Proc. Natl. Acad. Sci. USAVol. 74, No. 10, pp. 4341-4345, October 1977Biochemistry

Elongation factor Tu resistant to kirromycin in an Esherichia colimutant altered in both tufgenes(protein biosynthesis/EF-Tu GTPase regulation/EF-Tu peptide pattern/tufB gene product/dominance of kirromycin sensitivity)

ECKHARD FISCHER*t, HEINZ WOLFt, KLAUS HANTKEt, AND ANDREA PARMEGGIANI** Ecole Polytechnique, Laboratoire de Biochimie, (Laboratoire associ6 no 240 du CNRS) F 91128 Palaiseau-Cedex, France; and t Institut fur Biologie II,Lehrbereich Mikrobiologie, Universitat Tfibingen, D 7400 Tfibingen, German Federal Republic

Communicated by Bernard D. Davis, August 3,1977

ABSTRACT A mutant of Escherichia coli is described thatdisplays kirromycin resistance in a cell-free system by virtue ofan altered elongation factor Tu (EF-Tu). In poly(U)directedpoly(Phe) synthesis the kirromycin resistance of the crystallizedenzyme ranged between a factor of 80 and 700, depending ontemperature. Similarly, kirromycin-induced EF-Tu GTPaseactivity uncoupled from ribosomes and aminoacyl-tRNA re-quired correspondingly higher concentrations of the antibiotic.Resistance of EF-Tu to kirromycin is a consequence of a modi-fied enzyme structure as indicated by its altered fingerprintpattern.

P1 transduction experiments showed that the kirromycin-resistant EF-Tu is coded by an altered tufB gene (tufBl). Theknown existence of two genes coding for EF-Tu would interferewith the recognition of a mutant altered in only one of thosegenes, if the mutation were recessive. Because kirromycinblocks EF-Tu release from the ribosome, kirromycin sensitivityis dominant, as shown by the failure of a mixed EF-Tu popula-tion to express resistance in vitro. Therefore, phenotypic ex-pression of kirromycin resistance in vivo appears to be onlypossible if the EF-Tu mutant lacks an active tufA gene, aproperty likely to be inherited from the parental D22 strain: Theobservations that introduction of a tufA+ region makes the re-sistant strain sensitive to the antibiotic and that transductionof tufB1 into a recipient other than E. coli D22 yields kirro-mycin-sensitive progeny support these conclusions.

In addition to its well-established function in the elongationcycle, there is increasing evidence that elongation factor Tu(EF-Tu) is involved in other important biological processes ofthe bacterial cell. EF-Tu has been reported to participate in theregulation of rRNA synthesis (1), to represent a subunit of theQ,# replicase (2), and to be associated with the bacterial cellmembrane with still undefined functions (3, 4). No correlationhas yet been made between these EF-Tu functions and the twogenes, tufA and tufB, that have been shown to code for EF-Tuin the Escherichia coil genome (5).The isolation of well-characterized mutants with altered

EF-Tu may improve our insight into the mechanism of actionof this multifunctional protein. So far, reports have been pub-lished on temperature-sensitive EF-Tu (6, 7) and on a tufB-coded EF-Tu structurally altered in a region irrelevant to itsfunction (8). The recent finding of kirromycin, an inhibitor ofprotein biosynthesis acting on EF-Tu (9-13), has offered thepossibility of directly selecting E. coil mutants carrying anEF-Tu resistant to an agent that affects a region critical for theactivities of this elongation factor. In this communication wedescribe the isolation of such a mutant and some characteristicsof the altered EF-Tu.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby marked"advertisement" in accordance with 18 U. S. C. §1734 solely to indicatethis fact.

4341

MATERIALS AND METHODSThe E. coli strains used in this work are: D22 F- (envA ampArpsL his proB trp, lysogenic for X; refs. 14 and 15); D2216 F-(tufBl derived from D22); D2216Rifr F- (rpoB derived fromD2216); AT1371 (thi-1 argE3 his-4 proA2 pan-4 mtl-I xyl-5galK2 lacYl tsx-29; ref. 16); KL228 Hfr (thi-1 leu-6 sup-54lacYl or Z4 gal-6; ref. 17); and BW113 Hfr (metBl Valr; ref.17).The kirromycin-resistant mutant was isolated from E. coil

D22 (14, 15), a K12 strain with an altered membrane permeableto kirromycin. The mutagenized cell population was screened6-60 hr after treatment with N-methyl-N'-nitro-N-nitroso-guanidine (18) by plating on tryptone/NaCl agar plates (18)supplemented with kirromycin at 0.1, 0.2, and 0.4 mg/ml.Bacteria from resistant colonies were replicated on platescontaining increasing concentrations of kirromycin (0.1-3mg/ml). Slowly growing colonies were selected and cultivatedat 340 in minimal medium of Davis and Mingioli (19), sup-plemented with L-histidine, L-proline, and L-tryptophan at 50mg/liter and thiamine-HCI at 10 mg/liter. The cells wereharvested in late logarithmic phase by centrifugation, andwashed twice in cold buffer (25 mM Tris-HCI, pH 7.8/10mMMgCI2). Cell extracts were prepared by sonication of a cellsuspension (0.5-1 g/ml) in standard buffer (60mM Tris-HCI,pH 7,8/30 mM NH4CI/30 mM KCI/10 mM MgCI2/2 mMdithiothreitol); cell debris was spun down at 40,000 X g and thesupernatant (S30) was examined for kirromycin resistance inpoly(U)-directed poly(Phe) synthesis.To prepare purified EF-Tu, bacteria were grown in minimal

medium in a 200-liter fermentor that was inoculated with 10liters of culture. After 71/2 hr of growth at 340 with strain D22or 23 hr with strain D2216, the cells were harvested in the latelogarithmic phase and stored as a paste at -65°. The 40,000 Xg (S30) and 105,000 X g (S100) supernatants of the cell extract,as well as pure EF-Tu, elongation factor Ts (EF-Ts), elongationfactor T (EF-T = EF-Tu-EF-Ts), elongation factor G (EF-G),and NH4CI-washed ribosomes, were isolated as described (11,20,21) except that with EF-Tu from strain D2216, an additionalchromatography on DEAE-Sephadex A50 had to be appliedto accelerate crystallization. One microgram of EF-Tu, EF-Ts,EF-T, and EF-G was taken to correspond to 24, 35, 15, and 12pmol, respectively, and 1 A260 unit of ribosomes, to 25 pmol(21). Resistance of EF-Tu to kirromycin was determined by theconcentration of the antibiotic needed for either a 50% inhi-bition of poly(U)-directed poly(Phe) synthesis (9) or a 50% in-

Abbreviations: EF-Tu, elongation factor Tu; EF-Ts, elongation factorTs; EF-T, elongation factor T, the complex formed by EF-Tu andEF-Ts (EF-Tu-EF-Ts); EF-G, elongation factor G.

Dow

nloa

ded

by g

uest

on

Janu

ary

26, 2

021

4342 Biochemistry: Fischer et al.

duction of GTPase activity (21) with EF-Tu alone. Proteinconcentrations were determined by the method of Lowry etal. (22), with bovine serum albumin as standard. Allof the otherbiological components and materials, including the labeledproducts, were the same as used before (9-12, 20, 21).To map the mutation, phage P1 transduction and conjugation

experiments were performed by utilizing a rifampicin-resistantstrain of D2216 (D2216Rifr) selected after spontaneous muta-tion (18). For the transfer of tufB+, (BW113 X D2216Rifr)rifampicin-sensitive progeny were selected from Strrand Vairrecombinants and assayed for kirromycin resistance in ewo andin the cell-free system; for the transfer of tufA+, (KL228 XD2216Rifr) streptomycin-sensitive progeny were selected fromAmpr and His+ recombinants and assayed for kirromycin re-sistance in vivo and in vitro.

RESULTSSelection of Strain D2216 Resistant to Kirromycin in

Cell-Free System. The standard E. coli strains, such as K12, B,A19, and MRE600, are resistant in vivo to kirromycin (minimalinhibitory concentration in agar diffusion, >1 mg/ml; refs. 23and 24), whereas their cell-free systems are highly sensitive [50%inhibition of poly(U)-directed poly(Phe) synthesis at 0.5 AMkirromycin; ref. 9]. The resistance to the antibiotic is thus basedon membrane impermeability (23). The choice of the envelopemutant D22 for our experiments was suggested by its suscep-tibility to kirromycin in vivo (minimal inhibitory concentrationin agar diffusion, <160 ,ug/ml; ref. 24). Because spontaneousreversion of kirromycin permeability of this strain occurs at arate of about 10-10, the separation of these revertants is crucialfor the selection of mutants with altered EF-Tu. For this pur-pose, we reasoned that a mutation inducing resistance to anantibiotic affecting the basic activities of EF-Tu may reducethe rate of protein synthesis and consequently slow down cellgrowth. This consideration prompted us to select kirromycin-resistant colonies with retarded growth. Of the 29 colonies se-lected for resistance to kirromycin after a segregation time of6-24 hr, none yielded a resistant extract, whereas such in vitrokirromycin resistance was found in 1 of the 6 resistant coloniesselected after 36-60 hr of segregation. Cultivation in tryp-tone/NaCl medium showed a generation time of 75 min for themutant versus 45 min for the parental strain; cultivation inminimal medium increased the corresponding values to 270 and100 min. Although in vitro kirromycin resistance of this strainis apparently associated with a reduced growth rate, themechanism of this phenomenon is not clear; the specific activityof the modified EF-Tu seems substantially unchanged inpoly(U)-directed poly(Phe) synthesis.EF-Tu Is the Cell Component Resistant to Kirromycin:

Functional and Structural Characterization. In poly(U)-directed poly(Phe) synthesis the S30 from the resistant strainD2216 was found to be about 1/80 as sensitive to kirromycin asthat from the parental strain. Experiments with ribosomes andS100 prepared from either strain showed that the cell compo-nent mediating kirromycin resistance was located in the su-pernatant fraction (not illustrated). Preincubation of kirromycinwith S100 from the resistant strain did not affect the action ofthe antibiotic on a sensitive EF-Tu, excluding S100 activitiesin enzymatic degradation of kirromycin as a possible cause forresistance. When pure crystalline EF-Tu from the mutant re-placed the crude extract supernatants, kirromycin resistanceincreased to a factor of 100; a 50% inhibition of poly(U)-directedpoly(Phe) synthesis was obtained at 20 ,M kirromycin versus0.2MM found for the sensitive EF-Tu from either parental strainor E. coli B (Fig. IA). The presence of EF-Ts was irrelevant to

60

CL

100

50

50

0 -7 -6 -5 -4

Log [kirromycinlFIG. 1. Sensitivity to kirromycin of EF-Tu from strain D22 or

D2216 in polypeptide synthesis, and dominance of kirromycin sen-

sitivity in vitro. (A) Reaction mixtures contained, in 75 ,l of standardbuffer, 48 pmol of EF-Tu, 25 pmol of 70 ribosomes, 2.5 Ag of poly(U),4 nmol of tRNA, 2 ;g of a mixture of aa-tRNA synthetases, 450 pmolof [14C]phenylalanine (69 Ci/mol), 24 pmol of EF-G, 75 nmol of ATP,75 nmol of GTP, and kirromycin as indicated. The reactions werestarted by adding EF-Tu, aa-tRNA synthetases, tRNA, [14C]phe-nylalanine, ATP, and GTP. The samples were incubated for 10 minat 300 and assayed for incorporation of [14C]phenylalanine into hottrichloroacetic acid-insoluble material (9). 0, Kirromycin-sensitiveEF-Tu; 0, kirromycin-resistant EF-Tu.

(B) Reaction mixtures and assay conditions were the same as in A,except that 60 pmol of EF-Tu from either strain D22 or D2216 or 60pmol of EF-Tu mixtures from either origin was incubated with 12pmol of 70S ribosomes and 12 pmol of EF-G; 100% of activity corre-

sponds to 45 pmol of [14C]poly(Phe) with EF-Tu from strain D22 andto 40 pmol with EF-Tu from strain D2216. 0, Kirromycin-sensitiveEF-Tu; 0, kirromycin-resistant EF-Tu; V, A, and +, mixtures ofkirromycin-sensitive and -resistant EF-Tu containing 5096, 80%, or

90% of kirromycin-resistant EF-Tu, respectively.

kirromycin resistance (not shown), but it decreased the effectof the antibiotic on the sensitive EF-Tu, as expected from thecompetition between EF-Ts and kirromycin for EF-Tu [50%inhibition of poly(Phe) synthesis at 0.5 ,M kirromycin, as al-ready reported; refs. 9-12].When resistant and sensitive EF-Tu were mixed, sensitivity

was dominant, provided that the sensitive EF-Tu was presentin a quantity sufficient to block the ribosomal binding sites (Fig.1B). The residual activity in polypeptide synthesis (about 20%)observed in the EF-Tu mixture containing 80% of resistantEF-Tu does not contradict this conclusion but represents theinitial turnover of peptide bond formation that can occur untilall ribosomes are blocked by the kirromycin-sensitive EF-Tu.In this experiment, the molar ratio of EF-Tu to ribosomes was5:1, in line with the present knowledge of in vivo conditions (4,25). Dominance of sensitive EF-Tu is a consequence of kir-romycin action which inhibits protein synthesis by blockingEF-Tu release from the ribosome after enzymatic binding ofaminoacyl-tRNA and subsequent GTP hydrolysis (9-12).With increasing temperature up to 40°, both control and

mutant EF-Tu became progressively less susceptible to theantibiotic. This loss of sensitivity was more pronounced withthe mutant EF-Tu, so that at 400 a concentration of the anti-biotic 700-fold higher than with the parental strain was nec-

essary for a 50% inhibition of its activity (Fig. 2).

Proc. Nati. Acad. Sci. USA 74 (1977)

Dow

nloa

ded

by g

uest

on

Janu

ary

26, 2

021

B Proc. Natl. Acad. Sci. USA 74 (1977) 4343

-6

c

0

E20

-

.0

4Eg0oI_vy

30 40 50Temperature, 0C

FIG. 2. Effect of temperature on kirromycin sensitivity of EF-Tufrom strain D22 and strain D2216. The concentrations of kirromycinneeded for a 50% inhibition of poly(U)-directed poly(Phe) synthesisat the indicated temperatures were determined in a series of kir-romycin concentrations of 3 X 10-8, 10-7, 3 X 10-7, 10-6, 3 X 10-6,and 10-5 M with EF-Tu from strain D22 (0) and of 3 X 10-6, 10-5,3 X 10-5, 10-4, 3 X 10-4, and 10-3M for EF-Tu from strain D2216 (0).In either case, activity of a sample without kirromycin was taken as100%. Reaction mixtures contained, in 75 pl of standard buffer, 36pmol of EF-Tu, 12 pmol of 70S ribosomes, 1.2 sg of poly(U), 2 nmolof tRNA, 2 ,g of a mixture of aa-tRNA synthetases, 100 pmol of[14C]phenylalanine (522 Ci/mol), 36 pmol of EF-G, 75 nmol ofATP,and 75 nmol of GTP. Reactions were started by adding EF-Tu, aa-tRNA synthetases, tRNA[14C]phenylalanine, ATP, and GTP. After10 min of incubation at the indicated temperatures, activity was de-termined as [14C]phenylalanine incorporation into hot trichloroaceticacid-insoluble material.

One of the most striking effects of kirromycin is its abilityto substitute for aminoacyl-tRNA and ribosomes in the acti-vation of the center for GTP hydrolysis located on EF-Tu(9-12). In the presence of the antibiotic there is a significantGTPase reaction with EF-Tu alone. When this critical activitywas tested with the altered EF-Tu, the kirromycin concentra-tion needed for a 50% activation corresponded to that requiredfor 50% inhibition of poly(U)-directed poly(Phe) synthesis andas well was 80- to 100-fold higher than with EF-Tu from strainD22 (Fig. 3).On polyacrylamide slab gel electrophoresis without sodium

dodecyl sulfate, the EF-Tu from the mutant disclosed a slightbut significant increase in mobility, whereas in the presence ofthe detergent no difference from that of the control EF-Tucould be observed (not illustrated).To learn the extent of the structural alteration, we compared

the fingerprint of the kirromycin-resistant with that of thesensitive EF-Tu. Although there are some uncertainties, in thearea of the electrophoretically slowly moving spots, concerningposition and fluorescence intensity of some tryptic peptides,a clear difference between the fingerprint of the resistantEF-Tu from D2216 and that of sensitive one from either D22or E. coli B (the latter not shown) could be observed in onepeptide (see arrows in Fig. 4 A and B). The RF value of thechanged peptide was 0.69 versus 0.61 for the control; its ratioin electrophoretic mobility was 0.94. The fingerprint of a

mixture of the two hydrolysates emphasizes the clear differencebetween the two peptides (Fig. 4C).We summarize briefly other important properties of the

mutant EF-Tu that are not illustrated. The Kd of EF-Tu forGTP at 0° is normally decreased more than 200-fold in the

030 0E~~~

0. 1

N

~200. 0I-

10/

0-I--0

0 -7 -5 -3log [kirromycinJ

FIG. 3. Stimulation of EF-Tu GTPase activity by kirromycin.Reaction mixtures contained, in 75 ,l of standard buffer, 36 pmol ofEF-Tu from strain D22 (0) or strain D2216 (0), 80 pmol of [y-32P]-GTP (870 Ci/mol), and kirromycin as indicated. The reactions werestarted with 1[y-32P]GTP and, after 10-min incubation at 30°, theywere stopped by addition of 75 Al of 1 M HC104/3mM KH2PO4 andchilled in ice. The GTPase activity was determined as liberation ofinorganic phosphate (21).

presence of kirromycin (from 0.3 ,uM to 1.3 nM; ref. 11) butappears to be much less affected with EF-Tu from the mutant.Whereas kirromycin protects EF-Tu against heat denaturation(11), there is essentially no protection with the mutant EF-Tu.The phenotypic expression of kirromycin resistance in strain

D2216, despite the dominance of the kirromycin-sensitiveEF-Tu in vitro, and the different fingerprint patterns of mutantand wild-type EF-Tu are in apparent contrast to the existenceof two gene copies coding for EF-Tu in E. coil. These findingstherefore suggest the existence of only one active tuf gene inthe mutant.

Genetic Evidence for the Existence of Only One Active tufGene in Strains D22 and D2216. In the preceding section weconcluded that phenotypic expression of kirromycin resistancevia EF-Tu is only compatible with the absence of kirromycin-sensitive EF-Tu. The experiments carried out to investigate thegenetic constitution of EF-Tu in strains D22 and D2216 supportthis conclusion. To localize whether tufA or tufB codes forkirromycin resistance, we examined the cotransducibility ofthis property by general transduction with phage P1 lysates ofstrain D2216 made resistant to rifampicin (strain D2216Rifr).The resulting transductants of strain D22, selected by meansof the transmitted rifampicin resistance, revealed at the sametime kirromycin resistance (15 of 15). Because phage P1 co-transduces markers within 1.8 min distance of the recombi-nation map of the E. coli chromosome, the gene coding forkirromycin resistance must be located in the proximity of therpoB locus which codes for rifampicin resistance. Thus, thekirromycin resistant EF-Tu is a product of the tufB gene(tufBl). A P1 cotransduction of kirromycin resistance into E.coli AT1371 (16), a recipient other than the parental strain D22,was not effective: all of the selected rifampicin-resistantprogeny were kirromycin-sensitive both in vivo (20 of 20) andin the cell-free system (6 of 6). This result emphasizes the dif-ference between strain D22 and strain AT1371 in the geneticconstitution of EF-Tu.To confirm the existence of only one functional tuf gene in

strain D2216 we devised two mating experiments in which ei-

Biochemistry: Fischer et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

26, 2

021

4344 Biochemistry: Fischer et al.

<= < Cococ > E c°m >a.) co

I

-Jen_4.OL :

co._-

0 90 80 70 60 50

d*

L

min

KL 228x

D 2216Rifr

BW 113x

D 2216Rif r

(3D rig n eFIG. 4. Fingerprint patterns of kirromycin-sensitive and kir-

romycin-resistant EF-Tu. EF-Tu from either strain D22 or D2216 was

extensively dialyzed against H20 (24 hr) and then heated for 1 hr at100°. Tryptic digestion was carried out at 370 by pH-Stat titration(pH 8.5) with 25 mM NaOH using an autoburette (Radiometer, Co-penhagen, DK). To start the digestion, 1.5% (wttrypsin/wtEF-Tu) oftrypsin (TPCK-treated, Merck, Darmstadt, GFR) was added to theEF-Tu suspension. When the liberation of acid reached a plateau,additional trypsin was added, 1% after 2 hr and 0.5% after 4 hr. Thereaction was stopped after 5-6 hr, and the peptide digest was lyoph-ilized and redissolved in 10 mM NH4HCO3 (about 20 mg of peptidesper ml). Samples containing 7-10 nmol of EF-Tu were applied topaper sheets 49 X 50 cm (2043b Mgl, Schleicher & Schull, Dassel,GFR) and chromatographed for 16 hr at 240 in n-butanol/pyridine/acetic acid/water, 90:60:18:72 ( vol/vol). Electrophoresis of the driedchromatograms was carried out for 2.5 hr at 240 at 3 kV, 0,25 amp(electrophorator D, Gilson, Villiers-le-Bel, F) in 1.25% pyridine/1.25%acetate buffer. The peptides were visualized by spraying with 0.005%fluorescamine/acetone solution (28) and radiating with UV light (360nm). (A) Kirromycin-sensitive EF-Tu. (B) Kirromycin-resistantEF-Tu. (C) Equimolar mixture of kirromycin-sensitive and -resistantEF-Tu.

ther of the two gene copies coding for EF-Tu were transferredto the mutant. For transmission of tufB+ we chose E. coliBW1 13, a valine-resistant Hfr strain which transfers tufB+early and tufA+ late (17), starting at about 7 min of the E. coli

recombination map (Fig. 5). In spite of a possible transmissionof tufA+, the selective isolation of tufB+ recombinants was

FIG. 5. Chromosomal configuration of the progeny obtained bythe conjugation of KL228 X D2216Rifr and of BW113 X D2216Rifr.The origin of the introduced chromosomal parts is indicated by ar-rowheads. The putative chromosomal configuration of the recombi-nants is illustrated by the double lines.

maintained by screening for streptomycin-resistant progeny.An additional transfer of tufA+ was thus made possible onlyby a crossing-over between the very close tufA and rpsL, thelocus of streptomycin resistance. The Valr,Strr recombinantsof this conjugation-which were rifampicin-sensitive, sup-porting an exchange of tufBl-were sensitive to kirromycinin vvo (20 of 20) and in poly(U)-directed poly(Phe) synthesis(8 of 8). As a donor of tufA+, we used the Hfr strain E. coliKL228, which transfers ilv early and his late (17), startingchromosomal transfer at the replication origin of 83 min (Fig.5). An additional transmission of tufB+ is highly improbable,implying an almost entire transfer of the donor chromosomeas well as an additional double crossing-over if selected forampicillin-resistant recombinants. The ampr,his + progeny-which were streptomycin-sensitive, confirming the exchangeof tufAl region-were sensitive to kirromycin, both in vivo (15of 15) and in the cell-free system (6 of 6). This kirromycinsensitivity associated with transmittance of tufAl confirms thedominance of sensitive EF-Tu.

It is interesting that introduction of either tufA+ or tufB+into strain D2216Rifr restored a cell growth rate comparableto that observed with strain D22. This supports an associationof the EF-Tu alteration mediating kirromycin resistance withthe slowed growth rate.

DISCUSSIONIn each round of the elongation cycle, EF-Tu catalyzes theenzymatic binding of aminoacyl-tRNA to the ribosome withconcomitant hydrolysis of GTP. This is followed by the releaseof the EF-Tu-GDP complex from the ribosome, which allowsthe transfer of the peptidyl moiety from peptidyl-tRNA to thenewly bound aminoacyl-tRNA (for review, see ref. 26). Therecently discovered antibiotic kirromycin (23) inhibits theelongation cycle by acting specifically on EF-Tu (9-12). Kir-romycin affects allosterically the interaction of EF-Tu withGTP, aminoacyl-tRNA, and ribosomes (9-12); in its presence,each ligand can interact efficiently with EF-Tu even in absenceof the normally needed effectors. As a result, the EF-Tu. ribo-some complex fails to dissociate after GTP hydrolysis, inhibitingpeptide bond formation and recyclization of EF-Tu.

Given this mechanism, the large excess (4- to 14-fold) ofEF-Tu over ribosomes (4, 25) in E. coli clearly favors thedominance of kirromycin sensitivity over kirromycin resistance.The phenotypic expression of kirromycin resistance should

Proc. Natl. Acad. Sci. USA 74 (1977)

Dow

nloa

ded

by g

uest

on

Janu

ary

26, 2

021

Proc. Nati. Acad. Sci. USA 74 (1977) 4345

therefore not be possible in the presence of a mixed EF-Tupopulation, because the sensitive EF-Tu fraction is sufficientto block the entire ribosomal population. Because neither spe-cific functions nor structural differences of EF-Tu transcribedfrom either tufA or tufB gene and synthesized in vitro havebeen reported (5, 8, 27), it is justified to assume that both genescontribute to the formation of the EF-Tu population. If a moreactive translation of tufA gene (27) holds true in vivo, tufA geneproducts should predominate in the EF-Tu population. Theapparent contrast between these conclusions and the phenotypicexpression of kirromycin resistance mediated by the tufB codedEF-Tu of strain D2216 can be reconciled with the existence ofE. coli strains carrying a single functional tuf gene. Alterna-tively, the selection in a single step of a mutant strain with onlykirromycin-resistant EF-Tu requires either the simultaneousmutation of both EF-Tu structural genes or the modificationof only one tuf gene and blockage of the other gene copy. Thesimultaneous congruent modification of both tuf genes is ahighly improbable event. Similarly, a blockage of tufA syn-chronous with a mutation of tufB at a distance of 16 min is hardto imagine, because nitrosoguanidine increases the probabilityof secondary mutations only within the 2 min range of thereplication region. Even the possibility that kirromycin resis-tance is not due to an alteration of the tuf genes themselves butto a post-transcriptional event involving enzymatic conversionof EF-Tu is not consistent with our data. The neutralization ofresistance by the exchange of the tufBl as well as the tufAlregion would postulate two loci in proximity to the tuf genesalso for this hypothetic enzyme.The results from our biochemical and genetic experiments

all support the existence of only one active tuf gene in strainsD2216 and D22. Functionally, no indication of a mixed pop-ulation was found in the kinetics of the activation of EF-Tu-dependent GTP hydrolysis uncoupled from aminoacyl-tRNAand ribosomes. Moreover, the fingerprint of the EF-Tu fromthe mutant did not reveal the presence of the two peptide typesseen in the 1:1 mixture of kirromycin-sensitive and -resistantEF-Tu.

Although we do not offer a direct proof that E. coli D22 andthe derived mutant D2216 lack an active tufA gene, the exis-tence of only one functional tuf gene, as a prerequisite for theexpression of kirromycin resistance, is supported by our geneticexperiments. Cotransduction of tufBl from E. coli D2216Rifrinto D22 induces kirromycin resistance whereas with anotherrecipient, E. coli AT1371, it results in kirromycin-sensitiveprogeny. In conjugation experiments, kirromycin resistance inthe mutant strain was neutralized by substituting the tufAlgene copy. It cannot be decided at this time whether, in strainsD22 and D2216, tufA is deleted or inactive. Inactivation ofgenes can be caused by mutation, insertion sequence elements,or integration of certain phages (note that strain D22 is lysogenicfor X). The substitution of tufAl by conjugation confirms theresults obtained with the phage P1 transduction of tufBl geneinto E. coli AT1371, showing that additional presence of anintact tuf gene is sufficient to make the resistant cell sensitiveto the antibiotic. It is demonstrated hereby that in vivo bothtufA and tufB gene products can participate in protein bio-synthesis.The existence of E. coli strains with only one active tuf gene

can be an excellent tool to elucidate the role of the two EF-Tugene copies. Moreover, the isolation of an EF-Tu resistant tothe action of a specific agent offers additional possibilities forthe investigation of the activity of this elongation factor in thetranslation of proteins at the ribosomal level as well as for

delimiting its functions in other biological processes in thebacterial cell.

After completion of this manuscript, we learned from Dr. L. Boschand Mr. J. v.d. Klundert (State University, Leiden) that they had alsosucceeded in isolating a kirromycin-resistant mutant with alteredEF-Tu. We wish to thank them for the kind exchange of information.We are indebted to Dr. G. Mosig (Vanderbilt University, Nashville)and Dr. G. Sander for suggestions and criticism during the preparationof this manuscript. This work was supported in part by Grant SFB 76of the Deutsche Forschungsgemeinschaft and Grant 76.71.186 of theDelegation Generale a la Recherche Scientifique et Technique. E.F.is a recipient of a postdoctoral fellowship of the Deutsche For-schungsgemeinschaft.

1. Travers, A. A., Kamen, R. I. & Schleif, R. F. (1970) Nature 228,748-751.

2. Blumenthal, T., Landers, T. A. & Weber, K. (1972) Proc. Natl.Acad. Sci. USA 69,1313-1317.

3. Weissbach, H., Redfield, B. & Kaback, H. R. (1969) Arch, Bio-chem. Biophys. 135,66-74.

4. Jacobson, G. R. & Rosenbusch, J. P. (1976) Nature 261, 23-26.5. Jaskunas, S. R., Lindahl, L., Nomura, M. & Burgess, R. R. (1975)

Nature 257,458-462.6. Lupker, J. H., Verschoor, G. J., De Rooij, F. W. M. & Bosch, L.

(1974) Proc. Natl. Acad. Sci. USA 71,460-463.7. Young, F. S. & Neidhardt, F. C. (1975) Abstr. A. Meeting Am.

Soc. Microbiol., 75th Annual Meeting, New York, 161.8. Pedersen, S., Blumenthal, R. M., Reeh, S., Parker, J., Lemaux,

P., Laursen, R. A., Nagarkatti, S. & Friesen, J. D. (1976) Proc.Nati. Acad. Sci. USA 73, 1698-1701.

9. Wolf, H., Chinali, G. & Parmeggiani, A. (1974) Proc. Natl. Acad.Sci. USA 71, 4910-4914.

10. Parmeggiani, A., Wolf, H. & Chinali, G. (1976) in Ribosomes andRNA Metabolism, eds. Zelinka, J. & Balan, J. (Publishing Houseof the Slovak Academy of Sciences, Bratislava), Vol. 2, pp.283-290.

11. Chinali, G., Wolf, H. & Parmeggiani, A. (1977) Eur. J. Biochem.75,55-65.

12. Wolf, H., Chinali, G. & Parmeggiani, A. (1977) Eur. J. Biochem.75,67-75.

13. Brown, S. & Blumenthal, T. (1976) J. Biol. Chem. 251, 2749-2753.

14. Normark, S., Boman, H. G. & Matsson, E. (1969) J. Bacteriol. 97,1334-1342.

15. Monner, D. A., Jonsson, S. & Boman, H. G. (1971) J. Bacteriol.107,420-432.

16. Bachmann, B. J., Low, K. B. & Taylor, A. L. (1976) Bacteriol. Rev.40, 116-167.

17. Low, K. B. (1973) J. Bacteriol. 113,798-812.18. Miller, J. H. (1972) Experiments in Molecular Genetics (Cold

Spring Harbor Laboratory, Cold Spring Harbor, NY).19. Davis, B. D. & Mingioli, E. S. (1950) J. Bacteriol. 60, 17-28.20. Parmeggiani, A., Singer, C. & Gottschalk, E. M. (1971) in

Methods in Enzymology, eds. Moldave, K. & Grossman, L.(Academic Press, New York), Vol. 20, pp. 291-302.

21. Sander, G., Marsh, R. C., Voigt, J. & Parmeggiani, A. (1975)Biochemistry 14, 1805-1814.

22. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J.(1951) J. Biol. Chem. 193,265-275.

23. Wolf, H. & Zahner, H. (1972) Arch. Mikrobiol. 83, 147-154.24. Fischer, E. (1976) Ph.D Dissertation, Universitit Tubingen,

GFR.25. Furano, A. V. (1975) Proc. Natl. Acad. Sci. USA 72, 4780-

4784.26. Lucas-Lenard, J. & Lipmann, F. (1971) Annu. Rev. Biochem.

40,409-448.27. Pedersen, S., Reeh, S. V., Parker, J., Watson, R. J., Friesen, J. D.

& Fiil, N. P. (1976) Mol. Gen. Genet. 144,339-343.28. Vandekerckhove, J. & Van Montagu, M. (1974) Eur. J. Biochem.

44,279-288.

Biochemistry: Fischer et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

26, 2

021