon hypotheses concerning the mechanism bywhich protein synthesis is inhibited by chloram-phenicol. In the course of these studies therehave been many variations in conditions of test-ing, including different strains of organisms,different media and small as well as large inocula.The particular results selected for inclusion inthis report were obtained under conditionsdesigned to eliminate in so far as possible theinfluence of non-specific factors. In our experience

561

on August 17, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

HOPPS, WISSEMAN, HAHN, SMADEL, AND HO

(A) NHCOCHCI2

02N koeC ¢-CH2OH

OH H

(C) H

H~~~~~02N H0

Figure 1. Structure of chloramphenicol and one

of its cyclic derivatives.

all substances tested have uniformly failed toreverse specifically the growth inhibiting propertyof chloramphenicol.

MATERIALS AND METHODS

Test organisms and media. Escherichia colistrain B, obtained from Dr. Mark Adams, was

fully adapted to growth in the synthetic testmedium by serial transfer prior to its use in theexperiments. In like manner, a phenylalanine-requiring mutant of E. coli strain B/r (strainC-2), furnished by Dr. Joseph Gots, was adaptedto phenylalanine-fortified synthetic medium. Thebasic test medium employed in most experimentswith the E. coli strains was that of Fisher andArmstrong (1947) with glucose substituted forglycerol. However, in some experiments indicatedin the text below, the synthetic medium of Hooket al. (1946) was employed.

Leuconostoc citrovorum ATCC 8081 was main-tained on micro-inoculum agar (Difco) whileactual reversal studies were carried out in citrovo-rum broth (Difco).

Test substanees. Stock solutions of the variousamino acids and other substances were preparedin distilled water in a concentration 10 times thatdesired in the test media. Citrovorum factor was

donated by Dr. T. H. Jukes of Lederle Labora-tories in the form of the calcium salt. Chloram-phenicol, a gift of Parke, Davis and Co., wasdissolved in distilled water to give a stock solu-tion containing 1000 ,ug/ml. All solutions weresterilized by passage through a Seitz filter.

General procedure for reversal experiments withstrains of E. coli. The following general procedurewas employed for all the experiments withstrains of E. coli, except as indicated in theexploratory studies. Ten ml of a 16 hr culturegrown in the appropriate synthetic medium wastransferred to a 1-L flask containing 500 ml offresh medium. This culture was incubated in awater bath at 34 C and growth was followedturbidimetrically at 420 my until it attained theearly logarithmic phase. (The number of cellspresent in an early logarithmic phase culture didnot influence the minimal inhibitory concentra-tion of chloramphenicol; preliminary experimentsin which E. coli was tested in the conventionalserial dilution type of sensitivity test indicatedthat the minimal inhibitory concentration ofchloramphenicol after incubation for 18 hr at37 C was constant, regardless of the size ofinoculum which was varied between 5 X 101organisms per ml and 5 X 106 organisms per ml.)Then, 8.0 ml aliquots were quickly distributed tosterile cuvettes which contained 1.0 ml of testsubstance and 1.0 ml of chloramphenicol, each 10times the desired final concentration. Each seriesof tubes contained a constant amount of testsubstance but the final chloramphenicol con-centration was varied from 0.039 to 10 ,ug/ml in 2-fold increments of concentration. A control tubecontaining all additions except chloramphenicolwas included in each series. A similar series oftubes containing chloramphenicol in whichdistilled water was substituted for the metaboliteto be tested for reversing action was prepared forcomparison and was run simultaneously withevery series containing test substances. Thecuvettes were then incubated at 34 C in a waterbath equipped with forced circulation. Opticaldensities of each cuvette were read in a ColemanModel 14 spectrophotometer at a wave length of420 m,u at 30-min intervals over a 3-hr periodor until the maximum stationary growth phasewas attained.

Procedure for reversal studies with L. citrovorum.Two flasks of citrovorum factor medium (DifcoManual-9th edition) were inoculated withwashed (3 times) L. citrovorum strain 8081.

562 [VOL. 72

on August 17, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

MODE OF ACTION OF CHLORAMPHENICOL. IV

TABLE 1Failure of L-phenylalanine or L-tryptophan to reverse inhibition of growth of Escherichia coli strain B by

a sub-bacteriostatic concentration of chloramphenicol

Chloramphenicol 3-phenylalanine (growth rate) L-tryptophan (growth rate)

pg/mi 0 pg/ml 2000 pg/ml 0 pg/ml 1000 pg/mlControl ....................... 0 0.294 0.340 0.270 0.315Chloramphenicol ............... 1.0 0.183 0.220 0.171 0.206

Relative growth rate* ........ 62 65 63 65

Relative = growth rate with chloramphenicol x 100.growth rate of control

The medium in one flask contained 0.3 mugcalcium citrovorum per ml, the minimal amountyielding optimal growth under the conditions ofthe test, while the medium in the other flaskcontained 30 mpg calcium citrovorum per ml.The cells from each flask were collected bycentrifugation, washed three times and re-suspended in citrovorum factor broth. Twoseries of cuvettes, one containing a final con-centration of 0.3 and the other containing 30mpgg calcium citrovorum per ml, were inoculatedwith a sufficient amount of washed cell suspensionto give an initial optical density of 0.150 at awavelength of 620mpin the Coleman instrument.Each series contained a chloramphenicol-freecontrol and a range of chloramphenicol concen-trations as described in the preceding section. Thetubes were incubated at 34 C and the turbiditywas measured at 30 min intervals over a periodof 5 hr.

RESULTS

Reversal studies with E. coli strain B. Explora-tory experiments were designed to study theeffect of r-phenylalanine and L-tryptophan on thegrowth retarding action on E. coli strain B of achloramphenicol concentration well below thatrequired to inhibit growth completely. The lowantibiotic concentration was employed to in-crease the chances of detecting minor reversingactions. Hook's (1946) synthetic medium wasemployed here. A 16-hr culture of the organismgrown in this medium was washed by centrifuga-tion and resuspended in sufficient fresh mediumto give an optical density in the Beckman modelDU spectrophotometer of approximately 0.030in a 1-cm cell at 420 m,. This diluted suspensionwas divided into four 100-ml aliquots; additionswere then made to give cultures of the following

composition: (1) basal medium alone, (2) basalmedium containing 1.0 pug chloramphenicol/ml,(3) basal medium with L-phenylalanine (2000,ug/ml) or L-tryptophan (1000 pug/ml), and(4) basal medium containing chloramphenicol(1.0,g/ml) and I-phenylalanine (2000 pg/ml) ori-tryptophan (1000 pug/ml). The cultures wereplaced in a 34 C waterbath and after a 10-15min equilibration period, were found to have re-sumed the logarithmic phase of growth. Sampleswere removed from each culture at 10-min inter-vals over a period of 1.5-2.0 hr for optical densitymeasurements. The slope of the straight linewhich resulted from plotting log optical densityagainst time was used as a measure of the growthrate. The average of the results of three separateexperiments each with r-phenylalanine andL-tryptophan are recorded in table 1.The concentration of chloramphenicol em-

ployed in these experiments permitted the testorganism to grow in the basal medium at a rate62-63 per cent of that of the uninhibited culturein the same medium. The addition of phenyl-alanine or tryptophan to the basal mediumregularly caused the growth rate of E. coli strainB to increase up to 15 or 20 per cent over thatobserved in the basal medium alone. However,despite the stimulatory action of the aminoacids on growth, the relative magnitude of thechloramphenicol effect on growth remainedunaltered; that is, 1.0 pAg/ml of chloramphenicolin the amino acid-fortified medium still reducedthe rate of growth of the test organism to 65 percent of that of the control culture, which in thisinstance was a drug-free culture in the fortifiedmedium. This value does not differ significantlyfrom the one obtained in the unfortified basalmedium. Had a specific reversal of chlorampheni-col action occurred in the fortified medium, the

19561 563

on August 17, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

HOPPS, WISSEMAN, HAHN, SMADEL, AND HO

relative growth rate in the presence of 1.0 ,g/mlof the antibiotic in this medium should have beensignificantly higher. Since this is not the case,these experiments give no indication of anyspecific reversing action that could be ascribed toantagonism of an antimetabolite.More extensive studies were then performed

with these same amino acids, as well as with anumber of other natural metabolites. The methodemployed here was more sensitive and re-producible' than that employed in the explora-tory studies described above. It was based on adetermination of the 50 per cent effective dose(ED5o), that is, the amount of chloramphenicolrequired to reduce the rate of growth of a log-arithmic phase culture of the test organism to 50per cent of the rate of the corresponding anti-biotic-free control. This value was determinedsimultaneously in every instance both in thepresence and in the absence of the compoundbeing tested for reversing action. The methodwas broadly patterned after those described byKohn and Harris (1941) and by Joslyn andGalbraith (1950); special modifications, however,were introduced to cope with factors, other thanreversal of antimetabolic action, which have beensuggested by several investigators as potentiallycapable of influencing this type of experiment.These modifications were: (1) elimination orminimization of the lag phase which may intro-duce serious variations into the results, (2)minimization of the effects due to bacterialdegradation of the antibiotic (Smith and Worrel,1950) by reducing the time of exposure of thedrug to the microorganisms, (3) permitting onlya few cell divisions to occur during the experi-ment so as to give little opportunity for over-growth of resistant variants or mutants (Fosterand Pittillo, 1953), (4) compensation for non-specific effects of added nutrients on either therate or the maximal final levels of growth at-tained.Growth stops very rapidly when concentrationsI The competitive antagonism which exists

between p-aminobenzoic acid and sulfadiazinewas readily demonstrated by this method in pre-liminary studies. Furthermore, it would be an-ticipated that minimal degrees of reversal, of thekind expected in the case of non-competitive an-tagonism, would be detectable at the lower endof the dose-response curve in the form of diver-gence of the curve obtained in the fortified mediumfrom that obtained in the basal medium.

of chloramphenicol down to the bacteriostaticlevel are introduced into an actively growingculture of E. coli (Bozeman et al., 1954; Wissemanet al., 1954). Over a broad range of chlorampheni-col concentrations below the minimum bacterio-static level, growth proceeds more slowly than inthe antibiotic-free control but it is still expo-nential in character over most of the observationperiod so that the logarithm of the opticaldensity plotted against time forms a straightline; its slope bears an inverse relation to chlor-amphenicol concentration. The slope (K), whichis an expression of the growth rate, was calcu-lated by the equation

K= log D, -log DIT2-Tl (1)

in which DI and D2 are the optical densities atthe beginning and end of the straight segment ofthe growth curve and T1 and T2 are correspondingtimes in hours. Then, in order to compensate forany general stimulation or retardation of growthcaused by the addition of the metabolite beingtested, the relative growth rate at each chlor-amphenicol concentration in a given test wasexpressed as per cent of the growth rate exhibitedby the corresponding antibiotic-free controlculture, i. e., the one which contained all addi-tions except the antibiotic. This was calculatedby the equation.

Relative growth rate = K xX1Koontrol

(2)

in which Kc, is the slope of the growth curve forany given chloramphenicol concentration andKoontro1 is the slope of the growth curve of thecontrol. When the relative growth rates of thepartially inhibited cultures are plotted againstthe logarithm of the corresponding concentra-tions of chloramphenicol, a nearly symmetricalsigmoid curve is obtained (figure 2) which hasthe general form of dose-response curves givenby many biological systems (Bliss, 1950). Thecentral portion of the curve approaches linearityand, hence, the 50 per cent growth inhibition levelof chloramphenicol is readily determined graph-ically.

Phenylalanine, tryptophan, aspartic acid,glutamic acid, glycine and p-aminobenzoic acidwere tested for their capacity to reverse thegrowth inhibition effect of chloramphenicol onE. coli strain B. The results obtained, in each

564 [VOL. 72

on August 17, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

MODE OF ACTION OF CHLORAMPHENICOL. IV

DL-PHENYLALANINE CONC:

_-. 31.2 ug/mL*-e 1200 ,U,g/ml.

TABLE 2Failure of selected metabolites to reverse inhibition

of growth of Escherichia coli strain B bychloramphenicol

Test Compound Concen- Reversaltration Ratio*

pg/ml

L-Aspartic acid.100 0.95L-Glutamic acid. 100 1.09Glycine.100 1.04p-Aminobenzoic acid 10 1.11DL-Phenylalanine. 100 1.10DL-Phenylalanine.. . . 1000 1.04L-Tryptophan. 100 1.07

0 20 40 60 80 100GROWTH: PERCENT CONTROL

Figure 2. Dose-response curves of cultures ofEscherichia coli strain C-2 to chloramphenicol inthe presence of phenylalanine.

instance the average of three or more separateexperiments, are recorded in table 2 and are

expressed as the ratio

EDO0 (CM) with metaboliteED5o (CM) without metabolite

Thus, the ratio obtained in a test with a sub-stance having no reversing effect would be 1.0;the ratio for a substance which had a reversingaction would be significantly greater than 1.0.It is clear from the table, in which all the ratiosare near 1.0, that none of the substances causedany significant reversal of growth inhibition.

Reversal studies with phenylaanine employinga phenylalanine-requiring mutant of E. coli.Extensive studies with phenylalanine were

performed using a phenylalanine-requiring mutantof E. coli as test organism since experience hasshown that demonstration of antagonism may

prove unsuccessful except with an organismdependent upon an extrinsic source of themetabolite under consideration (Woolley, 1952).The effect of chloramphenicol on the growth ofthis mutant was tested over a range of antibioticconcentrations which extended from one whoseeffect was barely detectable to one which causedalmost complete inhibition of growth; two con-

centrations of DL-phenylalanine were employed,i. e., (1) 31.2 ,ug/ml which was 5 times theminimal level that gave optimal growth under the

I I I I I * Reversal ratio =ED50 (CM) with test compound

ED50 (CM) without test compound

conditions of the tests and (2) 1200 ;ug/ml whichwas 192 times the minimal optimal concentra-tion. The dose-response curves for each seriesare presented in figure 2. The ED50(CM) was1.22 i 0.09 ,ug/ml for the lower phenylalaninelevel and 1.24 :1: 0.11 ,ug/ml for the higher con-centration, giving a ratio of 1.01. The differencebetween the two values is small and statisticalanalysis showed that it is well within the limitsof accuracy of the method. The possibility re-mained that a noncompetitive antagonism notapparent at the ED5o(CM) level might bedemonstrable at chloramphenicol concentrationswell below the ED50(CM) level. However, frominspection of the dose-response curves it is seenthat, even at the lowest antibiotic levels givingperceptible growth retardation, the two curvesare almost identical. This impression is sub-stantiated by a probit analysis of the data whichindicated that the differences were not significantand that the two curves were essentially super-imposable (95 per cent confidence limits).

Reversal studies with citrovorum factor andLeuconostoc citrovorum. Three-tenths m,ug ofcitrovorum factor per ml were sufficient foroptimal growth of this test organism. TheED5o(CM) for this organism was the same re-gardless of whether the medium contained thisminimal concentration of citrovorum factor or100 times as much, i. e., 30 m,ug per ml.

DISCUSSION

The failure of selected amino acids to reversethe action of chloramphenicol under the de-

0Z

10Oz 50

-J 2.50Uz 1.25a-, .6250:-J .312-

8 .156-J

.078

qla

_-

5651956]

on August 17, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

HOPPS, WISSEMAN, HAHN, SMADEL, AND HO

scribed experimental conditions is in discord withthe idea that the drug acts as an antimetaboliteof an amino acid. Despite the negative nature ofthe findings, they bear significantly on thetheories concerning the mode of action of chlor-amphenicol.

Particular effort was made to employ testconditions which were heavily weighted in favorof obtaining reversal effects. For example, thewidest possible range of chloramphenicol concen-trations employed should have permitted thedetection even of minimal antagonistic effects ofa noncompetitive nature when the proper aminoacids were supplied in sufficiently high concen-trations. Furthermore, the present experimentswere carried out with bacteria in the logarithmicphase of growth, i. e., under cultural conditionsalmost identical with those under which theaction of chloramphenicol on bacteria proteinsynthesis has been demonstrated (Wissemanet al., 1954).While it may be inferred that antagonism of

amino acids is not the mechanism by whichchloramphenicol inhibits protein formation, thepossibility is not excluded that the drug mightbe a biological antagonist of some other essentialsubstance, for example, a catalyst which per-haps is involved in protein synthesis. Until suchaction is actually demonstrated, however, thepossibility must also be considered that themechanism of action of chloramphenicol is funda-mentally different and not based on antimetabo-lism of structural analogues.

SUMMRYThe presence of phenylalanine, tryptophan,

glycine, aspartic acid and p-aminobenzoic acidin logarithmic growth phase cultures of Escher-ichia coli strain B failed to reverse the inhibitoryeffect of chloramphenicol under a variety ofexperimental conditions. Similarly, no reversalwas demonstrable in systems which employed aphenylalanine-requiring mutant of E. coli or inLeuconostoc citrovorum which requires citrovorumfactor for growth. Observations which lead to thebelief that chloramphenicol does not act as asimple amino acid antagonist are discussed.

REFERENCESBERGMANN, E. D. AND SICHER, S. 1952 Mode

of action of chloramphenicol. Nature, 170,931-932.

Buss, C. I. 1950 The design of biologicalassays. Ann. N. Y. Acad. Sci., 52, 817-888.

BOZEMAN, F. M., WISSEMAN, C. L., JR., HOPPS,H. E., AND DANAUSKAS, J. X. 1954 Actionof chloramphenicol on T-1 bacteriophage.I. Inhibition of intracellular multiplication.J. Bacteriol., 67, 530-536.

DUNITZ, J. D. 1952 The crystal structure ofchloramphenicol and bromamphenicol. J.Am. Chem. Soc., 74, 995-999.

FISHER, K. C. AND ARMSTRONG, F. H. 1947 Theeffects of sulfathiazole and of propyl car-bamate on the rate of oxygen consumptionand growth in Escherichia coli. J. Gen.Physiol., 30, 263-278.

FOSTER, S. W. AND PITTLLO, R. F. 1953 Rever-sal by complex natural materials of growthinhibition caused by antibiotics. J. Bac-teriol., 65, 361-367.

GALE, E. F. AND FOLiSB, J. P. 1953 The assimi-lation of amino acids by bacteria. 19. Theinhibition of phenylalanine incorporation inStaphylococcus aureus by chloramphenicoland p-chlorophenylalanine. Biochem. J.(London), 55, 730-735.

HAHN, F. E., HAYES, J. E., WISSEMAN, C. L.,JR., Hopps, H. E. AND SMADEL, J. E. 1956Mode of action of chloramphenicol. VI.Relation between structure and activity inthe chloramphenicol series. Antibiotics andChemotherapy, 6, 531-543.

HIrcvMNGS, G. H. AND ELION, G. B. 1955 TheLactobacillus casei screening test. Investiga-tion of diverse systems for cancer chemo-therapy screening. Cancer Research, Sup-plement No. 3, 66-68.

HooK, A. E., BEAD, D., TAYLOR, A. R., SHARP,D. G., AND BEARD, J. W. 1946 Isolation andcharacterization of the T2 bacteriophage ofEscherichia coli. J. Biol. Chem., 165, 241-258.

JOSLYN, D. A. AND GALBRATH, M. 1950 Aturbidimetric method for the assay of anti-biotics. J. Bacteriol., 59, 711-716.

KOHN, H. I. AND HARRIS, J. S. 1941 On themode of action of the sulfonamides. I.Action on Escherichia coli. J. Pharmacol.and Exptl. Therap., 73, 343-382.

MARTIN, G. J. 1951 Biological antagonism.The Blakiston Co., Philadelphia.

MAXWELL, R. E. AND NICKEL, V. S. 1954 Theantibacterial activity of the isomers of chlor-amphenicol. Antibiotics and Chemotherapy,4, 289-295.

MENTZER, C., MEUNIER, P., AND MOLHO-LACROIX,L. 1950 Faits de synergie et d'antagonism

566 [voL. 72

on August 17, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

MODE OF ACTION OF CHLORAMPHENICOL. IV

entre la chloromyce6tine et divers aminoacides vis-A-vis des cultures d'E. coli. Compt.rend. soc. biol., 230, 241-243.

MOLHO, D. AND MOLHO-LACROIX, L. 1952 ftudecomparee de l'antagonisme entre quelquesd6rivds de la ph6nylalanine et la chloromyce-tine, la B2-thi6nylalanine et la fl-ph6nyls6rine.Bull. Soc. Chim. Biol., 34, 99-107.

ROBLIN, R. O., JR. 1954 Metabolite antago-nists. Ann. Rev. Biochem., 23, 501-526.

SMITH, G. N. AND WORREL, C. S. 1950 Thedecomposition of chloromycetin (chloram-phenicol) by microorganisms. Arch. Biochem.28, 232-241.

SMITH, G. N. 1952 The influence of chloromy-cetin decomposition products on the growthof Escherichia coli and their effects on revers-ing the growth inhibitory action of the anti-biotic. Arch. Biochem. and Biophys., 40,314-322.

SWENSEID, M. E., WRIGHT, P. D., AND BETHELL,F. H. 1952 A growth factor for L. citro-vorum synthesized by hemopoietic tissue re-

versing aminopterin and chloromycetin in-hibition. Proc. Soc. Exptl. Biol. Med., 80,689-690.

TAYLOR, E. P. 1953 A tasteless derivative ofchloramphenicol. J. Pharm. and Pharma-col., 5, 254-256.

U. S. Patent Number 2,587,641. Nov. 4, 1952(Ring derivatives of chloramphenicol).

WISSEMAN, C. L., JR., SMADEL, J. E., HAHN, F.E., AND Hopps, H. E. 1954 Mode of actionof chloramphenicol. I. Action of chloram-phenicol on assimilation of ammonia and onsynthesis of proteins and nucleic acids inEscherichia coli. J. Bacteriol., 67, 662-673.

WOOLLEY, D. W. 1950 A study of non-competi-tive antagonism with chloromycetin and re-lated analogues of phenylalanine. J. Biol.Chem., 185, 293-305.

WOOLLEY, D. W. 1952 A study of antimetabolites.John Wiley and Sons, Inc., New York.

WORK, T. S. AND WORK, E. 1948 The basis ofchemotherapy. Interscience Publishers, Inc.,New York.

19561 567

on August 17, 2019 by guest

http://jb.asm.org/

Dow

nloaded from