JOURNAL OF BACTERIOLOGY, Feb. 1989, p. 880-8860021-9193/89/020880-07$02.00/0Copyright © 1989, American Society for Microbiology

Vol. 171, No. 2

In Vitro Protein Synthesis by the Moderate Halophile Vibriocosticola: Site of Action of CF- IonsC. G. CHOQUET,1 M. KAMEKURA,2 AND D. J. KUSHNER'*

Department of Biology, University of Ottawa, Ottawa, Ontario, Canada KIN 6N5,1and Noda Institute for Scientific Research, Noda, Chiba 278, Japan2

Received 10 August 1988/Accepted 26 October 1988

In vitro protein synthesis in Vibrio costicola [poly(U)-directed incorporation of phenylalanine] was studied.The extent of protein synthesis was limited by the number of ribosomes present. Density gradient centrifugationexperiments suggested that, after runoff of ribosomes from the artificial messenger, the 50S subunit was unableto attach to the 30S-messenger complex. As shown previously (M. Kamekura and D. J. Kushner, J. Bacteriol.160:385-390, 1984), Cl- ions inhibited protein synthesis; indeed, the highest rate of synthesis took place in thelowest attainable Cl- concentration (37 mM). The inhibitory effects were partly reversed by glutamate andbetaine, both of which are concentrated within cells of V. costicola. The strongest reversal was seen when bothglutamate and betaine were present. Cl- ions car! prevent binding of ribosomes to poly(U) and displaceribosomes already bound to this artificial messenger. The effects of Cl- ions on binding were also reversed byglutamate and betaine. Cl- ions did not affect accuracy of translation; they were shown previously (Kamekuraand Kushner, J. Bacteriol. 160:385-390, 1984) not to affect phenylalanyl-tRNA synthetase. It was also foundthat washing ribosomes with inhibitory NaCl concentrations did not interfere with their ability to carry outprotein synthesis later in optimal (low) salt concentrations. On the contrary, these ribosomes were more activethan before they were washed. We conclude that the main site of action of Cl- in the system studied is on thebinding of ribosomes to the mRNA.

Vibrio costicola is a moderate halophile which grows inmedia containing a wide range (0.4 to 3.5 M) of NaClconcentrations, with its optimum growth occurring at aconcentration of around 1 M NaCl (7). Although its cell-associated concentration of Na+ and K+ ions is generallyequal to the external Na+ concentration (15), in vitro proteinsynthesis [poly(U)-dependent phenylalanine incorporation]by cellular extracts of this bacterium, after growth in thepresence of 1 M NaCl, was found to have maximal activity in0.1 to 0.2 M concentrations of NaCl, KCl, or NH4C1 and no,or very little, activity in 0.6 M concentrations of these salts.Amino acid incorporation dependent on "endogenous"mRNA had similar responses to these salts (19).The inhibitory effects were shown to be due mainly to the

C1- ions, which are normally excluded from the cells: V.costicola grown in the presence of 1.0 M NaCl contains only0.2 M C1- (6). The effects of higher concentrations of theseions were partly reversed by a number of organic sub-stances, including glycine betaine (betaine) and glutamate,which are accumulated within the cells.The site of action of C1- has not yet been identified. It is

not phenylalanyl-tRNA synthetase, which is barely inhibitedby as much as 0.9 M NaCl (6). In this paper, we reportfurther studies of the in vitro protein-synthesizing system ofV. costicola, with special reference to the site of action ofCl- ions.

MATERIALS AND METHODSBacterial cultures. V. costicola NRC 37001 was grown and

harvested as described previously (6).Preparation of cellular extracts and ribosomes. S-30 ex-

tract, that is, the cellular material remaining in the superna-tant after centrifugation for 30 min at 30,000 x g (Beckmantype 42.1 rotor), was prepared as described previously (6).

* Corresponding author.

The S-150 extract and the ribosomes were obtained bycentrifuging the S-30 extract at 150,000 x g for 3 h at 4°C.The soluble S-150 fraction and the ribosomal pellet werethen dialyzed or washed and stored following the procedureof Kamekura and Kushner (6).The extraction buffer used in preparation of the cellular

extracts and the ribosomes contained the following: 124 mMNH4C1, 20 mM MgCl2, 10 mM Tris hydrochloride (pH 7.6),3 mM spermidine-trihydrochloride, and 6 mM P-mercapto-ethanol.

In vitro protein synthesis. The poly(U)-directed incorpora-tion of [14C]phenylalanine into hot trichloroacetic acid-insoluble material was measured as described earlier (6)except for the following changes to the reaction mixture.Because of the inhibitory action of Cl-, we substitutedammonium glutamate, MgSO4, and Tris sulfate for NH4Cl,MgCl2, and Tris hydrochloride, previously used in thepreparation of the reaction mixture; 0.6 M glutamate andsulfate anions have been shown to be less inhibitory toprotein synthesis than Cl- ions (6). Therefore, the reactionmixture (0.4 ml) contained the following: 15 mM phosphoe-nolpyruvate, 2 mM ATP, 1 mM GTP, 124 mM ammoniumglutamate, 18 mM Mg2+ (MgSO4), 7.5 mM reduced glutathi-one, 82 mM Tris sulfate (ph 7.6), 1.2 mg of poly(U) per ml,1.96 ,M [14C]phenylalanine (512 mCi/mmol; Dupont, NENResearch Products), various concentrations of solutes, and0.2 volume (80 pu) of S-30 or 70 p.l of S-150 and 10 pI (4 A260units) of ribosomes. These changes resulted in a reactionmixture containing only 37 mM Cl- ions after adding 0.2volume of extract.

Fidelity of translation. Fidelity of translation of the in vitroprotein synthesis system at different NaCl concentrationswas studied by measuring the incorporation of phenylalanine(256 mCi/mmol; NEN Research Products), leucine (276mCi/mmol; NEN Research Products), or valine (225 mCi/

880

IN VITRO PROTEIN SYNTHESIS BY V. COSTICOLA 881

mmol; ICN Biochemicals Canada Ltd.) in the presence andabsence of poly(U).

Binding of poly(U) to ribosomes. A reaction mixture (0.4ml) containing 15 mM phosphoenolpyruvate, 2 mM ATP, 1mM GTP, 124 mM ammonium glutamate, 18 mM Mg2+(MgSO4), 7.5 mM reduced glutathione, 82 mM Tris sulfate(pH 7.6), 1.96 ,uM [12C]phenylalanine, ribosomes (20 A260units), and various concentrations of solutes (as indicated inthe legend to each figure) was incubated for 2 min at 30'C.Poly[5_-3H]uridylic acid (2 to 4 Ci/mmol of UMP; NENResearch Products) was added to a final concentration of 4iiCi/ml, and the reaction mixture was incubated for another2 min. Then 300 ,ul of the reaction mixture was loaded on topof a linear sucrose gradient containing the same solute(s) asthe reaction mixture: 124 mM ammonium glutamate, 18 mMMg2+, 82 mM Tris sulfate (pH 7.6), and various concentra-tions of solutes.The presence and nature of a solute in a sucrose gradient

changes the density and viscosity of the gradient. Therefore,the conditions of centrifugation were different for eachgradient. In the binding experiments described in this paper,when different concentrations and combinations of NaCl,sodium glutamate, and betaine were used, the range of thesucrose gradient and the temperature of centrifugation wereadjusted to obtain similar ribosomal profiles under eachcondition. This information is given in the legend to eachfigure. The centrifugation itself was always done at 25,000rpm for 16 h in an SW-41 Beckman rotor.

After centrifugation, the gradients were fractionated (0.13ml per fraction), and each fraction was diluted with 2 ml ofH20; 1 ml was added to 4 ml of Universal Cocktail (ICNBiochemicals Canada Ltd.) and the radioactivity wascounted (2000CA LSC, Packard Instruments Canada Ltd.),while the remainder was used to read the A260 (Spectronic1001; Milton Roy Co.).For locating the 70S, 50S, and 30S ribosomal particles,

parallel experiments were carried out in which purifiedribosomes were centrifuged in appropriate gradients in thepresence of 18 or 1 mM Mg2+; with 1 mM Mg2+, the 70Sribosomes dissociated into 50S and 30S subunits (18). (Notethat we are using 70S, 50S, and 30S to indicate the usualforms of ribosomes, though the actual "S" values deter-mined for V. costicola were slightly different: 64S, 48S, and28S [18].)

RESULTS

Importance of different parts of the protein-synthesizingsystem. Time course experiments showed that, even whenactive incorporation of ['4C]phenylalanine took place, therate of incorporation fell quickly (Fig. 1). Adding morepoly(U) or more ['4C]phenylalanine at 9 min did not restorethe rate of incorporation; adding more energy source (ATP,GTP, and phosphoenolpyruvate) caused some stimulation,but the greatest effect was obtained by adding fresh ribo-somes (Fig. 1). This suggests that ribosomes are not recycledin this system.As expected, no incorporation of phenylalanine occurred

if the S-150 fraction or the ribosomes were omitted (notshown).

Effects of Cl- ions on protein synthesis. In these experi-ments, a new incubation mixture was prepared that con-tained no Cl- ions (see Materials and Methods). Adding theextract itself brought the Cl- concentration to 37 mM. Theactivities of extracts measured in this system were substan-tially (about fivefold) higher than those measured in the

30

0

x

- 15

CL

10

5

00 10 15 20 25

TIME (min )FIG. 1. Importance of different parts of the protein-synthesizing

system. Symbols: Time course experiments showing the effects ofadding more poly(U), 0; more [14C]phenylalanine, 0; more energysource (ATP, GTP, and phosphoenolpyruvate), O; or more ribo-somes, *, at 9 min of incubation; control (no addition), A. Theamounts of all constituents added were the same as those originallypresent.

system used previously (6), which contained 279 mM Cl-.We found that all concentrations of NaCl added to the newincubation system inhibited protein synthesis (see legend toFig. 2). The inhibition was due to the Cl- rather than to the

100 200

150-80

;100

~60 4

01o 40 0 0.2 0.4 0.6 0.8 1.0

Na-glutamate (M)

20

00 0.2 0.4 0.6 0.8 1.0

NaCI (M)FIG. 2. Effects of sodium glutamate concentrations on in vitro

protein synthesis by S-30 of V. costicola at increasing NaCl concen-trations. The concentrations of added sodium glutamate were thefollowing: A, 0; O, 0.2 M; 0, 0.4 M; 0, 0.6 M; *, 0.9 M. Results areexpressed as percentage of controls (without NaCl). (Inset) Effect ofsodium glutamate concentrations on in vitro protein synthesis: 100%activity represents incorporation of 20 pmol of [14C]phenylalanine(22,000 dpm) into hot trichloroacetic acid-insoluble material per 10,ul of S-30.

VOL. 171, 1989

?n

882 CHOQUET ET AL.

TABLE 1. Poly(U)-directed incorporation of ["4C]phenylalanine,["4C]leucine, and [14C]valine by 10 ,ul of S-30 extracts of V.

costicola at increasing C1- concentrations

Amt of amino acid incorporatedConcn of Poly(U)C1- (MM) Pl()(ml

Phe Leu Val

0 + 47.3 0.5 0.0- 0.6 0.2 0.2

100 + 26.3 0.2 0.0- 0.4 0.0 0.2

300 + 10.8 0.0 0.0- 0.2 0.1 0.0

500 + 1.5 0.0 0.0- 0.3 0.0 0.0

1,000 + 0.0 0.0 0.0- 0.0 0.0 0.2

Na+, since adding different concentrations of sodium gluta-mate usually stimulated protein synthesis (Fig. 2, inset).

Fidelity of translation. Table 1 shows the effects of dif-ferent NaCl concentrations on the fidelity of translation ofthe protein-synthesizing system used here. In addition tothat of phenylalanine, we studied incorporation of valine andleucine whose codons (GUU and UUG, respectively) differin only one nucleotide from that of phenylalanine (UUU). Inthe absence of poly(U), there was virtually no endogenousincorporation of any of the amino acids. In the presence ofpoly(U), only phenylalanine was significantly incorporated.Increasing concentrations of NaCl reduced the extent ofphenylalanine incorporation. Since the presence of poly(U)did not stimulate the incorporation of leucine or valine withincreasing NaCl concentrations [ratio of amino acid incor-porated in the presence and absence of poly(U) did notincrease], the fidelity of this translation system is not af-fected by increasing Cl- ions.

Effects of glutamate and betaine on inhibition by Cl ions.Since cells grown in 1 M NaCl contain about 100 mMglutamate (6), the concentration of glutamate (124 mM) usedhere may be considered an endogenous value. All soluteconcentrations shown in the experiments described belowwere added to solutions containing this level of glutamate.We observed that all added concentrations of glutamate

stimulated in vitro protein synthesis and counteracted theinhibitory effects of NaCl. The effects were generally greaterin higher glutamate concentrations (Fig. 2).

It was shown earlier that betaine can partly overcome theinhibitory action of Cl- ions, even in the absence of gluta-mate (6). In the presence of 124 mM glutamate, the effectincreased with increasing betaine concentration (Fig. 3) upto the highest betaine concentration (1.2 M) studied. Thephysiological concentration of betaine is about 0.5 M inthese cells (M. Klein and D. J. Kushner, unpublished data),a value similar to that reported by Imhoff and Rodriguez-Valera (5). This concentration causes a near-maximal stim-ulation of activity and has a definite sparing action on Cl-inhibition. Adding more glutamate, in the presence of 0.5 Mbetaine, further decreased the inhibition by Cl- (Fig. 4).

Binding of poly(U) to ribosomes and ribosomal subunits. Inan attempt to locate the site of action of Cl- ions, the effectsof NaCl and other substances on the binding of radiolabeled

100O

60 ~50

o 40 040 ~~~~~~~~~~00.3 0.6 0.9 1.2betaine (M)

00

0 0.2 04 M 0.8 1.0

NaCI (M)FIG. 3. Effects of betaine concentrations on in vitro protein

synthesis by S-30 of V. costicola at increasing NaCI concentrations.The concentrations of betaine were as follows: A, 0; O, 0.3 M; *,0.6 M; *, 0.9 M; 0, 1.2 M. Results are expressed as percentagecontrols (without NaCl). (Inset) Effect of betaine concentrations onin vitro protein synthesis: 100% activity represents incorporation of20 pmol of ["4C]phenylalanine (22,000 dpm) into hot trichloroaceticacid-insoluble material per 10 p.l of S-30.

poly(U), an artificial mRNA, to the ribosomes of V. costi-cola was studied.

Density gradient centrifugation showed that [3H]poly(U)cosedimented with ribosomes. When the S-150 fraction wasomitted from the reaction mixture, poly(U) cosedimented

100

80

>. 60

V

0

40

20

20

0 0.2 0.4 0.6 0.8 1.0

NaCI (M)FIG. 4. Effects of betaine or sodium glutamate concentration or

both on in vitro protein synthesis by S-30 of V. costicola atincreasing NaCl concentrations. Symbols: A, no added salts; O, 0.2M sodium glutamate; 0, 0.5 M betaine; *, 0.2 M sodium glutamateand 0.5 M betaine. Results are expressed as percentage of controls(without NaCl). Activity at 100% (without NaCl) is the incorpora-tion of [14C]phenylalanine by 10 ,ul of S-30 extract: 20 pmol in theabsence of added solute; 25 pmol in the presence of 0.2 M sodiumglutamate; 24 pmol in the presence of 0.5 M betaine; and 29 pmol inthe presence of 0.2 M glutamate and 0.5 M betaine.

J. BACTERIOL.

IN VITRO PROTEIN SYNTHESIS BY V. COSTICOLA 883

I

E0

C~4

6i0

25

20

15

10

0

0~0CL

x

0 20 40 60 80 0 20 40 60 80

FRACTION #FIG. 5. Sucrose density gradients (15 to 30%; at 40C) showing binding of [3H]poly(U) to V. costicola ribosomes (A) with S-150 present in

the reaction mixture and (B) without S-150. O.D., Optical density.

with, and was presumably bound to, the 70S ribosomes andthe 30S subunits (Fig. 5B). However, with the additionalpresence of the S-150 fraction, poly(U) cosedimented onlywith the 30S subunits (Fig. 5A). This suggests that factors inthe S-150 fraction, probably elongation factors, caused theribosomes to run off the artificial messenger. It also suggeststhat, after runoff, although the 30S subunit was still able toattach to the messenger, the 50S subunit could not attach tothe 30S-messenger complex. In further studies of poly(U)binding to ribosomes, the S-150 fraction was omitted.We found that 0.6 M NaCl, which completely inhibits in

vitro protein synthesis, also completely inhibits binding ofribosomes and 30S subunits to poly(U) (Fig. 6A and B).

In the presence of 0.6 M sodium glutamate, a concentra-tion which best protects in vitro protein synthesis, significantbinding was still detected in the presence of 0.6 M NaCl (Fig.6D), although not to the same extent as in its absence (Fig.6C). The amount of binding detected in the presence of 0.6 Msodium glutamate alone confirmed that it is Cl-, and notNa', that is inhibitory to the binding between ribosomes andpoly(U).

Similar results were obtained for betaine. In the presenceof 1.2 M betaine, a concentration which affords maximalprotection, poly(U) cosedimented with both 70S and 30Sribosomal particles (Fig. 6E). With the addition of 0.6 MNaCl, a smaller but significant amount of poly(U) was foundassociated with each ribosomal particle (Fig. 6F).We next studied the ability of Cl- to cause the dissociation

of already bound poly(U) from ribosomes. A reaction mix-ture containing 0.6 M sodium glutamate was first incubatedat 30'C for 2 min. Following this treatment, the bindingpattern was the same as shown previously (Fig. 6), but aftermaking the reaction mixture 0.6 M in NaCl and incubating itfor an additional 2 min, the amount of poly(U) bound to theribosomal particles decreased to what had been observed

previously when both salts were present at the beginning ofthe reaction (Fig. 6). Therefore, not only does Cl- inhibit thebinding of poly(U) to the ribosomes, but also it causes therelease of already bound poly(U).Although the sedimentation behavior of the ribosomes

does not change at high Cl- ion concentrations (18), someribosomal proteins may still be lost (R. M. Wydro, Ph.D.thesis, University of Ottawa, Ottawa, Ontario, Canada,1977). However, this loss did not damage the protein-synthesizing ability of the ribosomes. Ribosomes washed in0.6 M NaCl could carry out more active protein synthesis inthe standard reaction mixture. The increased activity variedfrom 60 to 100% in different experiments. The mechanism ofthis increase has not yet been explored further, but theresults do show that Cl- ions cause no permanent damage tothe ribosomes.

DISCUSSIONThe poly(U)-directed incorporation of phenylalanine by

extracts of V. costicola is rapid at first but slows and stopswithin 15 min. The number of ribosomes present seems to bethe limiting factor. Some of our experiments (Fig. SA) havesuggested that, after the runoff of the 70S ribosome from theartificial messenger, the 50S subunit resulting from thedissociation of the ribosome cannot reattach to the 30S-mRNA complex. Whether it is due to the 50S- or the30S-mRNA complex is now known. However, major pro-portions of inactive ribosomes have been reported for othercell-free systems carrying out poly(U)-directed protein syn-thesis, including those of Eschericia coli (2) and Halobacte-rium halobium (14). It is also known that "ribosomes (of E.coli) that reach the end of poly(U) cease synthesizing poly-phenylalanine" (2).Our results confirm our earlier findings that Cl- is inhibi-

tory to poly(U)-directed in vitro protein synthesis by V.

VOL. 171, 1989

884 CHOQUET ET AL.

Ec0

0

B

70S/IIlIs1 I

IiI II

I II \I \I I

I \501 \50 s

FRACTION n

EC0

N

d0

0Y

0~

0

0 20 40 60 80 0 20 40 60 80

FRACTION **FIG. 6. Sucrose density gradients showing binding of [3H]poly(U) to V. costicola ribosomes. The sucrose gradient range and the

temperature of centrifugation (in parentheses) are given after the identity of the solute present in the reaction mixture and the sucrosegradient. (A) No added solute: 15 to 30% sucrose gradient (40C); (B) 0.6 M NaCi: 5 to 20% (40C); (C) 0.6 M sodium glutamate: 5 to 20%(12.50C); (D) 0.6 M sodium glutamate and 0.6 M NaCl: 5 to 20% -(12.50C); (E) 1.2 M betaine: 5 to 20% (7.50C); (F) 1.2 M betaine and 0.6 MNaCl: 5 to 20% (7.50C).

costicola (6). However, it has no effect on the fidelity oftranslation of this particular system. This may be a reflectionof the ability of V. costicola to grow in a wide range of NaClconcentrations; although the concentration of Cl- in cells

grown in the presence of 1.0 M NaCl is only 0.2 M, cellsgrown in the presence of 3.0 M NaCl contain 1.5 M Cl- (9).In contrast, fidelity of translation in cell-free protein-synthe-sizing system from the extremely halophilic t4rchaebacterium

J. BACTERIOL.

IN VITRO PROTEIN SYNTHESIS BY V. COSTICOLA 885

0.5

0.4

0.3

0.2

0.1

0.00 20 40 60 80 0 20 40 60 80

FRACTION U

H. halobium, which requires almost saturated solutions forgrowth, is increased by the presence of NaCl (1).

It was shown previously that the phenylalanyl-tRNAsynthetase was not responsible for Cl- inhibition of poly(U)-directed in vitro protein synthesis by V. costicola (6). Wehave now shown that Cl- can prevent attachment of an

artificial mRNA [poly(U)] to both the complete 70S ribo-somes and the 30S ribosomal subunits. A similar inhibitionby Cl- has been demonstrated for the in vitro proteinsynthesis system of reticulocyte lysates (17). In this eucary-otic cell-free system, 200 mM Cl- inhibited the binding ofmRNA to reticulocyte ribosomal particles (80S and 40S).Peumans et al. (12) have also shown that, in a pea cell-freesystem, 160 mM Cl- inhibited the initiation of proteinsynthesis by acting on one or more soluble initiation factors,some of which are involved in the binding of the mRNA.

Cl-, being a strong anion, might interfere with protein-protein and protein-nucleic acid associations which involveionic interactions (17), as well as induce intramolecularchanges of proteins by interfering with the internal ionicinteractions (12). It is also possible that Cl- interfered withthe protein-nucleic acid association involving the ribosomesand the mRNA, especially since it was shown that Cl- couldcause the release of already bound poly(U). We cannotknow, however, whether Cl- also has an inhibitory effect onthe initiation factors themselves or even on their associationwith the ribosomes. In presence of the high Mg2' concen-tration found in this system, the initiation factors are notneeded for poly(U)-directed in vitro protein synthesis (10),and the association between the ribosomes and poly(U) cantake place without them.The proteins and the RNA molecules that constitute the

ribosomes are held together by protein-protein and protein-nucleic acid associations. We considered the possibility thatCl-, by removing one or more ribosomal protein, couldprevent the binding of poly(U) to the ribosomes. But thisdoes not occur. For reasons not yet clear, ribosomes washed

in 0.6 M NaCl are more active than those not washed at highsalt concentration.

Is the inhibition of the binding of the mRNA to theribosomes the only site of action of Cl- on protein synthe-sis? There may well be others: since these and other cells(see reference 6) are so concerned with keeping Cl- ions out,presumably at some cost of energy, it must be suspected thatthese can have many toxic sites of action. Furthermore, Cl-sensitivity in halophilic eubacteria, and probably marinebacteria, is not restricted to protein synthesis. Althoughinformation of this subject is scarce, preliminary studieshave shown that Cl- can have an inhibitory effect on somecellular enzymes (reviewed in reference 9). Whether' it canalso affect other vital processes such as DNA replication ortranscription is still unknown.Our work has emphasized the importance of betaine in the

physiology of halophilic eubacteria. This substance accumu-lates in every moderate halophile so far examined, as well asin some nonhalophiles under osmotic stress (3, 5), suggestingthat it can act as a compatible solute in these bacteria. It hasalso been shown to relieve the effects of high NaCl concen-trations on active transport in V. costicola (8), to relieve theinhibition of respiration caused by high NaCl concentrationsin the moderate halophile Bal (13), and to protect theenzyme glutamine synthetase from inhibition by high saltconcentrations in a halotolerant and a marine strain ofcyanobacteria (16).Glutamate may also play an important role in V. costicola.

Synthesis of glutamate can be osmotically induced in anumber of nonhalophilic and marine bacteria (4, 11). In V.costicola itself, glutamate is the major known organic anionand, especially important to this discussion, stimulates invitro protein synthesis.Our results suggest that both of these substances could

play an important role in protecting the cellular proteinmachinery against the toxic action of Cl- in V. costicola.Glutamate might counteract the action of Cl- by occupying

EC0

C.,

0.0

25

20

15

0I0>1o x

0Z5

0

VOL. 171, 1989

886 CHOQUET ET AL.

some of the Cl- binding sites. Betaine, however, is a dipolarion at physiological pH and therefore could not act as acompetitive anion to Cl-. Therefore, it probably acts in adifferent manner than glutamate, perhaps by stabilizing thepoly(U)-ribosome complex at high CF- ion concentrations.

Regardless of the manner in which betaine and glutamateprotect the in vitro protein-synthesizing system, their com-bined presence is much more beneficial than either alone(Fig. 4). Since both would be present in living cells, thisprotection, whatever its mechanism, could be very impor-tant for these bacteria.

ACKNOWLEDGMENT

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

LITERATURE CITED1. Bayley, S. T., and E. Griffiths. 1967. Codon assignments and

fidelity of translation in a cell-free protein-synthesizing systemfrom an extremely halophilic bacterium. Can. J. Biochem.46:937-944.

2. Bilgin, N., L. A. Kirsebom, M. Ehrenberg, and C. G. Kurland.1988. Mutations in ribosomal proteins L7/L12 perturb EF-G andEF-Tu functions. Biochimie. 70:611-618.

3. Galinski, E., and H. G. Truper. 1982. Betaine, a compatiblesolute in the extremely halophilic phototrophic bacterium Ec-tothiorhodospira halochloris. FEMS Microbiol. Lett. 13:357-360.

4. Imhoff, J. F. 1986. Osmoregulation and compatible solutes ineubacteria. FEMS Microbiol. Rev. 39:57-66.

5. Imhoff, J. F., and F. Rodriguez-Valera. 1984. Betaine is the maincompatible solute of halophilic eubacteria. J. Bacteriol. 160:478-479.

6. 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.

7. Kushner, D. J. 1978. Life in high salt and solute concentration:halophilic bacteria, p. 317-368. In D. J. Kushner (ed.), Micro-bial life in extreme environments. Academic Press, Inc.(London), Ltd., London.

8. Kushner, D. J., F. Hamaide, and R. A. MacLeod. 1983. Devel-opment of salt-resistant active transport in a moderately halo-philic bacterium. J. Bacteriol. 153:1163-1171.

9. Kushner, D. J., and M. Kamekura. 1988. Physiology of halo-philic eubacteria, p. 109-140. In F. Rodriguez-Valera (ed.),Halophilic bacteria. CRC Press, Inc., Boca Raton, Fl.

10. Lucas-Lenard, J., and F. Lipmann. 1967. Initiation of polyphe-nylalanine synthesis by N-acetylphenylalanine-sRNA. Bio-chemistry 57:1050-1057.

11. Makemson, J. C., and J. W. Hastings. 1979. Glutamate functionsin osmoregulation in a marine bacterium. Appl. Environ. Mi-crobiol. 38:178-180.

12. Peumans, W. J., A. R. Carlier, and B. M. Delaey. 1982.Inhibition of protein synthesis by Cl-: different sensitivity ofpea and cereal initiation factors. Biochem. Physiol. Pflanz.177:1-8.

13. Rafaeli-Eshkol, D., and Y. Avi-Dor. 1968. Studies on halotoler-ance in a moderately halophilic bacterium. Effect of betaine onsalt resistance of the respiratory system. Biochem. J. 109:687-691.

14. Saruyama, H., and K. H. Nierhaus. 1985. A highly poly(U)-dependent poly(Phe) synthesis system for the extreme halophilearchaebacterium Halobacterium halobium. FEBS Lett. 183:390-394.

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

16. Warr, S. R. C., R. H. Reed, and W. D. P. Stewart. 1984.Osmotic adjustment of cyanobacteria: the effects of NaCl, KCl,sucrose and glycine betaine on glutamine synthetase activity ina marine and a halotolerant strain. J. Gen. Bacteriol. 130:2169-2175.

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

18. Wydro, R., M. Kogut, and D. J. Kushner. 1975. Salt response ofribosomes of a moderately halophilic bacterium. FEBS Lett.60:210-215.

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

J. BACTERIOL.