EFFECT OF CHLORAMPHENICOL ON THE SURVIVAL OF ESCHERICHIA COLIIRRADIATED WITH ULTRAVIOLET LIGHT'

HIROMU OKAGAKIDepartment of Anatomy and Cytochemistry Laboratory, Yamaguchi Medical School, Ube, Japan

Received for publication July 20. 1959

Chloramphenicol is known to inhibit proteinsynthesis in different microorganisms withoutaffecting the syntheses of nucleic acids (Galeand Folkes, 1953; Wisseman et al., 1954; Aronsonand Spiegelman, 1958). Harold and Ziporin(1958) and Drakulic and Errera (1959) reportedthat the treatment of Escherichia coli withchloramphenicol immediately after ultravioletirradiation prevented the resumption of de-oxyribonucleic acid (DNA) synthesis. Doudneyand Haas (1958) showed that a considerablerecovery of colony forming ability in ultravioletirradiated E. coli strain B occurred after irradia-tion in the absence of nitrogen source or in thepresence of chloramphenicol. A similar observa-tion was reported by Gillies and Alper (1959).

Contrary to such observations, a differentresponse to the same antibiotic was encounteredwhen one employed several other strains ofE. coli; the ultraviolet irradiated bacteria maydie when they are held in media containingchloramphenicol. The nature and the mechanismof this phenomenon is the subject of this paper.

MIATERIALS AND METHODS

Bacterial strains. E. coli strains 15, 15 T-(athymine-auxotroph employed by Cohen andBarner, 1954; to be designated 15 T-(Cohen))and 15 T-(another thymine-auxotroph employedby Zamenhof and Griboff, 1954; designated15 T-(Zamenhof)) wereobtained from theDepart-ment of Zoology, Columbia University; strain15 h-(requiring histidine) and B/r from theDepartment of Genetics, Osaka University. StrainB, originating from the stock used by Fujisawaand Sibatani (1954) has been maintained in thisschool. The strain 15-0S21 is a back mutant of15 T-(Cohen) produced in the course of thisstudy. Cultures were maintained on agar slantsand transferred at 4- to 6-week intervals.

1 Aided in part by a grant from the ScientificResearch Funds of the Ministry of Educationawarded to Professor A. Sibatani.

Growth and irradiation of bacteria. In most ofthe experiments Tris (tris(hydroxymethyl)amino-methane)-glucose medium (0.1 M Tris buffer (pH7.4), NaCl, 92mM; KCl, 40 mM; NH4Cl, 21 mM;CaCL2, 0.1 mM; MgSO4, 0.41 mM; KH2PO4,0.64 mM; and glucose, 0.4 per cent) and Tris-glucose-Casamino acid medium (the above me-dium fortified with 0.1 per cent Casamino acid(Difco)) were employed. The media were sup-plemented with appropriate compounds whenrequired for the auxotrophs or for particularexperimental conditions. Bacteria were grownfor 14 to 16 hr at 37 C under constant aeration,harvested, suspended in appropriate growthmedia at a density of about 3 X 108 cells per mland incubated for 3 to 6 hr until the late log-arithmic phase was reached, the growth beingfollowed by optical density measurement withKaken (Tokyo) 570 m, filter. The density ofviable cells at the expected phase was about2 to 3 X 109 cells per ml. The bacteria werethen harvested, washed with ice cold 0.1 Mphosphate buffer (pH 7.0), suspended in thesame buffer at a density of about 3 X 109 cellsper ml and kept at 0 C. Ten ml of bacterialsuspension were irradiated in a petri dish of8.5 cm diameter with ultraviolet light from aToshiba (Mazda) 15 w germicidal lamp at adistance of 1 m under contant stirring. Theirradiated suspension was diluted with 10 volumesof experimental medium with or without chloram-phenicol (100 ,g per ml) and incubated at 37 Cwith constant aeration. Samples were taken atintervals for viable cell counts or biochemicalanalyses. Necessary cautions were observed toavoid the photoreactivation in all manipulations.

Analytical methods. Viable cells were estimatedon agar plates. Schneider's method was usedfor the determination of nucleic acids and pro-tein. DNA-P and ribonucleic acid (RNA)-Pwere measured by diphenylamine test (Burton,1956) and the conventional orcinol test, re-spectively. Protein was measured by the Folin'sreaction (Lowry et al., 1951).

277

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

OKAGAKI

RESULTS

Strai-ns of E. coli susceptible to the killing effect ofchloramphenicol after ultraviolet irradiation. Whenthe ultraviolet irradiated cells of E. coli strain15 were suspended in a growth medium containingchloramphenicol and incubated with aeration,there was a phase of initial increase in viablecount, lasting some 20 min without the con-comitant synthesis of DNA ;2 such a recoveryfrom the primary lethal effect of ultravioletirradiation was followed by a steep decrease inviable count which continued for some 120min of incubation and then leveled off (figures1 and 5). Similar but more distinct reduction ofsurvivors was observed with strains 15 T-(Cohen)and 15 T-(Zamenhof) (figure 6 and table 1),and a less distinct one with strain B/r (figure 2).In strain 15 h-, the decrease started from thebeginning of incubation without a sign of re-covery or a delay (figures 8 and 9).As for strain B, the effects of the antibiotic

were different. In Tris-glucose-Casamino acidmedium containing chloramphenicol, the numberof survivors occasionally decreased during theinitial 20 min and then started to increase (figure3, A and B, bottom curves), but did not exceedthat of the control growing in Tris-glucose-Casamino acid medium. It is obvious that thisincrease was not due to the proliferation ofsurviving bacteria, but was a reflection of therevival of once killed bacteria. When the additionof chloramphenicol was made at 40 min afterincubation began in Tris-glucose-Casamino acidmedium, the number of survivors significantlyincreased and surpassed that of the control(figure 3, A). In a medium lacking nitrogen3 but

2 The initial recovery from the primary lethaleffect of ultraviolet irradiation and the beginningof the death in the chloramphenicol medium wasparalleled to some extent by the change of viablecount in the growth medium without chloram-phenicol. However, in the latter case this second-ary death after the recovery from the primarydeath was of short duration and the normalbacterial growth set in some 90 min after theultraviolet irradiation. This phenomenon wasstudied by Barner and Cohen (1956) who ascribedthe secondary death to the effect of unbalancedgrowth.

3 When irradiated bacteria were incubated inphosphate buffer with glucose, there was a markedrecovery of initially "killed" bacteria, i. e., therestoration of colony forming ability. This phe-

z0

>en

0

4

IJL

z5:

CO)

0-i

0

-_F-

0 60 120 180

MINUTESFigure 1. Effect of chloramphenicol oIn the

survival of ultraviolet-irradiated Escherichia colistrain 15. Incubation in Tris-glucose-Casaminomedium for 3 hr. Ultraviolet irradiation for 100sec. The irradiated bacteria were suspended inTris-glucose-Casamino medium with or withoutchloramphenicol and incubated. 0, Nonirradiatedcontrol; A, the same, with chloramphenicol; 0,irradiated; A, irradiated, with chloramphenicol.

containing chloramphenicol, the antibiotic evi-dently acted to increase survivors considerablyeven when addition was made at zero time ofincubation (figure 3, B). The situation is socomplicated that the mode of action of chloram-phenicol cannot be accounted for simply, but atleast it suggests that there was little, if any,trend of strain B being killed by chloramphenicolafter ultraviolet irradiation.The change in DNA, RNA, and protein con-

tent of the culture was followed in the experi-

nomenon was observed most markedly in Escher-ichia coli strain B, to a lesser extent in strains15, 15 T-(Cohen), 15 T-(Zamenhof) and veryslightly in strain B/r, but not at all in strain 15h- as illustrated in figures 2, 3, 5, 8, 11, 12, and 14.These observations will be discussed in a separatepaper (Okagaki, unpublished data).

278 [VOL. 79

A ---A-

12

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

CHLORAMPHENICOL DEATH OF IRRADIATED E. COLI

mentcontr4mediuperio(min

z0

0

Figof uliB/r. Ifor 3irradi

Casanbufferchlora

of strain B presented in figure 3, A. In a light employed. The addition of chloramphenicolol culture with Tris-glucose-Casamino acid at zero time of incubation prevented the resump-im the DNA synthesis was resumed after a tion of DNA synthesis, but not if added 40d of complete inhibition lasting for some 30 min after incubation. These observations confirmof incubation by the dose of ultraviolet the results of Harold and Ziporin (1958) and of

Drakuli6 and Errera (1959). The RNA synthesiscontinued at a reduced rate for 80 min afterthe addition of chloramphenicol then leveled off.Protein synthesis was always completely inhibitedimmediately upon addition of chloramphenicol.

_ ____0 Similar effects of chloramphenicol were observedin ultraviolet irradiated E. coli strain 15 h- in-cubated in Tris-glucose medium containing his-tidine.

In unirradiated bacteria, the increase in opticaldensity and cell multiplication was blocked bychloramphenicol (figure 1). In this case, theeffect of the antibiotic was bacteriostatic butnever bactericidal with any strains of E. coli

o 60 120 ISO 240 tested, whereas it may be bactericidal with theultraviolet irradiated bacteria of certain strains.

MINUTES This curious phenomenon, the further decreaserure 2. Recovery of chloramphenicol death of colony forming ability in the ultraviolettraviolet irradiated Escherichia coli strain irradiated bacteria as caused by the postirradia-ncubation in Tris-glucose-Casamino medium tion incubation in media containing chloram-hr. Ultraviolet irradiation for 100 sec. Theated bacteria were incubated in Tris-glucose- phenicol, may be termed "chioramphenicolnino medium (0), or in 0.1 M phosphate death" of the ultraviolet irradiated bacteria,with 0.02 M glucose (0) or with glucose and or simply as "ultraviolet-chloramphenicol death."Lmphenicol (A). Effects of the composition of incubation medium

0

-I

0

4

cr -2

IL

CD)

(l:

to

.j

-4.

I t

A

Il0 60 120 180 0 60 120 ISO 240

MINUTES

Figure S. Effect of chloramphenicol on the ultraviolet irradiated Escherichia coli strain B. Incubationin Tris-glucose-Casamino medium for 3 hr. (A) Irradiated for 30 sec and incubated in 0.1 M phosphatebuffer with 0.02 M glucose (0) or in Tris-glucose-Casamino medium (0). Chloramphenicol was added(as indicated by the arrows) to the latter culture at 0 (A) or at 40 min (V) of incubation. (B) Irradiatedfor 100 sec and incubated in phosphate buffer with glucose (0) or with glucose and chloramphenicol(A), in Tris-glucose-Casamino medium (0) or in this medium with chloramphenicol (A).

-I

0r-

I-uaCD2

5o

0-J

1960] 279

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

280

0

z0

C)4

L-

z

(I)

C,0>

-I

- 2

OKAGAKI

0

z0

z

Co0

4c

-3_

0 60 120 180MINUTES

Figure 4. Modification of the ultraviolet-chlor-amphenicol death of Escherichia coli strain 15 bythe incubation in deficient media or in the presenceof arsenate. Incubation in Tris-glucose mediumfor 5 hr. Ultraviolet irradiation for 100 sec. Theirradiated bacteria were incubated in Tris-glucosemedium with chloramphenicol (A), in phosphatebuffer with glucose and chloramphenicol (A), inphosphate buffer with chloramphenicol (O), or inTris-glucose medium with chloramphenicol and0.01 M arsenate (A).

upon the ultraviolet-chloramphenicol death. TheTris-glucose medium and the phosphate bufferwith glucose were compared for efficiency inproducing ultraviolet-chloramphenicol death withstrain 15 (figure 4). The decrease in viable countwas a little more pronounced in the first mediumthan the second. With strain 15 T-(Cohen),however, no difference was observed betweenthe two media (figure 6), and also the presenceor absence of Casamino acid and thymine inthe medium had no effect (figures 6 and 14).Ultraviolet-chloramphenicol death was thereforenot appreciably dependent upon the nitrogensource in the medium. In the presence of chloram-phenicol, the thymineless death (Cohen andBarner, 1954) of strain 15 T-(Cohen) was blocked.When chloramphenicol was employed in com-

bination with a medium lacking energy source,the viable count of ultraviolet irradiated bacteriaof strain 15 did not decrease for the initial 120min of incubation (figure 4). Metabolic inhibitors,

-I

-2

-3 _-I

[VOL. 79

-a------ --_

tz1

0 60 120 180

MINUTESFigure 5. Recovery, and the effect of 2,4-dini-

trophenol on the chloramphenicol death, in ultra-violet irradiated Escherichia coli strain 15. Incu-bation in Tris-glucose medium for 5 hr. Thebacteria were irradiated for 100 sec and incubatedin phosphate buffer with glucose (0) or in Tris-glucose medium with chloramphenicol (A). 2,4-Dinitrophenol was added to the latter medium(final 5 X 10-5 M) at 0 (U) or at 40 min (+) ofincubation.

such as 0.01 M arsenate and 5 X 10-5 M 2,4-dinitrophenol, also partially relieved the bacteriafrom the ultraviolet-chloramphenicol death (fig-ures 4 and 5).A death analogous to the chloramphenicol death in

the ultraviolet irradiated strain 15 h-. Strain 15 h-exhibited a similar decrease in viable count afterultraviolet irradiation, when its protein synthesiswas inhibited by histidine deficiency (figure 7).However, death was prevented in a mediumlacking all sorts of nitrogen sources (figure 8).Unirradiated cells of this strain exhibited a veryslow decrease in viable count in the histidinedeficient growth medium, but this death wasnot so remarkable as the one occurring afterthe ultraviolet irradiation (figure 7). When his-tidine was supplied to the Tris-glucose mediumat 60 min of incubation, the death of the irra-diated bacteria was promptly arrested and theviable count started to increase parallel to theone with zero time addition of histidine (figure 7).Death of the ultraviolet irradiated histidine-

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

CHLORAMPHENICOL DEATH OF IRRADIATED E. COLI

0r

-I

z0

IL

z

Ul)

CD0

-2

-31

-4

-51

0 60 120 180

MINUTES

Figure 6. The survival and the chloramphenicoldeath of ultraviolet irradiated Escherichia colistrain 15 T-(Cohen) in different media. Incubationin Tris-glucose-Casamino medium with 2,ug perml thymine for 3 hr. The bacteria were ultravioletirradiated for 100 sec and incubated in Tris-glucose-Casamino medium with 2 jug per ml thy-mine (0), in Tris-glucose-Casamino medium only(X), in phosphate buffer with glucose and chloram-phenicol (A), in Tris-glucose-Casamino mediumwith thymine and chloramphenicol (A) or inTris-glucose-Casamino medium with chloram-phenicol (V). The broken line represents thethymineless death (Cohen and Barner, 1954).

auxotroph caused by histidine deprivation mightbe taken to be analogous with the ultraviolet-chloramphenicol death, since the protein syn-theses were largely inhibited in both cases; but itis evidently different from the latter in itsdependence upon some other nitrogen source.

When the effects of histidine deprivation andthose of chloramphenicol were compared in

Tris-glucose medium, the difference emerged

early after ultraviolet irradiation. As shown infigures 8 and 9, in the case of histidine depriva-tion the decrease in viable count began after ashort lag and its curve followed the courseparallel to, but at a somewhat higher level than,the one with chioramphenicol added at zero timeof incubation. But the addition of chloram-phenicol at 60 min tended to diminish the gapbetween the two curves. This experiment sug-gested the presence of an intracellular storageof histidine at the beginning of incubation andthe occurrence of some protein synthesis whichmay have continued during the course of incuba-tion though at a much diminished rate. Needlessto say, chloramphenicol caused the death ofbacteria of this strain regardless of the presenceor absence of histidine (figures 8 and 9). It maytherefore be assumed that this difference is not anessential one.The change in the amount of DNA, RNA, and

protein was followed in the ultraviolet irradiatedcellsof strain 15 h- both in histidine-supplementedand hiistidine-deficient media (figure 10). In theformer medium, the DNA synthesis resumedafter some 60 min of incubation after ultravioletirradiation, whereas RNA and protein synthesesbegan immediately. In the medium lacking his-tidine, both protein and DNA syntheses wereblocked, and RNA synthesis was inhibited al-most completely. When the irradiated bacteriawere subjected to histidine deficiency for 30or 60 min preceding the addition of histidine,the resumption of DNA synthesis in the sup-plemented medium was further delayed andproceeded at a much diminished rate or failed tooccur. The RNA and protein syntheses werealso impaired by the preliminary incubationin the histidine-deficient medium. It is thus shownthat the ultraviolet irradiated bacteria lost theability to grow during the postirradiation incuba-tion without histidine.

Restoration of the irradiated bacteria from thestate susceptible to ultraviolet-chloramphenicol death.The extent of the ultraviolet-chloramphenicoldeath seemed to be most pronounced when theaddition of chloramphenicol was made at zerotime of incubation of ultraviolet irradiatedbacteria (figure 8). It may be assumed that thebacteria would recover, during the period theywere incubated in a chloramphenicol-free medium,from the irradiation lesion which makes themsusceptible to the ultraviolet-chloramphenicol

1960] 281

".X-"

x

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

TABLE 1Effect of Casamino acid present in the preincubation medium on the survival ratio and the following

chloramphenicol death, in ultraviolet irradiated Escherichia coli strains 15,16 T-(Zamenhof), and 15 T-(Cohen)

Bacteria were grown in the media with or without Casamino acid for 3 to 63 hr until the late logarith-mic phase was reached, harvested, resuspended in buffer, ultraviolet irradiated for 100 sec, and incu-bated for 120 min in 0.1 M phosphate buffer containing 0.02 M glucose and 100 ug chloramphenicolper ml.

Survival

Bacterial Strain Incubation Medium Chloramphen- Ultraviolet-Choramiphen-0 Timne after icol death~120 iclDahRtoultraviolet min after ul-

traviolet

15 TGCb 2.41 0.694 0.29TGc 3.62d 0. 399d 0.11

15T-(Zamenhof) TG + thymine (2,ug/ml) 1.06 0.0765 0.07215T-(Cohen) TGC + thymine (2 jg/ml) 0.233 0.0033 0.014

TG + thymine (2 ug/ml) 0.826 0.894 1.1

a Column 4/column 3.b Tris-glucose-Casamino acid medium.c Tris-glucose medium.d Geometrical mean of 3 experiments.

z

0

0

_j

0 6 0 120 180 240

MINUTES

Figure 7. Effect of histidine deprivation on

ultraviolet irradiated Escherichia coli strain 15h. Incubation in Tris-glucose-Casamino mediumfor 4 hr. Ultraviolet irradiation for 100 sec. Theirradiated bacteria were incubated in Tris-glucosemedium (0) and L-histidine was added (as indi-cated by the arrows) at 0 (0) or at 60 min (a) ofincubation (final 30 pg per ml). A, Nonirradiatedcontrol in Tris-glucose-Casamino medium; A,

the same, in Tris-glucose medium.

z0<0

4_

> A

0

-4

l0 60 120 ISO 240

MINUTES

Figure 8. Effect of chloramphenicol on ultra-violet irradiated Escherichia coli strain 15 h-.Incubation in Tris-glucose-Casamino mediumfor 3 hr. The bacteria were irradiated for 100 secand incubated in Tris-glucose medium containing30 pAg per ml L-histidine (-) or in phosphate bufferwith glucose (0). Chloramphenicol was added(as indicated by the the arrows) at 0 (A) or at 60min (U) of incubation with the former medium,or added at 60 min (0) to the latter medium.

[VOL. 79282 OKAGAKI

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

CHLORAMPHENICOL DEATH OF IRRADIATED E. COLI

Figuchlorarichia c(CasamiirradiaglucosetainingChloraiarrows)(0) aft

7_I

A

z2 ,1

z - 2

-Jn

0 3J9

283

I I:I I !

w i~~~---x-=- - *. .~~~~~~~~~~~~~r

0 60 120MINUTES OF ADDITIONOF CHLORAMPHENICOL

ISO

Figure 11. Effect of chloramphenicol added atdifferent intervals after the outset of postirradi-

o 60 1 20 IS8O° 240 ation incubation in phosphate buffer with glucose,MINUTES in Escherichia coli strain 15 h-. Incubation in

Tris-glucose-Casamino medium for 3 hr. Ultra-,re 9. Effects of histidine deprivation and of violet irradiation for 100 sec. Chloramphenicolrnphenicol on ultraviolet irradiated Escher- was added to the aliquots taken thereof at inter-oli strain 15 h-. Incubation in Tris-glucose- vals indicated by arrows and they were incubatedino medium for 3 hr. The bacteria were separately for additional 3 hr each. The survivors

of respective fractions at 1 hr (A), 2 hr (V) andmedifor 100ksec ansidincuaed in Trisn- 3 hr (U) after the addition of chloramphenicol are

plotted below the original surviving fractionsboth histidine and chloramphenicol (A). (0). The broken lines connecting these points

mphenicol was added (as indicated by the represent the change in the susceptibility to the) to the former culture at 0 (A), or 60 min ultraviolet-chloramphenicol death during theber, the outset of incubation. course of postirradiation incubation.

DNA

'p

_. A/,.- -

V . mr16--

* A U

0 60 120 ISO

RNA

I I I* A-

0 60 120 ISO

'p

PROTEIN ,/

I

IfI

I

/~~~ .

,

1It I* vA-

I

0 6s0 120 IS0MINUTES

Figure 10. Effect of the histidine deficient incubation of ultraviolet irradiated Escherichia coli strain15 h- on the subsequent syntheses of DNA, RNA, and protein in the presence of histidine. Incubationin Tris-glucose-Casamino medium for 3 hr. The bacteria were irradiated for 100 sec and incubated inTris-glucose medium lacking histidine (0). L-Histidine was added (as indicated by the arrows) at 0

(-), 30 (A) and 60 min (-) of incubation.

Oi

z0p04

Ii

U.

0z

U)

00-i

-2

-3

-4

400r

30F

200h

z0

4

I-z

z0U

$100

19601

Or

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

[VOL. 79

-I *r1J-U,

,

/ " | I,

0 60 120MINUTES OF ADDITIONOF CHLORAMPHENICOL

Figure 12. Effect of chloramphenicol added atvarious intervals after the outset of post-irradi-ation incubation in phosphate buffer with glucose,in Escherichia coli strain 15. An experiment similarto the one presented in figure 11 using strain 15.The bacteria incubated for 4.5 hr in Tris-glucosemedium served as the material for irradiation.Markings are the same as in figure 11. Solid linerepresents the recovery in phosphate buffer withglucose and broken lines show the change in thesurviving fractions after the treatment withchloramphenicol for 2 (V) and 3 (-) hr.

death. In order to detect such restoration,ultraviolet irradiated bacteria of strain 15 h-were incubated in phosphate buffer with glucose,chloramphenicol being added at intervals tosamples of the culture medium, which were

further incubated for 3 hr. The sensitivity tochloramphenicol slowly diminished while theirradiated bacteria resided in the chloramphenicol-free medium (figure 11). Similar experimentswere carried out with strain 15 using eitherphosphate buffer with glucose (figure 12) or

MINUTES OF ADDITIONOF CHLORAMPHENICOL

Figure 12. Effect of chloramphenicol added atvarious intervals after the outset of postirradi-ation incubation in Tris-glucose medium, inEscherichia coli strain 15. Incubation in Tris-glucose medium for 5 hr. The Tris-glucose mediumwas employed for the medium of incubatingirradiated cells. Results are expressed in the sameway as in figure 12.

Tris-glucose medium (figure 13). In this strain,a transient recovery from the state sensitiveto the ultraviolet-chloramphenicol death was

observed after 20 min incubation. Curiouslyenough, such a recovery was again reversed bythe second wave of chloramphenicol sensitivity,which, however, seemed to be largely eliminatedin 2 hr incubation without chloramphenicolafter the ultraviolet irradiation. It seems thatthe ultraviolet irradiation lesion causing thesensitivity to chloramphenicol is gradually re-

paired during incubation in the absence of thiscompound.

Effects of the composition of the preincubationmedia on the ultraviolet-chloramphenicol death.

284

o

OKAGAKI

Or

- Iz0

C.)4

UL.

z

5:

n(0

0-J

z0

U

4

LL

CDz5:

C',

CD0-J

-' IiII

I I~, II1I

'I, I

,

.3 _ ,tl-3

21

-4

60

-4

120

n I

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

CHLORAMPHENICOL DEATH OF IRRADIATED E. COLI

OrOrz0

iL -

z

5:

5i3-2CD

0-J

-3I

0 60 120

z

8400IL

I CDI80 z

MINUTESFigure 14. Repression of chloramphenicol death

in Escherichia coli strain 15 T-(Cohen) grown inthe medium lacking amino acids prior to theultraviolet irradiation. Incubation for 5 hr inTris-glucose medium supplemented with thymine(2,g per ml) but lacking Casamino acid. Thebacteria were irradiated for 100 sec and incubatedin phosphate buffer with glucose (0), phosphatebuffer with glucose and chloramphenicol (A), orTris-glucose medium with 2 /g per ml thymine andchloramphenicol (A). Casamino acid was added(as indicated by the arrows) at 0 (v) or 40 min(v) of incubation (final concentration of 1 mgper ml). The upper curve represents the recoveryof the primarily killed bacteria.

In the experiment with strain 15 T-(Cohen)presented in figure 6, the bacteria were incubated,prior to irradiation, in Tris-glucose-Casaminoacid medium supplemented with thymine, for 3hr. If the medium was not fortified with Casaminoacid (1 mg per ml), the bacteria of this strain,which exhibited slow growth in this medium,were not killed by chloramphenicol treatmentafter ultraviolet irradiation, even in the presenceof Casamino acid (figure 14). Parent strain15 and strain 15 T-(Zamenhof) died even ifthey had been incubated in the Tris-glucosemedium lacking amino acids, in which bacterialgrowth was usually retarded (table 1).

It may be suspected that strain 15 T-(Cohen)is actually a double mutant of the parent strain15 concerning both dependence on thymine forgrowth and the loss of susceptibility to theultraviolet-chloramphenicol death under certainnutritional conditions. A back mutant of 15

Co

CD

0-J

-5

6

EA

D

F

B

C

0 60 120 ISO

MINUTESFigure 15. Effect of Casamino acid in the incu-

bation media on the subsequent ultraviolet-chloramphenicol death, in Escherichia coli strain15-OS21. Details of the experiment are summarizedin table 2.

T-(Cohen), being a prototroph but distinct fromstrain 15 by the second property mentioned, hasnow been produced in the following way. Thebacteria harvested at the late logarithmic phasewere irradiated with ultraviolet light, suspendedin Tris-glucose medium lacking thymine, andincubated for 15 hr with aeration; then theback mutants were separated twice on chemi-cally defined agar plates; finally 4 colonies werepicked and tested. Strain 15-OS21, thus obtained,showed the properties closest to strain 15 T-(Cohen) with respect to sensitivity to ultravioletlight and the resistance to ultraviolet-chloram-phenicol death following incubation (figure 15and tables 2 and 3), but did not require thyminefor growth.

Further experiments were carried out with

19601 285

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

TABLE 2

Effect of incubation with or without Casamino acid on the sensitivity to chloranmphenicol death afterultraviolet irradiation, in Escherichia coli strain 15-OS21

Bacteria were grown in the TGCa medium (containing Casamino acid, 1 mg/ml) or in the TGb medium(lacking Casamino acid) until the late logarithmic phase was reached, then various modifications ofthe medium with or without Casamino acid were made as listed, and the bacteria were harvested, ultra-violet irradiated for 100 sec and incubated in 0.1 M phosphate buffer containing 0.02 M glucose and 100mg chloramphenicol per ml.

Bacterial GrowthLater Modification during Period of Course of OccurrenceModification Ultraviolet- of Ultravio-

Initial Incubation (% Increase) Chloram- let-Chloram-Medium ______________ ___________________ phenicol phenicol__________________________________________ _______-Death in Death

Addition to incubation medium Transfer to the 2nd Viable Optical figure 15incubation medium cell density

TGC, 3 hr B ++TGC, 3 hr TG, 40 min 14 30 F ++TG, 5 hr _ ATG, 5 hr TGC, 40 min 59 43 C ++TG, 5 hr Casamino acid (1 mg/ml), 36 20 D +

40 minTG, 5 hr Casamino acid + CP (100 0 0 E

,ug/ml) 40 min

a Tris-glucose-Casamino acid medium.Tris-glucose medium.

strain 15-OS21 on the effects of Casamino acidin the medium (table 2 and figure 15). Cells ofthis strain did not show a significant ultraviolet-chloramphenicol death unless they had been inbrief contact with casamino acid for 40 min.The sensitivity to the ultraviolet-chloramphenicaldeath of Casamino acid-treated bacteria washardly lost by the removal of amino acids fromthe culture medium for 40 min. This fact sug-gested that the sensitivity was not due to thestorage of amino acid within the cell. The addi-tion of Casamino acid to the culture in Tris-glucose medium was less effective in inducingthe sensitivity to the ultraviolet-chloramphenicaldeath than the transfer of the Tris-glucosemedium cultured bacteria to the Casamino acidcontaining one. Chloramphenicol in the mediumprevented the aquisition of sensitivity inducedby Casamino acid. It may be assumed that theinduction of this sensitivity in these strainsinvolves the synthesis of new protein(s) ratherthan a loss of some compound effective in pro-tecting the bacteria from ultraviolet-chloram-phenicol death. It seems that the loss of inducedsensitivity during the 5 hr incubation withoutCasamino acid was due to the extensive dilutionof such protein(s).The effect of individual amino acids in the

medium was tested for the induction of sen-sitivity to ultraviolet-chloramphenicol death(table 3). The bacteria of strain 15-OS21 weresuspended in Tris-glucose medium supplementedwith a single amino acid, and incubated untillate-middle to late logarithmic growth phase.Some of the amino acids used slightly acceleratedbacterial growth, but L-isOleucine, L-threonine,and L-cysteine were inhibitory. The bacteriawere then harvested, irradiated with ultravioletlight, and incubated for 2 hr in phosphate buffercontaining both glucose and chloramphenicol.As presented in the death ratio, a number ofamino acids were effective in inducing sensitivityto the ultraviolet-chloramphenicol death. Aminoacids ineffective at 1 mm were glycine, serine,isoleucine, aspartic and glutamic acids, tyrosine,tryptophan, and histidine.

Sensitivity to ultraviolet - chloramphenicoldeath was also induced by pyrimidines added tothe medium, but not by adenine.The effect of chloramphenicol on the synthesis

of nucleic acids in the postirradiation incubationwas examined with strain 15-OS21 after incuba-tion either for 3 hr in the Tris-glucose-Casaminoacid medium or for 5 hr in the Tris-glucosemedium. Only in the former case were bacteriakilled during the postirradiation incubation in

286 OKAGAKI (VOL. 79

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

CHLORAMPHENICOL DEATH OF IRRADIATED E. COLI

TABLE 3Effect of amino acids and nucleic acid bases in the incubation medium on the survival ratio and the

chloramphenicol death in ultraviolet irradiated Escherichia coli strain 15-OS21Bacteria grown in the Tris-glucose medium containing various compounds listed were ultraviolet

irradiated for 100 sec and incubated for 120 min in 0.1 M phosphate buffer containing 0.02 M glucose and100,ug chloramphenicol per ml.

SurvivalChloram- Occurrence

Compounds Tested Conc Incubation phenicol of Chloram-0 Time after 120 min with chlor- Death phenicolultraviolet amphenicol after Ratioa Death

ultraviolet

NoneCasamino acid (Difco)

Glycine

DL-Alanine

DL-ValineL-Leucine

L-Isoleucine

DL-Serine

L-ThreonineL-CysteineL-MethionineDL-Aspartic acidL-Glutamic acidL-LysineL-Phenylalanine

L-TyrosineL-TryptophanL-HistidineCytosine

Thymine

UracilAdenine

mg/ml

1

mM

101

202210151

202111211

10111110.110.111

hr

53.53

5555555654.75

565666555656565

a See table 1.I Geometrical means of 3 experiments each.c GI = growth inhibited.

1. 94b

0.3610. 303b

0.2440.6611.021.682.960.04140.1373.561.930.9902.37GIcGI

0.07200.05140.2550.2540.1770.1030.6220.7720.09958.571.083.010.7581.812.54

3.07b0.01640.0148b

0.07221.090.02190.2520.4460.00550.009858.755.650.8254.08

0.04630.1620.2790.02300.01420.03050.9735.250.2090.6121.800.1891.710.4536.61

1.60.0450.049

0.301.60.0210.150.150.130.0722.52.90.831.7

0.643.21.10.0910.0800.0291.66.82.10.0711.70.0631.60.252.6

++

++

++

++++++

++

++++

++

++

the Tris-glucose-Casamino acid medium con-taining chloramphenicol. There was no significantdifference of the change in nucleic acids betweenthe two conditions. In both cases, DNA synthesiswas completely blocked, and RNA synthesiscontinued for 30 min and then leveled off as inthe experiments with strain B and 15 h-. Theseresults showed that the difference in the induced

sensitivity to ultraviolet-chloramphenicol deathwas not significantly reflected in the ability ofnucleic acids syntheses during the postirradiationincubation in the presence of chloramphenicol.

DISCUSSION

Ultraviolet-chloramphenicol death and proteinsynthesis. Chloramphenicol death of ultraviolet

19601 287

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

OKAGAKI

irradiated bacteria, i. e., the progressive decreasein the colony forming ability of some strains ofE. coli as caused by chloramphenicol during thepostirradiation incubation, may be one of theconsequences of the complete blockade of proteinsynthesis effected by the antibiotic. Barner andCohen (1956) noted that in strain 15 the post-irradiation treatment with 5-methyltryptophandid not kill the bacteria. According to Cohen andMunier (1959) and Munier and Cohen (1959) thisanalogue does not completely arrest the proteinsynthesis, but seems to produce inactive proteinmolecules without itself being incorporated intothem. The results with 5-methyltryptophan are

therefore quite compatible with the presentobservations, and suggest the absence of in-corporating amino acids into polypeptide macro-

molecules, rather than failure of producing new

functioning molecules, is the crucial conditionfor the occurrence of the ultraviolet-chloram-phenicol death.

Ultraviolet-chloramphenicol death and DNA syn-

thesis. The DNA synthesis in ultraviolet irra-diated bacteria is inhibited for some period, butit soon resumes; this resumption is prevented,however, if the protein synthesis is inhibitedthroughout the postirradiation incubation (Ha-rold and Ziporin, 1958; Drakulic and Errera,1959). These observations are in agreementwith those of the present study with strains B,15-OS21 and 15 h-. The occurrence of chloram-phenicol death of ultraviolet irradiated bacteriasimply depends upon the presence of chloram-phenicol and apparently has no relation to theresumption of DNA synthesis. Also, the additionof the antibiotic to the medium 40 min afterpostirradiation incubation did not interfere withthe resumption of DNA synthesis, but increasedthe rate of death. Ultraviolet-chloramphenicoldeath must, however, involve some sort of activeprocess, because it requires an energy supply forits development.

Effects of chloramphenicol on survival of E. colistrain B. The effect of chloramphenicol is radicallydifferent in E. coli strain B and in strain 15and its various mutants, or strain B/r. StrainB is characterized by greater sensitivity of colonyforming ability to the action of ultraviolet light,and by its marked restoration in a nitrogen-freemedium, or by chloramphenicol treatment. Thislast point corroborates the observations ofDoudney and Haas (1958) and Gillies and Alper

(1959). The resumption of DNA synthesis inultraviolet irradiated bacteria is prevented underconditions which induce the restoration. However,such restoration is probably not related to themutation frequency decline as described byDoudney and Haas (1958). According to Doudneyand Haas (1958, 1959) the mutation fixationinvolves the restorable irradiation-induced changein ribonucleic acid precursors containing pyri-dines in vivo rather than the change in the pre-formed nucleic acid molecules. This changebecomes fixed in the newly formed ribonucleicacid molecules which then transfers the wronginformation to the DNA molecules synthesizedlater, upon completion of ultraviolet inducedinhibition of DNA synthesis. It seems that theprevention by chloramphenicol of the resump-tion of DNA synthesis is not directly involved inthe mutation frequency decline as observed byDoudney and Haas.Nature of lethal effect of ultraviolet irradiation.

The lethal effect of ultraviolet irradiation instrain B might be due to the damage of DNAmolecules rather than the nucleic acid precursors.The observed restoration would then imply thatthe irradiation lesion in DNA molecules mayprogressively be repaired during the postirradia-tion incubation without DNA synthesis (Barnerand Cohen, 1956). Then normal resumption ofDNA synthesis in ultraviolet irradiated bacteriawould fix the lethal change in replicating DNAmolecules, by forming abnormal DNA on themodified template. Since the restoration pro-ceeded at least until 3 hr after irradiation, asignificant portion of the lesion in DNA moleculeswould still have persisted as early as 40 minafter irradiation, when DNA synthesis was juststarting under the conditions employed. Additionof chloramphenicol at this moment did not affectDNA synthesis and therefore would have fixedthe lethal change and have affected the progressof restoration. Actually, however, the restorationoccurred at a greatly enhanced rate. It is there-fore unlikely that the ultraviolet induced reduc-tion of viable count in E. coli strain B is causedby the damage in DNA molecules.

In the strains employed other than B, theultraviolet-chloramphenicol death obscures thesituation, and it is only with strains 15 T-(Cohen)and 15-0S21 that chloramphenicol was seen tobring about some restoration of ultravioletirradiated bacteria under conditions which effec-

288 [VOL. 79

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

CHLORAMPHENICOL DEATH OF IRRADIATED E. COLI

tively prevented ultraviolet - chloramphenicoldeath. It can not be determined at the momenthow far the lethal effect of ultraviolet irradia-tion in these strains is based upon the samemechanism as in the strain B.

Comparison of ultraviolet-chloramphenicol deathwith the death by histidine deficiency. With re-gard to ultraviolet-chloramphenicol death inappropriate strains, chloramphenicol may eitheramplify the irradiation damage, or else it initiates,in the ultraviolet-irradiated bacteria, an entirelynew sort of killing process which is effectivelyblocked or has no possibility to occur in non-irradiated bacteria. Likewise, in the irradiatedstrain 15 h-, a histidine auxotrophy, the histidinedeprivation may cause an almost complete cessa-tion of protein synthesis, but the killing effectwas manifested only when the nitrogen sourcewas present. On the other hand, the killing effectof chloramphenicol on irradiated bacteria ofdifferent strains does not depend on the nitrogensource.

In nonirradiated bacteria, amino acid deficiencyin auxotrophic strains may lead to the inhibitionof RNA synthesis (Sands and Roberts, 1952;Gros and Gros, 1958). In the absence of a nitrogensource, the conditions for such an inhibition ofRNA synthesis may be lacking, because theprecursors of RNA cannot be produced. It issuggested that death by histidine deficiency ofultraviolet irradiated bacteria may be relatedto the inhibition of RNA synthesis, whereasultraviolet-chloramphenicol death is unrelated.

Histidine deficiency in ultraviolet irradiatedstrain h- results in almost complete cessation ofRNA synthesis, as in nonirradiated bacteria.However, the addition of the required aminoacid would immediately release the RNA syn-thesis from the inhibition in nonirradiated bac-teria (Gros and Gros, 1958), whereas in irradiatedstrain h- the ability to synthesize RNA, as wellas DNA and protein, seems to be irreversiblyimpaired by histidine deficiency of relativelyshort duration. It is highly likely that the irra-diated strain 15 h- rapidly becomes incapableof further growth during the period of histidinedeficiency. In contrast to this, ultraviolet irra-diated cells of strains 15 h- and 15-OS21 did notlose the ability to synthesize RNA during theinitial period of chloramphenicol treatment, whenthey were losing the colony forming ability.Moreover, the effects of chloramphenicol on the

synthetic ability of irradiated bacteria of dif-ferent strains under different conditions of in-duction were quite similar whether or not thebacteria were susceptible to ultraviolet-chloram-phenicol death. Thus, the mechanism of ultra-violet-chloramphenicol death is distinct from,that of the death of ultraviolet irradiated strain15 h- by histidine deficiency. The latter may alsodiffer from the lethal mechanism of ultraviolet-irradiation alone, which allows bacterial growthto continue for a substantial time. It follows thathistidine deficiency in ultraviolet irradiated strain15 h- probably initiates a new sort of damage.Whether a similar situation holds with theultraviolet-chloramphenicol death can not bedetermined at the moment.

Nature of induction of the sensitivity to chloram-phenicol in irradiated bacteria. In certain strains(15 T-(Cohen) and 15-0S21) the susceptibilityto the chloramphenicol death of the ultravioletirradiated bacteria could be modified by thechange in composition of the medium used forthe incubation prior to irradiation. Thus somesingle amino acid or pyrimidines may induce inthese strains a state which effects the killingwhen they were subsequently subjected to ultra-violet irradiation, followed by chloramphenicoltreatment. The uninduced bacteria resist thesame treatment. The process of such an induc-tion does not take more than 40 min and obviouslyinvolves protein synthesis, and its effect canbe eliminated only slowly during growth in theabsence of inducing substance. Consequently,the inducing compounds would be participatingin the formation of some new sort(s) of protein(s)rather than in the negative feedback mechanismoperating to repress certain enzymes. In otherstrains (15, 15 T-(Zamenhof), 15 h-, and B/r),such protein(s) would be constitutive becausethey do not require a special compound inincubation media to be killed by chloramphenicolupon irradiation with ultraviolet light. Furtherit is inferred that strain B cannot produce suchprotein(s), because the bacteria are not killedby the postirradiation treatment with chloram-phenicol.The whole situation now suggests some paralle-

lism to the formation of enzymes. The protein(s)necessary for the ultraviolet-chloramphenicoldeath is either constitutive, inducible, or deficientaccording to the genetic constitution of thebacteria. There seems to be some specificity in

1960] 289

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

OKAGAKI

compounds having inducing activity, since notall amino acids tested were effective. However,it is impossible at the moment to define theinducing activity in terms of the chemicalspecificity. Since the chloramphenicol deathoccurs in ultraviolet irradiated bacteria but notin normal ones, the elucidation of its naturewould be relevant to the understanding of thebiological effects of ultraviolet irradiation.

ACKNOWLEDGMENTS

The author expresses his appreciation toProfessor A. Sibatani for his guidance at allstages of this investigation; to Drs. F. J. Ryanand D. Nakada, Columbia University, and Mr.M. Sekiguchi, Osaka University, for the generoussupply of bacterial strains used; and to ProfessorB. Osogoe and Professor R. Hayashi of thismedical school for their valuable advice andencouragement. The kind help in various waysof Dr. H. K. Kihara, Messrs. Y.-T. Tchoe, andK. Kimura of this laboratory, of Dr. S. Osawa,Nagoya University, and of Professor T. Yanagita,Chiba University, is deeply acknowledged.

SUMMARY

A comparative study was made on the effectof chloramphenicol in the postirradiation incu-bation of ultraviolet irradiated bacteria using 7strains of Escherichia coli. The antibiotic ap-parently acted as a bactericidal agent (hence itseffect is called the chloramphenicol death) on theultraviolet irradiated cells of strains 15, 15T-(Cohen), 15 T-(Zamenhof), 15 h-, 15-OS21(a prototrophic back mutant of 15 T-(Cohen))and B/r, but not in strain B. In strain B, chlor-amphenicol enhanced recovery from the lethalaction of ultraviolet irradiation, especially whenadded at some interval after the irradiation.

Chloramphenicol death occurred when eachone of the following conditions was met: (a) thepreliminary ultraviolet irradiation, (b) thepresence of chloramphenicol in the incubationmedium with or witlhout nitrogen source, (c) thepresence of energy source in the incubationmedium, and (d) the presence of a certainmetabolic system within the cell, which may beeither constitutive or inducible by the incubationwith certain compounds, according to the strainsemployed.

Induction of the susceptibility to the ultra-violet-chloramphenicol death was realized by a

number of amino acids or pyrimidines addedsingly or in combination with the incubationmedium in strains 15-0S21 and probably also in15 T-(Cohen). Strains 15, 15 T-(Zamenhof), and15 h- did not require such induction for theultraviolet-chloramphenicol death.Death of the irradiated bacteria, analogous to

the chloramphenicol death, was also observed inhistidine deficient incubation, instead of additionof chloramphenicol, with ultraviolet-irradiatedbacteria of strain 15 h-, a histidine auxotroph.Unlike the ultraviolet-chloramphenicol death,this required a nitrogen source in the medium.In this case, the ability to synthesize nucleic acidsand proteins seems to be lost irreversibly by ahistidine deficiency of relatively short duration.The nature of various types of death and

recovery therefrom after ultraviolet irradiation indifferent strains of E. coli was discussed. It wasconcluded that in strain B the bulk of the lethaleffect was not due to the damage in deoxyribo-nucleic acid molecules. The lethal effect ofhistidine deficiency in irradiated strain 15 h-seems to be distinct from both the primary lethaleffect of the ultraviolet-irradiation, which can berestored under certain nutritional conditions invarious strains, and the ultraviolet-chloram-phenicol death.

REFERENCESARONSON, A. Z. AND SPIEGELMAN, S. 1958 On

the use of chloramphenicol inhibited systemfor investigating RNA and protein synthesis.Biochim. et Biophys. Acta, 29, 214-215.

BARNER, H. D. AND COHEN, S. S. 1956 Therelation of growth to the lethal damageinduced by ultraviolet irradiation in Escher-ichia coli. J. Bacteriol., 71, 149-157.

BURTON, K. 1956 A study of the conditions andmechanism of the diphenylamine reaction forthe colorimetric estimation of deoxyribo-nucleic acid. Biochem. J., 62, 315-323.

COHEN, G. N. AND MUNIER, R. 1959 Effects desanalogues structuraux d'aminoacides sur lacroissance, la synthese de protdines et lasynthUse d'enzymes chez Escherichia coli.Biochim. et Biophys. Acta, 31, 347-356.

COHEN, S. S. AND BARNER, H. D. 1954 Studieson unbalanced growth in Escherichia coli.Proc. Natl. Acad. Sci. U. S., 40, 885-893.

DOUDNEY, C. 0. AND HAAS, F. L. 1958 Modi-fication of ultraviolet-induced mutationfrequency and survival in bacteria by post-irradiation treatment. Proc. Natl. Acad.Sci. U. S., 44, 390-401.

[VOL. 79290

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

CHLORAMPHENICOL DEATH OF IRRADIATED E. COLI

DOUDNEY, C. 0. AND HAAS, F. L. 1959 Mutationinduction and macromolecular synthesis inbacteria. Proc. Natl. Acad. Sci. U. S., 45,709-722.

DRAKULIC, M. AND ERRERA, M. 1959 Chloram-phenicol-sensitive DNA synthesis in normaland irradiated bacteria. Biochim. etBiophys. Acta, 31, 459-463.

FUJISAWA, Y. AND SIBATANI, A. 1954 Is thereany quantitative relationship between thesynthesis and the breakdown of nucleic acidsin living cells? Experientia, 10, 178.

GALE, E. F. AND FOLKES, J. P. 1953 The as-similation of amino acids by bacteria. 15.Action of antibiotics on nucleic acid andprotein synthesis in Staphylococcus aureus.Biochem. J., 53, 493-498.

GILLIES, N. E. AND ALPER, T. 1959 Reductionin the lethal effects of radiations on Escher-ichia coli B by treatment with chlorampheni-col. Nature, 183, 237-238.

GROS, F., AND GROS, F. 1958 R6le des acidesamin6s dans la synthese des acides nucleiqueschez Escherichia coli. Exptl. Cell Research,14, 104-131.

HAROLD, F. M. AND ZIPORIN, Z. Z. 1958 Syn-

thesis of protein and of DNA in Escherichiacoli irradiated with ultraviolet light. Bio-chim. et Biophys. Acta, 29, 439-440.

LOWRY, 0. H., ROSENBROUGH, N. J., FARR, A.L., AND RANDALL, R. J. 1951 Proteinmeasurement with the Folin phenol reagent.J. Biol. Chem., 193, 265-275.

MUNIER, R. AND COHEN, G. N. 1959 Incorpo-ration d'analogues structuraux d'aminoacidesdans les proteines bact6riennes au cours deleur synthese in vivo. Biochim. et Biophys.Acta, 31, 378-391.

SANDS, M. K. AND ROBERTS, R. B. 1952 Theeffects of a tryptophan-histidine deficiencyin a mutant of Escherichia coli. J. Bacteriol.,63, 505-511.

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, 622-673.

ZAMENHOF, S., AND GRIBOFF, G. 1954 Incorpo-ration of halogenated pyrimidines into thedeoxyribonucleic acids of Bacterium coli andits bacteriophages. Nature, 174, 306-307.

2911960]

on August 24, 2019 by guest

http://jb.asm.org/

Dow

nloaded from