- Email: [email protected]
Expression of bovine mitochondrial elongation factor Ts in Escherichia coli and characterization of the heterologous complex formed with prokaryotic elongation factor Tu
Biochimica et Biophysica Acta 1352 Ž1997. 102–112
Expression of bovine mitochondrial elongation factor Ts in Escherichia coli and characterization of the heterologous complex formed with prokaryotic elongation factor Tu Hong Xin a , Karen Leanza a , Linda Lucy Spremulli
a,b,)
a
b
Department of Chemistry, UniÕersity of North Carolina, Campus Box 3290, Caphel Hill, NC 27599-3290, USA Lineberger ComprehensiÕe Cancer Research Center, UniÕersity of North Carolina, Campus Box 3290, Chapel Hill, NC 27599-3290, USA Received 3 October 1996; accepted 16 December 1996
Abstract When bovine mitochondrial elongation factor Ts ŽEF-Ts mt . is expressed in Escherichia coli, it forms a tightly associated complex with E. coli EF-Tu ŽEF-Tu Eco Ø Ts mt .. This complex is active in polyŽU.-directed polymerization and this activity is inhibited by kirromycin. The EF-Tu Eco Ø Ts mt complex does not bind guanine nucleotides detectably and is not dissociated to a significant extent by either GDP or GTP. A portion of the EF-Tu Eco Ø Ts mt complex can be dissociated by aa-tRNA in the presence of GTP. The heterologous complex cannot be dissociated completely in the presence of either the 8 M urea or 8 M guanidine hydrochloride, suggesting that EF-Ts mt has an unusually tight interaction with E. coli EF-Tu. The EF-Tu Eco Ø Ts mt complex can be dissociated by denaturation using 2 M guanidine thiocyanate. Free EF-Ts mt can then be purified and renatured. The refolded EF-Ts mt is active in stimulating the activity of expressed mitochondrial EF-Tu ŽEF-Tu mt . in polyŽU.-directed polymerization. Almost all the EF-Ts mt molecules appear to refold into a conformation which can interact with EF-Tu mt . Protease mapping of EF-Ts mt indicates that the first 54 residues fold into an independent domain. Analysis of deletion derivatives of EF-Ts mt indicates that extensive regions of this factor are required for its tight interaction with EF-Tu. q 1997 Elsevier Science B.V. All rights reserved. Keywords: Protein synthesis; Elongation; Mitochondrion; Organelle; Translation
1. Introduction Elongation factor Tu ŽEF-Tu. is responsible for promoting the binding of aminoacyl-tRNA Ž aa-tRNA.
Abbreviations: EF-Tu, elongation factor Tu; EF-Ts, elongation factor Ts; EF-Ts mt, mitochondrial elongation factor EF-Ts; EF-Tu Eco Ø Ts mt , heterologous complex between E. coli EF-Tu and EF-Ts mt ; aa-tRNA, aminoacyl-tRNA; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis. ) Corresponding author. Fax: q1 Ž919. 9663675; E-mail: [email protected]
to the A-site of the ribosome during protein biosynthesis w1x. EF-Tu forms a ternary complex with GTP and aa-tRNA prior to interaction with the ribosome. Following codon:anticodon interaction on the ribosome, EF-Tu is released in the form of an EF-Tu Ø GDP complex w2x. The release of GDP from EF-Tu is facilitated by a second elongation factor ŽEF-Ts.w3x. In E. coli, EF-Ts binds EF-Tu promoting the dissociation of GDP and the formation of an intermediate EF-Tu Ø Ts complex. This complex is, in turn, dissociated by GTP resulting in the formation of an EF-Tu Ø GTP complex which subsequently binds aa-tRNA
0167-4781r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved. PII S 0 1 6 7 - 4 7 8 1 Ž 9 7 . 0 0 0 0 3 - 1
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
forming a new ternary complex. In Thermus thermophilus, a dimeric complex ŽEF-Tu Ø Ts. 2 is formed by the interaction of two EF-Tu molecules with a stable EF-Ts dimer w4x. In contrast to the E. coli EF-Tu Ø Ts complex, the T. thermophilus complex is not dissociated to a significant extent by either GDP or GTP alone w5x. Mammalian mitochondrial EF-Tu and EF-Ts ŽEFTu mt and EF-Ts mt . have been purified as a tightly associated complex ŽEF-Tu Ø Ts mt . from bovine liver w6,7x. This complex has a number of features that distinguish it from the corresponding E. coli elongation factors. First, EF-Tu Ø Ts mt is quite stable and cannot be dissociated even in the presence of high concentrations of guanine nucleotides. In this respect, the mitochondrial factors differ significantly from the corresponding E. coli factors and share some resemblance to the thermophilic EF-Tu and EF-Ts. Second, no detectable amounts of intermediates equivalent to EF-Tu Ø GDP and EF-Tu Ø GTP can be observed in the mammalian mitochondrial system. Finally, significant binding of guanine nucleotides to EF-Tu Ø Ts mt can only be detected in the presence of aa-tRNA w7x. The cDNAs for human and bovine EF-Tu mt and EF-Ts mt have been cloned and sequenced w8,9x. The sequence of bovine EF-Tu mt is 56% identical to that of E. coli EF-Tu. This factor has been expressed in E. coli as a His-tagged protein. The purified factor is active in polyŽU.-directed polymerizationw8x. The sequence for EF-Ts mt is less than 30% identical to the sequences of the corresponding prokaryotic factors w9x. EF-Ts mt has also been expressed in E. coli as a His-tagged protein. When it is prepared from bacteria under native conditions, it is obtained as a 1:1 complex with E. coli EF-Tu. This heterologous complex is designated EF-Tu Eco Ø Ts mt w9x. The EF-Tu Eco Ø Ts mt complex is active in polymerization. In the current study, we have examined the properties of this heterologous complex and have developed procedures for the preparation of active EF-Ts mt free from EF-Tu. 2. Materials and methods 2.1. Materials Trypsin Žtype III. and Sephadex G50 were purchased from Sigma Chemical Company. Pure nitro-
103
cellulose blotting membranes were purchased from Schleicher and Schuell. Nitrocellulose filter paper ŽType HA, 0.45 m m. was from Millipore Corporation. w 3 HxGDP was from Dupont-New England Nuclear. Ni-NTA resin was from QIAGEN. Oligonucleotide primers were made in the Lineberger Comprehensive Cancer Research Center at the University of North Carolina. The pET expression system was from Novagen. E. coli ribosomes and elongation factors were purified as described before w10,11x. w 14 CxPhe-tRNAphe was prepared as described previously w12x. Expressed EF-Tu mt and the EF-Tu Ø Ts mt complex purified from bovine liver were kindly provided by Velinda Woriax ŽDepartment of Chemistry, University of North Carolina.. 2.2. Buffers Binding buffer: 50 mM Tris-HCl Ž pH 7.6., 60 mM KCl, 7 mM MgCl 2 , 7 mM b-mercaptoethanol, 0.1 mM phenylmethylsulfonylfluoride and 10% glycerol. Imidazole elution buffer: 50 mM Tris-HCl Ž pH 7.6., 40 mM KCl, 5 mM b-mercaptoethanol, 0.15 M imidazole and 10% glycerol. Buffer A: 20 mM Hepes-KOH ŽpH 7.0., 40 mM KCl, 1 mM MgCl 2 , 0.1 mM EDTA and 10% glycerol. High salt buffer: 50 mM Tris-HCl ŽpH 7.6., 1 M NH 4Cl, 5 mM b-mercaptoethanol, 10 mM imidazole and 10% glycerol. Denaturation buffer: 50 mM Tris-HCl ŽpH 7.6., 60 mM KCl, 0.1 M b-mercaptoethanol and the indicated amount of denaturing reagent. Renaturation Buffer: 50 mM Tris-HCl ŽpH 7.6., 60 mM KCl, 0.1 M b-mercaptoethanol. 2.3. Construction of plasmids for the expression of portions of EF-Tsmt PCR was used to add a NdeI cutting site to the 5X end and a XhoI cutting site to the 3X end of the bovine liver cDNA encoding various deletion mutants of EF-Ts mt . PCR amplification of the desired region of the EF-Ts mt gene was carried out as described before w9x. The cDNAs were then digested with NdeI and XhoI and cloned into the NdeI and XhoI sites of the pET24cŽq. vector. This vector provides an in-frame sequence coding for a histidinetag at the COOH-terminus. The resulting constructs were first transformed into E. coli DH5 a . Plasmid
104
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
DNA from each construct was prepared from 4 ml of cells using the QIAGEN mini-preparation kit. A portion Ž1r10. of each DNA preparation was then transformed into E. coli BL21ŽDE3. for expression. Expression was carried out as described previously w9x. Analysis of the expression of EF-Ts mt was examined under denaturing conditions and under native conditions as described w9x except that the protein inhibitor cocktail tablets ŽBoehringer Mannheim. were included in the buffers used for preparations made under native conditions. The EF-Tu Eco Ø Ts mt complex was purified as described w9x. Protein concentrations were determined by the Micro-Bradford method according to standard protocols ŽBio-Rad Inc... 2.4. Gel filtration analysis of GDP-binding E. coli EF-Tu Ž200 pmol., EF-Tu Eco Ø Ts mt Ž200 pmol, 15.0 m g., or a no protein control were incubated for 10 min at 378C in a reaction mixture Ž 250 m l. containing 10 m M w 3 HxGDP Ž1460 cpmrpmol., 10 mM MgCl 2 , 50 mM Tris-HCl Ž pH 7.8. , 1 mM dithiothreitol and 50 mM NH 4Cl. These samples were applied to a 15-cc Ž0.9 cm = 23 cm. Sephadex G-50 gel filtration column equilibrated in 50 mM Tris-HCl ŽpH 7.8., 10 mM MgCl 2 , 1 mM dithiothreitol and 50 mM NH 4Cl. The column was washed with the same buffer at 0.5 mlrmin, and 0.28-ml fractions were collected. Aliquots Ž0.15 ml. of various fractions were counted in 10 ml of Scintisafe Econo I liquid scintillant. 2.5. Effect of GDP, GTP and Phe-tRNA phe on the stability of EF-Tu Eco6Tsmt The EF-Tu Eco Ø Ts mt complex Ž100 m g. was mixed at 48C for 1 h with 50 m l of a 50% slurry of Ni-NTA resin equilibrated in binding buffer. The Ni-NTA resin was then put into a mini-column made from a yellow Eppendorf pipette tip and washed twice with 200 m l of buffer containing 10 mM Tris-HCl ŽpH 7.6., 10 mM MgCl 2 , 60 mM KCl, 1 mM DTT and 10% glycerol. The column was then washed with 50 m l of the buffer above containing 1 mM GDP or 1 mM GTP as indicated. The protein retained on the column was eluted with 5 aliquots of 50 m l Imidazole elution buffer. An aliquot of 4 m l of protein eluted at each step was analyzed by sodium dodecyl
sulfate polyacrylamide gel electrophoresis ŽSDSPAGE.. To test the effect of Phe-tRNAphe on EF-Tu Eco Ø Ts mt , 4.8 m g of complex Ž 50 pmol. was mixed with 60 m l of a 50% slurry of Ni-NTA resin at 48C for 1 h in binding buffer. The resin was washed with high salt buffer Ž400 m l. and buffer A Ž400 m l. to remove the unbound proteins, and was then poured to a mini-column made from a yellow Eppendorf pipette tip. A aliquot of 2 = binding buffer Ž30 m l. containing 6.7 m M Phe-tRNA and 0.66 mM GTP was added to the resin and incubated at 48C for 30 min. The resin was then washed with 100 m l of buffer containing 50 mM Tris-HCl ŽpH 7.6., 10 mM MgCl 2 , 60 mM KCl, 7 mM b-mercaptoethanol, 10% glycerol and 1 mM GTP where indicated. The protein retained on the column was eluted with three aliquots of 50 m l each of imidazole elution buffer. An aliquot Ž16 m l. of protein eluted at each step was analyzed by SDS-PAGE. 2.6. Denaturation of the EF-Tu Eco6Tsmt complex and renaturation of EF-Tsmt Small-scale denaturation experiments were carried out using 5 m g of the EF-Tu Eco Ø Ts mt complex in 200 m l denaturation buffer containing the indicated amount of the various denaturing reagents Ž3–8 M urea, 2–8 M guanidine-HCl, 1–6 M guanidine thiocyanate or 0.01–0.25% N-lauroylsarcosine. . These mixtures were incubated at room temperature for 20 min. The dissociation of the EF-Tu Eco Ø Ts mt complex was monitored by examining the retention of E. coli EF-Tu along with EF-Ts mt on a Ni-NTA column. For this analysis, 30 m l of a 50% slurry of Ni-NTA resin was added to the samples above and the mixtures were incubated at room temperature for one h. The resin was then loaded into a mini-column made from a yellow Eppendorf tip. The resin was washed twice with 400 m l of the starting buffer. Proteins retained on the column were eluted with 50 m l imidazole elution buffer, and 16 m l of the eluted material was analyzed by SDS-PAGE. Large-scale denaturation of the EF-Tu Eco Ø Ts mt complex and renaturation of EF-Ts mt began with the complex obtained from 4 g of cells Žfrom 1 liter cell culture. . The initial steps followed the strategy outlined for the purification of the EF-Tu Eco Ø Ts mt com-
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
plex described previously w9x. Once the EF-Tu Eco Ø Ts mt complex was bound to the Ni-NTA resin, the resin was placed in an empty 15-cc column and washed with high salt buffer Ž60 ml.. Denaturation buffer Ž10 ml. containing 2 M guanidine thiocyanate was added to the column and the resin was incubated with this buffer at room temperature for 20 min. The column was then washed with 15 ml denaturation buffer containing 2 M guanidine thiocyanate to remove E. coli EF-Tu. The EF-Ts mt retained on the column was renatured by mixing the resin with 10 ml renaturation buffer followed by washing with 50 ml of dilution buffer containing 50 mM Tris-HCl ŽpH 7.6. and 60 mM KCl. EF-Ts mt was eluted with 0.5 ml imidazole elution buffer and dialyzed against buffer containing 10 mM Tris-HCl ŽpH 7.6., 40 mM KCl, 7 mM MgCl 2 and 10% glycerol for 1.5 h with three changes of the solution. 2.7. Trypsin digestion of free EF-Tsmt and the EFTu Eco6Tsmt complex A stock solution of trypsin Ž10 mgrml in 50 mM Tris-HCl, pH 7.8. was prepared one day before use and kept at 48C overnight. A series of 10-fold dilutions was made immediately before use in 50 mM Tris-HCl ŽpH 7.8.. The EF-Tu Eco Ø Ts mt complex Ž5 m g in 2 m l of buffer A. or free EF-Ts mt Ž5 m g in 2.3 m l of buffer A. was added into 10 m l containing the indicated amount of trypsin and incubated at 278C for 15 min. Reactions were stopped by the addition of 4 m l buffer A and 4 m l of buffer containing 63 mM Tris-HCl Ž pH 6.8., 0.1 M b-mercaptoethanol, 2% SDS, 10% glycerol, 0.5% bromophenol blue, and then heated at 908C for 10 min. Samples Ž6 m l. were analyzed by SDS-PAGE and stained with silver w13x. 2.8. Poly(U)-directed polymerization assay The activity of EF-Tu Eco Ø Ts mt was measured by its ability to catalyze the polyŽ U. -directed polymerization of phenylalanine on E. coli ribosomes as described previously w6x. The activity of renatured EF-Ts mt was detected using the polymerization assay described above with varying amounts of expressed protein and 3 pmol of EF-Tu mt expressed in E. coli w8x. The effect of kirromycin on the polymerization activity of the EF-Tu Eco Ø Ts mt complex was tested
105
by adding different amounts of kirromycin Ž5 = 10y6–4 = 10y9 M. to each reaction. The percentage of activity of each reaction was calculated in comparison with a positive control without kirromycin.
3. Results and discussion 3.1. Interactions of the EF-Tu Eco6Tsmt complex with guanine nucleotides The E. coli EF-Tu Ø Ts complex can be readily dissociated by either GDP or GTP w14x. In contrast, neither the T. thermophilus ŽEF-Tu Ø Ts. 2 complex nor the bovine liver EF-Tu Ø Ts mt complex can be dissociated to a significant extent by high concentrations of guanine nucleotides w6,7,15x. To compare the EF-Tu Eco Ø Ts mt complex with E. coli EF-Tu Ø Ts and with native EF-Tu Ø Ts mt purified from bovine liver, the basic features of this heterologous complex were investigated. These features included its ability to bind guanine nucleotides and its susceptibility to dissociation by GTP, GDP and aa-tRNA. The ability of EF-Tu to bind guanine nucleotides is generally measured using a nitrocellulose filter binding assay w16x. The E. coli EF-Tu Ø Ts complex is readily dissociated in the presence of guanine nucleotides. As a result, incubation of E. coli EF-Tu Ø Ts with guanine nucleotides results in the formation of E. coli EF-Tu Ø GDP or EF-Tu Ø GTP complexes which are readily detected by a nitrocellulose filter binding assay when labeled nucleotides are used. In contrast, very little guanine nucleotide binding can be detected when EF-Tu Eco Ø Ts mt is incubated with labeled GDP or GTP, suggesting that the heterologous complex is not being dissociated to a significant extent by guanine nucleotides Ždata not shown.. This result also suggests that the EF-Tu Eco Ø Ts mt complex does not bind to guanine nucleotides directly. However, it was possible that the heterologous complex bound GDP but was not retained by the nitrocellulose membrane. To examine this possibility, GDP-binding was analyzed using a gel filtration assay. As indicated in Fig. 1, the binding of w 3 HxGDP to E. coli EF-Tu can be readily detected by this method. In contrast, no binding of GDP to the EF-Tu Eco Ø Ts mt complex can be observed. These observations suggest that the heterologous complex does not bind guanine
106
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
dure in the absence of guanine nucleotides. With buffers containing 1 mM GDP, only a trace of dissociation of the EF-Tu Eco Ø Ts mt complex was observed ŽFig. 2B. . No release of E. coli EF-Tu was observed in the presence of GTP ŽFig. 2C.. These results indicate that the EF-Tu Eco Ø Ts mt complex resembles the native bovine mitochondrial EF-Tu Ø Ts mt complex which cannot be readily dissociated by either GDP or GTP. The effect of aa-tRNA on the EF-Tu Eco Ø Ts mt
Fig. 1. Use of gel filtration to examine nucleotide binding to the EF-Tu Eco Ø Ts mt complex. EF-Tu Eco Ø Ts mt was incubated with w 3 HxGDP and nucleotide binding was examined by following the elution position of the labeled nucleotide during chromatography on Sephadex G50 Ž`.. As a control E. coli EF-Tu was incubated with w 3 HxGDP and nucleotide binding was examined by chromatography on Sephadex G50 ŽB.. Chromatography of w 3 HxGDP alone on the Sephadex G50 column is also shown ŽI..
nucleotides to a significant extent and is probably not readily dissociated by guanine nucleotides. The EFTu Ø Ts mt complex purified from bovine liver also fails to bind guanine nucleotides at a detectable level, possibly because it is not easily dissociated by either GDP or GTP w6x. The ability of GDP and GTP to promote the release of E. coli EF-Tu from the heterologous EFTu Eco Ø Ts mt complex was then tested directly. For these experiments, EF-Tu Eco Ø Ts mt was bound to a Ni-NTA affinity column, then washed with buffers lacking guanine nucleotides or with buffers containing 1 mM GDP or 1 mM GTP. As shown in Fig. 2A, no dissociation of the heterologous complex is observed when the complex is subjected to this proce-
Fig. 2. Failure of GDP or GTP to dissociate the EF-Tu Eco Ø Ts mt complex. The EF-Tu Eco Ø Ts mt complex was bound to Ni-NTA resin through the His-tag on EF-Ts mt as outlined in Section 2. A: The resin was washed with aliquots of a control buffer lacking guanine nucleotides Žlanes 1 and 2. and then eluted with buffer containing imidazole Žlanes 3 and 4.. Proteins present in these samples were analyzed by SDS-PAGE for the presence of E. coli EF-Tu and EF-Ts mt using silver staining. B: The resin was washed with buffer containing 1 mM GDP Žlanes 1 and 2.. The protein still remaining on the column was then eluted with 2 aliquots of a buffer containing imidazole Žlanes 3 and 4.. Aliquots of several fractions were analyzed by SDS-PAGE for the presence of E. coli EF-Tu and EF-Ts mt . C: The resin was washed with buffer containing 1 mM GTP Žlanes 1 and 2.. The protein still remaining on the column was then eluted with buffer containing imidazole Žlanes 3 and 4.. Aliquots of various fractions were analyzed by SDS-PAGE for the presence of E. coli EF-Tu and EF-Ts mt .
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
107
3.2. Effect of kirromycin on the EF-Tu Eco6Tsmt complex
Fig. 3. Effect of GTP and Phe-tRNA on the interaction between E. coli EF-Tu and EF-Ts mt . An aliquot of the EF-Tu Eco Ø Ts mt complex was bound to Ni-NTA resin through the His-tag on EF-Ts mt as outlined in Section 2. A: The resin was washed with buffer containing Phe-tRNA but lacking GTP Žlanes 1 and 2. and then eluted sequentially with aliquots of buffer containing imidazole Žlanes 3 and 4.. Proteins present in these samples were analyzed by SDS-PAGE for the presence of E. coli EF-Tu and EF-Ts mt . B: The resin was washed with buffer containing both Phe-tRNA and GTP Žlanes 1 and 2.. The protein still remaining on the column was then eluted with buffer containing imidazole Žlanes 3 and 4.. An aliquot of various fractions was analyzed by SDS-PAGE for the presence of E. coli EF-Tu and EF-Ts mt .
complex was investigated using a similar strategy. The EF-Tu Eco Ø Ts mt complex was bound to the NiNTA affinity column, followed by the incubation with Phe-tRNAphe in the presence or absence of GTP. As shown in Fig. 3, close to 30% of the E. coli EF-Tu was dissociated from EF-Ts mt under these conditions. No dissociation of the complex was observed when GTP was removed from the buffers ŽFig. 3A.. This observation suggests that both aatRNA and GTP must be present in order to promote the dissociation of a significant amount of the EFTu Eco Ø Ts mt complex. The order of binding of the GTP and aa-tRNA cannot be determined with the approach used here. However, in the presence of aa-tRNA, the E. coli EF-Tu Ø GTP complex binds aa-tRNA forming the ternary complex. Coupling these two reactions allows us to detect the net dissociation of the EF-Tu Eco Ø Ts mt complex.
The binding of kirromycin to E. coli EF-Tu is mutually exclusive with EF-Ts w17x. The polymerization activity of E. coli EF-Tu Ø Ts is inhibited by kirromycin, since this antibiotic freezes EF-Tu on the ribosome in a GTP-like conformation blocking the elongation cycle. The effect of kirromycin on the EF-Tu Eco Ø Ts mt complex was tested by examining the concentration of this antibiotic required to inhibit the activity of the complex in polyŽ U. -directed polymerization. As shown in Fig. 4, the activity of EFTu Eco Ø Ts mt can be inhibited by kirromycin at a concentration very similar to that required for the inhibition of the homologous E. coli EF-Tu Ø Ts complex. This result suggests that the ternary complex ŽEF-Tu Ø GTP Ø aa-tRNA. forms during the elongation cycle, and that EF-Ts mt dissociates from E. coli EF-Tu during each round of elongation. The bovine EF-Tu Ø Ts mt complex is not inhibited by kirromycin w6x. Two explanations had been suggested for this observation w8x. First, it was possible
Fig. 4. Effect of kirromycin on the activity of the EF-Tu Eco Ø Ts mt complex in polyŽU.-directed polymerization. Reaction mixtures were prepared as described w9x and contained 78 ng of the EF-Tu Eco Ø Ts mt complex ŽB. or 35 ng of the E. coli EF-Tu Ø Ts complex Ž`. and the indicated amount of kirromycin. Control reactions containing no kirromycin resulted in the polymerization of 6.3 and 4.3 pmol of phenylalanine for the EF-Tu Eco Ø Ts mt complex and E. coli EF-Tu Ø Ts complex respectively and were used to define the 100% points.
108
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
that the tight interaction of EF-Ts mt with EF-Tu altered the basic features of the elongation cycle so that kirromycin could no longer freeze EF-Tu on the ribosome. A second explanation was that EF-Tu mt could not bind kirromycin due to the presence of a threonine residue at position 380 Žequivalent to A375 in E. coli EF-Tu.. Mutation of this residue in prokaryotic EF-Tu renders the factor resistant to kirromycin. The observations made here suggest that the ability of EF-Ts mt to bind tightly to EF-Tu does not preclude sensitivity of the elongation cycle to kirromycin. Thus, the kirromycin resistance of mitochondrial protein synthesis is an intrinsic property of EF-Tu mt . 3.3. Dissociation of EF-Tsmt from E. coli EF-Tu by denaturing reagents The effect of denaturing reagents on the dissociation of the EF-Tu Eco Ø Ts mt complex was examined using several concentrations of different denaturing reagents. EF-Ts mt could not be released completely from E. coli EF-Tu by urea even at concentrations of 8 M Ž Fig. 5A.. Guanidine hydrochloride was somewhat more effective than urea. However, even 8 M guanidine hydrochloride could not completely dissociate this complex ŽFig. 5B.. These results suggest that the heterologous complex is unusually stable and that the interaction between E. coli EF-Tu and EFTs mt is quite tight. In the presence of 2 M guanidine thiocyanate Ž Fig. 5C. or 0.2% N-lauroylsarcosine ŽFig. 5D., EF-Ts mt could be almost completely released from E. coli EF-Tu. EF-Ts mt could not be recovered from the Ni-NTA affinity column at concentrations of guanidine thiocyanate higher than 4 M, suggesting that higher concentrations of this denaturing reagent might cause the release of the His-tagged protein from the resin or might damage the Ni-NTA resin. Once conditions had been established for the denaturation of the EF-Tu Eco Ø Ts mt complex, procedures were developed for the large-scale preparation of EF-Ts mt . The EF-Tu Eco Ø Ts mt complex was bound to the Ni-NTA affinity column. The complex was denatured by treatment with buffer containing 2 M guanidine thiocyanate and E. coli EF-Tu was removed from the column by washing with the denaturing solution. The EF-Ts mt retained on the resin was
Fig. 5. Denaturation of the EF-Tu Eco Ø Ts mt complex. The EFTu Eco Ø Ts mt complex Ž5 m g. was treated with the indicated concentrations of urea Žpanel A., guanidine hydrochloride Žpanel B., guanidine thiocyanate Žpanel C. or N-lauroylsarcosine Žpanel D.. The EF-Ts mt component and any associated EF-Tu was recovered by Ni-NTA chromatography. The material retained by the resin was eluted with imidazole and an aliquot was analyzed by SDS-PAGE followed by silver staining. The higher molecular weight band is EF-Tu while the lower molecular weight band is EF-Ts.
renatured by fast dilution of the resin to reduce the concentration of the denaturing reagent w18x. Refolded EF-Ts mt was then eluted with imidazole. About 400–600 m g of EF-Ts mt could be purified from one liter of E. coli cell culture by this procedure. This value represents a yield of 20–30% of the EF-Ts mt in the starting complex. The major loss of the protein is thought to happen in the renaturation step, in which 0.1 M b-mercaptoethanol is used to avoid the improper formation of disulfide bonds. High concentrations of b-mercaptoethanol may damage the Ni-NTA resin, resulting in the premature release of EF-Ts mt . Analysis of the purified EF-Ts mt on SDS-PAGE indicates that it is free of E. coli EF-Tu ŽFig. 6.. A small amount of a low molecular weight impurity can be observed in these preparations. This component is probably a contaminant from the original preparation since it does not cross-react with antibodies raised
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
109
Fig. 6. Purity of EF-Ts mt . Lane 1: SDS-PAGE analysis of the EF-Tu Eco Ø Ts mt complex Ž0.5 m g.. Lane 2: The EF-Tu Eco Ø Ts mt complex Ž ; 4 mg. bound on the Ni-NTA column was denatured by incubation with 2 M guanidine thiocyanate as described in Section 2. The E. coli EF-Tu component was washed from the column. The remaining EF-Ts mt was renatured on the Ni-NTA resin and then eluted with imidazole. An aliquot of the renatured protein Ž0.18 m g. was analyzed by SDS-PAGE.
against EF-Tu Ø Ts mt from bovine liver Ždata not shown.. The activity of the renatured EF-Ts mt was evaluated by its ability to stimulate the activity of EF-Tu mt in the polyŽ U.-directed polymerization. EF-Ts mt enhances the activity of EF-Tu mt by 4- to 6-fold ŽFig. 7.. A molar ratio of EF-Tu mt to EF-Ts mt required for maximal activity Žslightly less than 1:1. suggests that a considerable fraction of the renatured EF-Ts mt has refolded into a conformation that can interact with EF-Tu mt . 3.4. Proteolytic digestion of EF-Tsmt T. thermophilus EF-Ts has a trypsin-sensitive site at Lys48 and cleavage at this position results in the
Fig. 7. Stimulation of the activity of EF-Tu mt by renatured EF-Ts mt . The indicated amount of renatured EF-Ts mt was used to stimulate the activity of EF-Tu mt Ž144 ng. in polyŽU.-directed polymerization.
Fig. 8. Trypsin digestion of free EF-Ts mt and the EF-Tu Eco Ø Ts mt complex. A: EF-Ts mt Ž5 m g. was digested with the indicated amount of trypsin and the resulting products were analyzed by SDS-PAGE. Lane 1, no trypsin; lane 2, 0.01 ng; lane 3, 0.1 ng; lane 4, 1 ng; lane 5, 10 ng. The smaller protein whose position is indicated by the arrow is a contaminant in the particular EF-Ts mt preparation used here. It does not cross-react with antibodies raised against the EF-Tu Ø Ts mt complex Ždata not shown.. B: The EF-Tu Eco Ø Ts mt complex Ž5 m g. was digested with the indicated amount of trypsin, followed by the analysis of the samples on SDS-PAGE. Lane 1, no trypsin; lane 2, 0.01 ng; lane 3, 0.1 ng; lane 4, 1 ng; lane 5, 10 ng.
formation of a large COOH-terminal fragment w4x. This COOH-terminal fragment cannot interact tightly with EF-Tu. E. coli EF-Ts has similar trypsin-sensitive sites at Lys51 and Lys52 w19x. The large COOHterminal fragment released by trypsin has very low activity in promoting guanine nucleotide exchange with EF-Tu. As shown in Fig. 8, when free EF-Ts mt is digested with a low level of trypsin, a single major cut site was observed. The resulting product from EF-Ts mt has a molecular mass of about 30 kDa. This fragment is retained by Ni-NTA resins and must, therefore, be derived from the COOH-terminal end which carries the His-tag. These data, along with comparative sequence analysis suggest that the cleavage observed most likely occurs at Arg54 leaving a large COOH-terminal fragment. At higher concentrations of trypsin, EF-Ts mt is degraded rapidly, indicating that many of the 29 potential trypsin cleavage sites can be attacked when enough of the protease is present Ždata not shown. .
110
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
The effect of trypsin digestion on the EF-Tu Eco Ø Ts mt complex was then examined ŽFig. 8. . The single major cut of EF-Ts mt was observed following digestion of the complex, suggesting that this trypsin-sensitive site is located on the surface of EF-Ts mt both when it is free in solution and when it is present in the complex with E. coli EF-Tu. This idea is compatible with the crystal structure of the E. coli EF-Tu Ø Ts complex w20x. EF-Tu in the EF-Tu Eco Ø Ts mt complex was also cleaved producing a single major band. The cleavage in EF-Tu probably occurs in the effector region around Arg58 and the major band observed is derived from the COOH-terminus of EF-Tu w21x. 3.5. Deletion mutations of EF-Tsmt The resistance of the EF-Tu Ø Ts mt and the EFTu Eco Ø Ts mt complexes to dissociation by guanine
nucleotides raises the question of the nature of the interactions between these two factors. The three-dimensional structure of EF-Tu is known in both the GTP and GDP bound conformations w22–24x. EF-Tu folds into three domains ŽI, II and III.. Domain I binds guanine nucleotides while all three domains participate in binding aa-tRNA w25x. The three-dimensional structure of the E. coli EF-Tu Ø Ts complex shows that EF-Ts interacts with both Domain I and Domain III of EF-Tu w20x. This observation is in agreement with previous results w26,27x. An alignment of the sequences of EF-Ts from prokaryotes and organelles suggests that the primary sequence of EF-Ts mt can be divided into three regions ŽFig. 9. . The first 96 residues Žthe N-segment. are the most highly conserved region among the sequences known. The 3-dimensional structure of E. coli EF-Ts indicates that first 50–60 residues of the
Fig. 9. Deletion mutants of EF-Ts mt . A: The primary sequence of EF-Ts mt was divided into three regions according to the sequence similarity scores relative to EF-Ts from different organisms. B: Diagrammatic representation of deletion derivatives prepared from EF-Ts mt . The His-tag on the mutated proteins is indicated by the solid ellipse. The dashed line indicates that the middle section is not present in the native protein of T. thermophilus EF-Ts ŽT. ther EF-Ts., or has been deleted from of EF-Ts mt ŽTs-NC.. The table on the right summarizes the properties of the derivatives. The stability was assessed by the sensitivity of the derivative to proteases as measured by the ability to isolate the intact protein from cell extracts under native conditions. The ability of the derivative to interact with E. coli EF-Tu was determined by co-retention on Ni-NTA resins. The activity was measured by the ability to stimulate the activity of EF-Tu mt in polyŽU.-directed polymerization.
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
N-segment fold into an independent domain w20x. The middle section Žresidues 97 to 190. is referred to as the middle Ž M.-segment This region is present in the mammalian EF-Ts mt , E. coli, S. platensis, and S. citri EF-Ts w28–30x. It is not present in T. thermophilus EF-Ts or in chloroplast EF-Ts from G. sulphuraria w4,31x. The last 93 residues Ž191–283. are referred to as the C-segment. In E. coli EF-Ts, secondary structural elements from the M-segment and the C-segment are interwoven in 3 dimensions w20x. A comparison of the 3-dimensional structure of E. coli EF-Ts and the dimeric form of EF-Ts from T. thermophilus w20,32x indicates that the two structures have both structural similarities and differences. The dimer interface in T. thermophilus is formed by two three-stranded antiparallel b-sheets. A similar fold is observed internally in E. coli EF-Ts. Thus, in many ways, the dimeric form of EF-Ts found in T. thermophilus is similar to a monomer unit of E. coli EF-Ts. EF-Ts mt is the same size as E. coli EF-Ts and is likely to form the internal b-sheet interface observed in the E. coli factor. EF-Ts mt is 29% identical to E. coli EF-Ts in primary sequence. Alignment of the sequences of these two proteins is somewhat tenuous in places due to the low conservation of primary sequence. Analysis of their alignment indicates that there will be both similarities and differences between these two factors. The precise positions of insertions and deletions in the alignment varies with different alignment programs and as a function of the set of sequences used in the alignment Žsee Refs. w4x and w9x.. An examination of individual residues of E. coli EF-Ts and EF-Ts mt suggests that more differences will be found in the regions of EF-Ts that interact with Domain III of EF-Tu than in the regions interacting with Domain I. To begin an analysis of the structure and organization of EF-Ts mt , a series of deletion derivatives of EF-Ts mt were constructed ŽFig. 9.. The abilities of these derivatives to interact with EF-Tu and to stimulate the activity of EF-Tu mt in polymerization were tested. The proteins produced from three of the deletion derivatives ŽSN, N and MC. could be purified under non-denaturing conditions and appear to be stably folded. The NM and NC proteins were degraded rapidly in vitro and could not be purified under non-denaturing conditions ŽFig. 9.. The rapid
111
degradation of these derivatives implies that they are not folded correctly resulting in a sensitivity to proteases. The construct lacking the middle region might have conceivably resulted in the synthesis of a derivative of EF-Ts mt closer to T. thermophilus EF-Ts than to E. coli EF-Ts based on primary sequence alignment alone. The observation that deletion of the middle of EF-Ts mt does not allow the production of a stably folded derivative suggests that this region is probably woven into the structure of other regions of EF-Ts as observed for the E. coli factor. The SN fragment could be purified to near homogeneity under non-denaturing conditions, suggesting the very NH 2-terminal region of EF-Ts mt has a relatively compact structure as suggested by the proteolytic mapping data. The abilities of the deleted derivatives of EF-Ts mt to stimulate the activity of EF-Tu mt in polyŽU. -directed polymerization of phenylalanine were also tested. The results of these studies indicated that none of these proteins had any effect on the activity of EF-Tu mt ŽFig. 9 and data not shown.. Unlike the normal EF-Ts mt , none of the deletion derivatives formed a stable complex with E. coli EF-Tu ŽFig. 9 and data not shown.. This analysis suggests that the tight interaction between EF-Ts mt and E. coli EF-Tu or EF-Tu mt does not reside in a single small region of the mitochondrial factor. Rather, this tight interaction appears to arise from numerous segments of the structure of EF-Ts mt acting in concert. Acknowledgements This work has been supported in part by funds provided by the National Institutes of Health Ž Grant GM32734., U.S.A. References w1x Ravel, J.M., Shorey, R.L., Garner, C.W., Dawkins, R.C. and Shive, W. Ž1969. Cold Spring Harbor Symposia Quant. Biol. 34, 321–330. w2x Bosch, L., Kraal, B., Van der Meide, P. and Van Noort, J. Ž1983. Prog. Nuc. Acid Res. Mol. Biol. 30, 91–126. w3x Miller, D. and Weissbach, H. Ž1970. Biochem. Biophys. Res. Commun. 38, 1016–1022. w4x Blank, J., Nock, S., Kreutzer, R. and Sprinzl, M. Ž1996. Eur. J. Biochem. 236, 222–227
112
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
w5x Arai, K., Ota, Y., Nakamura, S., Hennele, C., Oshima, T. and Kaziro, Y. Ž1978. Eur. J. Biochem. 92, 509–519. w6x Schwartzbach, C. and Spremulli, L. Ž1989. J. Biol. Chem. 264, 19125–19131. w7x Schwartzbach, C. and Spremulli, L. Ž1991. J. Biol. Chem. 266, 16324–16330. w8x Woriax, V., Burkhart, W. and Spremulli, L. Ž1995. Biochim. Biophys. Acta 1264, 347–356. w9x Xin, H., Burkhart, W. and Spremulli, L. Ž1995. J. Biol. Chem. 270, 17243–17249. w10x Eberly, S.L., Locklear, V. and Spremulli, L.L. Ž1985. J. Biol. Chem. 260, 8721–8725. w11x Graves, M., Breitenberger, C. and Spremulli, L. Ž1980. Arch. Biochem. Biophys. 204, 444–454. w12x Ravel, J. and Shorey, R.L. Ž1971. Methods Enzymol. 20C, 306–316. w13x Blum, H., Beier, H. and Gross, H.S. Ž1987. Electrophoresis 8, 93–99. w14x Arai, K., Kawakita, M. and Kaziro, Y. Ž1974. J. Biochem. 76, 293–306. w15x Arai, K., Arai, N., Nakamura, S., Oshima, T. and Kaziro, Y. Ž1978. Eur. J. Biochem. 92, 521–531. w16x Hachmann, J., Miller, D.L. and Weissbach, H. Ž1971. Arch. Biochem. Biophys. 147, 457–466. w17x Swart, G. and Parmeggiani, A. Ž1987. Biochemistry 26, 2047–2054. w18x Plomer, J.J. and Gafni, A. Ž1993. Biochim. Biophys. Acta 1163, 89–96. w19x Bogestrand, S., Wiborg, O., Thirup, S. and Nyborg, J. Ž1995. FEBS Lett. 368, 49–54.
w20x Kawashima, T., Berthet-Colominas, C., Wulff, M., Cusack, S. and Leberman, R. Ž1996. Nature 379, 511–518 w21x Jurnak, F. Ž1985. Science 230, 32–36. w22x Kjeldgaard, M., Nissen, P., Thirup, S. and Nyborg, J. Ž1993. Curr. Biol. 1, 35–49. w23x Berchtold, H., Reshetnikova, L., Reiser, C., Schirmer, N., Sprinzl, M. and Hilgenfeld, R. Ž1993. Nature 365, 126–132. w24x Kjeldgaard, M. and Nyborg, J. Ž1992. J. Mol. Biol. 223, 721–742. w25x Nissen, P., Kjeldgaard, M., Thirup, S., Polekhina, G., Reshetnikova, L., Clark, B. and Nyborg, J. Ž1995. Science 270, 1464–1472. w26x Peter, M., Reiser, C., Schirmer, N., Kiefhaber, T., Ott, G., Grillenbeck, N. and Sprinzl, M. Ž1990. Nuc. Acids Res. 18, 6889–6893. w27x Hwang, Y., Carter, M. and Miller, D. Ž1992. J. Biol. Chem. 267, 22198–22205. w28x Chevalier, C., Saillard, C. and Bove, J. Ž1990. J. Bact. 172, 2693–2703. w29x An, G., Bendiak, D.S., Mamelak, L.A. and Friesen, J.D. Ž1981. Nuc. Acids Res. 9, 4163–4172. w30x Sanangelantoni, A.M., Calogero, R.C., Buttarelli, F.R., Gualerzi, C. and Tiboni, O. Ž1990. FEMS Micro. Lett. 66, 141–146. w31x Kostrzewa, M. and Zetsche, K. Ž1993. Plant Mol. Biol. 23, 67–76. w32x Jiang, Y., Nock, S., Nesper, M., Sprinzl, M. and Sigler, P. Ž1996. Biochemistry 35, 10269–10278.
Expression of bovine mitochondrial elongation factor Ts in Escherichia coli and characterization of the heterologous complex formed with prokaryotic elongation factor Tu Hong Xin a , Karen Leanza a , Linda Lucy Spremulli
a,b,)
a
b
Department of Chemistry, UniÕersity of North Carolina, Campus Box 3290, Caphel Hill, NC 27599-3290, USA Lineberger ComprehensiÕe Cancer Research Center, UniÕersity of North Carolina, Campus Box 3290, Chapel Hill, NC 27599-3290, USA Received 3 October 1996; accepted 16 December 1996
Abstract When bovine mitochondrial elongation factor Ts ŽEF-Ts mt . is expressed in Escherichia coli, it forms a tightly associated complex with E. coli EF-Tu ŽEF-Tu Eco Ø Ts mt .. This complex is active in polyŽU.-directed polymerization and this activity is inhibited by kirromycin. The EF-Tu Eco Ø Ts mt complex does not bind guanine nucleotides detectably and is not dissociated to a significant extent by either GDP or GTP. A portion of the EF-Tu Eco Ø Ts mt complex can be dissociated by aa-tRNA in the presence of GTP. The heterologous complex cannot be dissociated completely in the presence of either the 8 M urea or 8 M guanidine hydrochloride, suggesting that EF-Ts mt has an unusually tight interaction with E. coli EF-Tu. The EF-Tu Eco Ø Ts mt complex can be dissociated by denaturation using 2 M guanidine thiocyanate. Free EF-Ts mt can then be purified and renatured. The refolded EF-Ts mt is active in stimulating the activity of expressed mitochondrial EF-Tu ŽEF-Tu mt . in polyŽU.-directed polymerization. Almost all the EF-Ts mt molecules appear to refold into a conformation which can interact with EF-Tu mt . Protease mapping of EF-Ts mt indicates that the first 54 residues fold into an independent domain. Analysis of deletion derivatives of EF-Ts mt indicates that extensive regions of this factor are required for its tight interaction with EF-Tu. q 1997 Elsevier Science B.V. All rights reserved. Keywords: Protein synthesis; Elongation; Mitochondrion; Organelle; Translation
1. Introduction Elongation factor Tu ŽEF-Tu. is responsible for promoting the binding of aminoacyl-tRNA Ž aa-tRNA.
Abbreviations: EF-Tu, elongation factor Tu; EF-Ts, elongation factor Ts; EF-Ts mt, mitochondrial elongation factor EF-Ts; EF-Tu Eco Ø Ts mt , heterologous complex between E. coli EF-Tu and EF-Ts mt ; aa-tRNA, aminoacyl-tRNA; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis. ) Corresponding author. Fax: q1 Ž919. 9663675; E-mail: [email protected]
to the A-site of the ribosome during protein biosynthesis w1x. EF-Tu forms a ternary complex with GTP and aa-tRNA prior to interaction with the ribosome. Following codon:anticodon interaction on the ribosome, EF-Tu is released in the form of an EF-Tu Ø GDP complex w2x. The release of GDP from EF-Tu is facilitated by a second elongation factor ŽEF-Ts.w3x. In E. coli, EF-Ts binds EF-Tu promoting the dissociation of GDP and the formation of an intermediate EF-Tu Ø Ts complex. This complex is, in turn, dissociated by GTP resulting in the formation of an EF-Tu Ø GTP complex which subsequently binds aa-tRNA
0167-4781r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved. PII S 0 1 6 7 - 4 7 8 1 Ž 9 7 . 0 0 0 0 3 - 1
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
forming a new ternary complex. In Thermus thermophilus, a dimeric complex ŽEF-Tu Ø Ts. 2 is formed by the interaction of two EF-Tu molecules with a stable EF-Ts dimer w4x. In contrast to the E. coli EF-Tu Ø Ts complex, the T. thermophilus complex is not dissociated to a significant extent by either GDP or GTP alone w5x. Mammalian mitochondrial EF-Tu and EF-Ts ŽEFTu mt and EF-Ts mt . have been purified as a tightly associated complex ŽEF-Tu Ø Ts mt . from bovine liver w6,7x. This complex has a number of features that distinguish it from the corresponding E. coli elongation factors. First, EF-Tu Ø Ts mt is quite stable and cannot be dissociated even in the presence of high concentrations of guanine nucleotides. In this respect, the mitochondrial factors differ significantly from the corresponding E. coli factors and share some resemblance to the thermophilic EF-Tu and EF-Ts. Second, no detectable amounts of intermediates equivalent to EF-Tu Ø GDP and EF-Tu Ø GTP can be observed in the mammalian mitochondrial system. Finally, significant binding of guanine nucleotides to EF-Tu Ø Ts mt can only be detected in the presence of aa-tRNA w7x. The cDNAs for human and bovine EF-Tu mt and EF-Ts mt have been cloned and sequenced w8,9x. The sequence of bovine EF-Tu mt is 56% identical to that of E. coli EF-Tu. This factor has been expressed in E. coli as a His-tagged protein. The purified factor is active in polyŽU.-directed polymerizationw8x. The sequence for EF-Ts mt is less than 30% identical to the sequences of the corresponding prokaryotic factors w9x. EF-Ts mt has also been expressed in E. coli as a His-tagged protein. When it is prepared from bacteria under native conditions, it is obtained as a 1:1 complex with E. coli EF-Tu. This heterologous complex is designated EF-Tu Eco Ø Ts mt w9x. The EF-Tu Eco Ø Ts mt complex is active in polymerization. In the current study, we have examined the properties of this heterologous complex and have developed procedures for the preparation of active EF-Ts mt free from EF-Tu. 2. Materials and methods 2.1. Materials Trypsin Žtype III. and Sephadex G50 were purchased from Sigma Chemical Company. Pure nitro-
103
cellulose blotting membranes were purchased from Schleicher and Schuell. Nitrocellulose filter paper ŽType HA, 0.45 m m. was from Millipore Corporation. w 3 HxGDP was from Dupont-New England Nuclear. Ni-NTA resin was from QIAGEN. Oligonucleotide primers were made in the Lineberger Comprehensive Cancer Research Center at the University of North Carolina. The pET expression system was from Novagen. E. coli ribosomes and elongation factors were purified as described before w10,11x. w 14 CxPhe-tRNAphe was prepared as described previously w12x. Expressed EF-Tu mt and the EF-Tu Ø Ts mt complex purified from bovine liver were kindly provided by Velinda Woriax ŽDepartment of Chemistry, University of North Carolina.. 2.2. Buffers Binding buffer: 50 mM Tris-HCl Ž pH 7.6., 60 mM KCl, 7 mM MgCl 2 , 7 mM b-mercaptoethanol, 0.1 mM phenylmethylsulfonylfluoride and 10% glycerol. Imidazole elution buffer: 50 mM Tris-HCl Ž pH 7.6., 40 mM KCl, 5 mM b-mercaptoethanol, 0.15 M imidazole and 10% glycerol. Buffer A: 20 mM Hepes-KOH ŽpH 7.0., 40 mM KCl, 1 mM MgCl 2 , 0.1 mM EDTA and 10% glycerol. High salt buffer: 50 mM Tris-HCl ŽpH 7.6., 1 M NH 4Cl, 5 mM b-mercaptoethanol, 10 mM imidazole and 10% glycerol. Denaturation buffer: 50 mM Tris-HCl ŽpH 7.6., 60 mM KCl, 0.1 M b-mercaptoethanol and the indicated amount of denaturing reagent. Renaturation Buffer: 50 mM Tris-HCl ŽpH 7.6., 60 mM KCl, 0.1 M b-mercaptoethanol. 2.3. Construction of plasmids for the expression of portions of EF-Tsmt PCR was used to add a NdeI cutting site to the 5X end and a XhoI cutting site to the 3X end of the bovine liver cDNA encoding various deletion mutants of EF-Ts mt . PCR amplification of the desired region of the EF-Ts mt gene was carried out as described before w9x. The cDNAs were then digested with NdeI and XhoI and cloned into the NdeI and XhoI sites of the pET24cŽq. vector. This vector provides an in-frame sequence coding for a histidinetag at the COOH-terminus. The resulting constructs were first transformed into E. coli DH5 a . Plasmid
104
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
DNA from each construct was prepared from 4 ml of cells using the QIAGEN mini-preparation kit. A portion Ž1r10. of each DNA preparation was then transformed into E. coli BL21ŽDE3. for expression. Expression was carried out as described previously w9x. Analysis of the expression of EF-Ts mt was examined under denaturing conditions and under native conditions as described w9x except that the protein inhibitor cocktail tablets ŽBoehringer Mannheim. were included in the buffers used for preparations made under native conditions. The EF-Tu Eco Ø Ts mt complex was purified as described w9x. Protein concentrations were determined by the Micro-Bradford method according to standard protocols ŽBio-Rad Inc... 2.4. Gel filtration analysis of GDP-binding E. coli EF-Tu Ž200 pmol., EF-Tu Eco Ø Ts mt Ž200 pmol, 15.0 m g., or a no protein control were incubated for 10 min at 378C in a reaction mixture Ž 250 m l. containing 10 m M w 3 HxGDP Ž1460 cpmrpmol., 10 mM MgCl 2 , 50 mM Tris-HCl Ž pH 7.8. , 1 mM dithiothreitol and 50 mM NH 4Cl. These samples were applied to a 15-cc Ž0.9 cm = 23 cm. Sephadex G-50 gel filtration column equilibrated in 50 mM Tris-HCl ŽpH 7.8., 10 mM MgCl 2 , 1 mM dithiothreitol and 50 mM NH 4Cl. The column was washed with the same buffer at 0.5 mlrmin, and 0.28-ml fractions were collected. Aliquots Ž0.15 ml. of various fractions were counted in 10 ml of Scintisafe Econo I liquid scintillant. 2.5. Effect of GDP, GTP and Phe-tRNA phe on the stability of EF-Tu Eco6Tsmt The EF-Tu Eco Ø Ts mt complex Ž100 m g. was mixed at 48C for 1 h with 50 m l of a 50% slurry of Ni-NTA resin equilibrated in binding buffer. The Ni-NTA resin was then put into a mini-column made from a yellow Eppendorf pipette tip and washed twice with 200 m l of buffer containing 10 mM Tris-HCl ŽpH 7.6., 10 mM MgCl 2 , 60 mM KCl, 1 mM DTT and 10% glycerol. The column was then washed with 50 m l of the buffer above containing 1 mM GDP or 1 mM GTP as indicated. The protein retained on the column was eluted with 5 aliquots of 50 m l Imidazole elution buffer. An aliquot of 4 m l of protein eluted at each step was analyzed by sodium dodecyl
sulfate polyacrylamide gel electrophoresis ŽSDSPAGE.. To test the effect of Phe-tRNAphe on EF-Tu Eco Ø Ts mt , 4.8 m g of complex Ž 50 pmol. was mixed with 60 m l of a 50% slurry of Ni-NTA resin at 48C for 1 h in binding buffer. The resin was washed with high salt buffer Ž400 m l. and buffer A Ž400 m l. to remove the unbound proteins, and was then poured to a mini-column made from a yellow Eppendorf pipette tip. A aliquot of 2 = binding buffer Ž30 m l. containing 6.7 m M Phe-tRNA and 0.66 mM GTP was added to the resin and incubated at 48C for 30 min. The resin was then washed with 100 m l of buffer containing 50 mM Tris-HCl ŽpH 7.6., 10 mM MgCl 2 , 60 mM KCl, 7 mM b-mercaptoethanol, 10% glycerol and 1 mM GTP where indicated. The protein retained on the column was eluted with three aliquots of 50 m l each of imidazole elution buffer. An aliquot Ž16 m l. of protein eluted at each step was analyzed by SDS-PAGE. 2.6. Denaturation of the EF-Tu Eco6Tsmt complex and renaturation of EF-Tsmt Small-scale denaturation experiments were carried out using 5 m g of the EF-Tu Eco Ø Ts mt complex in 200 m l denaturation buffer containing the indicated amount of the various denaturing reagents Ž3–8 M urea, 2–8 M guanidine-HCl, 1–6 M guanidine thiocyanate or 0.01–0.25% N-lauroylsarcosine. . These mixtures were incubated at room temperature for 20 min. The dissociation of the EF-Tu Eco Ø Ts mt complex was monitored by examining the retention of E. coli EF-Tu along with EF-Ts mt on a Ni-NTA column. For this analysis, 30 m l of a 50% slurry of Ni-NTA resin was added to the samples above and the mixtures were incubated at room temperature for one h. The resin was then loaded into a mini-column made from a yellow Eppendorf tip. The resin was washed twice with 400 m l of the starting buffer. Proteins retained on the column were eluted with 50 m l imidazole elution buffer, and 16 m l of the eluted material was analyzed by SDS-PAGE. Large-scale denaturation of the EF-Tu Eco Ø Ts mt complex and renaturation of EF-Ts mt began with the complex obtained from 4 g of cells Žfrom 1 liter cell culture. . The initial steps followed the strategy outlined for the purification of the EF-Tu Eco Ø Ts mt com-
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
plex described previously w9x. Once the EF-Tu Eco Ø Ts mt complex was bound to the Ni-NTA resin, the resin was placed in an empty 15-cc column and washed with high salt buffer Ž60 ml.. Denaturation buffer Ž10 ml. containing 2 M guanidine thiocyanate was added to the column and the resin was incubated with this buffer at room temperature for 20 min. The column was then washed with 15 ml denaturation buffer containing 2 M guanidine thiocyanate to remove E. coli EF-Tu. The EF-Ts mt retained on the column was renatured by mixing the resin with 10 ml renaturation buffer followed by washing with 50 ml of dilution buffer containing 50 mM Tris-HCl ŽpH 7.6. and 60 mM KCl. EF-Ts mt was eluted with 0.5 ml imidazole elution buffer and dialyzed against buffer containing 10 mM Tris-HCl ŽpH 7.6., 40 mM KCl, 7 mM MgCl 2 and 10% glycerol for 1.5 h with three changes of the solution. 2.7. Trypsin digestion of free EF-Tsmt and the EFTu Eco6Tsmt complex A stock solution of trypsin Ž10 mgrml in 50 mM Tris-HCl, pH 7.8. was prepared one day before use and kept at 48C overnight. A series of 10-fold dilutions was made immediately before use in 50 mM Tris-HCl ŽpH 7.8.. The EF-Tu Eco Ø Ts mt complex Ž5 m g in 2 m l of buffer A. or free EF-Ts mt Ž5 m g in 2.3 m l of buffer A. was added into 10 m l containing the indicated amount of trypsin and incubated at 278C for 15 min. Reactions were stopped by the addition of 4 m l buffer A and 4 m l of buffer containing 63 mM Tris-HCl Ž pH 6.8., 0.1 M b-mercaptoethanol, 2% SDS, 10% glycerol, 0.5% bromophenol blue, and then heated at 908C for 10 min. Samples Ž6 m l. were analyzed by SDS-PAGE and stained with silver w13x. 2.8. Poly(U)-directed polymerization assay The activity of EF-Tu Eco Ø Ts mt was measured by its ability to catalyze the polyŽ U. -directed polymerization of phenylalanine on E. coli ribosomes as described previously w6x. The activity of renatured EF-Ts mt was detected using the polymerization assay described above with varying amounts of expressed protein and 3 pmol of EF-Tu mt expressed in E. coli w8x. The effect of kirromycin on the polymerization activity of the EF-Tu Eco Ø Ts mt complex was tested
105
by adding different amounts of kirromycin Ž5 = 10y6–4 = 10y9 M. to each reaction. The percentage of activity of each reaction was calculated in comparison with a positive control without kirromycin.
3. Results and discussion 3.1. Interactions of the EF-Tu Eco6Tsmt complex with guanine nucleotides The E. coli EF-Tu Ø Ts complex can be readily dissociated by either GDP or GTP w14x. In contrast, neither the T. thermophilus ŽEF-Tu Ø Ts. 2 complex nor the bovine liver EF-Tu Ø Ts mt complex can be dissociated to a significant extent by high concentrations of guanine nucleotides w6,7,15x. To compare the EF-Tu Eco Ø Ts mt complex with E. coli EF-Tu Ø Ts and with native EF-Tu Ø Ts mt purified from bovine liver, the basic features of this heterologous complex were investigated. These features included its ability to bind guanine nucleotides and its susceptibility to dissociation by GTP, GDP and aa-tRNA. The ability of EF-Tu to bind guanine nucleotides is generally measured using a nitrocellulose filter binding assay w16x. The E. coli EF-Tu Ø Ts complex is readily dissociated in the presence of guanine nucleotides. As a result, incubation of E. coli EF-Tu Ø Ts with guanine nucleotides results in the formation of E. coli EF-Tu Ø GDP or EF-Tu Ø GTP complexes which are readily detected by a nitrocellulose filter binding assay when labeled nucleotides are used. In contrast, very little guanine nucleotide binding can be detected when EF-Tu Eco Ø Ts mt is incubated with labeled GDP or GTP, suggesting that the heterologous complex is not being dissociated to a significant extent by guanine nucleotides Ždata not shown.. This result also suggests that the EF-Tu Eco Ø Ts mt complex does not bind to guanine nucleotides directly. However, it was possible that the heterologous complex bound GDP but was not retained by the nitrocellulose membrane. To examine this possibility, GDP-binding was analyzed using a gel filtration assay. As indicated in Fig. 1, the binding of w 3 HxGDP to E. coli EF-Tu can be readily detected by this method. In contrast, no binding of GDP to the EF-Tu Eco Ø Ts mt complex can be observed. These observations suggest that the heterologous complex does not bind guanine
106
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
dure in the absence of guanine nucleotides. With buffers containing 1 mM GDP, only a trace of dissociation of the EF-Tu Eco Ø Ts mt complex was observed ŽFig. 2B. . No release of E. coli EF-Tu was observed in the presence of GTP ŽFig. 2C.. These results indicate that the EF-Tu Eco Ø Ts mt complex resembles the native bovine mitochondrial EF-Tu Ø Ts mt complex which cannot be readily dissociated by either GDP or GTP. The effect of aa-tRNA on the EF-Tu Eco Ø Ts mt
Fig. 1. Use of gel filtration to examine nucleotide binding to the EF-Tu Eco Ø Ts mt complex. EF-Tu Eco Ø Ts mt was incubated with w 3 HxGDP and nucleotide binding was examined by following the elution position of the labeled nucleotide during chromatography on Sephadex G50 Ž`.. As a control E. coli EF-Tu was incubated with w 3 HxGDP and nucleotide binding was examined by chromatography on Sephadex G50 ŽB.. Chromatography of w 3 HxGDP alone on the Sephadex G50 column is also shown ŽI..
nucleotides to a significant extent and is probably not readily dissociated by guanine nucleotides. The EFTu Ø Ts mt complex purified from bovine liver also fails to bind guanine nucleotides at a detectable level, possibly because it is not easily dissociated by either GDP or GTP w6x. The ability of GDP and GTP to promote the release of E. coli EF-Tu from the heterologous EFTu Eco Ø Ts mt complex was then tested directly. For these experiments, EF-Tu Eco Ø Ts mt was bound to a Ni-NTA affinity column, then washed with buffers lacking guanine nucleotides or with buffers containing 1 mM GDP or 1 mM GTP. As shown in Fig. 2A, no dissociation of the heterologous complex is observed when the complex is subjected to this proce-
Fig. 2. Failure of GDP or GTP to dissociate the EF-Tu Eco Ø Ts mt complex. The EF-Tu Eco Ø Ts mt complex was bound to Ni-NTA resin through the His-tag on EF-Ts mt as outlined in Section 2. A: The resin was washed with aliquots of a control buffer lacking guanine nucleotides Žlanes 1 and 2. and then eluted with buffer containing imidazole Žlanes 3 and 4.. Proteins present in these samples were analyzed by SDS-PAGE for the presence of E. coli EF-Tu and EF-Ts mt using silver staining. B: The resin was washed with buffer containing 1 mM GDP Žlanes 1 and 2.. The protein still remaining on the column was then eluted with 2 aliquots of a buffer containing imidazole Žlanes 3 and 4.. Aliquots of several fractions were analyzed by SDS-PAGE for the presence of E. coli EF-Tu and EF-Ts mt . C: The resin was washed with buffer containing 1 mM GTP Žlanes 1 and 2.. The protein still remaining on the column was then eluted with buffer containing imidazole Žlanes 3 and 4.. Aliquots of various fractions were analyzed by SDS-PAGE for the presence of E. coli EF-Tu and EF-Ts mt .
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
107
3.2. Effect of kirromycin on the EF-Tu Eco6Tsmt complex
Fig. 3. Effect of GTP and Phe-tRNA on the interaction between E. coli EF-Tu and EF-Ts mt . An aliquot of the EF-Tu Eco Ø Ts mt complex was bound to Ni-NTA resin through the His-tag on EF-Ts mt as outlined in Section 2. A: The resin was washed with buffer containing Phe-tRNA but lacking GTP Žlanes 1 and 2. and then eluted sequentially with aliquots of buffer containing imidazole Žlanes 3 and 4.. Proteins present in these samples were analyzed by SDS-PAGE for the presence of E. coli EF-Tu and EF-Ts mt . B: The resin was washed with buffer containing both Phe-tRNA and GTP Žlanes 1 and 2.. The protein still remaining on the column was then eluted with buffer containing imidazole Žlanes 3 and 4.. An aliquot of various fractions was analyzed by SDS-PAGE for the presence of E. coli EF-Tu and EF-Ts mt .
complex was investigated using a similar strategy. The EF-Tu Eco Ø Ts mt complex was bound to the NiNTA affinity column, followed by the incubation with Phe-tRNAphe in the presence or absence of GTP. As shown in Fig. 3, close to 30% of the E. coli EF-Tu was dissociated from EF-Ts mt under these conditions. No dissociation of the complex was observed when GTP was removed from the buffers ŽFig. 3A.. This observation suggests that both aatRNA and GTP must be present in order to promote the dissociation of a significant amount of the EFTu Eco Ø Ts mt complex. The order of binding of the GTP and aa-tRNA cannot be determined with the approach used here. However, in the presence of aa-tRNA, the E. coli EF-Tu Ø GTP complex binds aa-tRNA forming the ternary complex. Coupling these two reactions allows us to detect the net dissociation of the EF-Tu Eco Ø Ts mt complex.
The binding of kirromycin to E. coli EF-Tu is mutually exclusive with EF-Ts w17x. The polymerization activity of E. coli EF-Tu Ø Ts is inhibited by kirromycin, since this antibiotic freezes EF-Tu on the ribosome in a GTP-like conformation blocking the elongation cycle. The effect of kirromycin on the EF-Tu Eco Ø Ts mt complex was tested by examining the concentration of this antibiotic required to inhibit the activity of the complex in polyŽ U. -directed polymerization. As shown in Fig. 4, the activity of EFTu Eco Ø Ts mt can be inhibited by kirromycin at a concentration very similar to that required for the inhibition of the homologous E. coli EF-Tu Ø Ts complex. This result suggests that the ternary complex ŽEF-Tu Ø GTP Ø aa-tRNA. forms during the elongation cycle, and that EF-Ts mt dissociates from E. coli EF-Tu during each round of elongation. The bovine EF-Tu Ø Ts mt complex is not inhibited by kirromycin w6x. Two explanations had been suggested for this observation w8x. First, it was possible
Fig. 4. Effect of kirromycin on the activity of the EF-Tu Eco Ø Ts mt complex in polyŽU.-directed polymerization. Reaction mixtures were prepared as described w9x and contained 78 ng of the EF-Tu Eco Ø Ts mt complex ŽB. or 35 ng of the E. coli EF-Tu Ø Ts complex Ž`. and the indicated amount of kirromycin. Control reactions containing no kirromycin resulted in the polymerization of 6.3 and 4.3 pmol of phenylalanine for the EF-Tu Eco Ø Ts mt complex and E. coli EF-Tu Ø Ts complex respectively and were used to define the 100% points.
108
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
that the tight interaction of EF-Ts mt with EF-Tu altered the basic features of the elongation cycle so that kirromycin could no longer freeze EF-Tu on the ribosome. A second explanation was that EF-Tu mt could not bind kirromycin due to the presence of a threonine residue at position 380 Žequivalent to A375 in E. coli EF-Tu.. Mutation of this residue in prokaryotic EF-Tu renders the factor resistant to kirromycin. The observations made here suggest that the ability of EF-Ts mt to bind tightly to EF-Tu does not preclude sensitivity of the elongation cycle to kirromycin. Thus, the kirromycin resistance of mitochondrial protein synthesis is an intrinsic property of EF-Tu mt . 3.3. Dissociation of EF-Tsmt from E. coli EF-Tu by denaturing reagents The effect of denaturing reagents on the dissociation of the EF-Tu Eco Ø Ts mt complex was examined using several concentrations of different denaturing reagents. EF-Ts mt could not be released completely from E. coli EF-Tu by urea even at concentrations of 8 M Ž Fig. 5A.. Guanidine hydrochloride was somewhat more effective than urea. However, even 8 M guanidine hydrochloride could not completely dissociate this complex ŽFig. 5B.. These results suggest that the heterologous complex is unusually stable and that the interaction between E. coli EF-Tu and EFTs mt is quite tight. In the presence of 2 M guanidine thiocyanate Ž Fig. 5C. or 0.2% N-lauroylsarcosine ŽFig. 5D., EF-Ts mt could be almost completely released from E. coli EF-Tu. EF-Ts mt could not be recovered from the Ni-NTA affinity column at concentrations of guanidine thiocyanate higher than 4 M, suggesting that higher concentrations of this denaturing reagent might cause the release of the His-tagged protein from the resin or might damage the Ni-NTA resin. Once conditions had been established for the denaturation of the EF-Tu Eco Ø Ts mt complex, procedures were developed for the large-scale preparation of EF-Ts mt . The EF-Tu Eco Ø Ts mt complex was bound to the Ni-NTA affinity column. The complex was denatured by treatment with buffer containing 2 M guanidine thiocyanate and E. coli EF-Tu was removed from the column by washing with the denaturing solution. The EF-Ts mt retained on the resin was
Fig. 5. Denaturation of the EF-Tu Eco Ø Ts mt complex. The EFTu Eco Ø Ts mt complex Ž5 m g. was treated with the indicated concentrations of urea Žpanel A., guanidine hydrochloride Žpanel B., guanidine thiocyanate Žpanel C. or N-lauroylsarcosine Žpanel D.. The EF-Ts mt component and any associated EF-Tu was recovered by Ni-NTA chromatography. The material retained by the resin was eluted with imidazole and an aliquot was analyzed by SDS-PAGE followed by silver staining. The higher molecular weight band is EF-Tu while the lower molecular weight band is EF-Ts.
renatured by fast dilution of the resin to reduce the concentration of the denaturing reagent w18x. Refolded EF-Ts mt was then eluted with imidazole. About 400–600 m g of EF-Ts mt could be purified from one liter of E. coli cell culture by this procedure. This value represents a yield of 20–30% of the EF-Ts mt in the starting complex. The major loss of the protein is thought to happen in the renaturation step, in which 0.1 M b-mercaptoethanol is used to avoid the improper formation of disulfide bonds. High concentrations of b-mercaptoethanol may damage the Ni-NTA resin, resulting in the premature release of EF-Ts mt . Analysis of the purified EF-Ts mt on SDS-PAGE indicates that it is free of E. coli EF-Tu ŽFig. 6.. A small amount of a low molecular weight impurity can be observed in these preparations. This component is probably a contaminant from the original preparation since it does not cross-react with antibodies raised
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
109
Fig. 6. Purity of EF-Ts mt . Lane 1: SDS-PAGE analysis of the EF-Tu Eco Ø Ts mt complex Ž0.5 m g.. Lane 2: The EF-Tu Eco Ø Ts mt complex Ž ; 4 mg. bound on the Ni-NTA column was denatured by incubation with 2 M guanidine thiocyanate as described in Section 2. The E. coli EF-Tu component was washed from the column. The remaining EF-Ts mt was renatured on the Ni-NTA resin and then eluted with imidazole. An aliquot of the renatured protein Ž0.18 m g. was analyzed by SDS-PAGE.
against EF-Tu Ø Ts mt from bovine liver Ždata not shown.. The activity of the renatured EF-Ts mt was evaluated by its ability to stimulate the activity of EF-Tu mt in the polyŽ U.-directed polymerization. EF-Ts mt enhances the activity of EF-Tu mt by 4- to 6-fold ŽFig. 7.. A molar ratio of EF-Tu mt to EF-Ts mt required for maximal activity Žslightly less than 1:1. suggests that a considerable fraction of the renatured EF-Ts mt has refolded into a conformation that can interact with EF-Tu mt . 3.4. Proteolytic digestion of EF-Tsmt T. thermophilus EF-Ts has a trypsin-sensitive site at Lys48 and cleavage at this position results in the
Fig. 7. Stimulation of the activity of EF-Tu mt by renatured EF-Ts mt . The indicated amount of renatured EF-Ts mt was used to stimulate the activity of EF-Tu mt Ž144 ng. in polyŽU.-directed polymerization.
Fig. 8. Trypsin digestion of free EF-Ts mt and the EF-Tu Eco Ø Ts mt complex. A: EF-Ts mt Ž5 m g. was digested with the indicated amount of trypsin and the resulting products were analyzed by SDS-PAGE. Lane 1, no trypsin; lane 2, 0.01 ng; lane 3, 0.1 ng; lane 4, 1 ng; lane 5, 10 ng. The smaller protein whose position is indicated by the arrow is a contaminant in the particular EF-Ts mt preparation used here. It does not cross-react with antibodies raised against the EF-Tu Ø Ts mt complex Ždata not shown.. B: The EF-Tu Eco Ø Ts mt complex Ž5 m g. was digested with the indicated amount of trypsin, followed by the analysis of the samples on SDS-PAGE. Lane 1, no trypsin; lane 2, 0.01 ng; lane 3, 0.1 ng; lane 4, 1 ng; lane 5, 10 ng.
formation of a large COOH-terminal fragment w4x. This COOH-terminal fragment cannot interact tightly with EF-Tu. E. coli EF-Ts has similar trypsin-sensitive sites at Lys51 and Lys52 w19x. The large COOHterminal fragment released by trypsin has very low activity in promoting guanine nucleotide exchange with EF-Tu. As shown in Fig. 8, when free EF-Ts mt is digested with a low level of trypsin, a single major cut site was observed. The resulting product from EF-Ts mt has a molecular mass of about 30 kDa. This fragment is retained by Ni-NTA resins and must, therefore, be derived from the COOH-terminal end which carries the His-tag. These data, along with comparative sequence analysis suggest that the cleavage observed most likely occurs at Arg54 leaving a large COOH-terminal fragment. At higher concentrations of trypsin, EF-Ts mt is degraded rapidly, indicating that many of the 29 potential trypsin cleavage sites can be attacked when enough of the protease is present Ždata not shown. .
110
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
The effect of trypsin digestion on the EF-Tu Eco Ø Ts mt complex was then examined ŽFig. 8. . The single major cut of EF-Ts mt was observed following digestion of the complex, suggesting that this trypsin-sensitive site is located on the surface of EF-Ts mt both when it is free in solution and when it is present in the complex with E. coli EF-Tu. This idea is compatible with the crystal structure of the E. coli EF-Tu Ø Ts complex w20x. EF-Tu in the EF-Tu Eco Ø Ts mt complex was also cleaved producing a single major band. The cleavage in EF-Tu probably occurs in the effector region around Arg58 and the major band observed is derived from the COOH-terminus of EF-Tu w21x. 3.5. Deletion mutations of EF-Tsmt The resistance of the EF-Tu Ø Ts mt and the EFTu Eco Ø Ts mt complexes to dissociation by guanine
nucleotides raises the question of the nature of the interactions between these two factors. The three-dimensional structure of EF-Tu is known in both the GTP and GDP bound conformations w22–24x. EF-Tu folds into three domains ŽI, II and III.. Domain I binds guanine nucleotides while all three domains participate in binding aa-tRNA w25x. The three-dimensional structure of the E. coli EF-Tu Ø Ts complex shows that EF-Ts interacts with both Domain I and Domain III of EF-Tu w20x. This observation is in agreement with previous results w26,27x. An alignment of the sequences of EF-Ts from prokaryotes and organelles suggests that the primary sequence of EF-Ts mt can be divided into three regions ŽFig. 9. . The first 96 residues Žthe N-segment. are the most highly conserved region among the sequences known. The 3-dimensional structure of E. coli EF-Ts indicates that first 50–60 residues of the
Fig. 9. Deletion mutants of EF-Ts mt . A: The primary sequence of EF-Ts mt was divided into three regions according to the sequence similarity scores relative to EF-Ts from different organisms. B: Diagrammatic representation of deletion derivatives prepared from EF-Ts mt . The His-tag on the mutated proteins is indicated by the solid ellipse. The dashed line indicates that the middle section is not present in the native protein of T. thermophilus EF-Ts ŽT. ther EF-Ts., or has been deleted from of EF-Ts mt ŽTs-NC.. The table on the right summarizes the properties of the derivatives. The stability was assessed by the sensitivity of the derivative to proteases as measured by the ability to isolate the intact protein from cell extracts under native conditions. The ability of the derivative to interact with E. coli EF-Tu was determined by co-retention on Ni-NTA resins. The activity was measured by the ability to stimulate the activity of EF-Tu mt in polyŽU.-directed polymerization.
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
N-segment fold into an independent domain w20x. The middle section Žresidues 97 to 190. is referred to as the middle Ž M.-segment This region is present in the mammalian EF-Ts mt , E. coli, S. platensis, and S. citri EF-Ts w28–30x. It is not present in T. thermophilus EF-Ts or in chloroplast EF-Ts from G. sulphuraria w4,31x. The last 93 residues Ž191–283. are referred to as the C-segment. In E. coli EF-Ts, secondary structural elements from the M-segment and the C-segment are interwoven in 3 dimensions w20x. A comparison of the 3-dimensional structure of E. coli EF-Ts and the dimeric form of EF-Ts from T. thermophilus w20,32x indicates that the two structures have both structural similarities and differences. The dimer interface in T. thermophilus is formed by two three-stranded antiparallel b-sheets. A similar fold is observed internally in E. coli EF-Ts. Thus, in many ways, the dimeric form of EF-Ts found in T. thermophilus is similar to a monomer unit of E. coli EF-Ts. EF-Ts mt is the same size as E. coli EF-Ts and is likely to form the internal b-sheet interface observed in the E. coli factor. EF-Ts mt is 29% identical to E. coli EF-Ts in primary sequence. Alignment of the sequences of these two proteins is somewhat tenuous in places due to the low conservation of primary sequence. Analysis of their alignment indicates that there will be both similarities and differences between these two factors. The precise positions of insertions and deletions in the alignment varies with different alignment programs and as a function of the set of sequences used in the alignment Žsee Refs. w4x and w9x.. An examination of individual residues of E. coli EF-Ts and EF-Ts mt suggests that more differences will be found in the regions of EF-Ts that interact with Domain III of EF-Tu than in the regions interacting with Domain I. To begin an analysis of the structure and organization of EF-Ts mt , a series of deletion derivatives of EF-Ts mt were constructed ŽFig. 9.. The abilities of these derivatives to interact with EF-Tu and to stimulate the activity of EF-Tu mt in polymerization were tested. The proteins produced from three of the deletion derivatives ŽSN, N and MC. could be purified under non-denaturing conditions and appear to be stably folded. The NM and NC proteins were degraded rapidly in vitro and could not be purified under non-denaturing conditions ŽFig. 9.. The rapid
111
degradation of these derivatives implies that they are not folded correctly resulting in a sensitivity to proteases. The construct lacking the middle region might have conceivably resulted in the synthesis of a derivative of EF-Ts mt closer to T. thermophilus EF-Ts than to E. coli EF-Ts based on primary sequence alignment alone. The observation that deletion of the middle of EF-Ts mt does not allow the production of a stably folded derivative suggests that this region is probably woven into the structure of other regions of EF-Ts as observed for the E. coli factor. The SN fragment could be purified to near homogeneity under non-denaturing conditions, suggesting the very NH 2-terminal region of EF-Ts mt has a relatively compact structure as suggested by the proteolytic mapping data. The abilities of the deleted derivatives of EF-Ts mt to stimulate the activity of EF-Tu mt in polyŽU. -directed polymerization of phenylalanine were also tested. The results of these studies indicated that none of these proteins had any effect on the activity of EF-Tu mt ŽFig. 9 and data not shown.. Unlike the normal EF-Ts mt , none of the deletion derivatives formed a stable complex with E. coli EF-Tu ŽFig. 9 and data not shown.. This analysis suggests that the tight interaction between EF-Ts mt and E. coli EF-Tu or EF-Tu mt does not reside in a single small region of the mitochondrial factor. Rather, this tight interaction appears to arise from numerous segments of the structure of EF-Ts mt acting in concert. Acknowledgements This work has been supported in part by funds provided by the National Institutes of Health Ž Grant GM32734., U.S.A. References w1x Ravel, J.M., Shorey, R.L., Garner, C.W., Dawkins, R.C. and Shive, W. Ž1969. Cold Spring Harbor Symposia Quant. Biol. 34, 321–330. w2x Bosch, L., Kraal, B., Van der Meide, P. and Van Noort, J. Ž1983. Prog. Nuc. Acid Res. Mol. Biol. 30, 91–126. w3x Miller, D. and Weissbach, H. Ž1970. Biochem. Biophys. Res. Commun. 38, 1016–1022. w4x Blank, J., Nock, S., Kreutzer, R. and Sprinzl, M. Ž1996. Eur. J. Biochem. 236, 222–227
112
H. Xin et al.r Biochimica et Biophysica Acta 1352 (1997) 102–112
w5x Arai, K., Ota, Y., Nakamura, S., Hennele, C., Oshima, T. and Kaziro, Y. Ž1978. Eur. J. Biochem. 92, 509–519. w6x Schwartzbach, C. and Spremulli, L. Ž1989. J. Biol. Chem. 264, 19125–19131. w7x Schwartzbach, C. and Spremulli, L. Ž1991. J. Biol. Chem. 266, 16324–16330. w8x Woriax, V., Burkhart, W. and Spremulli, L. Ž1995. Biochim. Biophys. Acta 1264, 347–356. w9x Xin, H., Burkhart, W. and Spremulli, L. Ž1995. J. Biol. Chem. 270, 17243–17249. w10x Eberly, S.L., Locklear, V. and Spremulli, L.L. Ž1985. J. Biol. Chem. 260, 8721–8725. w11x Graves, M., Breitenberger, C. and Spremulli, L. Ž1980. Arch. Biochem. Biophys. 204, 444–454. w12x Ravel, J. and Shorey, R.L. Ž1971. Methods Enzymol. 20C, 306–316. w13x Blum, H., Beier, H. and Gross, H.S. Ž1987. Electrophoresis 8, 93–99. w14x Arai, K., Kawakita, M. and Kaziro, Y. Ž1974. J. Biochem. 76, 293–306. w15x Arai, K., Arai, N., Nakamura, S., Oshima, T. and Kaziro, Y. Ž1978. Eur. J. Biochem. 92, 521–531. w16x Hachmann, J., Miller, D.L. and Weissbach, H. Ž1971. Arch. Biochem. Biophys. 147, 457–466. w17x Swart, G. and Parmeggiani, A. Ž1987. Biochemistry 26, 2047–2054. w18x Plomer, J.J. and Gafni, A. Ž1993. Biochim. Biophys. Acta 1163, 89–96. w19x Bogestrand, S., Wiborg, O., Thirup, S. and Nyborg, J. Ž1995. FEBS Lett. 368, 49–54.
w20x Kawashima, T., Berthet-Colominas, C., Wulff, M., Cusack, S. and Leberman, R. Ž1996. Nature 379, 511–518 w21x Jurnak, F. Ž1985. Science 230, 32–36. w22x Kjeldgaard, M., Nissen, P., Thirup, S. and Nyborg, J. Ž1993. Curr. Biol. 1, 35–49. w23x Berchtold, H., Reshetnikova, L., Reiser, C., Schirmer, N., Sprinzl, M. and Hilgenfeld, R. Ž1993. Nature 365, 126–132. w24x Kjeldgaard, M. and Nyborg, J. Ž1992. J. Mol. Biol. 223, 721–742. w25x Nissen, P., Kjeldgaard, M., Thirup, S., Polekhina, G., Reshetnikova, L., Clark, B. and Nyborg, J. Ž1995. Science 270, 1464–1472. w26x Peter, M., Reiser, C., Schirmer, N., Kiefhaber, T., Ott, G., Grillenbeck, N. and Sprinzl, M. Ž1990. Nuc. Acids Res. 18, 6889–6893. w27x Hwang, Y., Carter, M. and Miller, D. Ž1992. J. Biol. Chem. 267, 22198–22205. w28x Chevalier, C., Saillard, C. and Bove, J. Ž1990. J. Bact. 172, 2693–2703. w29x An, G., Bendiak, D.S., Mamelak, L.A. and Friesen, J.D. Ž1981. Nuc. Acids Res. 9, 4163–4172. w30x Sanangelantoni, A.M., Calogero, R.C., Buttarelli, F.R., Gualerzi, C. and Tiboni, O. Ž1990. FEMS Micro. Lett. 66, 141–146. w31x Kostrzewa, M. and Zetsche, K. Ž1993. Plant Mol. Biol. 23, 67–76. w32x Jiang, Y., Nock, S., Nesper, M., Sprinzl, M. and Sigler, P. Ž1996. Biochemistry 35, 10269–10278.