Euglena gracilis Chloroplast Elongation Factor Tu - The Journal of ...

2 downloads 0 Views 764KB Size Report
Tu (EF-T") from Euglena gracilis with guanine nu- cleotides and ...... L. Eberly and Richard L. Puetz for helpful discussions and careful reading of the manuscript.
Vol . 260,No. 15, Issue of July 25, pp. 8771-8776,1985 Printed in U.S.A.

CHEMISTRY THEJOURNAL OF BIOLOGICAL

0 1985 by The American Society of Biologied Chemists, Inc.

Euglena gracilis Chloroplast Elongation FactorTu INTERACTION WITH GUANINE NUCLEOTIDES AND AMINOACYL-tRNA* (Received for publication, January 28, 1985)

Sunil P. Sreedharan and LindaL. SpremulliS From the Department of Chemistry, Universityof North Carolina, Chapel Hill, North Carolina 27514

The interaction of the chloroplast elongation factorThis factor interacts with a number of ligands and other Tu (EF-T") from Euglena gracilis with guanine nu- protein factors during the elongation cycle. The binding of cleotides and aminoacyl-tRNA has been investigated. guanine nucleotides to EF-Tu and the formation of the terThe apparent dissociation constant at 37 "C for theEF- nary complex between EF-Tu, guanine nucleotides, and amiTILU-GDP complex is about 3 x 10" M and for the EF- noacyl-tRNA has been of considerable interest since these TI~u-GTPcomplex, it is about 1 order of magnitude interactions have important implications for protein biosynhigher. The sulfhydryl modifying reagent N-ethylmal- thesis. Studies with sulfhydryl-modifying reagents, such as eimide severely inhibits the polymerizationactivity of N-ethylmaleimide, have indicated that two cysteine residues Euglena EF-TIL,. In the presence of N-ethylmaleim- occupy critical positions in E. coli EF-Tu (2). One of the EFide,thedissociationconstantforthemodified TIL~OGDP complex is increased by an order of mag- cysteine residues is present in the guanine nucleotide-binding pocket, and the other is located in the aminoacyl-tRNAnitude. Conversely, both GDPandGTPprotectEFTIL^ from the modification. The polymerization activ- binding domain of EF-Tu (3). Modification of either of these ity of EF-T&, is also sensitive to the antibiotic kirro- residues leads to a loss of EF-Tu activity. Further information on ligand binding to E. coli EF-Tu and mycin. In the presence of kirromycin, the apparent dissociation constantfor the EF-T%,,*GTP complex is on allosteric interactions between ligand-binding sites has lowered 10-fold. The interaction of aminoacyl-tRNA been obtained from studies on the effects of the antibiotic with EF-TIL, was investigated by examining the abil- kirromycin. This antibiotic inhibits protein biosynthesis by ity of EF-TIL, to prevent the spontaneous hydrolysis specifically binding to EF-Tuand preventing its release from of Phe-tRNA and by gel filtration chromatography. the ribosome (4). It has been shown that kirromycin increases Thebindingofaminoacyl-tRNA to EF-T" occurs the binding affinity of EF-Tu for GTP (5) and that itinduces only in the presence ofGTP indicating the formation a second tRNA-binding site on the protein (6, 7). of the ternary complex EF-TwM*GTP*Phe-tRNA. The The nucleotide sequence of the gene for EF-TUfhlfrom effect of kirromycin on the interactionwas also inves- Eugkna has recently been determined (8), and the probable tigated. In the presence of kirromycin,no interaction amino acid sequence deduced from the DNA sequence indibetween EF-T&, and Phe-tRNA is observed, even in cates that EF-T&hlshares about 70% homology with E. coli the presence of GTP. EF-TU. The chloroplast factor has a single cysteine residue located in theaminoacyl-tRNA-binding pocket. In thepresent paper we have examined the binding of guanine nucleotides and aminoacyl-tRNA to EF-TG~, and the effects of N-ethylElongation factor Tu plays a pivotal role in the elongation maleimide and kirromycin on these interactions. cycle of prokaryotic and organellar protein biosynthesis. We have recently reported the purification and some of the cataEXPERIMENTAL PROCEDURES lytic properties of chloroplast EF-Tu' from Euglena gracilis (1). Our results indicate that the chloroplast factor has a Materials-Phospho(enol)pyruvate, pyruvate kinase, DEAE-Sephmolecular weight of about 50,000, similar to that of the arose CL-GB, Sephadex G-50 and G-75, bovine serum albumin, yeast Escherichia coli factor. In addition, EF"I'Qhl is as active on ribonucleic acid, and N-ethylmaleimide were purchased from Sigma. E. coli ribosomes as it is on chloroplast ribosomes, and its Scintiverse I and polyethylene glycol go00 were from Fisher. GDP activity in polymerization is stimulated by E. coli EF-Ts. and GTPwere obtained from P-L Biochemicals, Poly(U) and E. coli These results indicate that the organellar factor has many tRNA were from Boehringer Mannheim. ['HIGDP (7.8 Ci/mmol), [3H]GTP(7.5 Ci/mmol), and [''C]phenylalanine were purchased from similarities to theprokaryotic EF-Tu. New England Nuclear. HEPES was from Research Organics, and Of all the bacterial and organellar sources of EF-Tu re- Tris was obtained from Mallinckrodt Chemical Works. Nitrocellulose ported to date, E. coli EF-Tu has been the best characterized. membrane filter paper type HA (0.45-flrn pore size) was purchased ~~

~

~~~

*This workwas supportedin part by funds provided by the National Institutes of Health (Grant GM 24963). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 4 Recipient of United States Public Health Service Research Career Development AwardAM0083401. To whom correspondence should be addressed. The abbreviations used are: EF, elongation factor; HEPES, 4-(2hydroxyethy1)-1-piperazineethanesulfonicacid NEM, N-ethylmaleimide.

from Millipore Corp., and Polygram Cel 300 polyethyleneimine thin layer chromatography plates were from BrinkmannInstruments. ["CIPhe-tRNA (300-500 cpm/pmol) was prepared as previously described (9). Kirromycin was kindly donated by Dr. L. I. Slobin of the University of Mississippi. Kirromycin was stored a t -20 'C as a 3.1 mM solution in ethanol. Dilutions of the stock solution were made as previously reported (10). The purity of ['HIGDP and ['HIGTP was determined by ascending thin layer chromatography on polyethyleneimine plates using 1.5 M LiCl as the solvent. In all cases, GTP was treated with P-enolpyruvate and pyruvate kinase priorto use. Buffer I contains 25 mM HEPES-KOH, pH 7.6, 50 mM NHCI, 10 mM MgClz, 0.1 mM EDTA, 6 mM 2-mercaptoethanol, and 25% glycerol. Preparation of Chloroplast EF-Tu-Euglena EF-Tbhlwas prepared

8771

8772

ChloroplastElongation

as described previously (1)or by a modification of this procedure in which DEAE-Sephadex A-50 was replaced by DEAE-Sepharose CL6B in the preparative steps, and concentration of the samples at each stage of the purification was carried out by chromatography on DEAE-Sepharose rather than by dialysis against solid polyethylene glycol. In low salt buffers, chloroplast EF-Tu is retained by DEAESepharose and can be eluted with high salt buffer resulting in a concentration comparable to thatpreviously obtained with polyethylene glycol. When thesamples prepared by these two methods were tested in polymerization, the specific activity of E F - W , obtained by concentration against polyethylene glycol was 106 units/pg and that obtained by concentration with DEAE-Sepharose was 76 unitslpg. The activity of EF-TGhIin poly(U)-directed polymerization of ["c] phenylalanine was determined in the presence of saturating amounts of E. coli ribosomes, EF-G and EF-Ts, as reported previously (11). Inactivation of EF-Tu by N-Ethylmaleimide-EF-TGhl (0.3 pg) was incubated with the indicated concentration of N-ethylmaleimide for 10 min at 0 "C in a reaction mixture (45 p l ) containing 20 mM TrisHCI, pH 7.6, 2 mM MgCI,, 40 mM NH4Cl, and 1 mM 2-mercaptoethanol (from the sample). The reaction was stopped by addition of 2mercaptoethanol to a final concentration of 50 mM bringing the reaction volume to 50 pl. Incubation was continued for an additional 5 min at 0 "C, and the sample was assayed for EF-Tu activity in polymerization. To test for protection from inactivation by guanine nucleotides, EF-T&hl was incubated as specified below with 18 pM GDP or with 9 mM P-enolpyruvate, 5.9 pg of pyruvate kinase, and 0.2 mM GTP prior to treatment with NEM. Binding of GDP and GTP to EF-Tu-Studies on the binding of GDP to EF-T%,,Iwere carried out by incubation of the indicated amounts of GDP andEF-TGhl in areaction mixture (100 pl) containing 50 mM Tris-HCI, pH 7.6, 10 mM MgCl,, 100 mM NHdC1, and 7.5 mM 2-mercaptoethanol for 10 min at 37 "C. The binding of GTP to EF-TGh, was performed in a similar fashion except that varying amounts of GTP, 9 mM P-enolpyruvate, and 5.9 pg of pyruvate kinase were used in place of GDP. Following incubation, the reaction mixture was applied to a 12-ml (0.75 X 27 cm) Sephadex G-50 column equilibrated in Buffer I. The column was developed with Buffer I at a flow rate of 35 ml/h. Fractions (340 pl) were collected and transferred with 3.5 ml of distilled water into glass scintillation vials containing 10 ml of Scintiverse I, and counted. Free unbound [3H] GDP or [3H]GTPappeared well separated from the EF-Th,. guanine nucleotide complex which eluted in the void volume. Assuming that the binding reaction is bimolecular and that there is one guanine nucleotide binding site on EF-TGhI, the apparent dissociation constant (&) can be determined from the following expression, Kd

=

[Tu], [GXPI, [Tu. GXP]

where [Tu], is the concentration of free EF-Tu,,, [GXP], is the concentration of unbound 13H]GDPor [3H]GTP, and ITu.GXP] is the concentration of the complex. [Tu. GXP]was obtained from the amount of radioactivity eluting with E F - T G ~in~ the void volume of the column. [Tu], was determined from the original amount of EFTu added minus the amount of complex formed. The amount of active EF-Tuehl added was determined by the maximum amount of GDP that could be bound (generally about 40 to 50% of the EF-Tu was active). [GXP], was determined from the amount of guanine nucleotide originally added minus the amountpresent in the Tu.GXF' complex. The effect of kirromycin on guanine nucleotide binding was tested by treatment of EF-TGhI with 10 pM kirromycin for 10 min at 0 "c prior to the addition of guanine nucleotides. The effect of NEM on the binding of guanine nucleotides was tested by treatment of EFTG~Iwith 6 mM NEM for 10 min at 0 "C followed by the addition of 2-mercaptoethanol to a final concentration of 50 pM prior to the addition of the labeled guanine nucleotide as described above. Protection of Aminoacyl-tRNA by EF-Tu-The rate of hydrolysis of the aminoacyl-tRNA bond was determined by the incubation of , cpm/pmol) at 37 "C with 3.8 pg of either ["CIPhe-tRNA (1.4 p ~390 EF-TGhl or bovine serum albumin in a reaction mixture (100 p1) containing 50 mM Tris-HC1, pH 7.6, 10 mM MgCl,, 100 mM NH&l, 7.5 mM 2-mercaptoethanol, and either 0.1 mM GDP or 1 mM penolpyruvate, 12 pg of pyruvate kinase, and 0.2 mM GTP. An aliquot of 5 pl was removed every 10 min and precipitated with cold 5% trichloroacetic acid in the presence of 135 pg of yeast ribonucleic acid

Factor Tu TABLEI Dissociation constant for the EF-TuAPGDP comDler Total EF-TU~U ItM

3.7 35 35 87.5

Total GDP

EF-Ts,,GDP

nM

nM

M

3.5 X 10-7

45.6

114

Dissociation constant

2.5

76

10.2 13.5

X

10-7

3.4 x 10-7

as a carrier. The precipitate was collected on a nitrocellulose membrane filter and counted. The effect of kirromycin on the protection by EF-T&l of aminoacyl-tRNA from hydrolysis was tested by treating thefactor with 0.1 mM kirromycin prior to the incubation with ["CIPhe-tRNA and either GDP or GTP asdescribed above. Detection of an EF-Tu.GTP.Phe-tRNA Complex by Gel Filtration Chromatography-Ternary complex formation was examined by incubating 0.2 p~ [l4C]Phe-tRNA (390 cpm/pmol) in the presence or absence of 3.8 fig of EF-TQhl in a reaction mixture (100 pl) containing 50 mM Tris-HCI, pH 7.6, 10 mMMgC12, 100 mM NH&I,0.2 mM GTP, 1mM P-enolpyruvate, 12 pg of pyruvate kinase, and 7.5 mM 2mercaptoethanol for 10 min at 37 "C. The reaction mixture was applied to a 25-ml(1.1 X 26 cm) Sephadex G-75 column equilibrated in Buffer I. The column was developedwith Buffer I at a flow rate of 35 ml/h, and 570-p1 fractions were collected, transferred with 3.5 ml of distilled water into glass scintillatfon vials containing 10 ml of Scintiverse I, and counted. RESULTS

Euglena EF-T&hl, unlike E. coli EF-Tu, is stable in the absence of guanine nucleotides; thus, it can be readily prepared free from them. This property of the chloroplast factor has facilitated the examination of its interaction with both GDP and GTP. The binding of guanine nucleotides to EF-Tu from various sources has traditionally been assayed by the retention of the complex containing labeled nucleotides on nitrocellulose membrane filters (12). Recently it has been observed' that E. coli EF-Tu concentrated during preparation by dialysis against solid polyethylene glycol fails to bind quantitatively to nitrocellulose membranes. Our routine purification scheme for Euglena EF-T&hl (1) employs concentration of samples against solid polyethylene glycol. Under these conditions EF-TGhl is not retained by nitrocellulose filters (data not shown). If DEAE-Sepharose is used for the concentration of EF-Thhl (see "Experimental Procedures"), the factor is now retained by nitrocellulose membranes; however, EF-TGhlfails to give reproducible values for GDP binding by the membrane filter method. The results obtained by gel filtration are similar to those obtained by the membrane filter method but are much more reproducible. Thus, guanine nucleotide binding constants were obtained by incubating varying amounts of the factor with different concentrations of guanine nucleotides followed by chromatography on Sephadex G-50. The EF-T&hl.guanine nucleotide complex elutes as a sharp symmetrical peak well separated from unbound nucleotides during chromatography on Sephadex G-50 permitting the quantitative determinationof complex formation. The apparent dissociation constant for EF"I'&hl. GDP (Table I) is about 3 X lo-' M, and theapparent dissociation constant for the EF-Tuchl.GTP complex (Table II), at 2 X M, is about an order of magnitude higher than that observed for the GDP complex. E. coli EF-Tu forms a relatively stable EF-Tu.GTP.aminoacyl-tRNA complex which functions as an intermediate in the elongation cycle (13, 14). We have investigated the ability of EF-TG~] to form a ternary complex with GTP andamino-

' R. Leberman, manuscript in preparation.

Chloroplast Elongation Factor Tu

8773

acyl-tRNA by several methods. As indicated in Fig. 1, the and EF-Tuehl with GDP instead of GTP did not result in rate of hydrolysis of Phe-tRNA is reduced in the presence of complex formation (data not shown) indicating that the inEF-T&hl and GTP suggesting that the aminoacyl-tRNA is teraction of aminoacyl-tRNA with the chloroplast factor reinteracting with the factor under theseconditions. No protec- quired GTP. The effect of M F on the interaction of EF-TQhl, GTP, tion is observed when GTP is replaced by GDP suggesting that aminoacyl-tRNA interacts with the chloroplast elonga- and aminoacyl-tRNA was investigated by carrying out the tion factor only in the presence of GTP and that this inter- incubation and subsequent gel filtration chromatography at action involves the formation of a ternary complex. In order different Mg“ concentrations. In thepresence of 10 mM Mg“, to obtain physical evidence for the formation of such a com- the ternary complex elutes at the void volume. However, in plex, EF-T&hi was incubated with GTP and [14C]Phe-tRNA the presence of 1mM M$+ the ternarycomplex elutes slightly and themixture subsequently analyzed by gel filtration chro- behind the void volumeand itsprofile is broadened suggesting matography. As indicated in Fig. M , the elution position of that the ternary complex is less stable at lower concentrations the aminoacyl-tRNA is shifted to higher molecular weight of Mg2+ (data not shown). Investigations with Arterniu EF-1 values in the presence of EF-TQMand GTPindicating ternary (15) indicate that at low concentrations of M e , the cytocomplex formation. Subtraction of the values obtained when plasmic factor can form a binary complex with aminoacylPhe-tRNA is subjected to chromatography in the absence of tRNA. We have tested the chloroplast factor for such an E F - T G ~from the values obtained for the ternary complex interaction by looking for the formation of an EF-TQh1.Pheyields asymmetrical peak corresponding to an EF-TQhl. tRNA complex at 1 mM M$+ and at 10 mM M$+ in the GTP. Phe-tRNAcomplex (Fig. 2B). Incubation of Phe-tRNA absence of guanine nucleotides. No complex formation was observed at either Mg“ concentration (data not shown),again TABLE I1 indicating that astableinteraction between EF-TQ, and aminoacyl-tRNA occurs only in the presence of GTP. Dissociation constant for the EF-TucM.GTP complex E. coli EF-Tu contains two reactive cysteine residues (Cys Dissociation Total Total EF-fi,, GTP constant GTP EF-TUM 81 in the aminoacyl-tRNA binding pocket and Cys 137in the guanine nucleotide-binding site). An examination of the probRM M nM nM able amino acid sequence of Euglena EF-TGhI (8) indicates 2.2 10.2 x lo6 87.5 306 10.8 2.1 x 87.5 310 that the chloroplast factor has a single sulfhydryl in position 15.9 2.0 x 10-6 87.5 465 82 corresponding to residue 81 in E. coli EF-Tu.It was, therefore, of interest to examine the sensitivity of EF-T&hl to the sulfhydryl-modifying reagent NEM. As indicated in Fig. 3, the chloroplast factor is sensitive to inhibition by NEM suggesting that a nucleophilic group (probably cysteine 82) is susceptible to modification in EF-Tuchl.Although the residue we believe to be modified by NEM does not lie in the proposed guanine nucleotide-binding pocket ofEF”I’uehl, both GDP and GTP protect EF-Tuchlfrom NEM modification (Fig. 3). GDP appears to be slightly more effective in this protection than GTP, reflecting perhaps the tighter binding of GDP or some subtle conformational effect. The protection of EF-Tuchi from NEM inactivation by guanine nucleotides indicates that there may be an allosteric interaction between the nucleotide-binding site and the aminoacyl-tRNA-binding pocket. This interaction was investigated further by determining the effect of NEM modification TIME (min) on the ability of EF-TGhl tobind GDP and GTP. Thelevels FIG; 1. Protection of Phe-tRNA by EF-TG~.The rate of of NEM tested were sufficient to almost completely inhibit hydrolysis of [14C]Phe-tRNA was determined as described under the activity of EF-TQhl in polymerization. As indicated in “Experimental Procedures” in the presence (0)of EF-TUehI or in the Table 111, modification with NEM increases the dissociation presence of an equivalent amount of bovine serum albumin (0).

FIG. 2. Formation of a ternary complex containing EF-T-u, GTP, and aminoacyl-tRNA. A , the binding of [“CIPhe-tRNA to EF-Tueu was determined as described under “Experimental Procedures.” Reaction mixtures contained E F - m u (0)or were incubated in the absence of the factor (0).B, the values for [“CIPhe-tRNA obtained by chromatography in the absence of EFT G have ~ been subtracted from the Values obtained when Phe-tRNA was incubated with EF-TGhl. The resulting Values (0)represent the amount of PhetRNA bound to EF-TGM.

FRACTION NUMBER

FRACTION NUMBER

Chloroplast Elongation Factor Tu

8774

>

k > F 0

a

I

I

OO

0.5

I

1

1.0 0

05

1.0

N-ETHYLMALEIMIDE, (mM) FIG. 3. Guanine nucleotide protection of EF-Tu,,from NEM inactivation. A , EF-TUfhlwas treated with the indicated concentrations of NEM in the presence (0)or absence (0)of GDP as described under “Experimental Procedures” and assayed for EFT & h l activity in polymerization. B , E F - T ~ was M treated with NEM in the presence (W) or absence (0) of GTPas described under “Experimental Procedures” and assayed for activity in polymerization.

TABLE111 Dissociation constant for the EF-TucM. NEM.GDP complex Total EF-T&,d

Total GDP

EF-T,,ua, GDP

nM

flM

nM

35 35

14

2.6

21

Dissociation constant M

1.9 x 1.7 X

‘EF-TQhI was incubated with NEM prior to the addition of GDP nM as described under “Experimental Procedures.”

flM

constant of the EF-T&hl. GDP complex by an order of magnitude. Similarexperiments carried out to determine the effect of NEM modification on the binding of GTP to EFT & h l indicated that the dissociation constant for the EFT&hl.GTP complex is also increased by an order of magnitude; however, the resulting binding is extremely weak, limiting the accuracy of the measurement of the dissociation constant (data not shown). The effect of NEM modification on the binding of aminoacyl-tRNA to EF-T&hl could not be directly investigated due to the severe impairment of GTP binding resulting from this modification. Previous studies (1)have indicated that Euglena EF-T&hl is sensitive to the antibiotic kirromycin. This antibiotic has been shown to inhibit E. coli EF-Tu in polymerization by preventing itsrelease from the ribosome. Kirromycin has also been shown to influence the structuresof the guanine nucleotide and aminoacyl-tRNA binding sites on the E. coli factor (4, 16). The effect of kirromycin on the activity of EF-TGhl in guanine nucleotide binding was investigated. The levels of kirromycin tested were sufficient to almost completely inhibit the activity of EF-Tqhl in polymerization. Kirromycin has only a slight adverse effect on the binding of GDP to the factor (data not shown); however, the effect of the antibiotic on the binding of GTP to EF-TQhl is far more dramatic. Kirromycin substantially enhances the binding of GTP to

E F - n h l resulting in about a10-fold decrease in the apparent dissociation constant of the complex (Table IV). As indicated previously, treatment of EF-T&hl with NEM results in a large decrease in the binding of guanine nucleotides to EF-TGhl. Addition of kirromycin to E F - n h lprior to or after NEM inactivation, largely overcomesthe impairment to GTP binding resulting from the sulfhydryl modification. At limiting concentrations, the amount of GTP bound to NEM modified E F - n h l is enhanced by kirromycin to about the levelof GTP bound by the unmodified factor in the presence of the antibiotic, resulting in about a 100-fold decrease in the apparent dissociation constant of the modified complex. Kirromycin has only a small effect on the binding of GDP toNEM-modified EF-Tuebl (data not shown). Kirromycin converts E. coli EF-Tu into a conformation that resembles that of the factor when GTP is bound (4, 5). As a result, under these conditions the bacterial factor is able to bind aminoacyl-tRNA in thepresence of GDP andpossibly even in the absence of guanine nucleotides (4). Since kirromycin clearly affects the GTP-binding site on EF-TQhl, we have examined the effect of this antibiotic on the ability of the chloroplast factor to bind Phe-tRNA.As indicated in Fig. 4, in the presence of kirromycin, no protection of aminoacyltRNA by EF-T&hl can be detected in the presence of GDP or even in the presence of GTP, suggesting that the factor is unable to bind aminoacyl-tRNA under these conditions. The effect of kirromycin on the interaction of EF-T&h, with aminoacyl-tRNA in the presence or in the absence of guanine nucleotides was further investigated bygel filtration chromatography. Again, no complex formation was observed (data not shown). These results indicate that although kirromycin enhances GTP binding to EF-T&hl, it appears to prevent the subsequent binding of aminoacyl-tRNA. The effect of kirromycin on EF”I’&hl may, therefore, differ significantly from its effect on E. coli EF-Tu. TABLEIV Dissociation constant for the EF-TucM.kirromycin.GTP complex Total EF-Tbu“ nM

Total GTP

EF-%ua,GTP

Dissociation constant M

1.3 27 3.9 X 10-7 32 0.7 4.2 X 10-7 EF-TQhl was incubated with kirromycin prior to the addition of GTP asdescribed under “Experimental Procedures.” 35 53

TIME ( m i d FIG. 4. Protection of Phe-tRNA by EF-TGu in the presence of kirromycin. EF-T” was pretreated with kirromycin and then tested for interaction with aminoacyl-tRNA as described under “Experimental Procedures.” The rate of hydrolysis of [“CIPhe-tRNA was determined in the presence of kirromycin-treated EF-TQ~,with either GDP or GTP (0).0, represents the rate of hydrolysis in the presence of untreated EF-mhl andGTP.

Tu

Chloroplast FactorElongation DISCUSSION

The present report describes the first examination of the guanine nucleotide-binding constants for an organellar EFTu. The dissociation constant for the E F - n , , .GDP complex is 2 orders of magnitude higher than that reported for E. coli EF-Tu or Thermoactirwmyces thermophilis EF-Tu (3,17,18). This observation indicates that the chloroplast factor binds GDP much less tightly than do the prokaryotic factors. EFT u & ] resembles the prokaryotic EF-Tus in having a significantly higher dissociation constant (about 10-fold) for the GTP complex than for the GDP complex. This difference is similar to that observed for T.thermophilis EF-Tu. E. coli EF-Tu, on theother hand, shows a 100-fold decrease in affinity for GTP over GDP. In general, guanine nucleotide binding by the eukaryotic cytoplasmic factor (EF-1) is weaker than that observed for the prokaryotic factors. For example, the dissociation constant for the rabbit reticulocyte EF-lL- GDPcomplex is about 5X M (19). From this perspective the chloroplast factor resembles more closely its cytoplasmic counterparts than the bacterial EF-Tus. No generalizations can be made about the relative strength of GTP versus GDP binding for the cytoplasmic factors (10, 15, 20, 21). In E. coli a ternarycomplex can be readily detected between EF-Tu, GTP, and aminoacyl-tRNA, and this complex has been shown to function as an intermediate in the elongation cycle (22). It has proven more difficult to obtain evidence for the formation of a similar complex with the eukaryotic cytoplasmic factors and todemonstrate that such a complex plays a physiological role in protein biosynthesis (19). As indicated in this report, EF-TUchl will form a ternary complex in the presence of aminoacyl-tRNA and GTP. Complex formation occurs only in thepresence of GTP and canbe observed either directly by gel filtration chromatography or indirectly by the protection of aminoacyl-tRNA from spontaneous hydrolysis. In this respect, the chloroplast factor resembles the bacterial factor more than its cytoplasmic counterpart.

8775

The guanine nucleotide and aminoacyl-tRNA-binding sites on EF-Tu are spatially distinct (22, 23) but must be able to interact allosterically since occupancy of the aminoacyl-tRNA binding site is possible only when GTP is already bound to the factor. We have obtained evidence for an allosteric interaction between these two sites on EF-%hl by using the sulfhydryl-modifying reagent NEM. The DNA sequence of EF-TQ, (8)indicates that ithas asingle cysteine in aposition corresponding to cysteine 81 in the E. coli EF-Tu aminoacyltRNA-binding site. The corresponding cysteine residue located in the guanine nucleotide-binding pocket is absent.The binding of either GDP or GTP will, however,protect EF"I'&hl from NEM modification indicating that theaminoacyl-tRNA binding pocket of EF-"&hl has adifferent conformation when guanine nucleotides are bound to thefactor. In addition, NEM modification affects the conformation of the guanine nucleotide-binding site and reduces the binding affinity of EF"I'QhL for GDP and GTPby an order of magnitude. It is difficult to examine the effect of guanine nucleotide binding to free E. coli EF-Tu due to theinstability of the factor in the absence of GDP,butsubtle conformational differences havebeen detected when GDP is replaced by GTP. For example, the formation of E. coli EF-Tu. GTPfrom the EF-Tu GDPcomplex causes an increase in the reactivity of the sulfhydryl group present in the aminoacyl-tRNA-binding site (3, 24). We have previously reported that Euglena EF-TQhl is sensitive to the antibiotic kirromycin (1).In E. coli, it has been shown that kirromycin does not compete directly with either GTP or aminoacyl-tRNA for binding to the factor. In fact, studies with E. coli EF-Tu mutantshave led to thesuggestion that kirromycin actually binds to domain 11 of EF-Tu while the guanine nucleotide- and aminoacyl-tRNA-binding sites are located in domain I (26). However, the binding of kirromycin to E. coli EF-Tu does affect the guanine nucleotidebinding site. The EF-Tu. kirromycin complex has about a 100-fold higher affinity for GTP than does the free factor, and this complex can bind aminoacyl-tRNA forming a qua-

118

I

EF-Tumt (yeast) 82

EF-TuChl

(Euglena)

137

81

EF-TU

(E)

EF-TU (€.Coli) (X-Ray Crystallography)

6'

f32

a'

f3j

a*

6'

a3

65 ' a

a5

FIG. 5. Secondary structure analyses of portions of domain I of EF-Tus from several sources. The secondary structure predictions were derived using the program developed by R. S. MacWright and W. T. McAllister. The secondary structures for portions of domain I of E. coli EF-Tu (27) (residues 60-210), Euglena E F - T ~ (8) N (residues 60-211), and S. cereoisioe mitochondrial EF-Tu (28) (residues 97-242) are depicted along with the structure for E. coli EF-Tu domain I predicted from x-ray crystallography (23) (residues 65-200). The positions of cysteine residues are indicated above the predicted a-helices (a)and 8-strands(8).

8776

Chloroplast FactorElongation

Tu

ternary complex (4,5). In addition, the binding of kirromycin Tu, around cysteine 118. These comparisons suggest that common properties sharedby these factors have a strong basis to the bacterial factor freezes it into the EF-Tu.GTP-like intheir secondary structuresandthat evolutionary conconformation and allows interaction with aminoacyl-tRNA in the presence of GDP and possibly even in the absence of straints on function have maintained these structures from organism to organism. guanine nucleotides (4). Treatment of Euglena EF-TQ~, with kirromycin has a Acknowledgments-We thank Dr. John H. Harrisonfor kindly slightly adverse effect on its affinity for GDP but enhances providing the secondary structure analyses program and Dr. Susan G T P binding about 10-fold, bringing G T P binding up to the L. Eberly and Richard L. Puetz for helpful discussions and careful levels observed for GDP. In this respect, the interaction of reading of the manuscript. kirromycin with EF-T&hl resembles its interaction with E. REFERENCES coli EF-Tu. However, unlike the bacterial factor,kirromycinSreedharan, 1. S. P., Beck, C. M., and Spremulli, L.L. (1985) J. treated EF-Tuehl does not appear to bind aminoacyl-tRNA Biol. Chem. 260,3126-3131 even in the presence of GTP. Furthermore, no aminoacyl2. Miller,D.L., Hachmann, J., and Weissbach, H. (1971) Arch. tRNA binding to kirromycin-treated EF-TQhl isobserved in Biochem. Biophys. 144, 115-121 the absenceof guanine nucleotides or in presence the of GDP. 3. Arai, K., Kawakita, M., Nakamura, S., Ishikawa, I., and Kaziro, Y. (1974) J. Biochem. (Tokyo) 76,523-534 Again, EF-TQhl behaves quite unlike the bacterial factor in this respect. These results suggest that kirromycin binding to 4. Wolf, H., Chinali, G., andParmeggiani, A. (1977) Eur. J. Biochem. 75,67-75 EF-TuChl affects the guanine nucleotide-binding pocket and 5. Fasano, O., Bruns, W., Crechet, J.-B., Sander, G., and Parmegenhances G T P binding; however,the antibiotic fails trigger to giani, A. (1978) Eur. J. Biochem. 89,557-565 the conformational change required for the binding of ami6. van Noort, J. M., Duisterwinkel, F. J., Jon&, J., SedliEek, J., Kraal, B., and Bosch, L. (1982) EMBO J. 1 , 1199-1205 noacyl-tRNA. Thus,themechanism bywhich kirromycin 7. Van Noort, J. M., Kraal, B., Bosch, L., La Cour, F. M., Nyborg, inhibits the activityof EF-TQhl inpolypeptide synthesis may J., and Clark, B. F. C. (1984) Proc. Natl. Acad. Sci. U. S. A . be quite different from the mechanism by which it inhibits 81,3969-3972 the activity of the bacterial factors. 8. Montandon, P.-E., and Stutz, E. (1983) Nucleic Acids Res. 11, The results presented here indicate that chloroplast EF-Tu 5877-5892 resembles E. coli EF-Tu in many general features although 9. Ravel, J., and Shorey, R. (1971) Methods Enzymol. 20,306-316 several distinct differences have also been observed. Low 10. Beck. C. M.. and SDremulli. L. L.(1982) Arch. Biochem. Bioohvs. 215,414-424 resolution x-ray diffraction of E . coli EF-Tu has been perSuremulli. L. L. (19821 Arch. Biochem. Bioohvs. 214. 734-741 11. formed, and three domains based on electron density have 12. Miller, D.'L., and Weissbach, H. (1974) h h k o d s Enzymol. 3 0 , been observed (23, 25). Domain I belongs to the cy/P class of 219-232 structureand comprisesresidues60 to 240. The guanine 13. Gordon, J. (1968) Proc. Nutl. Acad. Sci. U. S. A. 59,179-183 nucleotide-binding site has been located in this domain as 14. hasShorey, R. L., Ravel,J. M., Garner, C. W., and Shive, W. (1969) J. Biol. Chem. 244,4555-4564 cysteine 81, implicated in aminoacyl-tRNA binding. Domains Roohol, K., and Moller,W. (1978) Eur. J. Biochem. 90,471-477 15. 11 and I11 have been less well characterized. We performed P. H.,Duisterwinkel, F. J., DeGraaf, J. M., 16. VanderMeide, secondary structure analyses for the amino acid residues of Kraal, B., Bosch, L., Douglass, J., and Blumenthal, T. (1981) portions of domain I of Euglena EF-TuChl,E. coli EF-Tu, and Eur. J. Biochem. 1 1 7 , l - 6 yeast mitochondrial EF-Tu using the program developed by 17. Miller, D. L., and Weissbach, H. (1970) Arch. Biochem. Biophys. 141,26-37 R. S. MacWright and W. T. McAllister based on the logic K., Arai,N., Nakamura, S., Oshima, T., andKaziro, Y. published by Garnier et al. (26). As indicated in Fig. 5, there 18. Arai, (1978) Eur. J. Biochem. 92,521-531 is good agreement between the secondary structure predic- 19. Carvalho, G., Carvalho, J. F., and Merrick, W. C. (1984) Arch. tions and the x-ray crystallographic information available for Biochern. Biophys. 234,603-611 domain I of E. coli EF-Tu (23). The proposed secondary 20. Nolan, R. D., Grasmuk, H., Hagenauer, G., and Drews, J. (1974) Eur. J. Biochem. 45,601-609 structures for domain I of Euglena EF-TGhl,E. coli EF-Tu, S., Iwasaki, K., and Kaziro,Y. (1977) J. Biochem. (Tokyo) S. cereuisiae reveal a common 21. Nagata, and mitochondrial EF-Tu from 82,1633-1646 organization and conservation of secondary structure espe- 22. Miller, D. L.,and Weissbach, H.(1977) in Molecular Mechanism cially at the guanine nucleotide- and aminoacyl-tRNA-bindof Protein Biosynthesis (Weissbach, H., andPestka, S., eds) pp. 323-373, Academic Press, New York ing sites. The major cy helices and p sheets indicated for E. 23. Clark, B. F. C., Kruse, T. A., La Cour, T. F. M., Nyborg, J., and coli EF-Tu are also predicted for the chloroplast and mitoRubin, I. R. (1982) FEBS Proc. Meet. 3,377-392 chondrial factors.Homologies in the primary structure of 24. Ohta, S., Nakanishi, M., Tsuboi, M., Arai,K., and Kaziro, Y. various nucleotide-binding proteins have revealed a core se(1977) Eur. J . Biochern. 78, 599-608 quence of Val-X-X-X-hydrophobic-hydrophobic-hydropho-25. Bosch, L., Kraal, B., Van der Meide, P. H., Duisterwinkel, F. J., and Van Noort, J. M. (1983) Prog. Nucleic Acid Res. Mol. Biol. bic-hydrophobic-Asn/Asp(29).In E. coli EF-Tu,thisse30,92-126 quence is located in the p4 sheet of domain 1 and the last J., Osguthorpe, D. J., andRohson, B. (1978) J. Mol. residue of the core sequence, asparagine 135, has been found 26. Garnier, Biol. 120, 97-120 to H-bond toO6 of the guanine base(29). Chloroplast EF-Tu 27. Arai, K., Clark, B. F. C., Duffy,L., Jones, M. D., Kaziro, Y., Laursen, R. A., L'Italein, J. J., Miller, D. L., Nagarkatti, S., from Euglena has this sequence as well, spanning residues Nakamura, S., Neilsen, K. M., Petersen, T. E., Takahashi, K., 128-136 and locatedin a n identical position in domain I. and Wade, M. (1980) Proc. Natl. Acad. Sci. U. S. A . 7 7 , 1326While the aminoacyl-tRNA binding on EF-Tu has not yet 1330 been well characterized, the secondary structure aroundcys- 28. Nagata, S., Tsunetsugu-Yokota, Y., Naito, A., and Kaziro, Y. teine 81 for E. coli EF-Tu is virtually identical to the struc(1983) Proc. Nutl. Acud. Sci. U. S. A . 80,6192-6196 tures of Euglena EF-Tuchl around cysteine82 and yeast EF- 29. Leberman, R., and Egner, U.(1984) EMBO J. 3,339-341 "