109
Biochimica et Biophysica Acta, 1132 (1992) 109-113 © 1992 Elsevier Science Publishers B.V. All rights reserved 0167-4781/92/$05.00
BBAEXP 90389
Short Sequence-Paper
Molecular cloning and DNA sequence analysis of cDNA encoding chicken homologue of the Bcl-2 oncoprotein Dominique L. Cazals-Hatem, Diane C. Louie, Shigeki Tanaka and John C. Reed Department of Pathology and Laboratory Medicine, Uniuersity of Pennsyh'ania, Philadelphia, PA (USA) (Received 4 May 1992)
Key words: Bcl-2; Oncoprotein; Mitochondrion; Nucleotide sequence; Apoptosis; Programmed cell death; (Chicken)
We have isolated a 2228 bp cDNA clone encoding a chicken homologue of the human Bcl-2 oncoprotein by low-stringency hybridization screening of a Agtl0 cDNA library derived from a chicken B-cell lymphoma. DNA sequence analysis of this cDNA revealed an open reading frame predicting a polypeptide of 232 amino acids and an M r of 25 839. The predicted protein is highly homologous to the human (73%) and mouse (70%) Bcl-2 proteins, and contains a hydrophobic stretch of amino acids within its carboxyl-end (213-229) consistent with an integral membrane protein. Areas of very high sequence homology shared by all three Bcl-2 proteins at the NH2-terminus (amino acids 1-33) and within the last 150 amino acids of these proteins suggest the presence of at least two evolutionarily conserved domains within the family of Bcl-2 proteins that may be important either for their targeting to mitochondria or their ability to block programmed cell death. The human bcl-2 proto-oncogene was originally discovered because of its involvement in the t(14;18) translocations found in the majority of non-Hodgkin B-cell lymphomas [1]. This cellular gene encodes a 26 kDa integral membrane protein that has recently been shown to reside in the inner membrane of mitochondria and that shares no significant sequence homology with other known proteins [2]. The predicted amino acid sequence of the human Bcl-2 protein does not suggest the presence of a positively charged amphipathic a-helical domain at its NH2-terminus, unlike many other cellular proteins that undergo import into mitochondria with concomitant proteolytic cleavage of the NHz-leader sequence [3]. Thus, p26-Bcl-2 may utilize other import pathways for its uptake into mitochondria. Bcl-2 is unique among oncoproteins described to date in that it contributes to the cellular accumulation recognized as cancer, not by augmenting the rate of cellular proliferation, but rather by enhancing cell survival. The specific biochemical mechanism by which p26-Bcl-2 accomplishes this remains enigmatic, but must somehow be related to its ability to markedly
Correspondence to: J.C. Reed, La Jolla Cancer Research Institute, Cancer Research Center, 10901 N Torrey Pines Road, La Jolla, CA 92037, USA. The sequence data in this paper have been submitted to the EMBL/Genbank Data Libraries under the accession number Zl1961.
interfere with programmed cell death (also termed 'apoptosis') [2,4]. The human Bcl-2 protein contains no sequence motifs that might suggest its biochemical function, including no nucleotide-binding, Ca2+-bind ing, or metal-binding regions. Likewise, molecular cloning and DNA sequence analysis of the mouse bcl-2 gene has added little additional information that might help to identify functionally important domains within Bcl-2 proteins, since the predicted mouse protein is 85% homologous to its human counterpart and this homology is distributed throughout the length of the proteins [5]. In an effort to define evolutionarily conserved regions within Bcl-2 proteins that may coincide with functionally important domains, we have attempted to clone bcl-2 gene homologs from non-mammalian species. Here we report the nucleotide and deduced amino acid sequences for a cDNA encoding a chicken homolog of Bcl-2. Comparisons of this avian and the previously published mammalian sequences revealed several evolutionarily conserved features that provide potentially important insights into the structure-function relations of these unique oncoproteins. For these experiments, a chicken B-cell lymphoma cDNA library constructed in Agtl0 phage (gift from M. Bishop) was screened by low-stringency hybridization using a 910 bp human bcl-2 cDNA probe pB4 which contains the open reading frame (ORF) for p26-Bcl-2 plus 64 bp of upstream and 129 bp of downstream untranslated sequence [6]. Approx. 20 000 phages/plate were hybridized with this 32p-labeled probe at 50°C in 6 × SSC containing 0.1% sodium dodecyl sulfate (SDS),
110 150 p~g/ml sonicated salmon sperm DNA, and 5 × Denhardt's solution. The final washes were performed for 30 min at 50°C in 2 × SSC, 0.1% SDS. A phage clone containing a 2.2 kb EcoRI insert was plaquepurified, and its insert was subcloned into the plasmid plBl31 (International Biotechnologies) for D N A sequence analysis. The complete nucleotide sequence was determined for both strands by the dideoxynucleotide chain termination method, using Sequenase version 2.0 (USB) [7]. Sequence comparisons of the chicken c D N A with the published sequences for human and mouse bcl-2 cDNAs and predicted proteins were accomplished with the aid of the NALIGN, PAL1GN and CLUSTAL computer programs [8,9]. Computer predictions of protein secondary structure, hydropathic index determination, and the prediction of sites and signatures in the protein sequence were performed using the GARNIER, SOAP and PROSITE programs, respectively [10-12]. The nucleotide and amino acid sequences of the chicken bcl-2 c D N A clone are presented in Fig. 1. The longest ORF within this 2228 bp c D N A was 699 nucleotides in length and began 62 bp from the 5' end of the cDNA, ending with a TAG stop codon at position 758. The translation initiation codon and surrounding bases shared 8 of 10 matches with the Kozak concensus sequence for eukaryotic translation start sites [13]. This ORF exhibited 75% and 72% homology at the nu1 62 116 170 224
278 332
386 440 494 548 602 656 71o 764 835 906 977 1048 1119 1190 1261 1332 1403 1474 1545 1616 1687 1758 1829 1900 1971
ATG M ATC I CCG P GCT A AGG R CGC R TCG S GTG V GAG E GTG V TGG W ATG M GTT V
GCT A CAC H CCC P GGA G TGC C CAG Q GGC G GAG E TTC F GAC D ATC I AGG R CTG L
cleotide level with the regions corres-ponding to the ORFs previously identified for human and mouse bcl-2 cDNAs, respectively. It is unlikely that this chicken bcl-2 c D N A contains the complete 3' untranslated regions of the full-length chicken bcl-2 mRNA since neither a poly(A) tail nor A A T A A A motifs for transcript termination and polyadenylation were found within the 1468 bp of sequence located 3' of the ORF. The 61 bp of sequence found 5' of the chicken bcl-2 cDNA's ORF also may not represent the full extent of 5' untranslated sequences in the full-length chicken bcl-2 mRNA, given that the mouse and human bcl-2 mRNAs begin approx. 1.5 kbp upstream of their ORFs [5,6]. Several ATTA and ATTTA motifs were located within the 3' untranslated portion of the chicken bcl-2 cDNA. Similar sequence motifs have been identified previously in human and mouse bcl-2 cDNAs and have been associated with rapid turnover of many mRNAs [14]. Otherwise, the untranslated regions within the chicken bcl-2 c D N A exhibited no striking sequence homology with the corresponding portions of mammalian bcl-2 cDNAs. The overall homology between the 3' untranslated region of the chicken bcl-2 c D N A and its human and mouse counterparts was 53% and 42%, respectively. To experimentally confirm the presence of the predicted open reading frame in the chicken c D N A clone (62-758 bp), the plBI-31 plasmid containing the 2.2 kb
AATTCCGCCCCCCCCCCCCCCTTCCTCTCCCCCTCCCACTCGCTGCTTCCCCTCGGAAACC CAC CCC GGG AGA AGA GGC TAC GAC AAC CGC GAG ATA GTG CTG H P G R R G Y D N R E I V L TAT ~ CTC TCG CAG CGG GGC TAC GAC TGG GCC GCC GGC GAG Y K L S Q R G Y D W A A G E GTG CCC CCG GCC CCG GCT CCC GCT GCT GCT CCC GCC GCG CTC V P P A P A P A A A P A A V GCC TCC TCC CAC CAC CGC CCG AGC CCC CCG GCT CGG CTG CTG A S S H H R P S P P A R L L C C C C G G C T G A G G G G C T G C G C C G C G C C T CCC G G C G T C C A C C T C P R L R G C A A P P G V H L GCC GGG GAC GAG TTC TCG CGC CGC TAC CAG AGG GAC TTC GCC A G D E F S R R Y Q R D F A CAG CTG CAC CTG ACG CCC TTC ACG GCC ACC GGC CGC TTC GTG Q L H L T P F T A T G R F V GAG CTC TTC CGT GAT GGG GTC AAC TGG GTC CGG ATC GTC GCC E L F R D G V N W V R I V A GGC GGC GTG ATG TGC GTC GAG AGC GTC AAC CGG GAG ATG TCG G G V M C V E S V N R E M S AAC ATT GCC ACC TGG ATG ACC GAG TAC CTG AAC CGG CAC CTG N I A T W S T E Y L N R I[ L CAG GAC AAC GGA GGA TGG GAT GCC TTT GTG GAA TTG IAC GGC Q D N G G W D A F V E L Y C CCT TTG TTC GAT TTC TCC TGG ATC TCT CTG AAG ACC ATC CTG P L F D F S W I S L K T I L GTG GGA GCT TGC ATC ACT CTT GGC GCT TAT CTT GGA CAT AAG V G A C I T L G A Y L G H K
AAG K GAC D GCT A CTA L GCC A CAG Q GCC A TTC F CCG P CAC H AAC N AGC S TAG ,
TAC Y 18 AGG R 36 GCT A 54 GTG V 72 CTG L 90 ATG M i0~ GTC V ]26 TTC F 144 CTG L 162 AAC N 180 AGT S 198 CTG L 216 AGT 232
CACCCAGTTTATCGTGGATTACAAAGTCTTTCAAAACAGCAAAATAATATTTTTTTTCTGATGCACAATGG TATTTTATGAGCA~CAACTTCCTGGCTAAAGATTAAATCAGCTATTACTGCCAAAGGAAATATCATTTA TTTTTACATTGCAGG~%AATTATTTATTA~TATATTTACATTTTA.ILCCTGCTATTTTCAGAAACTCTGC
AAGTTGTATCTGTCATCACATCATGTTGTTATTCTTAACTTTAATGAGCCCCAAAGCATTTTAGGTTCTTT TTTTAATTAATGCAG~GTCTGGATTATT~ACCTGCGAAGAGCAAGTATACTGACAAAGACCTACCTGCT
TACACTTAGGAAGGCAGTTACTTGAGCTGCTAAAGGCTGTGTAATTGAGTAGATTTCCTTGCAATTTTTGG ATGGCCGGGAGGGCTCTGAAAGAAAAGGTCAACGTGAACCAAGTTCTAATCCTGTCCTGCCAGCGAGAGCA GTCACAGAGAGTAAGCGGCTCAACGTGGTACATGCGTGCAGTCA~TGGCAACTTCTTGAAGGACACTTC GGTATCCCTACAGCTTGA~KATACAATTGCCTCTGCCTTGTAGACATGCGTTTTGAAGTTATCAGTCGTGCC ATCGCATCATAGTCATGATAGTATCATCCTATACCCTACTATGTGTCAGCTCAAATACTTTTCTTATAGAT GTGTGAAGTTAAATCTTTGGGGCTTGCAGGAAATTACTGTTCCAATGATTTTCTCAGCAGTAAAAAGTTTT CACGGCTACTGACACAGAAACATCAGGAGACCTTAAGTGACCTTAAGTGCCTGCTTCTACAGAGCAGAGAA GATCTGACCTGCGTCCCACTTTAGGGCTCATATCTGCAACTAATCCTGTTTACTTATGGAGTTCCTCCAGG GGTTTGAAACCAATTAAATGTGGTCTCCAAATCCCTGAAAGTTGAAAAAGATTTTAATTTTATTTGGCAAG AAACAGACCTCAGCTGAGTTACAAATTAGCTCTGACTATCCTTGTTTCTCAAACCAGACACCAAAATCAAG TTCATTTTCATAACACAAAAAGAAAGTGAGCTTTATCCTCCTGCCCCTCGCCCTCAAATTTATTTCATGCT AATAATTTCCTAAGACTGTCTACTGAAGAAAAAATGAAGAAGAAACTCCAACTCCTCTATCTGACACAGAC TCTCTGACACCTCGATCTCACAGGTGCCTACTGTCGTTTCAAGCGAAAACAGGGTGGAGGAAAAAGGGGAG
Fig. 1. Nucleotide sequence of a chicken bcl-2 c D N A and its predicted open reading frame. The sequence of a 2228 bp chicken bcl-2 c D N A was determined on both strands. A n open reading flame was found between 6 2 - 7 5 8 bp in the sequence. Predicted amino acids in the chicken Bcl-2 protein are shown in single-letter code.
111
1
2
3 4 5
chicken bcl-2 was cleaved at position I055 bp with the restriction endonuclease Pac-! and then in vitro transcribed using T7 RNA polymerase [15]. The resulting RNA was then in vitro translated in the presence of [35S]methionine (Amersham; 0.8 p~Ci/mmol) using rabbit reticulocyte lysates (from Promega), and the resulting proteins were subjected to SDS-PAGE analysis in 12% gels. As shown in Fig. 2 (lane 2), SDS-PAGE analysis of the resulting proteins revealed a prominent band at approx. 26 kDa. This approx. 26 kDa protein encoded by the chicken bcl-2 cDNA essentially comigrated in gels with the in vitro translated human Bcl-2 protein (lane 6). Translation of RNA samples prepared by in vitro transcription of plasmids lacking a cDNA insert (lane 4) or having the same chicken bcl-2 cDNA cloned in reversed orientation ('antisense'; lane 3) demonstrated the specificity of these results. These data in Fig. 2 thus confirm the presence of a bona fide ORF encoding a protein of approx. 26 kDa. Fig. 3 compares the predicted amino acid sequences of the chicken, human and mouse Bcl-2 proteins. The overall homology of the chicken Bcl-2 protein with its human and mouse counterparts is 73.3% and 70.3%, respectively. The chicken protein is slightly shorter (232 amino acids) than the human (239 amino acids) and the mouse (236 amino acids) proteins. The first 33 amino acids of these three Bcl-2 proteins are strikingly similar, having 81.8% identical matches and 97% homology when conservative amino acid substitutions are allowed. The conserved NH2-terminal sequence within the chicken Bcl-2 protein is followed by a region bearing little resemblance to the mammalian proteins
6
M W (kD) 106 8050-
3328Bcl-2
19-
Fig. 2. In vitro translation analysis of chicken cDNA. RNA was prepared by T7 RNA polymerase-mediated transcription of pIBI31plasmids containing: the 2.2 kbp chicken bcl-2 cDNA in sense orientation (lane 2); the chicken cDNA fragment in antisense orientation (lane 3); no cDNA (lane 4), or a 910 bp human bcl-2 cDNA (lane 6). In vitro translation of these RNAs was accomplished using rabbit reticulocyte lysates in reactions containing [3SS]methionine. Translation products were analyzed by SDS-PAGE (12% gels). The results shown for lanes 1 and 5 were derived by in vitro translation of no RNA (negative control) or RNA from the Brome Mosaic Virus (positive control).
CKBCL2PEP BC2A$HUMAN BC2A$MOUSE
MAHPGRRGYDNREIVLKYIHYKLSQRGYDWAAGEDRPPVPPAPAPAAAPA MAHAGRTGYDNREIVMKYIHYKLSQRGYEWDAGDVGAAPPGAAPAPGIFS MAQAGRTGYDNREIVMKYIHYKLSQRGYEWDAGDADAAPLGAAPTPGIFS
CKBCL2PEP BC2A$HUMAN BC2ASMOUSE
AVAAAGASSHHRPSPPARLLLVRCPRLRGCAA ....... PPGVHLALRQA SQPGHTPHPAASRDPVARTSPLQTPAAPGAAAGPALSPVPPVVHLALRQA FQPESNPMPAVHREMAARTSPLRPLVA---TAGPALSPVPPCVHLTLRRA .
.
.
.
.
.
.
**
..
.*
**
50 50 50
93 i00 97
***.**.*
CKBCL2PEP BC2A$HUMAN BC2ASMOUSE
GDEFSRRYQRDFAQMSGQLHLTPFTATGRFVAVVEELFRDGVNWVRIVAF GDDFSRRYRGDFAEMSSQLHLTPFTARGRFATVVEELFRDGVNWGRIVAF GDDFSRRYRRDFAEMSSQLHLTPFTARGRFATVVEELFRDGVNWGRIVAF
143 150 147
CKBCL2PEP BC2A$HUMAN BC2ASMOUSE
FEFGGVMCVESVNREMSPLVDNIATWMTEYLNRHLHNWIQDNGGWDAFVE FEFGGVMCVESVNREMSPLVDNIALWMTEYLNRHLHTWIQDNGGWDAFVE FEFGGVMCVESVNREMSPLVDNIALWMTEYLNRHLHTWIQDNGGWDAFVE
193 200 197
CKBCL2PEP BC2A$HUMAN BC2A$MOUSE
LYGNSMRPLFDFSWISLKTILSLVLVGACITLGAYLGHK LYGPSMRPLFDFSWLSLKTLLSLALVGACITLGAYLSHK LYGPSMRPLFDFSWLSLKTLLSLPWVGACITLGAYLGHK
232 239 236
Fig. 3. Comparison of predicted amino acid sequences of the chicken, human and mouse Bcl-2 proteins. The predicted amino acid sequences of the chicken, human and mouse Bcl-2 proteins were aligned using the CLUSTAL program, with adjustment of parameters to minimize gaps [9]. Identical matches are indicated by (*), and conservative replacements by (.).
112 (amino acids 35-82). Like the human and mouse Bcl-2 proteins, however, this region of the chicken Bcl-2 protein is rich in prolines, and therefore presumably devoid of higher order structures such as a-helicies or /3-pleated sheets. Interestingly, somatically acquired point mutations have been described previously within the portion of the human bcl-2 gene corresponding to this proline-rich, non-conserved region in some lymphomas having t(14;18) translocations [16], thus suggesting that this region of Bcl-2 can tolerate sequence alterations. Of further note for this non-conserved region is the presence of a serine residue at position 58 of the chicken Bcl-2 protein, which corresponds to a proline at 59 of the human protein (Fig. 3). Two tumors have been described with single base mutations predicted to produce proline-~ serine substitutions at position 59 of the human Bcl-2 protein [16]. Based on comparison with the chicken sequence reported here, therefore, a proline -~ serine substitution at position 59 of the human Bcl-2 protein may be of little consequence for the function of this oncoprotein, and indeed this prediction is supported by preliminary gene transfer data [unpublished]. Unlike the non-conserved region found at residues 34-82 of the chicken Bcl-2 protein, the last 150 amino acids of this protein are highly homologous with the corresponding regions of the human and mouse forms of the Bcl-2 protein. The three proteins contain 87% identical matches over this region, and this homology rises to 99% when conservative replacements are considered. Hydrophobicity plot analysis of the chicken Bcl-2 protein revealed the presence of a single predicted transmembrane domain at the carboxyl-end of the protein (residues 213-229), followed by two highly charged residues (histidine-231 and lysine-232) that presumably serve to anchor the protein in membranes. Though conservative amino acid substitutions are often permitted within the transmembrane segments of proteins (i.e., replacement of one hydrophobic amino acid with another), the transmembrane domain of the chicken Bcl-2 protein was highly homologous with the predicted transmembrane portions of the human and mouse Bcl-2 proteins (82.4% identical matches) and a stretch of 11 amino acids was perfectly conserved among the three species ( V G A C I T L G A Y L ) suggesting a role for this region beyond simply inserting the protein into membranes. Computer-assisted predictions of secondary structure suggest that the hydrophobic tails of the three Bcl-2 proteins assume an alphahelical conformation. Interestingly, the Bcl-2 proteins from all three species contain a conserved cysteine residue (position 222 in the chicken protein). The presence of cysteines within the transmembrane segments of several proteins has been shown to mediate non-covalent interactions with other proteins [17,18], and thus raises the possibility that this residue helps
Bcl-2 proteins associate with themselves or other proteins within the inner mitochondrial membrane. Though the chicken Bcl-2 protein is predicted to contain three other cysteines, only one of these is conserved among all three species (cysteine-151). Given that the human Bcl-2 protein contains only two cysteines, with one of these residing within membranes, it seems unlikely that intramolecular disulfide bonding contributes to Bcl-2 protein structure. Interestingly, this conserved cysteine (151, 158, 155 of chicken, human and mouse, respectively) is predicted to reside within a portion of the Bcl-2 protein that assumes an a-helical conformation (residues 149-177 in the chicken) and exhibits a high degree of evolutionary conservation at the primary amino acid level (97% exact matches for the three species). Since the local environment of Bcl-2 proteins are predicted to be highly hydrophilic where this conserved cysteine is located, it could potentially participate in disulfide bonding reactions with other proteins. Only biochemical studies of the purified proteins and site-specific mutagenesis however, can determine the functional significance of the two conserved cysteines found within Bcl-2 proteins. With the exception of five potential phosphorylation sites (24,98,120,198,209), five myristoylation sites (85,147,186,220,226) and one amidation site (4), the predicted chicken Bcl-2 protein contained no sequence motifs suggesting a biochemical function for this protein, including no potential glycosylation sites, kinase domains, or binding sites for nucleotides, calcium, or other ions (determined by the P R O S I T E program). The significance of the potential phosphorylation and myristoylation sites in the chicken Bcl-2 protein however, must be interpreted with caution since biochemical evaluations of the human protein have failed to reveal any phosphorylation or myristoylation of this mitochondrial protein (Ref. 19, and unpublished data). The majority of mitochondrial proteins that are encoded by nuclear genes contain within their NH 2termini a positively charged amphiphilic c~-helical region and a proteolytic cleavage site that are necessary of mitochondrial import and processing of these proteins [3]. Using the T R A N S P E P program [20] to search for analogous structures in the predicted Bcl-2 proteins for chicken, human, and mouse, we were unable to identify similarities with other mitochondrial proteins. This finding suggests that Bcl-2 may utilize an alternative pathway for import into mitochondria. Given that the NH2-termini of many mitochondrial proteins play a critical role in their uptake into this organelle, and the striking conservation noted here for the first 33 amino acids of Bcl-2 proteins from chicken, human and mouse, it will be of interest to experimentally assess the importance of the NHz-terminal domain of Bcl-2 proteins for their import into mitochondria. Also of potential inter-
113
est for mitochondrial import is the estimated isoelectric point (pl) of the chicken Bcl-2 protein. Many proteins that undergo import into mitochondria have high p l values [21], and the chicken Bcl-2 protein is no exception with its calculated p l of 8.91. The sequence comparisons reported here for the chicken, human and mouse Bcl-2 proteins should prove useful for designing mutagenesis strategies intended to assess the structural features of Bcl-2 proteins that permit their uptake by mitochondria and that account for their unique function as blockers of apoptosis. We thank John Lambris for assistance with computer analyses, Michael Bishop for the chicken cDNA library, and Yoshihide Tsujimoto and Carlo Croce of the human bcl-2 plasmid. This work was support by grant 3113 from the Council for Tobacco Research. Dr. Reed is a Scholar of the Leukemia Society of America. References I Tsujimoto, Y., Cossman, J., Jaffe, E. and Croce, C. (1987) Science 228, 1440-1443. 2 Hockenbery, D., Nufiez, G., Milliman, C., Schreiber, R.D. and Korsmeyer, S.J. (1990) Nature 348, 334-336. 3 Von Heijne, G., Steppuhn, J. and Herrmann, R.G. (1989) J. Biochem. 180, 535-545.
4 Vaux, D.L., Copy, S. and Adams, J.M. (1988) Nature 335,440-442. 5 Negrini, M., Silini, E., Kozak, C., Tsujimoto, Y. and Croce, C.M. (1987) Cell 49, 455-463. 6 Tsujimoto, Y. and Croce, C,M (1986) Proc. Natl. Acad. Sci. USA 83, 5214-5218. 7 Tabor, S. and Richardson, C.C. (1987) Proc. Natl. Acad. Sci. USA 84, 4767-4771. 8 Myers, E.W. and Miller, W. (1988) Babios 4, 11-17. 9 Higgins, D.G. and Sharp, P.J. (1989) Cabios 5, 151-153. 10 Garnier, J., Osguthorpe, D.J. and Robson, B. (1978) J. Mol. Biol. 120, 97-120. 11 Klein, P., Kanehisa, M. and Delisi, C. (1985) Biochim. Biophys. Acta 815, 468-476. 12 Bairoch, A. (1991) Nucleic Acids Res. 19, 2241-2245. 13 Kozak, M. (1991) J. Biol. Chem. 266, 19867-19870. 14 Shaw, G. and Kamen, R. (1986) Cell 46, 659-667. 15 Tabor, S. and Richardson, C.C. (1985) Proc. Natl. Acad. Sci. USA 82, 1074-1078. 16 Tanaka, S., Louie, D.C., Kant, J.A. and Reed, J.C. (1992) Blood 99, 229-237. 17 Fraser, C.M. (1989) J. Biol. Chem. 264, 9266-9270. 18 Fujii, J., Maruyama, K., Tada, M. and MacLennan, D.H. (1989) J. Biol. Chem. 264, 12950-12955. 19 Tsujimoto, Y., Ikegaki, N. and Croce, C.M. (1987) Oncogene 2, 3-7. 20 Gavel, Y. and Von Heijne, G. (1990) Prot. Eng. 4, 33-37. 21 Hartmann, C., Christen, P. and Jaussi, R. (1991) Nature 352, 762--763.
Biochimica et Biophysica Acta, 1132 (1992) 109-113 © 1992 Elsevier Science Publishers B.V. All rights reserved 0167-4781/92/$05.00
BBAEXP 90389
Short Sequence-Paper
Molecular cloning and DNA sequence analysis of cDNA encoding chicken homologue of the Bcl-2 oncoprotein Dominique L. Cazals-Hatem, Diane C. Louie, Shigeki Tanaka and John C. Reed Department of Pathology and Laboratory Medicine, Uniuersity of Pennsyh'ania, Philadelphia, PA (USA) (Received 4 May 1992)
Key words: Bcl-2; Oncoprotein; Mitochondrion; Nucleotide sequence; Apoptosis; Programmed cell death; (Chicken)
We have isolated a 2228 bp cDNA clone encoding a chicken homologue of the human Bcl-2 oncoprotein by low-stringency hybridization screening of a Agtl0 cDNA library derived from a chicken B-cell lymphoma. DNA sequence analysis of this cDNA revealed an open reading frame predicting a polypeptide of 232 amino acids and an M r of 25 839. The predicted protein is highly homologous to the human (73%) and mouse (70%) Bcl-2 proteins, and contains a hydrophobic stretch of amino acids within its carboxyl-end (213-229) consistent with an integral membrane protein. Areas of very high sequence homology shared by all three Bcl-2 proteins at the NH2-terminus (amino acids 1-33) and within the last 150 amino acids of these proteins suggest the presence of at least two evolutionarily conserved domains within the family of Bcl-2 proteins that may be important either for their targeting to mitochondria or their ability to block programmed cell death. The human bcl-2 proto-oncogene was originally discovered because of its involvement in the t(14;18) translocations found in the majority of non-Hodgkin B-cell lymphomas [1]. This cellular gene encodes a 26 kDa integral membrane protein that has recently been shown to reside in the inner membrane of mitochondria and that shares no significant sequence homology with other known proteins [2]. The predicted amino acid sequence of the human Bcl-2 protein does not suggest the presence of a positively charged amphipathic a-helical domain at its NH2-terminus, unlike many other cellular proteins that undergo import into mitochondria with concomitant proteolytic cleavage of the NHz-leader sequence [3]. Thus, p26-Bcl-2 may utilize other import pathways for its uptake into mitochondria. Bcl-2 is unique among oncoproteins described to date in that it contributes to the cellular accumulation recognized as cancer, not by augmenting the rate of cellular proliferation, but rather by enhancing cell survival. The specific biochemical mechanism by which p26-Bcl-2 accomplishes this remains enigmatic, but must somehow be related to its ability to markedly
Correspondence to: J.C. Reed, La Jolla Cancer Research Institute, Cancer Research Center, 10901 N Torrey Pines Road, La Jolla, CA 92037, USA. The sequence data in this paper have been submitted to the EMBL/Genbank Data Libraries under the accession number Zl1961.
interfere with programmed cell death (also termed 'apoptosis') [2,4]. The human Bcl-2 protein contains no sequence motifs that might suggest its biochemical function, including no nucleotide-binding, Ca2+-bind ing, or metal-binding regions. Likewise, molecular cloning and DNA sequence analysis of the mouse bcl-2 gene has added little additional information that might help to identify functionally important domains within Bcl-2 proteins, since the predicted mouse protein is 85% homologous to its human counterpart and this homology is distributed throughout the length of the proteins [5]. In an effort to define evolutionarily conserved regions within Bcl-2 proteins that may coincide with functionally important domains, we have attempted to clone bcl-2 gene homologs from non-mammalian species. Here we report the nucleotide and deduced amino acid sequences for a cDNA encoding a chicken homolog of Bcl-2. Comparisons of this avian and the previously published mammalian sequences revealed several evolutionarily conserved features that provide potentially important insights into the structure-function relations of these unique oncoproteins. For these experiments, a chicken B-cell lymphoma cDNA library constructed in Agtl0 phage (gift from M. Bishop) was screened by low-stringency hybridization using a 910 bp human bcl-2 cDNA probe pB4 which contains the open reading frame (ORF) for p26-Bcl-2 plus 64 bp of upstream and 129 bp of downstream untranslated sequence [6]. Approx. 20 000 phages/plate were hybridized with this 32p-labeled probe at 50°C in 6 × SSC containing 0.1% sodium dodecyl sulfate (SDS),
110 150 p~g/ml sonicated salmon sperm DNA, and 5 × Denhardt's solution. The final washes were performed for 30 min at 50°C in 2 × SSC, 0.1% SDS. A phage clone containing a 2.2 kb EcoRI insert was plaquepurified, and its insert was subcloned into the plasmid plBl31 (International Biotechnologies) for D N A sequence analysis. The complete nucleotide sequence was determined for both strands by the dideoxynucleotide chain termination method, using Sequenase version 2.0 (USB) [7]. Sequence comparisons of the chicken c D N A with the published sequences for human and mouse bcl-2 cDNAs and predicted proteins were accomplished with the aid of the NALIGN, PAL1GN and CLUSTAL computer programs [8,9]. Computer predictions of protein secondary structure, hydropathic index determination, and the prediction of sites and signatures in the protein sequence were performed using the GARNIER, SOAP and PROSITE programs, respectively [10-12]. The nucleotide and amino acid sequences of the chicken bcl-2 c D N A clone are presented in Fig. 1. The longest ORF within this 2228 bp c D N A was 699 nucleotides in length and began 62 bp from the 5' end of the cDNA, ending with a TAG stop codon at position 758. The translation initiation codon and surrounding bases shared 8 of 10 matches with the Kozak concensus sequence for eukaryotic translation start sites [13]. This ORF exhibited 75% and 72% homology at the nu1 62 116 170 224
278 332
386 440 494 548 602 656 71o 764 835 906 977 1048 1119 1190 1261 1332 1403 1474 1545 1616 1687 1758 1829 1900 1971
ATG M ATC I CCG P GCT A AGG R CGC R TCG S GTG V GAG E GTG V TGG W ATG M GTT V
GCT A CAC H CCC P GGA G TGC C CAG Q GGC G GAG E TTC F GAC D ATC I AGG R CTG L
cleotide level with the regions corres-ponding to the ORFs previously identified for human and mouse bcl-2 cDNAs, respectively. It is unlikely that this chicken bcl-2 c D N A contains the complete 3' untranslated regions of the full-length chicken bcl-2 mRNA since neither a poly(A) tail nor A A T A A A motifs for transcript termination and polyadenylation were found within the 1468 bp of sequence located 3' of the ORF. The 61 bp of sequence found 5' of the chicken bcl-2 cDNA's ORF also may not represent the full extent of 5' untranslated sequences in the full-length chicken bcl-2 mRNA, given that the mouse and human bcl-2 mRNAs begin approx. 1.5 kbp upstream of their ORFs [5,6]. Several ATTA and ATTTA motifs were located within the 3' untranslated portion of the chicken bcl-2 cDNA. Similar sequence motifs have been identified previously in human and mouse bcl-2 cDNAs and have been associated with rapid turnover of many mRNAs [14]. Otherwise, the untranslated regions within the chicken bcl-2 c D N A exhibited no striking sequence homology with the corresponding portions of mammalian bcl-2 cDNAs. The overall homology between the 3' untranslated region of the chicken bcl-2 c D N A and its human and mouse counterparts was 53% and 42%, respectively. To experimentally confirm the presence of the predicted open reading frame in the chicken c D N A clone (62-758 bp), the plBI-31 plasmid containing the 2.2 kb
AATTCCGCCCCCCCCCCCCCCTTCCTCTCCCCCTCCCACTCGCTGCTTCCCCTCGGAAACC CAC CCC GGG AGA AGA GGC TAC GAC AAC CGC GAG ATA GTG CTG H P G R R G Y D N R E I V L TAT ~ CTC TCG CAG CGG GGC TAC GAC TGG GCC GCC GGC GAG Y K L S Q R G Y D W A A G E GTG CCC CCG GCC CCG GCT CCC GCT GCT GCT CCC GCC GCG CTC V P P A P A P A A A P A A V GCC TCC TCC CAC CAC CGC CCG AGC CCC CCG GCT CGG CTG CTG A S S H H R P S P P A R L L C C C C G G C T G A G G G G C T G C G C C G C G C C T CCC G G C G T C C A C C T C P R L R G C A A P P G V H L GCC GGG GAC GAG TTC TCG CGC CGC TAC CAG AGG GAC TTC GCC A G D E F S R R Y Q R D F A CAG CTG CAC CTG ACG CCC TTC ACG GCC ACC GGC CGC TTC GTG Q L H L T P F T A T G R F V GAG CTC TTC CGT GAT GGG GTC AAC TGG GTC CGG ATC GTC GCC E L F R D G V N W V R I V A GGC GGC GTG ATG TGC GTC GAG AGC GTC AAC CGG GAG ATG TCG G G V M C V E S V N R E M S AAC ATT GCC ACC TGG ATG ACC GAG TAC CTG AAC CGG CAC CTG N I A T W S T E Y L N R I[ L CAG GAC AAC GGA GGA TGG GAT GCC TTT GTG GAA TTG IAC GGC Q D N G G W D A F V E L Y C CCT TTG TTC GAT TTC TCC TGG ATC TCT CTG AAG ACC ATC CTG P L F D F S W I S L K T I L GTG GGA GCT TGC ATC ACT CTT GGC GCT TAT CTT GGA CAT AAG V G A C I T L G A Y L G H K
AAG K GAC D GCT A CTA L GCC A CAG Q GCC A TTC F CCG P CAC H AAC N AGC S TAG ,
TAC Y 18 AGG R 36 GCT A 54 GTG V 72 CTG L 90 ATG M i0~ GTC V ]26 TTC F 144 CTG L 162 AAC N 180 AGT S 198 CTG L 216 AGT 232
CACCCAGTTTATCGTGGATTACAAAGTCTTTCAAAACAGCAAAATAATATTTTTTTTCTGATGCACAATGG TATTTTATGAGCA~CAACTTCCTGGCTAAAGATTAAATCAGCTATTACTGCCAAAGGAAATATCATTTA TTTTTACATTGCAGG~%AATTATTTATTA~TATATTTACATTTTA.ILCCTGCTATTTTCAGAAACTCTGC
AAGTTGTATCTGTCATCACATCATGTTGTTATTCTTAACTTTAATGAGCCCCAAAGCATTTTAGGTTCTTT TTTTAATTAATGCAG~GTCTGGATTATT~ACCTGCGAAGAGCAAGTATACTGACAAAGACCTACCTGCT
TACACTTAGGAAGGCAGTTACTTGAGCTGCTAAAGGCTGTGTAATTGAGTAGATTTCCTTGCAATTTTTGG ATGGCCGGGAGGGCTCTGAAAGAAAAGGTCAACGTGAACCAAGTTCTAATCCTGTCCTGCCAGCGAGAGCA GTCACAGAGAGTAAGCGGCTCAACGTGGTACATGCGTGCAGTCA~TGGCAACTTCTTGAAGGACACTTC GGTATCCCTACAGCTTGA~KATACAATTGCCTCTGCCTTGTAGACATGCGTTTTGAAGTTATCAGTCGTGCC ATCGCATCATAGTCATGATAGTATCATCCTATACCCTACTATGTGTCAGCTCAAATACTTTTCTTATAGAT GTGTGAAGTTAAATCTTTGGGGCTTGCAGGAAATTACTGTTCCAATGATTTTCTCAGCAGTAAAAAGTTTT CACGGCTACTGACACAGAAACATCAGGAGACCTTAAGTGACCTTAAGTGCCTGCTTCTACAGAGCAGAGAA GATCTGACCTGCGTCCCACTTTAGGGCTCATATCTGCAACTAATCCTGTTTACTTATGGAGTTCCTCCAGG GGTTTGAAACCAATTAAATGTGGTCTCCAAATCCCTGAAAGTTGAAAAAGATTTTAATTTTATTTGGCAAG AAACAGACCTCAGCTGAGTTACAAATTAGCTCTGACTATCCTTGTTTCTCAAACCAGACACCAAAATCAAG TTCATTTTCATAACACAAAAAGAAAGTGAGCTTTATCCTCCTGCCCCTCGCCCTCAAATTTATTTCATGCT AATAATTTCCTAAGACTGTCTACTGAAGAAAAAATGAAGAAGAAACTCCAACTCCTCTATCTGACACAGAC TCTCTGACACCTCGATCTCACAGGTGCCTACTGTCGTTTCAAGCGAAAACAGGGTGGAGGAAAAAGGGGAG
Fig. 1. Nucleotide sequence of a chicken bcl-2 c D N A and its predicted open reading frame. The sequence of a 2228 bp chicken bcl-2 c D N A was determined on both strands. A n open reading flame was found between 6 2 - 7 5 8 bp in the sequence. Predicted amino acids in the chicken Bcl-2 protein are shown in single-letter code.
111
1
2
3 4 5
chicken bcl-2 was cleaved at position I055 bp with the restriction endonuclease Pac-! and then in vitro transcribed using T7 RNA polymerase [15]. The resulting RNA was then in vitro translated in the presence of [35S]methionine (Amersham; 0.8 p~Ci/mmol) using rabbit reticulocyte lysates (from Promega), and the resulting proteins were subjected to SDS-PAGE analysis in 12% gels. As shown in Fig. 2 (lane 2), SDS-PAGE analysis of the resulting proteins revealed a prominent band at approx. 26 kDa. This approx. 26 kDa protein encoded by the chicken bcl-2 cDNA essentially comigrated in gels with the in vitro translated human Bcl-2 protein (lane 6). Translation of RNA samples prepared by in vitro transcription of plasmids lacking a cDNA insert (lane 4) or having the same chicken bcl-2 cDNA cloned in reversed orientation ('antisense'; lane 3) demonstrated the specificity of these results. These data in Fig. 2 thus confirm the presence of a bona fide ORF encoding a protein of approx. 26 kDa. Fig. 3 compares the predicted amino acid sequences of the chicken, human and mouse Bcl-2 proteins. The overall homology of the chicken Bcl-2 protein with its human and mouse counterparts is 73.3% and 70.3%, respectively. The chicken protein is slightly shorter (232 amino acids) than the human (239 amino acids) and the mouse (236 amino acids) proteins. The first 33 amino acids of these three Bcl-2 proteins are strikingly similar, having 81.8% identical matches and 97% homology when conservative amino acid substitutions are allowed. The conserved NH2-terminal sequence within the chicken Bcl-2 protein is followed by a region bearing little resemblance to the mammalian proteins
6
M W (kD) 106 8050-
3328Bcl-2
19-
Fig. 2. In vitro translation analysis of chicken cDNA. RNA was prepared by T7 RNA polymerase-mediated transcription of pIBI31plasmids containing: the 2.2 kbp chicken bcl-2 cDNA in sense orientation (lane 2); the chicken cDNA fragment in antisense orientation (lane 3); no cDNA (lane 4), or a 910 bp human bcl-2 cDNA (lane 6). In vitro translation of these RNAs was accomplished using rabbit reticulocyte lysates in reactions containing [3SS]methionine. Translation products were analyzed by SDS-PAGE (12% gels). The results shown for lanes 1 and 5 were derived by in vitro translation of no RNA (negative control) or RNA from the Brome Mosaic Virus (positive control).
CKBCL2PEP BC2A$HUMAN BC2A$MOUSE
MAHPGRRGYDNREIVLKYIHYKLSQRGYDWAAGEDRPPVPPAPAPAAAPA MAHAGRTGYDNREIVMKYIHYKLSQRGYEWDAGDVGAAPPGAAPAPGIFS MAQAGRTGYDNREIVMKYIHYKLSQRGYEWDAGDADAAPLGAAPTPGIFS
CKBCL2PEP BC2A$HUMAN BC2ASMOUSE
AVAAAGASSHHRPSPPARLLLVRCPRLRGCAA ....... PPGVHLALRQA SQPGHTPHPAASRDPVARTSPLQTPAAPGAAAGPALSPVPPVVHLALRQA FQPESNPMPAVHREMAARTSPLRPLVA---TAGPALSPVPPCVHLTLRRA .
.
.
.
.
.
.
**
..
.*
**
50 50 50
93 i00 97
***.**.*
CKBCL2PEP BC2A$HUMAN BC2ASMOUSE
GDEFSRRYQRDFAQMSGQLHLTPFTATGRFVAVVEELFRDGVNWVRIVAF GDDFSRRYRGDFAEMSSQLHLTPFTARGRFATVVEELFRDGVNWGRIVAF GDDFSRRYRRDFAEMSSQLHLTPFTARGRFATVVEELFRDGVNWGRIVAF
143 150 147
CKBCL2PEP BC2A$HUMAN BC2ASMOUSE
FEFGGVMCVESVNREMSPLVDNIATWMTEYLNRHLHNWIQDNGGWDAFVE FEFGGVMCVESVNREMSPLVDNIALWMTEYLNRHLHTWIQDNGGWDAFVE FEFGGVMCVESVNREMSPLVDNIALWMTEYLNRHLHTWIQDNGGWDAFVE
193 200 197
CKBCL2PEP BC2A$HUMAN BC2A$MOUSE
LYGNSMRPLFDFSWISLKTILSLVLVGACITLGAYLGHK LYGPSMRPLFDFSWLSLKTLLSLALVGACITLGAYLSHK LYGPSMRPLFDFSWLSLKTLLSLPWVGACITLGAYLGHK
232 239 236
Fig. 3. Comparison of predicted amino acid sequences of the chicken, human and mouse Bcl-2 proteins. The predicted amino acid sequences of the chicken, human and mouse Bcl-2 proteins were aligned using the CLUSTAL program, with adjustment of parameters to minimize gaps [9]. Identical matches are indicated by (*), and conservative replacements by (.).
112 (amino acids 35-82). Like the human and mouse Bcl-2 proteins, however, this region of the chicken Bcl-2 protein is rich in prolines, and therefore presumably devoid of higher order structures such as a-helicies or /3-pleated sheets. Interestingly, somatically acquired point mutations have been described previously within the portion of the human bcl-2 gene corresponding to this proline-rich, non-conserved region in some lymphomas having t(14;18) translocations [16], thus suggesting that this region of Bcl-2 can tolerate sequence alterations. Of further note for this non-conserved region is the presence of a serine residue at position 58 of the chicken Bcl-2 protein, which corresponds to a proline at 59 of the human protein (Fig. 3). Two tumors have been described with single base mutations predicted to produce proline-~ serine substitutions at position 59 of the human Bcl-2 protein [16]. Based on comparison with the chicken sequence reported here, therefore, a proline -~ serine substitution at position 59 of the human Bcl-2 protein may be of little consequence for the function of this oncoprotein, and indeed this prediction is supported by preliminary gene transfer data [unpublished]. Unlike the non-conserved region found at residues 34-82 of the chicken Bcl-2 protein, the last 150 amino acids of this protein are highly homologous with the corresponding regions of the human and mouse forms of the Bcl-2 protein. The three proteins contain 87% identical matches over this region, and this homology rises to 99% when conservative replacements are considered. Hydrophobicity plot analysis of the chicken Bcl-2 protein revealed the presence of a single predicted transmembrane domain at the carboxyl-end of the protein (residues 213-229), followed by two highly charged residues (histidine-231 and lysine-232) that presumably serve to anchor the protein in membranes. Though conservative amino acid substitutions are often permitted within the transmembrane segments of proteins (i.e., replacement of one hydrophobic amino acid with another), the transmembrane domain of the chicken Bcl-2 protein was highly homologous with the predicted transmembrane portions of the human and mouse Bcl-2 proteins (82.4% identical matches) and a stretch of 11 amino acids was perfectly conserved among the three species ( V G A C I T L G A Y L ) suggesting a role for this region beyond simply inserting the protein into membranes. Computer-assisted predictions of secondary structure suggest that the hydrophobic tails of the three Bcl-2 proteins assume an alphahelical conformation. Interestingly, the Bcl-2 proteins from all three species contain a conserved cysteine residue (position 222 in the chicken protein). The presence of cysteines within the transmembrane segments of several proteins has been shown to mediate non-covalent interactions with other proteins [17,18], and thus raises the possibility that this residue helps
Bcl-2 proteins associate with themselves or other proteins within the inner mitochondrial membrane. Though the chicken Bcl-2 protein is predicted to contain three other cysteines, only one of these is conserved among all three species (cysteine-151). Given that the human Bcl-2 protein contains only two cysteines, with one of these residing within membranes, it seems unlikely that intramolecular disulfide bonding contributes to Bcl-2 protein structure. Interestingly, this conserved cysteine (151, 158, 155 of chicken, human and mouse, respectively) is predicted to reside within a portion of the Bcl-2 protein that assumes an a-helical conformation (residues 149-177 in the chicken) and exhibits a high degree of evolutionary conservation at the primary amino acid level (97% exact matches for the three species). Since the local environment of Bcl-2 proteins are predicted to be highly hydrophilic where this conserved cysteine is located, it could potentially participate in disulfide bonding reactions with other proteins. Only biochemical studies of the purified proteins and site-specific mutagenesis however, can determine the functional significance of the two conserved cysteines found within Bcl-2 proteins. With the exception of five potential phosphorylation sites (24,98,120,198,209), five myristoylation sites (85,147,186,220,226) and one amidation site (4), the predicted chicken Bcl-2 protein contained no sequence motifs suggesting a biochemical function for this protein, including no potential glycosylation sites, kinase domains, or binding sites for nucleotides, calcium, or other ions (determined by the P R O S I T E program). The significance of the potential phosphorylation and myristoylation sites in the chicken Bcl-2 protein however, must be interpreted with caution since biochemical evaluations of the human protein have failed to reveal any phosphorylation or myristoylation of this mitochondrial protein (Ref. 19, and unpublished data). The majority of mitochondrial proteins that are encoded by nuclear genes contain within their NH 2termini a positively charged amphiphilic c~-helical region and a proteolytic cleavage site that are necessary of mitochondrial import and processing of these proteins [3]. Using the T R A N S P E P program [20] to search for analogous structures in the predicted Bcl-2 proteins for chicken, human, and mouse, we were unable to identify similarities with other mitochondrial proteins. This finding suggests that Bcl-2 may utilize an alternative pathway for import into mitochondria. Given that the NH2-termini of many mitochondrial proteins play a critical role in their uptake into this organelle, and the striking conservation noted here for the first 33 amino acids of Bcl-2 proteins from chicken, human and mouse, it will be of interest to experimentally assess the importance of the NHz-terminal domain of Bcl-2 proteins for their import into mitochondria. Also of potential inter-
113
est for mitochondrial import is the estimated isoelectric point (pl) of the chicken Bcl-2 protein. Many proteins that undergo import into mitochondria have high p l values [21], and the chicken Bcl-2 protein is no exception with its calculated p l of 8.91. The sequence comparisons reported here for the chicken, human and mouse Bcl-2 proteins should prove useful for designing mutagenesis strategies intended to assess the structural features of Bcl-2 proteins that permit their uptake by mitochondria and that account for their unique function as blockers of apoptosis. We thank John Lambris for assistance with computer analyses, Michael Bishop for the chicken cDNA library, and Yoshihide Tsujimoto and Carlo Croce of the human bcl-2 plasmid. This work was support by grant 3113 from the Council for Tobacco Research. Dr. Reed is a Scholar of the Leukemia Society of America. References I Tsujimoto, Y., Cossman, J., Jaffe, E. and Croce, C. (1987) Science 228, 1440-1443. 2 Hockenbery, D., Nufiez, G., Milliman, C., Schreiber, R.D. and Korsmeyer, S.J. (1990) Nature 348, 334-336. 3 Von Heijne, G., Steppuhn, J. and Herrmann, R.G. (1989) J. Biochem. 180, 535-545.
4 Vaux, D.L., Copy, S. and Adams, J.M. (1988) Nature 335,440-442. 5 Negrini, M., Silini, E., Kozak, C., Tsujimoto, Y. and Croce, C.M. (1987) Cell 49, 455-463. 6 Tsujimoto, Y. and Croce, C,M (1986) Proc. Natl. Acad. Sci. USA 83, 5214-5218. 7 Tabor, S. and Richardson, C.C. (1987) Proc. Natl. Acad. Sci. USA 84, 4767-4771. 8 Myers, E.W. and Miller, W. (1988) Babios 4, 11-17. 9 Higgins, D.G. and Sharp, P.J. (1989) Cabios 5, 151-153. 10 Garnier, J., Osguthorpe, D.J. and Robson, B. (1978) J. Mol. Biol. 120, 97-120. 11 Klein, P., Kanehisa, M. and Delisi, C. (1985) Biochim. Biophys. Acta 815, 468-476. 12 Bairoch, A. (1991) Nucleic Acids Res. 19, 2241-2245. 13 Kozak, M. (1991) J. Biol. Chem. 266, 19867-19870. 14 Shaw, G. and Kamen, R. (1986) Cell 46, 659-667. 15 Tabor, S. and Richardson, C.C. (1985) Proc. Natl. Acad. Sci. USA 82, 1074-1078. 16 Tanaka, S., Louie, D.C., Kant, J.A. and Reed, J.C. (1992) Blood 99, 229-237. 17 Fraser, C.M. (1989) J. Biol. Chem. 264, 9266-9270. 18 Fujii, J., Maruyama, K., Tada, M. and MacLennan, D.H. (1989) J. Biol. Chem. 264, 12950-12955. 19 Tsujimoto, Y., Ikegaki, N. and Croce, C.M. (1987) Oncogene 2, 3-7. 20 Gavel, Y. and Von Heijne, G. (1990) Prot. Eng. 4, 33-37. 21 Hartmann, C., Christen, P. and Jaussi, R. (1991) Nature 352, 762--763.
