Vol. 172, No. 6

Use of Natural mRNAs in the Cell-Free Protein-SynthesizingSystems of the Moderate Halophile Vibrio costicola

C. G. CHOQUETt AND D. J. KUSHNERt*

Department of Biology, University of Ottawa, Ottawa, Ontario, Canada KIN 6N5

Received 19 October 1989/Accepted 2 April 1990

In vitro protein synthesis was studied in extracts of the moderate halophile Vibrio costicola by using as

mRNAs the endogenous mRNA of V. costicola and the RNA of the R17 bacteriophage of Escherichwa coli.Protein synthesis (amino acid incorporation) was dependent on the messenger, ribosomes, soluble cytoplasmicfactors, energy source, and tRNAmet (in the R17 RNA system) and was inhibited by certain antibiotics. Theseproperties indicated de novo protein synthesis. In the V. costicola system directed by R17 RNA, a protein of thesame electrophoretic mobility as the major coat protein of the R17 phage was synthesized. Antibiotic action andthe response to added tRNA4et showed that protein synthesis in the R17 RNA system, but not in theendogenous messenger system, absolutely depended on initiation. Optimal activity of both systems was

observed in 250 to 300mM NH4' (as glutamate). Higher salt concentrations, especially those with Cl- as anion,were generally inhibitory. The R17 RNA-directed system was more sensitive to Cl- ions than the endogenoussystem was. Glycine betaine stimulated both systems and partly overcame the toxic effects of Cl- ions. Bothsystems required Mg2+, but in lower concentrations than the polyuridylic acid-directed system previouslystudied. Initiation factors were removed from ribosomes by washing with 3.0 to 3.5 M NH4C1, concentrationsabout three times as high as that needed to remove initiation factors from E. coli ribosomes. Washing with 4.0M NH4Cl damaged V. costicola ribosomes, although the initiation factors still functioned. Cl- ions inhibited theattachment of initiation factors to tRNAmet but had little effect on binding of initiation factors to R17 RNA.

Polyuridylic acid [poly(U)]-directed incorporation ofphenylalanine by cell extracts of the moderate halophileVibrio costicola is inhibited by Cl- ions (9), which preventribosome binding to the artificial mRNA (3).There are a number of differences between a translation

system directed by poly(U) and that directed by a naturalmRNA. Poly(U) has no Shine-Dalgarno sequence, whichnormally plays an important role in the initial attachment ofnatural mRNAs to ribosomes. At high Mg2" concentrations(18 mM), the initiation factors are not required in thepoly(U)-directed system ofEscherichia coli (15). Finally, theinitiator tRNA is not involved, since poly(U) has no AUGinitiator codon. Therefore, the initiation process of proteinsynthesis is not as well represented in an in vitro translationsystem directed by poly(U) as in one directed by a naturalmessenger.We wanted to study the effects of Cl- ions on protein

synthesis in V. costicola directed by a natural mRNA.Unfortunately, none have been isolated from this or anyother moderate halophile, although some studies have beendone with a mixture of unidentified endogenous mRNAs(38). We have now found that a natural mRNA whichfunctions well in the cell-free translation system of E. coli,that of bacteriophage R17 (2), can function in vitro with V.costicola. This study describes some properties of both V.costicola systems and studies the site of action of differentions.

* Corresponding author.t Present address: National Research Council, Division of Bio-

logical Sciences, Ottawa, Ontario, Canada KlA OR6.t Present address: Department of Microbiology, University of

Toronto, Toronto, Ontario, Canada M5S 1A8.

MATERIALS AND METHODS

Bacterial culture. V. costicola NRC 37001 was grown andharvested as described previously (9).

Preparation of cellular extracts, ribosomes, and initiationfactors. S-30 extract, that is, the cellular material remainingin the supernatant after centrifuging for 30 min at 30,000 x

gmax (Beckman type 42.1 rotor), was prepared as describedpreviously (9).The S-150 extract and the low-salt-washed ribosomes

were obtained by centrifuging the S-30 extract at 150,000 x

gmax for 3 h at 4°C. The soluble S-150 fraction and theribosomal pellet were then dialyzed or washed and storedaccording to the procedure of Kamekura and Kushner (9).A different low-salt extraction buffer than the one de-

scribed by Kamekura and Kushner (9) was used in thepreparation of the cellular extracts and of the ribosomes. Itcontained 124 mM ammonium glutamate, 8 mM magnesiumacetate, 10 mM Tris acetate (pH 7.6), 3 mM spermidine-trihydrochloride, and 6 mM P-mercaptoethanol.Crude initiation factors were prepared by following the

method described for E. coli (27), with some modifications toadapt it to V. costicola. Ribosomes were homogenized inextraction buffer (15 ml/10 g of wet weight of cells) in whichthe ammonium glutamate had been replaced by 3.0 to 3.5 MNH4Cl (high-salt extraction buffer) and left at 4°C overnight.These high-salt-washed ribosomes were spun down (150,000X gmax for 3 h at 4°C), washed once with low-salt extractionbuffer, and stored at -70°C. Solid (NH4)2SO4 (0.52 g/ml ofsupernatant) was added, and the solution was stirred in thecold for 2 h. The precipitated proteins were centrifuged(15,000 x gmax for 30 min), dissolved in extraction buffer (0.5ml/10 g of original wet weight of cells), dialyzed overnightagainst several liters of this buffer, clarified by centrifuga-tion, and then stored at -70°C.Growth of phage R17 and isolation of its RNA. E. coli

CSH39 was grown with maximum aeration at 37°C in 2-liter

3462

JOURNAL OF BACTERIOLOGY, June 1990, p. 3462-34680021-9193/90/063462-07$02.00/0Copyright © 1990, American Society for Microbiology

on May 26, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

USE OF NATURAL mRNAs IN CELL-FREE SYSTEMS 3463

Erlenmeyer flasks containing 250 ml of 4YT medium (32 g oftryptone, 20 g of yeast extract, 5 g of NaCl per liter [19]) upto an optical density of 0.4 (550 nm). The bacterial culturewas then made 5 mM in CaCl2 (19), infected with R17 at amultiplicity of 0. 15, and incubated for an additional 4 h. Withthis incubation time, titers of 7 x 1011 PFU/ml were rou-tinely obtained. For labeling the R17 RNA, 400 ,uCi of[6-3H]uridine (29 Ci/mmole; NEN Research Products) perliter of medium was added at the time of infection (1).The following operations were all carried out at 4°C. After

incubation, EDTA, MgSO4, lysozyme, and DNase wereadded to final concentrations of 50 mM, 200 mM, 100 ,ug/ml,and 2 ,ug/ml, respectively (35). The bacteriophage was thencollected by the method of Steitz (27). It was precipitatedovernight by adding 330 g of (NH4)2SO4 per liter of mediumand recovered by centrifugation (25,000 X gmax for 30 min).The pellet was well suspended in 100 ml of standard sodiumcitrate buffer (SSC) (lx SSC is 0.15 M NaCl plus 0.015 Msodium citrate [pH 7.2]) per liter of lysate, and the solutionwas clarified by centrifugation (25,000 X gmax for 10 min).The phage (20 ml per tube) was then pelleted at 100,000 xgm.x (Beckman type 42.1 rotor) for 4 h, and each pellet wasresuspended overnight in 2 ml of SSC.The phage was purified further by two successive sedi-

mentations through CsCl step gradients (39). After eachCsCl centrifugation, the visible phage band was removed,diluted to 20 ml with SSC, pelleted at 100,000 x g.,, for 4 hand resuspended in SSC (2 to 4 ml). The RNA was thenextracted by the method of Steitz (27), which includedsolubilization of the bacteriophage with 1% sodium dodecylsulfate (SDS), followed by phenol (SSC saturated) extractionand ethanol precipitation.

In vitro protein synthesis assays. For the R17 RNA-di-rected in vitro protein-synthesizing system, the completereaction mixture (50 ,ul) contained 15 mM phosphoenolpyru-vate, 2 mM ATP, 1 mM GTP, 225 mM ammonium glutamate(including the ammonium glutamate contributed by the cel-lular extracts and ribosomal preparations), 8 mM Mg2+ (asmagnesium acetate), 7.5 mM reduced glutathione, 82 mMTris acetate (pH 7.6), 1.25 ,uM tRNA"Ct (from E. coli MRE600; Boehringer Mannheim Biochemicals), 0.03 mM[14C]valine (225 mCi/mmole; ICN Biomedicals CanadaLtd.), 0.05 mM each of the other 19 amino acids, 100 p.g ofR17 RNA, 0.2 volume of S-30 (ca. 0.8 mg of protein) or 10 ,ulof S-150 (ca. 0.5 mg of protein) and 2 A260 units of ribosomesor 10 ,ul of S-150, 4 A260 units of NH4Cl-washed ribosomes,and 39 p,g of crude initiation factors.The complete reaction mixture (50 [lI) for the in vitro

translation system directed by endogenous mRNA of V.costicola contained 15 mM phosphoenolpyruvate, 2 mMATP, 1 mM GTP, 300 mM ammonium glutamate (includingthe ammonium glutamate contributed by the S-30), 8 mMMg2+ (as magnesium acetate), 7.5 mM reduced glutathione,82 mM Tris acetate (pH 7.6), [U-14C]amino acids (1.74mCi/mg; ICN Biomedicals Canada Ltd.) at a concentrationof 0.05 mM each (adjusted with unlabeled amino acids), anda 0.4 volume of S-30 (ca. 1.5 mg of protein).

In both systems, protein synthesis was followed by mea-suring the amount of radiolabeled amino acid(s) incorporatedinto hot trichloroacetic acid (TCA)-insoluble material. Reac-tion mixtures were incubated at 30°C for the times indicated.Reactions were stopped by adding 500 p.l of 5% TCA; themixtures were heated at 90°C for 20 min, cooled in ice, andfiltered (membrane filters [Millipore Corp.], type HA; poresize, 45 p.m). The filters were washed twice with 5 ml of 5%TCA, dried, and mixed with 4 ml of Universol Cocktail (ICN

Biomedicals Canada Ltd.). The radioactivity trapped on thefilters was then counted.

Assay of formylmethionyl-tRNA synthetase. The enzymereaction was performed in a mixture of the same composi-tion as that for R17 RNA-directed protein synthesis exceptthat R17 RNA was not included and that ['4C]valine and theother 19 amino acids were replaced by 0.07 mM of[14C]methionine (285 mCilmmole; Amersham Corp.). S-150was used as the source of enzyme. Incubation was at 30°Cfor 20 min. Synthetase activity was measured by determiningthe amount of radioactivity rendered insoluble by 500 p.l ofcold 10% TCA and then retained on Millipore filters (typeHA; pore size, 45 p.m) (5).

Binding of ['4C]fmet-tRNA or [3H]R17 RNA by initiationfactors. Binding of [14C]fmet-tRNA (29) and [3H]R17 (8) wasmeasured by determining the amount of radioactivity re-tained on nitrocellulose filters (BA 85; Schleicher & Schuell,Inc.). The reaction mixture contained the following: 225 mMammonium glutamate, 8 mM Mg2+ (as magnesium acetate),82 mM Tris acetate (pH 7.6), 7.5 mM reduced glutathione, 40p.g of crude initiation factors, and 11 pmol of [14C]fmet-tRNA (550 dpm/pmol) or 100 p.g of [3H]R17 RNA (900dpm/l,g). The reaction mixture (50 pl.) was incubated at 30°Cfor 15 min, cooled on ice, filtered, and washed twice with 3ml of ice-cold buffer (225 mM ammonium glutamate, 8 mMmagnesium acetate, 82 mM Tris acetate [pH 7.6], 6 mMP-mercaptoethanol). The filters were then dried, and theradioactivity was counted.

Synthesis of [14C]fmet-tRNA. The charging and formyla-tion reactions were done by the method of Voorma et al. (32)with modifications of the reaction mixture to adapt it toV. costicola. The reaction mixture (1 ml) contained thefollowing: 225 mM NH4 glutamate, 8 mM magnesium ace-tate, 82 mM Tris acetate (pH 7.6), 6 mM 3-mercaptoethanol,1.5 mM ATP, 10 mM phosphoenolpyruvate, S A260 units oftRNAMet, 10 p.g of folinic acid, 3.5 mM [14C]methionine(225 mCi/mmol; ICN Biomedicals Canada Ltd.), and 250 p.lof S-150 of V. costicola. Incubation was at 30°C for 15 min.The initiator tRNA, fmet-tRNA, was then isolated andpurified by phenol extraction and ethanol precipitation asdescribed by Voorma et al. (32).

Gel electrophoresis and autoradiogram. Following incuba-tion, the reaction mixture of the R17 RNA-directed proteinsynthesis system was mixed with a 0.125 volume of samplebuffer (250 mM Tris hydrochloride [pH 6.8], 5% SDS, 5%2-mercaptoethanol, 50% glycerol, 0.1% bromophenol blue)and 0.2 A260 units of R17, as carrier, and heated for 3 min at90°C (5). Fifty-microliter samples were used for electropho-resis.Gel slabs for SDS-polyacrylamide gel electrophoresis,

with linear gradients of polyacrylamide (10 to 15%) andglycerol (0 to 13%) (5), were prepared by the method ofLaemmli (12) and run at 25 mA. Proteins were fixed andstained with Coomassie blue R250 (0.1%) in water-methanol-glacial acetic acid (5:5:2). The gels were destained in asolution of 12.5% isopropanol and 10% acetic acid and thendried in vacuo on Whatman 3MM paper. The dried gels wereautoradiographed at -80°C with Kodak X-ray film (X-OmatAR) and with an intensifying screen.

RESULTS AND DISCUSSION

Characterization of cell-free translation systems directed bynatural mRNAs. The RNA of the R17 phage of E. coli wasable to act as a messenger for in vitro protein synthesis byextracts of V. costicola. The incorporation of valine, one of

VOL. 172, 1990

on May 26, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

3464 CHOQUET AND KUSHNER

00

0.5 1.0 2.0 30Ribosomes (A 260 units ) *--

I a a. . I

0 0.25 0.50 0.75 1.00 1.25

tRNAfm ( Mm )

0 4 8 12 16

I

,4v m.O

0-c

0 25 50 75 100R17 RNA (-jg-)o-

0 0.25 0.50 0.75 1.00S-150 (mg protein) n--

FIG. 1. Effects of concentrations of different components of theR17 RNA-directed protein-synthesizing system of V. costicola.Symbols: 0, ribosomes; 0, tRNAMe,; O, R17 RNA; *, S-150.

the predominant amino acids found in the coat protein of thisphage (16), was absolutely dependent on the presence of thisRNA, as well as on ribosomes and substances found in theS-150 supernatant (Fig. 1). Final incorporation increasedwith increasing concentrations of ribosomes, mRNA, andS-150 supernatant (Fig. 1). Added tRNAm et increased, butwas not essential for, incorporation (Fig. 1). Incorporationwas also absolutely dependent on a mixture of ATP, GTP,and phosphoenolpyruvate (not shown), which presumablyserved as an energy source in different steps of translation.Effects of individual components of this mixture were notstudied.

In an in vitro system containing ribosomes and the S-150fraction, incorporation was linear for the first 30 min; whenS-30 fractions, which contained both ribosomes and theS-150 fraction, were used, valine incorporation reached atleast as high a plateau as in the former system by 10 min (notshown). Consequently, in experiments to determine totalvaline incorporation, the reaction mixture was incubated for10 min in the presence of S-30 fractions and for 30 min in thepresence of S-150 fractions plus ribosomes.

SDS-polyacrylamide gel electrophoresis showed that thenewly synthesized proteins made in the presence of R17RNA comigrated with purified R17 coat protein (Fig. 2).

In vitro protein synthesis directed by endogenous mRNAsof V. costicola was also dependent on the amount of cellextract. Incorporation reached a plateau within 10 min ofincubation. This system did not respond to increasingamounts of tRNAmet (not shown).

Effect of protein synthesis inhibitors. Chloramphenicol and

FIG. 2. Autoradiogram of translated products of R17 RNA-di-rected in vitro protein synthesis by the S-30 fraction of V. costicola.The time of incubation (in minutes) is indicated above each lane. Firstlane on the right, SDS-gel electrophoresis of purified A and coatproteins of the R17 bacteriophage stained with Coomassie blue.

neomycin (10-' M), which act on peptide chain formation(10, 20), strongly (93 to 98%) inhibited protein synthesis inboth the R17-directed and endogenous systems. Kasugamy-cin (4 x 10-4 M) and aurin tricarboxylic acid (4 x 10-5 M),which inhibit initiation of translation (6, 18), inhibited R17-directed translation by 80% but had less effect (30 and 5%inhibition, respectively) on the endogenous system. Prein-cubation with RNase completely inhibited translation in bothsystems.The above experiments confirm that amino acid incorpo-

ration in both systems involves de novo protein synthesis.Both systems are more physiological than the poly(U)-directed system studied previously (3, 9), but there areimportant differences between them. In the R17-directedsystem, every ribosome that becomes involved in proteinsynthesis must first go through normal initiation. In contrast,the translation of endogenous mRNA presumably involvesmainly the completion of nascent polypeptides, as in thecell-free translation system directed by endogenous mRNAsof E. coli (6, 10). Preattached ribosomes can carry outelongation immediately without having first to go throughinitiation.The differences are reflected by the action of different

protein synthesis inhibitors: those (chloramphenicol andneomycin) which inhibit polypeptide elongation act similarlyon both systems, while those (kasugamycin and aurin tricar-boxylic acid) that primarily affect initiation act much morestrongly on the R17-directed system.Response of translation systems to different ions. The R17-

directed protein-synthesizing system required Mg2' andNH4' ions (ammonium glutamate) at concentrations of 8 and225 mM, respectively, for optimal activity (Fig. 3A). Potas-sium glutamate could partly replace ammonium glutamate(Fig. 3A). However, sodium glutamate, NaCl, KCl, orNH4Cl alone did not support any levels of protein synthesis(not shown).

J. BACTERIOL.

on May 26, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

USE OF NATURAL mRNAs IN CELL-FREE SYSTEMS 3465

aS

Z._

a00

cY

100

!! 80

,60" 40-

20

0 5 10 15 20Mg+* ( mM ) u

0S

-a

0 0.2 0.4 0.6 0.8 1.0

A Salt concentrotion (M) B Salt concentration ( M)

FIG. 3. Ionic requirements of the R17 RNA-directed protein-synthesizing system (A) and the in vitro endogenous translation system of V.costicola (S-30 fraction) (B). Symbols: A, ammonium glutamate; M, potassium glutamate; 0, sodiumn glutamate; A, NH4Cl; O, KCl; 0, NaCl.Inserted figures show the effects of Mg2" ions on in vitro protein synthesis. Results are expressed as percentage of maximum activity. lOO1 oactivity represents incorporation of 50 pmol of ['4C]valine into hot TCA-insoluble material per 10 ,ul of S-30 fraction in the R17 RNA systemand 13,500 dpm per 20 ,ul of S-30 fraction in the endogenous system.

The translation of endogenous mRNA also worked best ata Mg2" ion concentration of 8 mM but required slightly moreammonium glutamate (300 mM) than the R17 system foroptimal activity (Fig. 3B). Ammonium glutamate could bereplaced quite efficiently by potassium glutamate, partly byNH4Cl and KCl (at lower optimal concentrations), and not atall by sodium glutamate and NaCl (Fig. 3B).The low Mg2" ion requirement (8 mM) of both systems is

typical of translation systems directed by natural mRNAs(15, 17). In contrast, the poly(U)-directed in vitro system ofV. costicola required more than twice as much Mg2" (38).Adding NaCl, KCl, or NH4Cl to either system in the

presence of an optimal ammonium glutamate concentrationinhibited protein synthesis. For example, addition of 0.1 Mof these salts caused approximately 90% inhibition of trans-lation in the R17-directed system. Such an inhibition wasalso caused by 0.2 M sodium glutamate, 0.4 M ammoniumglutamate, or 0.6 M potassium glutamate.

All concentrations of added NaCl, KCl, NH4Cl, andammonium glutamate also inhibited translation of endoge-nous mRNAs, although this system was less sensitive thanthe R17-directed system. Thus, ca. 0.6 M NaCl was neededto cause 90% inhibition. However, 0.2 M added sodium andpotassium glutamate stimulated this system 125 and 100%,respectively; 0.4 M concentrations caused some stimulation,and higher concentrations were inhibitory (not shown).These results confirm the toxic nature of the Cl- ions on in

vitro protein synthesis by V. costicola, effects previouslyobserved with poly(U) as artificial mRNA (3, 9, 38). Thegreater sensitivity of the R17-directed system than theendogenous system to added Cl- ions suggests that theseions are more toxic to initiation than to elongation.The fact that sodium or potassium glutamate stimulated

translation directed by endogenous mRNA, in the presenceof optimal ammonium glutamate, must be due to the Na+and K+ ions, which are found in high concentrations in thesebacteria (25) and which would be expected to play animportant role in protein synthesis. This stimulation isprobably directed towards elongation rather than initiation.mRNA structure. We considered the possibility that inhi-

bition of the R17 RNA-directed system was due to changesin the secondary structure of the RNA with increasing saltconcentrations. Such changes could have resulted in ashielding of the normally available ribosome-binding sites.However, similar results (not shown) were obtained withpartially denatured formaldehyde-treated (13) R17 RNA.Thus, salts probably did not act by changing RNA structure.Lodish (13) had previously found that partially denaturedviral RNA was more active in stimulating in vitro proteinsynthesis by E. coli, but we did not observe such an increasein our system.

Effects of added betaine and glutamate. Glycine betaine(betaine) stimulated protein synthesis in both systems, espe-cially that directed by R17 RNA. Its presence also resultedin higher activity, in both systems, in the presence of NaClor KCI (Fig. 4A and B). Therefore, it seems that betaine canafford some protection against the toxic action of Cl- ions,as was previously seen with the poly(U) system (3, 9).Although glutamate also affords protection against Cl- ionsin the poly(U) system (3, 9), further addition of 0.2 M sodiumor potassium glutamate was inhibitory in the systems studiednow, especially the R17 RNA-directed system (Fig. 4A andB).Formylmethionyl-tRNA synthetase. We investigated this

enzyme as a possible site of action of Cl- ions. However, itsactivity was only inhibited 20% by as much as 1.0 M NaCl(not shown), and it is clearly not a primary site of Cl- iontoxicity. It was previodsly shown. (9) that phenylalanyl-tRNA synthetase was not a site of Cl- action in the poly(U)-directed system.

Isolation of crude iD,itiation factors of V. costicola. Initiationfactors were isolated from E. coli ribosomes by washing with1 M NH4Cl (27) but were not isolated from those of V.costicola. We found that washing the latter ribosomes with1.0 and 2.0 M NH4C1 did not decrease their ability to carryout in vitro protein synthesis. However, those washed with3.0, 3.5, and 4.0M NH4Cl had relative activities of 15, 6, and2%, respectively. Ribosomes treated with 3.0 and 3.5 MNH4Cl were active only if protein obtained during thewashing procedures (see Materials and Methods) was added

VOL. 172, 1990

on May 26, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

3466 CHOQUET AND KUSHNER

>1

0

0

0

B

04 0.6

NaCI (M)

0.4 0.6 0.8 1.0

NaCI (M)

FIG. 4. Effects of betaine alone or with sodium (or potassium) glutamate on the R17 RNA-directed protein-synthesizing system (A) andthe endogenous translation system of V. costicola (S-30 fraction) (B). Symbols: 0, no addition; A, 0.5 M betaine; *, 0.5 M betaine plus 0.2M potassium glutamate; 0, 0.5 M betaine plus 0.2M sodium glutamate. Results are expressed as percentage of control (without added solute).100%6o activity represents incorporation of 47 pmol of [14C]valine into hot TCA-insoluble material per 10 ,ul of S-30 in the R17 RNA system

and 14,200 dpm per 20 of S-30 in the endogenous system. Higher levels of betaine, up to 1.2 M, caused no further stimulation of activityin either system.

again (Table 1). Presumably, these proteins included theinitiation factors. However, a much lower incorporation wasobserved with 4.0 M NH4Cl-washed ribosomes recombinedwith the crude initiation factors isolated from them (Table 1).Reconstruction studies (not shown) between 3.5 and 4.0 MNH4Cl-washed ribosomes and their respective initiationfactors showed that this lack of incorporation was mainlydue to inactive 4.0 M NH4Cl-washed ribosomes rather thaninactive extracted initiation factors.

Binding of [14C]fnet-tRNA or [3HJR17 RNA by initiationfactors. Retention of ['4C]fmet-tRNA or [3H]R17 RNA onnitrocellulose filters was specifically dependent on the addi-tion of crude initiation factors (Table 2). This retention wasnot due to unspecific binding of the radiolabeled compoundsto proteins, since the addition of S-150 from V. costicola,rich in proteins, barely increased the amount of materialretained on the filter. Presumably, retention of [14C]fmet-tRNA was caused by the formation of a complex betweenthe tRNA and initiation factor-2 (IF2). This complex is notstable with IF2 isolated from E. coli and is normally stabi-lized by fixing with glutaraldehyde (29, 30). However, withthe initiation factors of V. costicola, fixing with 3% glutaral-dehyde did not increase the amount of ['4C]fmet-tRNAretained on the filters (not shown). The retention of [3H]R17

TABLE 1. R17 RNA-directed incorporation of [14C]valine (dpm)at different concentrations of added crude initiation factors

isolated from NH4CI-washed ribosomes of V. costicola

Incorporation (dpm)in ribosomesInitiation (4 A260 units) washed in:

factor' (,ug)3.0 M NH4Cl 3.5 M NH4CI 4.0 M NH4C1

0 4,238 1,746 72113 8,210 5,855 95326 11,210 9,775 1,00439 12,097 10,531 1,253

a Crude initiation factors isolated from the ribosomes washed at eachNH4Cl concentration.

RNA was presumably due to the formation of a complexbetween itself and IF3, as is the case with IF3 of E. coli (7,22, 33).

In the presence of initiation factors, ribosomes interferedwith the retention of both radioactive molecules. In E. coli,IF3, which is required for mRNA binding, has a greateraffinity for 30S ribosomal subunits than MS2RNA does (31).There are no data on the association of IF2 with fmet-tRNA,but it is known that the association constant of IF2 with 30Sribosomal subunits, in the presence of IF3 and IF1, is as highas that of IF3 (36). This may also be true for V. costicola, sothat the amounts of initiation factors available for the bindingof fmet-tRNA to IF2 and R17 RNA to IF3, in the absence ofribosomes, could decrease in the presence of ribosomes dueto the higher affinity of the initiation factors for the ribo-somes.These results also show that the retention of [14C]fmet-

tRNA and [3H]R17 RNA on nitrocellulose filters was asuitable method to study the effects of salts on the initiationfactors of V. costicola. This should, at least, give us someindication of the ability of these factors to recognize theirrespective ligand, namely fmet-tRNA for IF2 and mRNA forIF3, under different ionic conditions.Cl- ions had a strong inhibitory effect on the binding of

TABLE 2. Retention of [14C]fmet-tRNA and [3H]R17 RNAon nitrocellulose filters

Components ['4C]fmet-tRNA [3H]R17 RNAaddeda (dpm) (dpm)

No addition 100 152Ribosomes (4 A26 units) 124 232Initiation factors (40 p.g) 4,828 6,770S-150 (350 pLg of proteins) 120 440Ribosomes (4 A26 units) plus 743 2,736

initiation factors (40 ,ug)a The reaction mixture contained 225 mM ammonium glutamate, 8 mM

Mg2+, 82 mM Tris acetate (pH 7.6), 7.5 mM reduced glutathione, and 11 pmolof [14CJfmet-tRNA (550 dpmtpmol) or 100 pLg of [3H]R17 RNA (900 dpm/4g).

._

0

0

0

A

J. BACTERIOL.

on May 26, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

USE OF NATURAL mRNAs IN CELL-FREE SYSTEMS 3467

Betaine (M)

-

u

0

0

0

ol

Salt concentration ( M )FIG. 5. Effects of added solutes on the retention of [14C]fmet-

tRNA on nitrocellulose filters by crude initiation factors of V.costicola. Symbols: 0, NaCl; 0, sodium glutamate; l, betaine; *,0.6 M betaine at increasing NaCl concentrations. Results are ex-

pressed as percentages of controls (without added solute). 100%

activity represents the retention of 8 pmol of [14C]fmet-tRNA.

[14C]fmet-tRNA by initiation factors: 0.2 M NaCl caused a

90% inhibition, while 0.6 M sodium glutamate caused only a

25% inhibition (Fig. 5). Betaine alone had no effect on

binding over the whole concentration range studied, and itspresence did not protect against the action of Cl- ions.During the initiation of protein synthesis, the initiator tRNAis recognized by and bound to IF2 (29, 30), and therefore itis likely that Cl- ions may interfere with the function of thisfactor.

[3H]R17 RNA was retained on nitrocellulose filters in thepresence of initiation factors over a wide range of saltconcentrations. The results (not shown) had considerablescatter (from 80 to 120% of control values), but there was no

evidence that NaCl, glutamate, or betaine had any real effecton binding, which presumably depends on mRNA-IF3 inter-action.These experiments have identified initiation factors as yet

another site for the toxic action of Cl- ions in these bacteria.Peumans et al. (21) and Weber et al. (34) have shown that invitro eucaryotic translation systems were inhibited by Cl-ions, due to their action on one or more initiation factors.Although V. costicola seems more closely related to other

eubacteria, such as E. coli, than to the extremely halophilicarchaebacteria, it must deal with environments differentfrom those of most other eubacteria. Some of its enzymaticprocesses, including protein synthesis, may seem overly saltsensitive for bacteria that normally live in high salt concen-

trations. This apparent sensitivity is likely due to the Cl-ions, which are maintained in quite low concentrations inthese cells (11). Some of the sites of action of these ions havenow been further defined.We can now further compare the properties of the protein-

synthesizing system of V. costicola with those of the non-

halophile E. coli and of the extremely halophilic archaebac-terium Halobacterium cutirubrum, the most studied speciesof this type.Ribosomes of the extreme halophiles require high salt

(mainly high KCI) concentrations for stability. Those of V.costicola are stable in both high and low salt concentrations,while those of E. coli are stable only in low salt concentra-tions (37). Higher concentrations of NH4Cl are needed toremove initiation factors from the ribosomes of V. costicolathan from those of E. coli.NH4' ions usually stimulate protein synthesis more than

K+ ions (4, 14, 26), but the NH4' requirement in differentmicroorganisms may be quite different. E. coli needs about60 mM NH4+ for maximal activity (5, 28); our present resultsshow that V. costicola needs 250 to 300 mM, while halobac-teria may require as much as 3.0 M NH4' (23, 24), althoughin fact, these organisms contain much lower amounts of thision (11, 23). The halobacteria specifically require Cl- ions,which are found in high concentrations inside their cells, forprotein synthesis (23). As discussed elsewhere (11), this maybe one of the major physiological differences between aero-bic halophilic eubacteria and archaebacteria. The formercontinue to reveal physiological properties somewhat be-tween those of the latter and those of the nonhalophiliceubacteria, to which they are more closely related.

ACKNOWLEDGMENT

This work was supported by a grant to D.J.K. from the NaturalSciences and Engineering Research Council of Canada.

LITERATURE CITED1. Albrecht, J., B. W. Hoogendam, W. Rozenboom, N. J. Verhoef,

H. 0. Voorma, and L. Bosch. 1969. Chain initiation duringpolypeptide synthesis in cell-free bacterial systems programmedwith plant viral messengers. III. A comparison with the phageRNA-directed system. Biochim. Biophys. Acta 190:504-516.

2. Boedtker, H., and S. Stumpp. 1966. Some properties of viralribonucleic acid-dependent amino acid incorporation. Biochim.Biophys. Acta 114:355-370.

3. Choquet, C. G., M. Kamekura, and D. J. Kushner. 1989. In vitroprotein synthesis by the moderate halophile Vibrio costicola:site of action of Cl- ions. J. Bacteriol. 171:880-886.

4. Conway, T. W. 1964. On the role of ammonium or potassium ionin amino acid polymerization. Proc. Natl. Acad. Sci. USA51:1216-1220.

5. Ellis, S., and T. W. Conway. 1984. Initial velocity kineticanalysis of 30 S initiation complex formation in an in vitrotranslation system derived from Escherichia coli. J. Biol. Chem.259:7607-7614.

6. Grolimann, A. P., and M. L. Stewart. 1968. Inhibition of theattachment of messenger ribonucleic acid to ribosomes. Proc.Natl. Acad. Sci. USA 61:719-725.

7. Johnson, B., and M. Szekely. 1979. The binding site of IF-3 inMS2 RNA. Methods Enzymol. 60:343-350.

8. Kaempfer, R., and G. Jay. 1979. Binding of messenger RNA ininitiation of prokaryotic translation. Methods Enzymol. 60:332-343.

9. Kamekura, M., and D. J. Kushner. 1984. Effect of chloride andglutamate ions on in vitro protein synthesis by the moderatehalophile Vibrio costicola. J. Bacteriol. 160:385-390.

10. Kucan, Z., and F. Lipmann. 1964. Differences in chloramphen-icol sensitivity of cell-free amino acid polymerization systems.J. Biol. Chem. 239:516-520.

11. Kushner, D. J. 1988. What is the "true" internal environment ofhalophilic and other bacteria? Can. J. Microbiol. 34:482-486.

12. Laemmli, U. K. 1970. Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature (London)227:680-685.

13. Lodish, H. F. 1970. Secondary structure of bacteriophage f2

VOL. 172, 1990

on May 26, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

3468 CHOQUET AND KUSHNER

ribonucleic acid and the initiation of in vitro protein biosynthe-sis. J. Mol. Biol. 50:689-702.

14. Lubin, M. 1963. A priming reaction in protein synthesis. Bio-chim. Biophys. Acta 72:345-348.

15. Lucas-Lenard, J., and F. Lipmann. 1967. Initiation of polyphe-nylalanine synthesis by N-acetylphenylalanyl-sRNA. Proc.Natl. Acad. Sci. USA 57:1050-1057.

16. Min Jou, W., G. Haegeman, M. Ysebaert, and W. Fiers. 1972.Nucleotide sequence of the gene coding for the bacteriophageMS2 coat protein. Nature (London) 237:82-88.

17. Modelell, J., and B. D. Davis. 1968. Rapid inhibition of polypep-tide chain extension by streptomycin. Proc. Natl. Acad. Sci.USA 61:1279-1286.

18. Okuyama, A., N. Machiyama, T. Kinoshita, and N. Tanaka.1971. Inhibition by kasugamycin of initiation complex formationon 30S ribosomes. Biochem. Biophys. Res. Commun. 43:196-199.

19. Osborn, M., A. M. Weiner, and K. Weber. 1970. Large scalepurification of A-protein from bacteriophage R17. Eur. J. Bio-chem. 17:63-67.

20. Pestka, S. 1972. Studies on transfer ribonucleic acid-ribosomecomplexes. J. Biol. Chem. 247:4669-4678.

21. Peumans, W. J., A. R. Carlier, and B. Delaey. 1982. Inhibition ofprotein synthesis by Cl-: different sensitivity of pea and cerealinitiation factors. Biochem. Physiol. Pflanz. 177:1-8.

22. Sabol, S., M. A. G. Sfllero, K. Iwasaki, and S. Ochoa. 1970.Purification and properties of initiation factor F3. Nature (Lon-don) 26:1269-1273.

23. Sanz, J. L., I. Marin, M. A. Balboa, D. Urena, and R. Amils.1988. An NH4-dependent protein synthesis cell-free system forhalobacteria. Biochemistry 27:8194-8199.

24. Saruyama, H., and K. H. Nierhaus. 1985. A highly efficientpoly(U)-dependent poly(phe) synthesis system for the extremehalophile archaebacterium Halobacterium halobium. FEBSLett. 183:390-394.

25. Shindler, D. B., R. M. Wydro, and D. J. Kushner. 1977.Cell-bound cations of the moderately halophilic bacteriumVibrio costicola. J. Bacteriol. 130:698-703.

26. Spyrides, G. J. 1964. The effect of univalent cations on thebinding ofsRNA to the template-ribosome complex. Proc. Natl.Acad. Sci. USA 51:1220-1226.

27. Steitz, J. A. 1979. Prokaryotic ribosome binding sites. Methods

Enzymol. 60:311-321.28. Tai, P., J. W. Wallace, E. L. Herzog, and B. D. Davis. 1973.

Properties of initiation-free polysomes of Escherichia coli. Bio-chemistry 12:609-615.

29. Van der Hofsdtad, G. A. J. M., H. 0. Voorma, and L. Bosch.1979. Formation of a preribosomal binary complex consisting offmet-tRNA and IF-2 and its interaction with the ribosomes.Methods Enzymol. 60:215-224.

30. Van der Laken, K., H. Bakker-Steeneveld, B. Berkhout, andP. H. van Knippenberg. 1980. The role of the codon and theinitiation factor IF-2 in the selection of N-blocked aminoacyl-tRNA for initiation. Eur. J. Biochem. 104:19-23.

31. Vermeer, C., J. Boon, A. Talens, and L. Bosch. 1973. Binding ofthe initiation factor IF-3 to Escherichia coli ribosomes andMS2-RNA. Eur. J. Biochem. 40:283-293.

32. Voorma, H. O., R. Benne, and T. J. A. den Hertog. 1971.Binding of aminoacyl-tRNA to ribosomes programmed withbacteriophage MS2-RNA. Eur. J. Biochem. 18:451-462.

33. Wabba, A. J., K. Iwasaki, M. J. Miller, S. Sabol, M. A. G.Sillero, and C. Vasquez. 1969. Initiation of protein synthesis inEscherichia coli. II. Role of the initiation factors in polypeptidesynthesis. Cold Spring Harbor Symp. Quant. Biol. 34:291-299.

34. Weber, L. A., E. D. Hickey, P. A. Moroney, and C. Baglioni.1977. Inhibition of protein synthesis by C1-. J. Biol. Chem.252:4007-4010.

35. Webster, R. E., D. L. Engelhardt, and N. D. Zinder. 1967.Amber mutants and chain termination in vitro. J. Mol. Biol.29:27-43.

36. Weiel, J., and W. B. Hershey. 1982. The binding of fluorescein-labeled protein synthesis initiation factor 2 to Escherichia coli30 S ribosomal subunits determined by fluorescence polariza-tion. J. Biol. Chem. 257:1215-1220.

37. Wydro, R. M., M. Kogut, and D. J. Kushner. 1975. Saltresponse of ribosomes of a moderately halophilic bacterium.FEBS Lett. 60:210-215.

38. Wydro, R. M., W. Madira, T. Hiramatsu, M. Kogut, and D. J.Kushner. 1977. Salt-sensitive in vitro protein synthesis by amoderately halophilic bacterium. Nature (London) 269:824-825.

39. Yamamoto, K. R., B. M. Alberts, R. Benzinger, L. Lawhorne,and G. Treiber. 1970. Rapid bacteriophage sedimentation in thepresence of polyethylene glycol and its application to large-scalevirus purification. Virology 40:734-744.

J. BACTERIOL.

on May 26, 2021 by guest

http://jb.asm.org/

Dow

nloaded from