In vitro suppression as a tool for the investigation of translation initiation
In vitro suppression as a tool for the investigation of translation initiationVladimir A. Karginov, Sergey V. Mamaev and Sidney M. Hecht*
Departments of Chemistry and Biology, University of Virginia, Charlottesville, VA 22901, USA
Received June 9, 1997,Revised and Accepted August 5, 1997
ABSTRACT
An in vitro protein synthesizing system that employs rabbit reticulocyte lysates has been employed for protein production from mRNAs containing nonsense (UAG) codons in the presence of misacylated suppressor tRNAs.The system includes a misacylated Escherichia coli tRNAAlaCUA that functions at least as efficiently as any suppressor tRNA transcript reported to date and which has been shown not to be a substrate for (re)activation by alanyl-tRNA synthetase. Application of the optimized system for preparation of dihydrofolate analogs has also permitted analysis of competing mechanisms that control the sites(s) of translation initiation.
INTRODUCTION
Within the past several years increasing activity has been devoted to the preparation of proteins containing synthetic amino acids at single, predetermined positions. The strategy involves the use of misacylated suppressor tRNAs (1 -11 ) for read-through of nonsense codons incorporated in the mRNA at the position of interest (12 -19 ). As first shown by Hecht and co-workers (2 ,3 ), the misacylated tRNAs are accessible via T4 RNA ligase catalyzed condensation of aminoacylated pCpA (or pdCpA) derivatives with a tRNA or tRNA transcript lacking the 3'-terminal dinucleotide pCpA.
The foregoing strategy has been employed successfully for the preparation of numerous modified proteins and polypeptides (12 -28 ). It has also provided a vehicle for investigating the actual mechanism of peptide bond formation (2 ,3 ,10 ,14 ,16 ,29 -32 ) and for characterizing the parameters that control the read-through of nonsense codons (17 ,33 ,34 ).
Another process potentially amenable to analysis involves initiation of translation. In a series of investigations it has been shown convincingly by Kozak that translation initiation generally occurs at the AUG nearest the 5'-end of the mRNA (35 -40 ); there are well characterized exceptions to this rule, however. Most of the exceptions involve either reinitiation at a downstream AUG codon following translation termination in proximity to the first AUG (41 -43 ) or a process termed `leaky scanning' (38 -40 ,43 ), in which the occurrence of the first AUG in a suboptimal context permits bypass of this codon and initiation at a downstream AUG codon.
Presently we employ misacylated suppressor tRNAs in a protein synthesizing system containing multiple AUG codons at the 5'-end of a mRNA encoding dihydrofolate reductase (DHFR) to explore the mechanism of eukaryotic translation initiation. Initiation at the first AUG can be distinguished readily from reinitiation and leaky scanning; the use of misacylated suppressor tRNAs affords the opportunity to dissect the contribution of aminoacyl-tRNA structure to the process of translation.
MATERIALS AND METHODS
[35S]Methionine (1000 Ci/mmol) was purchased from Amersham Corp. Nuclease-treated rabbit reticulocyte lysate, Escherichia coli S-30 extract system for linear templates, wheat germ extract, DNA polymerase I (Klenow fragment), T4 DNA ligase, T7 RNA polymerase and restriction endonucleases were obtained from Promega Inc. Purified acylated bovine serum albumin (BSA) was from New England Biolabs. [beta]-NADPH and dihydrofolate were obtained from Sigma Chemical Co. Kits for plasmid isolation and for purification of proteins using Ni-NTA agarose were purchased from Qiagen Inc. Synthetic oligonucleotides were obtained from Cruachem Inc. or Midland Co. Escherichia coli competent cells were purchased from Stratagene Cloning Systems. AmpliScribe transcription kits were from Epicentre Technologies.
Ultraviolet spectral measurements were made using a Perkin-Elmer Lambda Array 3840 spectrophotometer equipped with a thermal control unit. Radioactivity measurements were made with a Beckman LS-100C liquid scintillation counter. Phosphorimager analysis was performed using a Molecular Dynamics 300E phosphorimager equipped with ImageQuant software.
Construction of plasmids carrying the dihydrofolate reductase gene (DHFR), construction of plasmids for run-off transcription of yeast tRNAPheCUA (-CA) and E.coli tRNAAlaCUA (-CA), run-off transcription of tRNAs and synthesis of misacylated tRNAs and of mRNAs were carried out as described (11 ,18 ).
In vitro suppression in a rabbit reticulocyte protein synthesizing system
In a typical experiment DHFR was synthesized in reaction mixtures (25-100 [mu]l total volume) that contained, per 100 [mu]l, 70 [mu]l methionine-depleted, nuclease-treated rabbit reticulocyte lysate, 80 [mu]Ci [35S]methionine, 2 [mu]l 1 mM amino acid mixture (lacking methionine), 8 [mu]g mRNA and 10 [mu]g misacylated suppressor tRNA. Reactions were incubated at 30oC for 1 h. In vitro translation of (control) pTHis15-derived mRNA was carried out without addition of misacylated tRNA. Aliquots (typically 1 [mu]l) were utilized for analysis by 20% SDS-PAGE (44 ). Autoradiography of the gels was carried out to determine the location of 35S-labeled proteins; quantification of the bands was carried out using a phosphorimager.
In vitro suppression in an E.coli S-30 extract protein synthesizing system
The syntheses were carried out in reaction mixtures (25-100 [mu]l total volume) that typically contained, per 100 [mu]l, 30 [mu]l S-30 extract, 40 [mu]l premix (45 ), 80 [mu]Ci [35S]methionine, 4 [mu]l 1 mM amino acid mixtures lacking methionine, 4 [mu]l 0.4 M Mg(OAc)2, 4 [mu]g plasmid DNA linearized with BamHI, 200 U T7 RNA polymerase and 10 [mu]g misacylated suppressor tRNA. Reactions were incubated at 37oC for 1 h. In vitro translation of pTHis15-derived mRNA was carried out without addition of misacylated tRNA. Aliquots from the reaction mixtures (typically 5 [mu]l) were withdrawn and precipitated with acetone for analysis by 20% SDS-PAGE, followed by autoradiography.
Purification of proteins
In vitro translation mixture containing a 35S-labeled protein was loaded onto a Ni-NTA agarose column (200 [mu]l) equilibrated with 50 mM sodium phosphate, pH 8.0, 300 mM NaCl, 20 mM imidazole and 100 [mu]g/ml BSA. The column was washed with five 100 [mu]l portions of this buffer, then the protein was eluted with five 100 [mu]l portions of 50 mM sodium phosphate, pH 8.0, containing 300 mM NaCl, 250 mM imidazole and 100 [mu]g/ml BSA. The amount of 35S-labeled products in individual fractions was determined by liquid scintillation counting.
RESULTS
The general scheme used to study translation initiation is presented in Figure 1 . The scheme involved in vitro transcription of a linearized plasmid DNA containing a modified DHFR gene with three ATG codons (at positions -9, 1 and 16) and a stop codon (TAG) at position 10 (or 27). The derived mRNA was employed for in vitro translation in the presence of a misacylated suppressor tRNA. Because the DHFR gene was modified to encode a fusion nonapeptide at its N-terminus that includes hexahistidine, the full-length protein product could be purified on a Ni-NTA column (46 ). The requisite plasmids containing the modified DHFR genes, as well as two additional plasmids required as controls, have been constructed starting with plasmid pTZRKE (Fig. 2 ; 34 ).
DISCUSSION
We have previously described overexpression of DHFR in a cell-free system. The strategy involved initial gene amplification via PCR, coupled with transcription and translation (GATT) (49 ). When employed with careful optimization of the transcription and translation reactions, ~1010 copies of DHFR could be elaborated for each DNAfol used initially. The present study extends this work by defining conditions suitable for read-through of nonsense codon UAG in the rabbit reticulocyte lysate system, thereby facilitating preparation of quantities of proteins containing synthetic amino acids at predetermined sites (12 -19 ).
Of particular interest was the finding during this optimization that E.coli suppressor tRNAAlaCUA, described previously by Hou and Schimmel (48 ), could function well in protein synthesis as a suppressor tRNA following activation by `chemical aminoacylation' (2 ,3 ,10 ). The ability of this tRNA to function in suppression could not be assessed in the earlier study because this tRNA is not a substrate for alanyl-tRNA synthetase; thus the present findings establish the ability of this species to function in the partial reactions of protein synthesis subsequent to aminoacylation. In addition to its favorable properties as a suppressor tRNA in comparison with other transcript so employed (Fig. 4 ; 8 ), it seems unlikely that tRNAAlaCUA could be reactivated enzymatically to a significant extent by any endogenous aminoacyl-tRNA synthetase after transferring the synthetic amino acid introduced via chemical aminoacylation. This would thereby preclude incorporation of a natural amino acid into a site in the protein intended for incorporation of a synthetic amino acid. This point was verified by inclusion of non-acylated, full-length tRNAAlaCUA in a protein synthesizing system containing pTZN2H4-derived mRNA (cf. Fig. 7 ). Less than 1% read-through of the UAG codon at position 10 was observed.
Initiation of protein synthesis in eukaryotes has been studied extensively (35 -43 ). The scanning model (36 -40 ) envisions entry of the 40S ribosomal subunit at the 5'-end of the mRNA, from which it travels linearly until the first AUG codon is encountered. Providing that this codon occurs in a favorable context, this AUG codon will constitute the unique site of translation initiation. Where the context in which this AUG codon occurs is suboptimal, it is postulated that some of the 40S ribosomal subunits may pass this codon and initiate translation at a downstream AUG codon. This process is termed `leaky scanning' (38 -40 ,43 ) and results in synthesis of more than one protein from a single mRNA (50 -53 ). A second mechanism by which a downstream AUG codon can be employed involves reinitiation of protein synthesis if the first AUG codon is followed by a termination codon (41 ,42 ).
The protein synthesizing system employed here for elaboration of DHFR mutants at positions 10 and 27 provides a unique opportunity to observe both mechanisms for downstream initiation of protein synthesis operating simultaneously. Leaky scanning is possible due to the suboptimal context of the first AUG codon (Fig. 9 ), which, for example, lacks both a purine in position -3 and a G in position +4 (38 -40 ,43 ). As a consequence, a shorter protein can be translated from the second AUG codon; in the present case this would be the codon for Met1 of wild-type DHFR. In fact, the larger protein `by-product' (protein 2) evident in Figures 7 and 8 has properties entirely consistent with those that would be expected for the leaky scanning product. In addition to its migration relative to molecular weight markers, it co-migrates with DHFR lacking the hexahistidine fusion peptide (elaborated from plasmid pTZRKE; cf. Fig. 8 , lanes 2-4 and 6). This protein is present both in translation mixtures containing wild-type mRNA (Fig. 7 , lane 3) as well as those containing mRNA with a stop codon in position 10 (Fig. 7 , lane 1). As anticipated, this protein does not bind to Ni-NTA agarose. Further, the ratio between levels of expression of the full-length DHFR product and protein 2 does not depend on the concentration of misacylated tRNA in in vitro suppression reactions (Fig. 8 ).
Figure 9. Nucleotide sequence and deduced amino acid sequence at the 5'-end of the DHFR gene. Amino acids are numbered starting from the first Met of wild-type DHFR; a hexahistidine sequence added to the N-terminus to facilitate affinity purification is underlined. Restriction sites are indicated by arrows. Also indicated are the sites of introduction of nonsense codons (positions 10 and 27) and the start of the DHFR sequence (bold arrow).
The third product observed in the in vitro suppression reactions (protein 3) is obviously the result of the stop and reinitiation mechanism. It appeared as the major band in the blank suppression reaction when no misacylated tRNA was added (Fig. 8 , lane 1) and the level of its synthesis decreased with increasing concentrations of misacylated tRNA (Fig. 8 ). Like the leaky scanning product, protein 3 disappears after Ni-NTA chromatography. Because the first Met codon after the UAG codon at position 10 appears in position 16, this protein probably results from reinitiation of translation of Met16; this is fully consistent with the relative mobilities of the full-length and truncated proteins derived from PAGE experiments that included molecular weight markers. Accordingly, the results presented above are fully consistent with the postulated rules of initiation of translation in eukaryotes.
The present system thus provides a very good model for the qualitative and quantitative study of alternative mechanisms for initiation of translation. Both postulated mechanisms can be observed simultaneously and it seems likely that they can be controlled independently by judicious choice of codon context (for the leaky scanning mechanism) and suppressor tRNA concentrations (for the reinitiation mechanism).
ACKNOWLEDGEMENTS
We thank Drs John Abelson and George Komatsoulis (California Institute of Technology) for the plasmid containing DNAfol and Drs Paul Schimmel (Massachusetts Institute of Technology) and Ya-Ming Hou (Thomas Jefferson University) for the plasmid that permitted us to elaborate E.coli tRNAAlaCUA (-CA). This work was supported by NIH Research Grant GM43328.
REFERENCES
1 Hecht,S.M., Alford,B.L., Kuroda,Y. and Kitano,S. (1978) J. Biol. Chem., 253, 4517-4520.MEDLINE Abstract
2 Heckler,T.G., Zama,Y., Naka,T. and Hecht,S.M. (1983) J. Biol. Chem., 258, 4492-4495.MEDLINE Abstract
45 Lesley,S.A., Brow,M.A.D. and Burgess,R.R. (1991) J. Biol. Chem., 266, 2632-2638.MEDLINE Abstract
46 Janknecht,R., de Martynoff,G., Lou,J., Hipskind,R.A., Nordheim,A. and Stunnenberg,H.G. (1991) Proc. Natl. Acad. Sci. USA, 88, 8972-8976.MEDLINE Abstract
47 Sampson,J.R. and Uhkenbeck,O.C. (1988) Proc. Natl. Acad. Sci. USA, 85, 1033-1037.MEDLINE Abstract