Isolation of a Full-length Complementary DNA Coding for Human Ela Subunit of the Pyruvate Dehydrogenase Complex*

Transcription

1 THE JOURNAL OF BIOLOGICAL CHEMISTRY by The American Society for Biochemistry and Molecular Biology, Inc. Vol. 263, No. 4, hue of February 5, pp ,1988 Printed in U. S. A. Isolation of a Full-length Complementary DNA Coding for Human Ela Subunit of the Pyruvate Dehydrogenase Complex* (Received for publication, May 21,1987) Linda De MeirleirS, Nevena MacKay, Anne Marie Lam Hon Wah, and Brian H. Robinson4 From the Department of Pediatrics and Biochemistry, University of Toronto and Research Institute, The Hospital for Sick Children, Toronto, Ontario M5G IX8, Canada A 1.5-kilobase cdna clone for human pyruvate dehydrogenase El was isolated from a Xgt ll expression library by screening with polyclonal antiserum to the Ela subunit of the porcine pyruvate dehydrogenase complex, a polyclonal antibody against bovine pyruvate dehydrogenase complex and a synthetic oligonucleotide based on the known amino acid sequence of the amino-terminal of the bovine pyruvate dehydrogenase-ela subunit. Nucleotide sequence analysis of the cdna revealed a 5 untranslated sequence of 72 nucleotides, a translated sequence of 1170 nucleotides, and a 3 untranslated sequence of 223 nucleotides with a poly(a) tail. The cdna structure predicts a leader sequence of 29 amino acids and a mature protein of 362 amino acids comprising an amino-terminal peptide identical to that of the bovine Ela subunit and three serine phosphorylation sites whose sequence was also identical to those in the bovine Ela subunit. The trans- tially from that described by Dahl et al. (Dahl, H. H., Hunt, S. M., Hutchison, W. M., and Brown, G. K. (1987) J. Biol. Chem. 262, ) by virtue of a frameslip between bases 390 and 594. This amended sequence is confirmed by the presence of additional restriction sites for the enzymes NueI and Hue11 at the beginning and end, respectively, of this section. The leader sequence is typical for mitochondrial enzymes being composed of a combination of neutral and basic residues. The amino acid composition is strik- ingly similar to that of the bovine protein. This cdna clone hybridizes with a 1.8-kilobase mrna on a Northern blot analysis of human fibroblasts, and a second minor band of 4.4 kilobases is also detected. Pyruvate dehydrogenase complex is a complex whose catalytic activity is maintained by three subunit enzymes: pyruvate dehydrogenase E, (EC ), dehydrolipoyl transacetylase E, (EC ), and dihydrolipoyl dehydrogenase E, (EC ) (1-4). The enzymatic activity of the total pyruvate dehydrogenase complex is regulated by a phosphorylation-dephosphorylation system. The phosphorylation sites are located in pyruvate * These studies were supported by the Canadian Medical Research Council and the Beta Sigma Phi Sorority. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked aduertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accessionumber(s) Present address: Dept. of Pediatrics, Laarbeeklaan, Brussels, Belgium. 5 TO whom correspondence should be addressed. dehydrogenase E, (5). Pyruvate dehydrogenase E, is a tetramer composed of two subunits each of a and p which have an M, of 41,000 and 36,000, respectively (5). Human pyruvate dehydrogenase E, deficiency has been documented in infants suffering from chronic lacticacidemia (6). Since the a-chain phosphorylation status determines enzymatic activity, we decided to undertake the cdna cloning of the subunit as a first step to the d Isease. understanding of the molecular basis of this MATERIALS AND METHODS Antibodies-The preparation of the bovine pyruvate dehydrogenase complex antibodies was previously described (6). Antibodies to the pyruvate dehydrogenase-ela protein were raised in rabbits to material isolated from sodium dodecyl sulfate-polyacrylamide gels of the whole porcine pyruvate dehydrogenase complex. IgG fractions were prepared by elution from Protein A-Sepharose columns. Oligonucleotide Probes-From the known amino acid sequence of lated sequence for the mature protein differs substan- the bovine pyruvate dehydrogenase-ela amino terminus (7), we designed a synthetic oligonucleotide 39-mer, based on preferred nucleotide utilization as described in Ref. 8, using codon utilization data with an overall probe target homology of 87%. The oligonucleotide 1991 was end-labeled with [y3 P]ATP. Hybridization and washing conditions were carried out according to standard procedures (9). Isolation of cdna Clones-A Xgtll human hepatoma cdna expression library (gift from J. OBrien, Univ. of California, San Diego) constructed from a cell line grown in BALB/c athymic mice (10) containing 8 X lo6 independent clones was screened with the two antibodies and subsequently an oligonucleotide probe as above. Antibody screening was performed as described by Young and Davis (11, 12). Nucleotide Sequence Analysis-DNA sequencing was performed by the Sanger dideoxy chain termination method on double-stranded templates (13, 14) using 2 -deoxyadenosine 5 -[a-%]thiotriphosphate (Du Pont-New England Nuclear). Both restriction fragments and exonuclease III/mung bean nuclease-generated deletion frag- ments (15) were subcloned into either plasmid psp64 or psp65 (Boehringer Mannheim) for sequencing. DNA was prepared by the alkaline lysis method (16), and a specific SP6 primer or the M13 reverse primer was used to allow sequencing from either end of the insert. In certain difficult regions oligonucleotide primers were synthesized and used to initiate sequencing. Northern Blotting Analysis-Total RNA from human skin fibroblasts was prepared using methods described by Maniatis et al. (17). Blot hybridization was performed after electrophoresis of the purified RNA through a 1% agarose gel containing formaldehyde (9). Nicktranslated 32P-labeled cdna for pyruvate dehydrogenase-elcu and for pyruvate carboxylase were used as probes. Chemicals-Restriction enzymes, T-4 polynucleotide kinase, and DNA polymerase were purchased from IBI and Pharmacia LKB Biotechnology Inc. The oligonucleotide was synthesized using a Pharmacia LKB Biotechnology Inc. gene assembler. RESULTS Screening of the Human Hepatoma Library-Out of 3 X IO6 plaques screened, 10 positive clones were identified with pyruvate dehydrogenase-ela antiserum and plaque-purified.

2 1992 cdna Clone of Pyruvate Dehydrogenase Ela FIG. 1. Restriction maps of the three pyruvate dehydrogenase-ela cdna clones isolated from the Xgtll library and the subcloning strategy used for sequencing. Three clones were isolated from the Xgtll library that were positive with Ela and pyruvate dehydrogenase complex antibodies A, B, and C. Clones A and B were also positive with the oligonucleotide probe constructed to the known NH2-terminal sequence of bovine pyruvate dehydrogenase-elm Subclones were constructed as shown below the clones A, B, and C and sequenced in both directions to obtain the full sequence of the identical clones A and I3 which was 1465 base pairs in length. Antisense oligonucleotides were synthesized to bases and to bases ; sense oligonucleotides were synthesized to bases and These oligonucleotides were used as primers to sequence through areas which gave inconsistent results with the SP6 forward and M13 reverse primers. Only three of these clones (A, B, and C) were found to be positive with pyruvate dehydrogenase complex antibody. When DNA was prepared from these isolates and the cdna inserts were produced by EcoRI digestion separated on a 1% agarose gel, two clones (A, B) had a size of 1.5 kb and one (C) had a size of 1 kb. Blotting and hybridization with the 39- mer oligonucleotide showed that only the two longer clones contained the amino-terminal sequence. The three clones were subcloned in psp64. Restriction mapping (Fig. 1) demonstrated that the two longer clones were identical, while the 1-kb clone was homologous but with a 0.5-kb segment missing at the 5 end. Nucleotide Sequence-The strategy for the sequencing of pyruvate dehydrogenase-e,a cdna is shown in Fig. 1. A series of subclones was constructed to facilitate the sequencing process either by cutting the cdna with restriction endonucleases or by exonuclease digestion and subcloning the fragments into SP64 or SP65. The total sequence obtained was 1465 base pairs long (Table I). There was an open reading frame of 1242 base pairs starting at the first base. There were two possible sites for the initiation codon, either at nucleotide 73 or 82. The two resulting possible contexts for initiation would then either be UGCUUCAUGG or AGGAAGAUGC, neither of them being close to the consensus sequence of CCACCAUGG proposed by Cozak (18). However, considerable variation does exist so that despite the greater similarity of the first possibility to the consensus, the initiation sequence for pig lipoamide dehydrogenase recently elucidated in our laboratory (GGCAAGAUGC(19)) resembles more the second possibility. For the purposes of this paper we will place initiation at the first site although this may have to be revised as further experimental data become available. Since the identity of the NH2-terminal sequence of the mature protein is known (7) this means that there is a leader sequence of either 26 or 29 amino acids depending on the choice of initiator codon. This sequence has the typical features of a mitochondrial leader sequence (20). The NHz-terminal amino acid sequence deduced from the cdna is identical with the known bovine pyruvate dehydrogenase-elm NH2-terminal sequence. The 39-mer oligonucleotide depicted in Fig. 2 which was used as a probe to confirm the identity of the long cdna clones was only 84% homologous with the human nucleotide sequence but was nevertheless close enough to hybridize under stringent conditions. The three sites where serine phosphorylation takes place at amino acids and are completely conserved with respect to the known sequence for The abbreviation used is: kb, kilobase(s). these sites from porcine heart pyruvate dehydrogenase-ela (21). After the stop codon there is a 219-base pair 3 untranslated region in which we could find a poly(a) addition signal system 26 nucleotides upstream from the poly(a) tail. The calculated molecular weight of the human Ela protein is 40,231 which is close to the 41,000 value obtained by physical characterization of the bovine subunit (5). Our sequence differs from that obtained by Dah1 et al. (22) in one large section to a degree that could not be explained as a polymorphism. The presence of an additional C at base position 390 and the absence of a T at base position 594 endows a frameshift that affects the sequence over a span of 68 amino acids. Because of this rather critical difference in the nucleotide sequence obtained we focused on the structure of the two sites responsible for this frameshift. The sequence GCCGGC begins at residue 388 and the sequence GGCGCT at residue 592, both sequences being unique to our proposed sequence and spanning the 5 and 3 frameshift sites, respectively. These two sequences are recognized as cut sites by the restriction endonucleases NueI and HueII, respectively. In common with the previously published sequence there is a NueI site at base position 126 and a Hue11 site at base position 112. In Fig. 3 the digests of the long cdna clone with NueI and HueII show that two cut sites are present for each restriction endonuclease as predicted by our sequence. Minor single base differences are also present in our sequence in the codons for amino acids 265,349, and 354. The discrepancy in the case of the codon for amino acid 265 is conservative while those for 349 and 354 represent a transition from alanine to proline and threonine to alanine, respectively. These differences may simply represent polymorphisms that are present in the gene pool. RNA Blot-Total RNA from control fibroblast cell lines and one pyruvate carboxylase-deficient cell line was examined by RNA blot analysis using the nick-translated Ela cdna as a probe. In subsequent blots pyruvate carboxylase cdna was used as a control (Fig. 4). Two bands were identified with the E1a cdna, one migrating slightly faster than the 28 S RNA, suggesting a transcript size of approximately 4.4 kb and a second one at 1.8 kb, which can be considered a right size mrna to encode a protein of M, 45,000. The cell line that has no mrna for pyruvate carboxylase has both the 1.8- and the 4.4-kb mrna species for the pyruvate dehydrogenase subunit. DISCUSSION We cloned a full-length cdna encoding a complete mature protein and, in addition, the leader sequence responsible for

3 cdna Clone of Pyruvate Dehydrogenase Ela 1993 TABLE I The nucleotide and corresponding amino acid sequence of plasmid cdna clones A and B encoding the E,a subunit of human pyruvate dehydrogenase complex Amino acids are numbered starting from -24 at the beginning of the open reading frame so that the first amino acid of the protein is designated number 1. The ochre stop codon occurs after amino acid number 390. Underlined are the two serine phosphorylation sites which match exactly those sequenced in the bovine protein Glu Thr Trp Gly His Pro Arq Arq Ala Ser Trp Val Val Arq Ser Arg Arg Cys Arq Hls Cys Leu Cy8 Phe Met Arg Lys Met Leu Ala GAG ACT TGG GGG CAC CCG CGT CGT GCC TCC TGG GTT GTG CGC CGC AGG TGC AGT CGC CAC TGC CTG TGC TTC MATG G ATG AGG CTC GCC Ala Val Ser Arg Val Leu Ser Gly Ala Set Gln Lys Pro Ala Ser Arg Val Leu Val Ala Ser Arq Asn Phe Ala Asn Asp Ala Thr Phe GCC GTC TCC CGC GTG CTG TCT GGC GCT TCT CAG AAG CCG GCA AGC AGA CTG GTG GTA tu TCC CGT AAT TTT GCA M T GAT GCT ACA TTT Glu Ile Lys Lys Cys Asp Leu His Arq Leu Glu Glu Gly Pro Pro Val Thr Thr Val Leu Thr Arq Glu Asp Gly Leu Lys Tyr Tyr Arg GAA ATT AAG AAA TGT GAC CTT CAC CGG CTG GAA GAA CCT GGC CCT GTC ACA ACA GTG CTC ACC AGG GGG GAG CTC GAT M A TAC TAC AGG Met Met Gln Thr Val Arg Arg Met Glu Leu Lys Ala Asp Gln Leu Tyr Lys Gln Lys Ile Ile Arg Gly Phe Cys His Leu Cys Asp Gly ATG ATG CAG ACT GTA CGC CGA ATG GAG AAA TTG GCA GAT CAG CTG TAT MA CAG AAA ATT ATT CGT GGT TTC UTGT C TTG TGT GAT GGT Gln Glu Ala Cys Cys Val Gly Leu Glu Gly Ala Ile Asn Pro Thr Asp His Leu Ile Thr Ala Tyr Arg Ala His Gly Phe Thr Phe Thr CAG GAA GCT TGC TGT GTG GGC CTG GAG GGC GCC ATC M C CCC ACA GAC CAT CTC ATC ACA GCC TAC CGG GCT TTT CAC ACT GGC TTC ACC Arg Gly Leu Ser Val Arg Glu Ile Leu Ala Glu Leu Thr Gly Arg Lys Gly Gly Cys Ala Lys Gly Lys Gly Gly Ser Met His Met Tyr CGG GGC CTT TCC GTC CGA GAA ATT CTC GCA GAG CTT ACA AM GGA GGA CGA GGT TGT GCT A M GGG AAA GGA GGA TCG ATG CAC ATG TAT Ala Lys Asn Phe Tyr Gly Gly Aan Gly Ile Val Gly Ala Gln Val Pro Leu Gly Ala Gly Ile Ala Leu Ala Cys Lys Tyr Asn Gly LYS GCC AAG AAC TTC TAC GGG GGC AAT GGC ATC GTG GGA GCG CAG GTG CCC CTG GGC GCT GGG ATT GCT CTA MGCC T GGA TGT AAA AAG TAT Asp Glu Val Cys Leu Thr Leu Tyr Gly Asp Gly Ala Ala Asn Gln Gly Gln Ile Phe Glu Ala Tyr Asn Met Ala Ala Leu Trp LYS Leu GAT GAG GTC TGCTG ACT TTA TAT GGC GAT GGT GCT GCT AAC CAG CAG GGC ATA TTC GAA GCT TAC AAC ATG GCA TTG GCT TGG AM TTA Pro Cys Ile Phe Ile Cys Glu Asn Arg Tyr Gly Met Gly Thr Ser Val Glu Arq Ala Ala Ala Ser Thr Asp Tyr Tyr Lys Arq Gly CCT TGT ATT TTC ATC TGT GAG MT MT CGC ~ A ~ G ~ G A G ACG TCT A GTT GAG AGA GCG GCA GCC AGC ACT GAT TAC TAC M G AGA GGC Asp Phe Ile Pro Gly Leu Arq Val Asp Gly Met Asp Ile Leu Cys Val Arg Glu Ala Thr Arg Phe Ala Ala Ala Tyr Cya Arg Ser Gly GAT TTC ATT CCT GGG CTG AGA GTG GAT GGA ATG GAT CTG ATC TGC GTC CGA GAG GCA ACA AGG TTT GCT GCT GCC TAT TGT AGA TCT GGG Ile Gln Glu Val Arq Ser Lys Ser Asp Pro Ile Met Leu Leu LYE Asp Arq Met Val Asn Ser Asn Leu Ala Ser Val Glu Glu Leu LYs ATT CAG GAA GTA AGA AGT AAG AGT GAC CCT ATT ATG CTT CTC AAG GAC AGG ATG GTG AAC AGC AAT CTT GCC AGT GTG GAA GAA CTA AAG Glu Ile Asp Val Glu Val Arq Lys Glu Ile Glu Asp Pro Gln Ala Phe Ala Ala Ala Asp Pro Glu Pro Pro Leu Glu Glu Leu Gly Tyr GAA ATT GAT GTGAA GTG AGG AAG GAG ATT GAG GAT CCT CAG GCC TTT GCC GCG GCC CCT GAT GAG CCA CCT TTG GAA GAG CTG GGC TAC His Ile Tyr Ser Ser Asp Pro Pro!he Glu Val Arq Gly Ala Asn Trp Gln Ile Lys Phe Lys Ser Val Ser OC CAC ATC TAC TCC AGC GAC CCA CCT TTT GAA GTT CGT GGT GCC AAT TGG CAG ATC M G TTT M G TCA GTC ACT TAA GGGGAGGAGAAGGAGAGGTTA I phe-ala-asn-asp-ala-thr-phe-glu-ile-lys-lys-cys-asp-leu-his arg-leu-glu-glu-gly-pro... mgcc-aat-gat-gcc-accm~gatg~-~tgt~gct~a Best Codon Usage Pmbe mgca-aat-gat-gct-aca-m~-aag"tg e l FIG. 2. Comparison of the sequence of the oligonucleotide probe with the actual sequence of the NH2 terminus of pyruvate dehydrogenase-ela. The oligonucleotide probe was designed according to the best codon usage protocal of Lathe (8). The actual homology of the probe with the sequence obtained from the cdna was 84% compared with the 87% homology predicted. the transport of the subunit to inner mitochondrial membrane. The combined use of two independent specific antibodies enabled us to rapidly select clones which were specific for pyruvate dehydrogenase-ela. The use of an oligonucleotide probe of 39 bases designed on the basis of preferred nucleotide utilization allowed us to confirm that two of the clones were pyruvate dehydrogenase-e,a specific and contained an NH2-terminal sequence, thus increasing the likelihood that we had isolated a full-length clone. The probe was designed with an overall target probe homology of 87%, but successful hybridization was obtained under stringent conditions in spite of the fact that the real homology was only 84% when the sequences were compared (8). When analyzing the sequence starting from the initiator codon we find two very closely located ATG triplets in the same reading frame as the amino terminal. Since pyruvate dehydrogenase-e1a is a mitochondrial enzyme, nuclear-encoded, synthesized in the cytoplasm as a larger precursor, it

4 1994 cdna Clone of Pyruvate Dehydrogenase E,CY " FIG. 3. Fragments produced in a digest of the 1.5-kb insert with NaeI and Had1 as seen after agarose gel electrophoresis. Aliquots (0.5 pg) of the 1.5-kh insert were incubated with 2 units of restriction endonuclease at 37 'C for 3 h. At the end of this time period the samples were separated by electrophoresis on a 1.8% agarose gel. Lane I, the digest with NaeI is predicted to show hands in a partial digest at 1465, 1339, 951, 388, 262, and 126 kb while in Lune 2 the digest with HaeII is predicted to show hands at 1465,1353, 873,592,480, and 112 kb. The hands were visualized by fluorescence in the presence of ethidium bromide. The position of the bands obtained with standard DNA is shown on the right, the size being given in base pairs. 28s- 9 -PC mrna FIG. 4. Northern blot of mrna from human skin fibroblasts using the Ela pyruvate dehydrogenase cdna probe. The experiment was carried out as described under "Materials and Methods." 18 and 285 rrna markers were visualized by ethidium bromide staining. The blot was probed first with pyruvate carboxylase (PC) cdna and then with the pyruvate dehydrogenase (PDH)E,N cdna. The picture shown is a composite obtained by an exact overlay of the two resulting x-ray films. Lune I, control cell line. Lane 2, mrna (-) pyruvate carboxylase-deficient cell lines. Lunes 3 and 4, mrna (+) pyruvate carboxylase-deficient cell lines. should contain a leader sequence. The sequence between the first or most upstream methionine and the first codon of the amino terminus is constituted in a similar fashion to other leader sequences described for mitochondrial enzymes. The total length of the leader is 29 amino acids. Four arginine residues are strategically located as in other documented leader sequences, the rest of the amino acids being either basic or neutral (20). It has been documented that the arginine residues are essential recognition features for proteolytic cleavage and import of the precursor protein into the mitochondria. Both the amino-terminal amino acid sequence of bovine heart pyruvate dehydrogenase-e,a (7) and the three serine binding sites of porcine heart pyruvate dehydrogenase- Ela (18) can be found with 100% homology, as would be expected for the essential features of the enzyme which are involved in phosphorylation/dephosphorylation control of activity. A tetradecapeptide identified in yeast pyruvate dehydrogenase-e,a LS a serine phosphorylation site is also similar in sequence to one of the bovine and human sites. Interestingly the serine binding sites (amino acids and ) seem to have some characteristics in common; they are preceded by arginine, start with tyrosine, and end with arginine. The explanation for this arrangement is not clear, but it is possible that the function of the arginines is either charge related or part of a recognition site for the kinase and phosphatase. The poly(a) addition signal ATTAAA about 30 nucleotides from the 3' poly(a) tail is different from the consensus sequence AATAAA but is sporadically found (12%) in mrnas from different species (23) and probably functions in stabilizing the 3' end of the mrna (24). Our sequence differs from that previously published by Dah1 et ul. (22), but the presence of one additional restriction site each for the enzymes NueI and Hue11 suggests that our sequence is the correct one. The amino acid composition (Table 11) of the mature protein is almost completely in agreement with the bovine sequence (5). Finally, a Northern blot indicated the presence of two species of mrna in fibroblast cell lines which hybridized with the cdna for human pyruvate dehydrogenase-ela, one being the exact size for the precursor of pyruvate dehydrogenase-e,a. The other species is more likely to be an unprocessed higher molecular weight transcript. Further study of cell lines for patients deficient in pyruvate dehydrogenase-e, using both Northern and Southern blot analysis may lead to the eventual elucidation of the TABLE I1 Amino acid composition of bovine kidney and human liver E,Npyruvate dehydrogenase Bovine kidney data from Barrera et al. (5), calculated from amino acid analyses and a M, of 41,000. Human liver composition derived from the cdna sequence beginning at the NH, terminus of the mature protein. Bovine kidney Human liver LY s His A% Asp + Asn Thr Ser Glu + Gln Pro GlY Ala 'hcys Val Met Ile 21.6 Leu TY r Phe Trp ND" 2 Not determined.

5 cdna Clone of Pyruvate Dehydrogenase Ela 1995 molecular events responsible for human pyruvate dehydrogenase-e, deficiency. REFERENCES 1. Reed, L. J. (1981) Curr. Top. Cell. Regul. 18, Randle, P. J. (1983) Philos. Trans. R. SOC. Lond. Biol. Sci. 302, Linn, T. E., Pettit, F. H., Hucho, F., and Reed, L. J. (1969) Proc. Natl. Acad. Sci. U. S. A. 64, McAllister, A., Allison, S. P., and Randle, P. J. (1973) Biochem. J. 134, Barrera, C. R., Namihira, G., Hamilton, L., Munk, P., Eley, M. H., Linn, T. C., and Reed, L. J. (1972) Arch. Biochem. Biophys. 148, McKay, N., Petrova-Benedict, R., Thoene, J., Bergen, B., Wilson, W., and Robinson, B. (1986) Eur. J. Pediatr. 144, Lawson, R., Aitken, A., and Yeaman, S. J. (1983) Biochem. Trans. SOC. 11, Lathe, R. (1985) J. Mol. Biol. 183, Davis, L. O., Dibner, M. D., and Battey, J. F. (1986) Basic Methods in Molecular Biology, pp , Elsevier Scientific Publishing Co., Inc., New York 10. De Wet, J. R., Fukushima, H., Dewji, N. N., Wilcox, E., O Brien, J. S., and Helinski, D. R. (1984) DNA 6, Young, R. A., and Davis, R. W. (1983) Science 222, Young, R. A., and Davis, R. W. (1983) Proc. Natl. Acad. Sci. U. S. A. 80, Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. U. S. A. 74, Korneluk, R. G., Quan, F., and Gravel, R. A. (1985) Gene (Amst.) 40, Guo, L. H., Yang, R. C. A., and Wu, R. (1983) Nucleic Acids Res. 11, Birnboim, H. C., and Doly, J. (1979) Nucleic Acids Res. 7, Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning, A Laboratory Manual, pp , Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 18. Kozak, M. (1986) Cell 44, Otulakowski, G., and Robinson, B. H. (1987) J. Biol. Chem. 262, Allison, D. S., and Schatz, G. (1986) Proc. Natl. Acad. Sci. U. S. A. 83, Sugden, L., and Broch, J. (1979) Biochem. J. 181, Dahl, H. H., Hunt, S. M., Hutchison, W. M., and Brown, G. K. (1987) J. Biol. Chem. 262, Bernstiel, M. L., Busslinger, M., and Strub, K. (1985) Cell 41, Weelers, M., and Stephenson, P. (1984) Science 226,