Expression of Escherichia coli Methionyl-tRNA Formyltransferase in Saccharomyces cerevisiae Leads to Formylation of the Cytoplasmic Initiator tRNA and Possibly to Initiation of Protein Synthesis with Formylmethionine

Strains and media. S. cerevisiae CKY473 (Gal⁺ ura3-52 leu2-3,112) was from C. Kaiser, Massachusetts Institute of Technology, Cambridge; UMY 543 (MATα ura3-52 trpΔ1 imt::TRP1 imt3::TRP1 imt4::TRP1) (4) was from A. S. Bystrom, Umea University, Umea, Sweden; EKY3 (ser1 ura3-52 trp1 leu2 his4 shm1::URA3 shm2::LEU2) (17) was from D. R. Appling, University of Texas at Austin, and JEL1 (MATα leu2 trp1 ura3-52 prb1-1122 pep4 Δhis3::PGAL1-GAL4) (27) was from J. C. Wang, Harvard University, E. coli XL1-Blue (recA1 endA1 gyr96 thi-1 hsdR17 supE44 relA1 lac [F′ proAB lacI^q ZΔM15 Tn10 (Tet^r)]) from Stratagene was used as the routine strain for cloning. Yeast strains were grown in YPD (1% Bacto yeast extract, 2% Bacto peptone, 2% dextrose) or in synthetic minimal medium SMRG (0.67% yeast nitrogen base without amino acids, 0.5% Casamino Acids, 2% raffinose, and 3% glycerol). For induction, the cells were grown in SMRG supplemented with 2% galactose. Plasmids containing URA3 and LEU2 markers were selected by growing in medium lacking uracil and leucine as required. The strain EKY3 was grown in medium containing 0.7% yeast nitrogen base without amino acids, 2% galactose (or dextrose), 375 mg of serine per liter, 20 mg of tryptophan per liter, 20 mg of histidine per liter, and 40 mg of leucine per liter. When required, potassium formate (pH 5.0) was supplemented to a final concentration of 100 mg/liter.

Construction of plasmids.The E. coli MTF gene was cloned in the yeast expression vector YEp352-GAL1 (37). The wild-type and mutant MTF genes were amplified by PCR with the plasmids pQE16FMTp and pQE16FMTpR42L (44, 45) as templates and the oligonucleotides 5′-ATCGGATCCAAAAAAAATGTCAGAATCACTACGTATTATT-3′ containing a BamHI site as the forward primer and 5′-CTCAGCTAATTAAGCTTAGTG-3′ containing a HindIII site as the reverse primer. The PCR products coding for MTF-six-His fusion proteins were digested with BamHI and HindIII and cloned into the respective sites of YEp352-GAL1 to obtain YEp.MTF WT (wild type) and YEp.MTF R42L. The gene for E. coli DEF was amplified by PCR with genomic DNA isolated from E. coli JM109 as a template and 5′-AGAGCTCGAGAAAAAAAATGTCAGTTTTGCAAGTGTTACATG-3′ containing a site for XhoI as the forward primer and 5′-ACTAGCATGCTTAGTGATGGTGATGGTGATGAGCCCGGGCTTTCAGACG-3′ coding for the six-His tag and containing a site for SphI as the reverse primer. The PCR product was digested with XhoI and SphI and cloned into the respective sites of plasmid pAS565, a pRS315-based vector (51). A NotI-ApaI fragment from this plasmid containing the GAL1 promoter and the DEF gene was then cloned into the 2μm vector pRS425 (7) at these sites to obtain pRS425.DEF. The E133A mutation in the E. coli DEF gene was introduced into this plasmid by the Quikchange mutagenesis procedure with the Pfu DNA polymerase (Stratagene) to generate pRS425.DEF E133A.

Immunoblot analysis.Cells from 1.5 ml of culture were collected and resuspended in 50 μl of water followed by the addition of an equal volume of lysis buffer (4% sodium dodecyl sulfate [SDS], 10 mM EDTA) and glass beads. The suspension was mixed vigorously for 2 min with a Vortex mixer, kept in boiling water for 5 min, and mixed again for 2 min. After centrifugation, the protein content in the clarified supernatant was estimated by using the Micro BCA (bicinchoninic acid) protein assay reagent kit (Pierce). Proteins (15 μg per lane) were resolved on SDS-12% polyacrylamide gels (22) and transferred onto Immobilon polyvinylidene difluoride (PVDF) membrane (Millipore) with a buffer (pH 8.3) containing 50 mM Tris, 40 mM glycine, 0.04% SDS, and 20% methanol. The membrane was blocked in 5% nonfat dry milk in TBS (20 mM Tris-HCl [pH 7.5], 140 mM NaCl, and 0.05% Tween 20) at room temperature for 60 min. The membrane was then washed three times for 10 min each with TBS and incubated with a monoclonal antibody against the four-His epitope (Qiagen; 1:2,000 dilution) for 60 min at room temperature. The membrane was subsequently washed three times for 10 min each and incubated with peroxidase-conjugated secondary anti-mouse antibody (New England Biolabs; 1:3,000 dilution) for 1 h at room temperature. After three washes for 10 min each, the bound antibodies were detected with enhanced chemiluminescence reagents (New England Biolabs). In some experiments, the membrane was also probed with polyclonal antiserum against MTF as described previously (26).

Isolation of total tRNA.A single colony of S. cerevisiae CKY473 was grown in 10 ml of YPD to saturation. Five milliliters of this culture was inoculated into 100 ml of YPD and grown for 24 h at 30°C. The cells were harvested and resuspended in 40 ml of 10 mM Tris-HCl (pH 8.0) and 1 mM EDTA (TE). An equal volume of phenol (equilibrated with TE) was added, and the contents were mixed vigorously in a shaker at room temperature for 30 min. The phases were separated by centrifugation, and the nucleic acids were recovered from the aqueous phase by ethanol precipitation. The pellet was washed with 70% ethanol and dissolved in 2 ml of TE. High-molecular-weight nucleic acids were removed by the addition of 0.5 ml of 5 M NaCl and centrifugation. Total tRNA was recovered from the supernatant by ethanol precipitation. The yield was ∼100 A ₂₆₀ units.

Aminoacylation and formylation of cytoplasmic initiator tRNA.Total tRNA isolated from yeast was aminoacylated with methionine with E. coli MetRS or aminoacylated with methionine and subsequently formylated with E. coli MTF (62). The reaction mixture was extracted with phenol equilibrated with 10 mM sodium acetate (pH 4.5). The tRNA was recovered by precipitation and used for electrophoresis on acid-urea polyacrylamide gels to provide markers of cytoplasmic $$mathtex$$\(tRNA_{i}^{Met}\)$$mathtex$$ , $$mathtex$$\(Met-tRNA_{i}^{Met}\)$$mathtex$$ , and $$mathtex$$\(fMet-tRNA_{i}^{Met}\)$$mathtex$$ .

Northern blot analysis of tRNAs.Total tRNAs were isolated from yeast cells (3 ml of culture) under acidic conditions (0.3 M NaOAc [pH 4.5]), and 0.1 A ₂₆₀ unit was subjected to electrophoresis on acid-urea-polyacrylamide gels and electroblotted on to Nytran-Plus membrane (Schleicher & Schuell) as described previously (57). The membrane was washed with 4× SET (1× SET is 0.15 M NaCl, 2 mM EDTA, and 0.03 M Tris-HCl [pH 8.0]) containing 1% SDS, baked at 70°C for 90 min, and prehybridized at 42°C in 4× SET containing 0.1% SDS and 10× Denhardt's solution. tRNAs were detected by hybridization to a 5′-³²P labeled oligonucleotide probe complementary to nucleotides 8 to 24 of the yeast cytoplasmic initiator tRNA.

Measurement of protein synthesis rates.Protein synthesis rates were measured essentially as described previously (16, 50). For in vivo labeling, the synthetic medium lacking methionine contained leucine (40 mg/liter), tyrosine (30 mg/liter), isoleucine (30 mg/liter), phenylalanine (50 mg/liter), glutamic acid (100 mg/liter), aspartic acid (100 mg/liter), valine (150 mg/liter), threonine (200 mg/liter), serine (400 mg/liter), histidine (20 mg/liter), lysine (30 mg/liter), tryptophan (20 mg/liter), arginine (20 mg/liter), 0.67% yeast nitrogen base without amino acids, 2% raffinose, 3% glycerol, and 2% galactose. The yeast strains were grown in 200 ml of this medium to the early log phase, and protein synthesis rates were measured over a period of 6 h. At various times, cells corresponding to 2 A ₆₀₀ units were harvested and resuspended in 300 μl of fresh medium. Easy Tag Express protein labeling mix (50 μCi; 1,175 Ci of l-[³⁵S]methionine per mmol; NEN) was added, and the mixture was incubated for 5 min at 30°C. To stop incorporation of [³⁵S]methionine into proteins, 1 ml of a solution containing 1.2 mg of methionine per ml and 0.5 mg of cycloheximide per ml was added, and the culture was immediately frozen on dry ice. Cells were lysed by the addition (150 μl) of a freshly prepared solution containing 2 N NaOH and 12% β-mercaptoethanol, and proteins were precipitated with trichloroacetic acid (TCA) as described previously (63). The pellet was washed twice with 1.5 ml of acetone and suspended in 300 μl of 1% SDS containing a few acid-washed glass beads (425 to 600 μm in size; Sigma), and the suspension was mixed vigorously for 2 min with a Vortex mixer and boiled for 5 min. The clear supernatant obtained after centrifugation was used for determining the TCA-precipitable radioactivity on filter paper discs. The total protein concentration in these samples was estimated with the BCA reagents (Pierce). The rate of protein synthesis is expressed as the cpm of radioactivity incorporated per minute into 1 μg of TCA-precipitable protein. To ensure that the acid-precipitable ³⁵S radioactivity measured reflected incorporation into protein and not into [³⁵S]Met-tRNA, the extracts were also spotted on filters soaked in 1 N NaOH and then used for TCA precipitation. The counts in these samples were essentially the same as those in samples directly analyzed by TCA precipitation.

Isolation of ribosomes.Ribosomes were isolated from the following yeast strains as described previously (48): S. cerevisiae CKY473 (grown in YPD for ∼44 h), S. cerevisiae CKY473/YEp.MTF WT+pRS425 (grown in SMRG+Gal, −Leu for ∼60 h), and S. cerevisiae CKY473/YEp.MTF WT+pRS425.DEF (grown in SMRG+Gal, −Leu for ∼48 h). The cells were suspended in 10 ml of cold buffer A (20 mM Tris-HCl [pH 7.5], 16 mM MgCl₂, 100 mM KCl, 0.2 mM EDTA, 12 mM β-mercaptoethanol). Acid-washed glass beads (5 g) were added, and the suspension was mixed vigorously for 1 min with a Vortex mixer followed by chilling on ice for 2 min; the vigorous mixing followed by chilling was repeated six times. The lysate was poured off, and the glass beads were washed twice with buffer A. The lysate and washings were pooled, cell debris was removed by centrifugation at 3,000 × g for 10 min at 4°C, and the supernatant was further clarified by centrifugation at 27,000 × g for 20 min at 4°C. The clear supernatant was subjected to ultracentrifugation at 100,000 × g for 130 min at 4°C. The S100 pellet was overlaid with 5 ml of buffer A and left overnight on ice. The suspension was centrifuged at 27,000 × g for 20 min at 4°C, and the clear supernatant was then subjected to ultracentrifugation as described before.

Isolation and two-dimensional gel electrophoresis of ribosomal proteins.Proteins were isolated from the ribosomes and subjected to two-dimensional electrophoresis by a protocol provided by John Warner, Albert Einstein College of Medicine, Bronx, N.Y. The ribosomal pellet was suspended in 5 ml of a mixture containing 10 mM Tris-HCl (pH 7.5), 100 mM NaCl, 6 mM β-mercaptoethanol, and 1 mM MgCl₂, followed by addition of 0.5 ml of 1 M MgCl_2. Glacial acetic acid (final concentration, 67%) was slowly added, and the solution was stirred for 15 min and clarified by centrifugation (17,000 × g, 20 min, 4°C). The supernatant was dialyzed against 1% acetic acid. Two-dimensional gel electrophoresis was performed with the Mini-PROTEAN II 2-D cell apparatus (Bio-Rad). The first-dimension gel was in capillary tubes (1-mm internal diameter by 7.5 cm long) containing 4% (wt/vol) acrylamide, 0.1% (wt/vol) bisacrylamide, 8 M urea, 0.057 M Bis-Tris (adjusted to pH 5.0 with acetic acid). The ribosomal proteins were lyophilized and dissolved in sample buffer containing 8 M urea, 10 μM β-mercaptoethanol, 10% glycerol, 1% acetic acid, and 0.01% basic fuchsin. The electrophoresis buffer contained 10 mM Bis-Tris adjusted to pH 4.0 with acetic acid (anode) and 0.179 M potassium acetate (pH 5.0) (cathode). The second dimension was on SDS-15% polyacrylamide gels.

N-terminal sequence analysis of RPL3, a yeast large ribosomal subunit protein.Proteins were electroblotted onto Immobilon membranes (Millipore) as described previously (29). After transfer, the membrane was washed three times (10 min each) in MilliQ water, stained with 0.1% Coomassie blue in 50% MeOH for 2 min, and destained in 50% MeOH-10% acetic acid and washed repeatedly with MilliQ water and air dried. The RPL3 protein on membrane was then subjected to Edman degradation at the Biopolymers Laboratory, Massachusetts Institute of Technology.

Regulated expression of E. coli MTF in S. cerevisiae.In an attempt at formylation of the yeast cytoplasmic $$mathtex$$\(Met-tRNA_{i}^{Met}\)$$mathtex$$ in vivo, we expressed the E. coli MTF in the cytoplasm of yeast under the control of a galactose-inducible promoter with a high-copy-number (2μm) plasmid. For this purpose, the genes coding for the wild-type MTF and an inactive R42L mutant of MTF were cloned as C-terminal six-His fusions in the yeast expression vector YEp352-GAL1. The expression of the enzyme in total protein extracts was analyzed by immunoblotting with a monoclonal antibody recognizing the four-His epitope (Fig. 2A). A signal corresponding to the size of purified MTF is detected in extracts of yeast cells containing the recombinant plasmids upon induction with galactose. Both wild-type and the R42L mutant MTFs are expressed, and they accumulate to similar levels (lanes 3 and 5). There was no detectable expression in the uninduced state (lanes 2 and 4). Similar results were obtained with polyclonal antibodies raised against E. coli MTF (data not shown). Taken together, the results indicate the expression and stable accumulation of both the wild-type and R42L mutant MTFs in the cytoplasm of yeast without significant proteolysis.

• Open in new tab
• Download powerpoint

FIG. 2.

(A) Immunoblot analysis of MTF expression in extracts from S. cerevisiae CKY473 transformants containing the indicated plasmids grown in SMRG medium in the absence (−) or presence (+) of 2% galactose. Histidine-tagged MTF overexpressed and purified from E. coli (50 ng) is present in lane 1. (B) Northern blot analysis of tRNAs isolated under acidic conditions (pH 4.5) from S. cerevisiae CKY473 transformants grown in the absence (−) or presence (+) of galactose and carrying the indicated plasmids. The blot was probed with an oligonucleotide complementary to the cytoplasmic initiator tRNA. Lanes 1 to 3 provide markers of S. cerevisiae CKY473 tRNA_i, Met-tRNA_i, and fMet-tRNA_i, respectively. Lane 1 contains total tRNA isolated under basic conditions (pH 8.0) from S. cerevisiae CKY473 cells. MRS, methionyl-tRNA synthase.

Evidence for formylation of the yeast cytoplasmic initiator tRNA in vivo.Initiator tRNA is present predominantly as $$mathtex$$\(fMet-tRNA_{f}^{Met}\)$$mathtex$$ in eubacteria and as $$mathtex$$\(Met-tRNA_{i}^{Met}\)$$mathtex$$ in archaebacteria and in the cytoplasm of eukaryotes (42, 47). We investigated the in vivo state of the cytoplasmic yeast initiator tRNA in cells expressing the six-His-tagged wild-type MTF and the R42L mutant MTF. Previous studies had shown that introduction of the polyhistidine tag does not affect the activity of MTF (44). The R42L mutant MTF is ∼1,400-fold less active in formylation of the E. coli initiator tRNA (45, 46) and is used as a negative control here.

Total tRNAs isolated from cells under acidic conditions (pH 4.5) were separated by electrophoresis on an acid-urea-polyacrylamide gel and subjected to Northern blot analysis with a probe specific to the yeast cytoplasmic initiator tRNA (Fig. 2B). For use as markers, total tRNA (lane 1) isolated from S. cerevisiae CKY473 under basic conditions (pH 8.0) was aminoacylated in vitro with E. coli MetRS (lane 2) or aminoacylated with MetRS and subsequently formylated with E. coli MTF (lane 3) and subjected to electrophoresis on the same gel. As for the E. coli initiator tRNA (57), all three forms of the yeast cytoplasmic initiator tRNA, the $$mathtex$$\(tRNA_{i}^{Met}\)$$mathtex$$ , the $$mathtex$$\(Met-tRNA_{i}^{Met}\)$$mathtex$$ , and the $$mathtex$$\(fMet-tRNA_{i}^{Met}\)$$mathtex$$ are separated clearly from one another (Fig. 2B, lanes 1, 2, and 3). The mobility under acidic conditions of S. cerevisiae initiator tRNA isolated from the cells expressing the wild-type MTF (Fig. 2B, lane 5) shows that up to 70% of the $$mathtex$$\(tRNA_{i}^{Met}\)$$mathtex$$ is formylated. There was no detectable accumulation of any $$mathtex$$\(fMet-tRNA_{i}^{Met}\)$$mathtex$$ in uninduced cells (lane 4) or in cells expressing the inactive R42L mutant of MTF (lane 7). In these cells, the initiator tRNA is present mostly (>70%) in the aminoacylated form.

Effect of formylation of the yeast cytoplasmic initiator tRNA on growth rates.The growth of S. cerevisiae CKY473 strain containing the YEp352-GAL1 vector alone or with the wild-type or the R42L mutant MTF genes was studied on agar plates in the absence or presence of the inducer galactose (Fig. 3). In the absence of galactose (uninduced), there was no detectable difference in growth (left). In the presence of galactose (induced), cells expressing the wild-type MTF grow significantly more slowly than cells expressing the R42L mutant MTF (right). The cells expressing R42L mutant MTF do not show any slow-growth phenotype compared to cells containing the empty vector YEp352-GAL1. In liquid minimal medium also (see Materials and Methods for SMRG medium), the cells expressing the wild-type MTF enzyme grow much more slowly, with a doubling time of 300 min compared to the doubling time of 120 min for cells containing vector alone or cells expressing the R42L mutant MTF (Fig. 4). Similar effects of expression of wild-type MTF enzyme on growth were seen in S. cerevisiae UMY543 (4), which contains a single functional copy of the cytoplasmic initiator tRNA gene instead of the normal four or five copies in most laboratory strains, and in S. cerevisiae JEL1 (27), which expresses larger amounts of MTF due to the presence of higher levels of the GAL4 activator protein (data not shown).

• Open in new tab
• Download powerpoint

FIG. 3.

Growth of S. cerevisiae strain CKY473 carrying the indicated plasmids on agar plates containing SMRG medium in the absence (left) or presence (right) of galactose.

• Open in new tab
• Download powerpoint

FIG. 4.

Growth rate of S. cerevisiae CKY473 transformants carrying plasmids YEp352-GAL1 (○), YEp.MTF WT (▪), and YEp.MTF R42L (▴), respectively, in SMRG minimal medium containing 2% galactose.

Effect of formylation of yeast cytoplasmic initiator tRNA on rates of protein synthesis in vivo.The effect of formylating the yeast cytoplasmic initiator tRNA on overall rates of protein synthesis was investigated by monitoring the rates of incorporation of [³⁵S]methionine into proteins in medium containing galactose. Cells were grown in medium lacking methionine. At various times in the mid log phase of growth, 2 A ₆₀₀ units of cells was recovered by centrifugation and suspended in the same medium, ³⁵S-labeled methionine and cysteine were added and incubated for a short period of time (5 min), and the incorporation of ³⁵S radioactivity into proteins was measured (see Materials and Methods). The rates of protein synthesis in strains expressing wild-type MTF were essentially the same as those in cells expressing the R42L mutant MTF or in cells containing the empty vector (Fig. 5). These results suggest that the slower growth rate of cells expressing the wild-type MTF is not due to changes in overall rates of protein synthesis.

• Open in new tab
• Download powerpoint

FIG. 5.

Rates of incorporation of labeled methionine and cysteine into proteins in S. cerevisiae strain CKY473 containing the indicated plasmids during mid-log phase of growth. Details are given in Materials and Methods.

Effect of formate on the growth of yeast strains expressing E. coli MTF.MTF utilizes N ¹⁰-formyl tetrahydrofolate (N ¹⁰-fTHF) as the formyl donor to formylate the initiator Met-tRNA. Therefore, the slow-growth phenotype of the CKY473 cells expressing an active MTF could be due to depletion of the cytoplasmic pool of N ¹⁰-fTHF, which is used in the cytoplasm for many biosynthetic reactions. The one-carbon unit in fTHF is readily generated by the enzyme N ¹⁰-fTHF synthase from formate. Thus, exogenous addition of formate to CKY473 cells expressing MTF should be sufficient to overcome any limitation of depleted N ¹⁰-fTHF pools and thereby rescue the slow-growth phenotype. To investigate this possibility, the growth of the CKY473 strains was studied in the presence or absence of formate (Fig. 6). The S. cerevisiae EKY3 strain was used as a positive control. This strain (see Materials and Methods), which carries disruptions in both the cytoplasmic and mitochondrial serine hydroxymethyltransferases (17, 31), has problems maintaining adequate pools of one-carbon units in the cytoplasm. As a consequence, this strain requires exogenous formate for optimal growth. As expected, this strain grows slowly in medium lacking formate, and addition of exogenous formate rescues the slow-growth phenotype. In contrast, there was no enhancement of growth of the CKY473 yeast cells expressing wild-type MTF in medium containing formate. In liquid minimal medium as well, the EKY3 strain grew much more slowly in medium lacking formate, with a doubling time of approximately 12 h compared to the rate in medium supplemented with formate, with a doubling time of 4 h (data not shown). In contrast, the doubling time of CKY473 cells expressing MTF was essentially unaffected by the presence or absence of formate in the medium. Thus, the slow-growth phenotype of the CKY473 cells expressing MTF is not due to depletion of cytoplasmic pools of N ¹⁰-fTHF pools.

• Open in new tab
• Download powerpoint

FIG. 6.

Effect of formate on the growth of S. cerevisiae strains EKY3 and CKY473 carrying the indicated plasmids.

Effect of coexpression of E. coli polypeptide DEF on growth rate of yeast cells expressing E. coli MTF.The lack of any effect on rates of protein synthesis coupled to the fact that a significant fraction of the initiator tRNA is formylated in CKY473 cells expressing wild-type MTF raises the possibility that proteins are being initiated at least partly with formylmethionine. The reason for slow growth could thus be due to accumulation of proteins containing formylmethionine at the N terminus, which interferes with posttranslational N-terminal processing of proteins. To investigate this possibility, we expressed the E. coli polypeptide DEF as a C-terminal six-His fusion protein in these strains of yeast. As a control, we also expressed an inactive E133A mutant as a C-terminal six-His fusion protein (34). Expression of wild-type DEF or the DEF E133A mutant alone did not have any detectable growth phenotype on CKY473 transformants (Fig. 7A, sectors 1 and 2). Coexpression of the wild-type DEF rescues the slow-growth phenotype of cells expressing wild-type MTF (compare sectors 4 and 3). However, coexpression of the inactive E133A mutant of DEF fails to rescue the slow growth of cells expressing the wild-type MTF (sector 5). As expected, cells expressing inactive mutants of both E. coli MTF (R42L) and DEF (E133A) grow normally (sector 6).

• Open in new tab
• Download powerpoint

FIG. 7.

(A) Effect of expression of E. coli DEF on growth of S. cerevisiae CKY473 on agar plates containing 2% galactose. The plasmids present in the strains in the individual sectors are as follows: YEp352-GAL1+pRS425.DEF WT (sector 1), YEp352-GAL1+pRS425.DEF E133A (sector 2), YEp.MTF WT+pRS425 (sector 3), YEp.MTF WT+pRS425.DEF WT (sector 4), YEp.MTF WT+pRS425.DEF E133A (sector 5), and YEp.MTF R42L+pRS425.DEF E133A (sector 6). The slow-growth phenotype of a yeast strain expressing wild-type MTF (sector 3) is rescued by the expression of a functional DEF (sector 4), but not by expression of an inactive E133A mutant DEF (sector 5). (B) Immunoblot analysis of the expression of MTF and DEF in extracts of various transformants by using the four-His monoclonal antibody. The numbering of the lanes corresponds to the extracts prepared from the respective yeast strains in sectors 1 to 6 in panel A.

Total proteins from the CKY473 transformants were analyzed for the expression of MTF and DEF proteins by immunoblot analysis with a monoclonal antibody directed against the four-His epitope (Fig. 7B). The results indicate that both the wild-type and the E133A mutant DEF are expressed and accumulate to similar levels (lanes 1, 2, and 4 to 6). The immunoblot result also shows that the expression level of MTF is not affected by coexpression of DEF (compare lanes 3 and 4). Acid-urea Northern analysis also indicates that coexpression of DEF has no effect on the extent of formylation of the cytoplasmic initiator tRNA (data not shown). Taken together, these results strongly suggest that rescue of the slow-growth phenotype of yeast cells expressing the E. coli MTF by DEF is due to the removal by DEF of the formyl group from the N-terminal formylmethionine in proteins produced in cells expressing MTF.

Attempts to detect formylmethionine at the N terminus of proteins.Several approaches were used to detect formylmethionine at the N terminus of proteins in CKY473 transformants expressing the E. coli MTF.

(i) Comparison of two-dimensional polyacrylamide gel electrophoresis patterns of ribosomal proteins.Many, but not all, of the ribosomal proteins in yeast contain an N-terminal acetyl group as a result of posttranslational N-terminal acetylation often following the removal of the initiating methionine (2). Mutant proteins that lack these N-terminal acetyl groups because of a defect in one of the N-acetyltransferases (NAT1p) migrate differently from the wild-type protein on two-dimensional gels (55). It was expected, therefore, that if some of the ribosomal proteins were to carry a formylmethionine at the N terminus instead of methionine or another amino acid, these proteins would also migrate differently on two-dimensional gels. To investigate this, we isolated ribosomes from CKY473 cells transformed with either the empty vector or the vector carrying the E. coli MTF gene and grown in the presence of galactose (inducing conditions). The ribosomal proteins were subjected to two-dimensional gel electrophoresis and visualized with Coomassie blue or silver stain. There was essentially no difference in the relative mobilities of the proteins isolated from the two sets of ribosomes (data not shown). Thus, if there is any formylmethionine at the N terminus of the yeast ribosomal proteins in cells expressing the E. coli MTF, it is below the detection level of the staining techniques.

(ii) N-terminal sequence analysis of RPL3, a yeast large ribosomal subunit protein.A second approach used N-terminal sequence analysis to investigate whether there was a formyl group blocking the N terminus of a protein known to normally have a free amino group. The protein of choice was RPL3, a yeast large subunit ribosomal protein, which is known to have a free amino group at its N terminus and which is well separated from other proteins on two-dimensional gels (18, 55). Ribosomal proteins from CKY473 were separated on two-dimensional gels, transferred to PVDF membranes by electroblotting, and visualized by staining with Coomassie blue. The tailing part of the spot corresponding to RPL3 in the first dimension, which is most likely to contain a blocked N terminus, was excised and used for N-terminal sequence analysis by the Edman degradation procedure. The sequences of RPL3 isolated from the CKY473 cells transformed with the empty vector and with the vector containing the E. coli MTF gene were both the same, S-H-R-K-Y-E. Importantly, there was little methionine released in the second cycle of sequencing, as would have been expected if there was a significant fraction of the RPL3 protein that had been initiated with formylmethionine.

The attempts to identify formylmethionine at the N terminus of proteins described above have involved analyses of only the ribosomal proteins. The possibility that proteins that carry formylmethionine at the N terminus will not be assembled into ribosomes has not been ruled out, but is extremely unlikely. Many of the ribosomal proteins in yeast are acetylated at the N terminus. Mutant proteins that lack the N-terminal acetyl group because of mutations in the N-terminal acetyltransferases are still assembled into ribosomes (2, 55). Thus, the presence or absence of an N-terminal acetyl group has no effect on assembly of a protein on the ribosome. It is, therefore, extremely unlikely that the presence or absence of a formyl group at the N terminus will have an effect on assembly of a protein on the ribosome.