75, No. 5, pp. 2165-2169, May 1978. Biochemistry. Large and small tumor antigens from simian virus 40 have identical amino termini mapping at 0.65 map units.
Proc. Natl. Acad. Sci. USA
Vol. 75, No. 5, pp. 2165-2169, May 1978 Biochemistry
Large and small tumor antigens from simian virus 40 have identical amino termini mapping at 0.65 map units (cell-free protein synthesis/NHrterminal acetylation/protein sequencing/model for early simian virus 40 proteins/mRNA splicing)
EVA PAUCHA*, ANDREW MELLOR*, ROBERT HARVEY*, ALAN E. SMITH*, RODNEY M. HEWICKt, AND MICHAEL D. WATERFIELDt * Translation and tProtein Chemistry Laboratories, Imperial Cancer Research Fund, Lincoln's Inn Fields, London WC2A 3PX, England Communicated by Paul Berg, February 13, 1978
ABSTRACT Large and small tumor (T)antigens of simian virus 40 were synthesized in vitro with L-cell extracts that had been treated by the method of Palmiter to prevent amino-terminal acetylation of nascent proteins. Partial amino-terminal amino acid sequences of both forms of T-antigen were determined and found to be identical. Methionine residues were located at positions 1 and 14, a lysine residue at position 3, and leucine residues at positions 5, 11, 13, 16, 17, and 19. These amino acid sequence data match perfectly the amino acid sequence predicted from a sequence of nucleotides in the E strand of simian virus 40 DNA which begins near the junction between HindII/III fragments A and C at about 0.65 map units. This strongly suggests that the sequence coding for the amino terminus of both proteins is located at this position. Furthermore, the data are consistent with a model for the synthesis of both forms of T-antigen that predicts that (f) small T-antigen is coded for by a sequence of nucleotides from the 5' end of the early region and (ii) large T-antigen is coded for by nucleotide sequences from two noncontiguous regions of simian virus 40 DNA.
Simian virus 40 (SV40) is a small double-stranded DNA virus which, in addition to its ability to reproduce lytically, can also cause tumors in animals and transform cells grown in culture (see refs. 1 and 2 for reviews). SV40-induced tumor cells, SV40-transformed cells, and cells productively infected with the virus all express an antigen, referred to as tumor antigen (T-Ag). T-Ag has been implicated as an important control element in productively infected cells, in which it is necessary for the synthesis and may influence the transcription of viral DNA. In transformed cells T-Ag appears to be necessary for both initiation and maintenance of transformation (1, 2). There is strong evidence from both genetic and biochemical studies that T-Ag is at least predominantly coded for by the E strand of the early region of SV40 DNA (3-6). When isolated from productively infected or transformed cells under conditions that minimize proteolytic degradation, two major forms of SV40 T-Ag are detected (5, 6). These are referred to as large-T and small-t and have apparent molecular weights of 90,000-100,000 and 15,000-20,000, respectively. Fingerprint analysis of both forms of SV40 T-Ag has shown that most of the [a5S]methionine-labeled tryptic peptides of small-t are also present in large-T (5, 6). Presumably, therefore, both forms of T-Ag share amino acid sequences and are coded for, at least in part, by a common region of SV40 DNA. Studies with specific deletion mutants of SV40 have shown that a deletion at 0.21 map units on the DNA gives a shorter form of large-T with apparently wild-type small-t (7), while several mutants having deletions in the region 0.54-0.59 give The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.' S. C. §1734 solely to indicate this fact.
apparently wild-type large-T (8) with siiall-t either absent (7) or present as putative smaller fragments (M. J. Sleigh, W. C. Topp, and J. Sambrook, personal communication; and unpublished data). A model based on these results has been pro-
posed for the synthesis of both forms of T-Ag (7). The model specifies that (i) small-t is coded for by a contiguous nucleotide sequence in SV40 DNA between approximately 0.65 and 0.55 map units, and (ii) the sequences coding for large-T begin at the same position, include a region that overlaps with the sequences coding for the amino-terminal half of small-t extending to about 0.59 map units, and, after an interruption in the sequence, continue from about 0.54 until 0.17 map units. A direct prediction of this model is that both large-T and small-t share the same amino-terminal amino acid sequence. We have tested this prediction by synthesizing both forms of T-Ag in vitro and examining their amino-terminal amino acid sequences. MATERIALS AND METHODS Materials. Pig heart citrate synthase (EC 4.1.3.7) (130 units/mg, 6.7 mg/ml) and oxaloacetate were obtained from Sigma. [3H]Leucine (specific activity 58 Ci/mmol), [3H]lysine (specific activity 67 Ci/mmol), [3H]proline (specific activity 105 Ci/mmol), [35S]methionine (specific activity 900 Ci/mmol), and [14C]acetyl-CoA (specific activity 58 mCi/mmol) were obtained from the Radiochemical Centre. The source of other reagents for cell-free synthesis and immunoprecipitation have been described (7, 9). Reagents for amino acid sequence analysis were obtained from Beckman (United Kingdom) or Pierce Chemicals (Illinois). Assay for Citrate Synthase Activity. Citrate synthase was collected by centrifugation (2 min, 10,000 X g) from a suspension in ammonium sulfate, dissolved in an equal volume of 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonate (Hepes) buffer (pH 7.0), and dialyzed for 4 hr at 40 against the same buffer. Incubations mixtures (25 ,u) contained 10 mM Hepes (pH 7.0), 1 mM oxaloacetate neutralized with KOH, 34 units of citrate synthase per ml, 0.1 ,uCi of [14C]acetyl-CoA, and in some cases 10,ul of a mixture containing L-cell extract and all the components necessary for cell-free protein synthesis (9). After incubation at 340, the reaction products were separated by chromatography on cellulose thin-layer chromatography sheets in a buffer containing isobutyric acid/i M NaOH/water (57:35:8) and subsequently detected by autoradiography. Cell-Free Synthesis and Inmunoprecipitation. mRNA was isolated from CV1 cells 72 hr after infection with SV40 (strain 777) and translated in a L-cell cell-free system as described (6, 9). Reaction mixtures (500-1250,gl) included 400,uCi of each Abbreviations: SV40, simian virus 40; T-Ag, tumor antigen; Hepes, 4-(2-hydroxethyl)-1-piperazineethanesulfonate; PTH, phenylthiohydantoin.
2165
2166
Biochemistry: Paucha et al.
tritiated amino acid per ml and 2.4 mCi of [a5S]methionine per ml. The endogenous pool of acetyl-CoA was depleted by adding 1 mM oxaloacetate and 34 units of citrate synthase per ml and the mixture, complete with the components necessary for protein synthesis, was preincubated for 10 min at 34°. After addition of mRNA (80 ,ug/ml) incubation was continued for 90 min at 340. Pretreatment of the cell-free extract reduced the subsequent yield of SV40 T-Ag by about 25%. Immunoprecipitation with antiserum against T-Ag was carried out essentially as described (7) except that the preprecipitation step with normal sheep serum was omitted. The immune complexes were collected by adsorption to protein A-bearing Staphylococci (10) and eluted with buffer containing 2% sodium dodecyl sulfate (NaDodSO4). The eluate was applied to a two-tier polyacrylamide gel (7) with a single loading slot. After electrophoresis, the gel was wrapped in Saran-wrap and the separated proteins were visualized by autoradiography (6 hr). Amino-Terminal Sequence Analysis of Radiolabeled Protein. Protein was eluted from the excised gel slices in 50 mM ammonium bicarbonate containing 0.1% NaDodSO4, 0.2% 2-mercaptoethanol, and 100-150 nmol of horse heart apomyoglobin. after the mixture was shaken for 18 hr at 370 the eluted protein was precipitated with trichloroacetic acid and collected on siliconized glass fiber filters as described (11). Protein was eluted from the filters with 98% formic acid and loaded directly into a Beckman 890C sequencer. The sequence of the protein was determined in 0.1 M Quadrol buffer by a program similar to that described by Brauer et al. (12). A mixture of phenylthiohydantoin (PTH)-amino acids (10 nmol of each derivative), corresponding to the radioactive amino acids used for in vitro synthesis, and PTH-norleucine as internal standard was added to each fraction from the sequencer. After conversion of the thiazolinone derivatives to PTH-amino acids (13) the fractions were analyzed by high-performance liquid chromatography on a Zorbax ODS reverse-phase column (25 X 0.46 cm, Du Pont), on a Waters high-pressure liquid chromatograph with the gradient elution system described by Zimmerman et al. (14). PTH-amino acids were detected by absorbance at 254 nm with an Altex analytical UV detector (model 153, Altex Scientific Inc., Berkeley, CA), equipped with a peak detector coupled to a fraction collector (Gilson, model TDC 80). Each PTHamino acid was collected separately and radioactivity was detected by liquid scintillation counting. RESULTS Conditions that Prevent NH2-Terminal Acetylation of Nascent Proteins. We have developed a microsequencing method to determine the amino-terminal amino acid sequence of radioactively labeled proteins. The method has been successfully applied both to proteins isolated from virus particles and to proteins synthesized in cell-free extracts (unpublished results). However, our initial attempts to determine the amino-terminal sequence of SV40 T-Ags using material either isolated from cells or synthesized in vitro proved negative and suggested that the amino terminus was chemically blocked. Palmiter (15) has recently shown that nascent ovalbumin is acetylated at its amino terminus but that the protein can be prepared in a form that is unblocked, and whose sequence can therefore be determined by synthesizing it in vitro in a reticulocyte cell-free extract that had been depleted of acetyl donor. In the same way we have found that SV40 large-T and small-t synthesized in L-cell extracts previously depleted of acetyl-CoA have unblocked amino termini. L-cell extracts were depleted of acetyl-CoA by pretreatment
Proc. Natl. Acad. Sci. USA 75 (1978)
--
1
2
3
4
5
A
6
FIG. 1. Conditions for removal of acetyl-CoA. Incubation mixtures contained [14C]acetyl-CoA (tracks 1-6), citrate synthase (tracks 2 and 4-6), oxaloacetate (tracks 3-6), and L-cell extract and the components necessary for cell-free protein synthesis (track 6). Incubation was for 10 min at 340 except for track 4 (30 sec, 00), and products were analyzed by chromatography, followed by autoradiography. 0, Origin; A, acetyl-CoA; B, presumptive citric acid.
with citrate synthase in the presence of oxaloacetate (15). This results in the conversion of the oxaloacetate to citric acid by transfer of an acetyl group from acetyl-CoA. Since we have not been able to label SV40 T-Ags or any other proteins synthesized in vitro to a sufficiently high specific activity to demonstrate that they are acetylated in in the cell-free system, we have no direct assay to test whether the pretreatment prevents subsequent acetylation of nascent proteins. However, as shown in Fig. 1, in the presence of both citrate synthase and oxaloacetate, exogenous [14C]acetyl CoA is converted to a more slowly migrating form, which we assume but have not yet proved to be citric acid. The conversion of acetyl-CoA to putative citric acid B t. 2 3 4
A
- ^_
L clr e -1
- -
_ -
__,
.--_ vp-i4M VP_1l
---
130 94
81
6C
-
.-.60
..,
-
4C
FIG. 2. Large-scale preparation and analysis of labeled large-T and small-t. (A) Part of an autoradiograph (6 hr) of large-T and small-t synthesized in a preincubated cell-free extract containing tritiated leucine, lysine, and proline and [35S]methionine and separated by electrophoresis on a layer (5 cm) of 15% polyacrylamide overlaid with a layer (5 cm) of 7.5% polyacrylamide. (B) Autoradiograph of large-T and small-t synthesized in vitro in the presence of labeled amino acids after elution and further analysis on a 15% polyacrylamide gel. Track 1, eluted large-T; track 2, a sample of the immunoprecipitate before separation on the preparative scale gel; track 3, eluted small-t; track 4, [14C]carboxymethylated protein markers (7). The numbers refer to the apparent molecular weights (in thousands) of the markers. Note that VP-1 nonspecifically precipitates during our immunoprecipitation procedure (6, 7).
Proc. Natl. Acad. Sci. USA 75 (1978)
Biochemistry: Paucha et al. 1i
A
2167 A
0
o
2
x
x
E 0.4
a
14
E
vU en
va w
1
1
0o 2
B
B
0 0 Il,
x
aE 1 I
32
0
x 30 x
0.4
E
D
I-
0 x
E 0.2
a I-
20 Step number
WIF
10
20
30
Step number
FIG. 3. Amino-terminal analysis of large-T. The sequence of large-T, labeled in vitro with tritiated leucine, lysine, and proline and [35S]methionine, was determined as described in the text. The radioactivity of material chromatographing with the marker PTHamino acid derivatives of leucine, lysine, proline, and methionine was measured in a scintillation counter, and the results were plotted as counts releAsed at each cycle for each amino acid. (A) Methionine; (B) lysine; (C) leucine; (D) proline. occurs at 00 in less than 30 sec and is not affected by the presence of L-cell extract and the components necessary for cell-free protein synthesis. We assume that during the preincubation step
(10 min at 34°) all endogenous acetyl-CoA present in the L-cell extract is similarly converted to an inactive form. NH2-Terminal Sequence Determination of Large-T and Small-t. SV40 large-T and small-t were synthesized in vitro by using a preincubated L-cell S30 and mRNA isolated from productively infected CV1 cells. A mixture containing [s5S]methionine and three or four tritiated amino acids was added as the radioactive label. The T-Ags were reacted with antisera raised in tumor-bearing hamsters and the immune complexes collected by adsorption to protein A present on the surface of Staphylococcus aureus (6, 7, 10). The antigens were separated on a preparative scale polyacrylamide gel, detected by autoradiography of the wet gel (Fig. 2A), excised, and eluted.
FIG. 4. Amino-terminal analysis of small-t. Small-t, labeled in vitro with tritiated leucine and lysine and [35S]methionine, was subjected to the same analysis described in Fig. 3 for large-T. The results are plotted as counts released at each cycle for each amino acid. (A) Methionine; (B) lysine; (C) leucine.
Samples of the eluted proteins were analyzed on further polyacrylamide gels. Fig. 2B shows that large-T and small-t prepared in this way migrate as single bands with their characteristic mobilities relative to markers, and appear to be essentially undegraded. The separated proteins were mixed with carrier apomyoglobin, and their sequences were determined as described in Materials and Methods. After each degradative cycle, the PTH-amino acid derivatives released were separated by high-pressure liquid chromatography. The derivative from apomyoglobin was detected by absorbance at 254 nm; since the sequence of apomyoglobin is known, this gives an independent and quantitative check at each cycle that degradation is proceeding normally. Radioactively labeled derivatives from the proteins synthesized in vitro were detected by measuring labeled material comigrating with marker PTH-amino acids. Fig. 3 shows the amino-terminal sequence data obtained with large-T labeled in vitro with [asS]methionine and tritiated leucine, lysine, and proline. There are clear peaks of methionine released at cycles 1 and 14 (Fig. 3A), of lysine at cycle 3 (Fig. 3B), and leucine at cycles 5, 11, 13, 16, 17, and 19 (Fig. 3C).
2168
Biochemistry: Paucha et al.
Proc. Natt. Acad. Sci. USA 75 (1978)
AUG GAU AAA GUU UUA AAC AGA GAG 1
2
Met
Asp
3
Lys
4
Va I
5
6
7
8
Leu
Asn
Arg
Glu
GAA UcU UUG CAG CUA AUG GAC CuU 9
10
11
12
13
14
15
16
Glu
Ser
Leu
Gin
Leu
Met
Asp
Leu
CUA GGU cUU GAA AGG AGU GCC UGG 17
18
19
20
21
22
23
24
Leu
Gly
Leu
Glu
Arg
Ser
Ala
Trp
GGG AAU AUU CCU CUG AUG AGA AAG 27 25 29 26 28 30 31 32 Gly Asn lle Pro Leu Met Arg Lys * ......@
* .... . .
*
......@
Fic(. 5. Nucleotide sequence of SV40 RNA and predicted amino acid sequence from the region 0.65-0.63 map units. The RNA sequence given is deduced from the DNA sequence determined by Volckaert et al (31) and Reddy et al. (personal communication). (See also ref. 30.) The amino acids underscored with a solid line have been detected in both large-T and small-t; those with a dotted line, in small-t only. Proline was not detected even though its specific activity was sufficiently high to expect a discernible peak of radioactivity if this amino acid were present in the first 25 residues of large-T. Fig. 4 shows the corresponding analysis of small-t, from material prepared in a separate experiment and labeled with I 5SImethionine and tritiated leucine and lysine. The data obtained for the first 25 residues are virtually identical to that obtained with large-T. In this case, however, the analysis was extended to 33 cycles. Further peaks of leucine at position 29 (Fig. 4C) and of methionine at positions 30 (Fig. 4A) are evident. A small, but discernible peak of lysine is present at position 32. DISCUSSION We have previously shown that large-T and small-t of SV40 can be synthesized in vitro to give proteins that are very closely related, if not identical to, T-Ags isolated from productively infected cells, as judged by specific immunoprecipitation, mobility on polyacrylamide gels, and [-5S~methionine tryptic peptide mapping (6; and unpublished data). However, the sequences of T-Ags synthesized in this way could not be determined by an automated Edman micromethod that we had previously used successfully to determine the amino-terminal sequence of SV40 VP1 unless we pretreated the L-cell extracts with citrate synthase and oxaloacetate to deplete the endogenous pool of acetyl-CoA (15). We assume from this result that the T-Ags synthesized in vitro under normal conditions are acetylated at their amino termini, but this has not been demonstrated directly. The amino-terminal sequence data obtained with large-T and small-t were virtually identical (Figs. 3 and 4) and strongly suggest that the two proteins share at least the first 25 amino acids. The position of the amino acids corresponds to that predicted from a nucleotide sequence in SV40 DNA beginning very close to the junction between HindII/III fragments A and C, at approximately 0.65 map units (ref. 31 and V. B. Reddy, B. Thimmappaya, R. Dhar, K. N. Subramanian, B. S. Zain, J. Pan, M. L. Celma, and S. M. Weissman, personal communi-
Fic. 6. Model for the location of the sequences coding for SV40 proteins. The location of sequences coding for the late proteins is taken from work with specific deletion mutants (16) and DNA sequence data (17, 18); that for the early proteins from the data presented here, work with specific deletion mutants (7), and DNA sequence data (W. Fiers and Reddy et al., personal communications).
cation). Fig. 5 shows that 9 of the first 20 predicted amino acids after the start at 0.65 map units are present in the correct positions at the amino terminus of both forms of T-Ag. Three additional amino acids have been detected in small-t between residues 25 and 33, and these also match the DNA sequence. No other nucleotide sequence in the entire early region of SV40 DNA is consistent with the amino acid sequence data. Since the chance of producing such a perfect correlation between amino acid sequence and DNA sequence on a random basis is statistically extremely remote, we believe these data unequivocally position the start of both forms of T-Ag on SV40 DNA. Fig. 6 shows the position of this start relative to other biological markers on the DNA. Fig. 6 also indicates that small-t is coded for by a contiguous sequence in SV40 DNA extending from 0.65 to about 0.55 map units. This is consistent with the nucleotide sequence of SV40 DNA, which indicates that it is possible to read an uninterrupted stretch of 174 codons from this region (31). The model is also consistent with the finding that SV40 cRNA, which is made in vitro with Escherichia coli RNA polymerase and which presumably represents a faithful, contiguous transcript of SV40 DNA, directs the cell-free synthesis of a polypeptide similar to small-t in immunoprecipitation properties, in polyacrylamide gel mobility, and in methionine tryptic peptide content (6). The finding that cRNA made from the DNA of three different mutants (dl884, d1885, and d1890; ref. 8) with specific deletions in the region 0.59-0.54 all make shortened forms of the protein related to small-t (unpublished data), as well as analysis of the T-Ags from cells infected with these and similar mutants (7; Sleigh et al., personal communication), also support the idea that the sequences coding for small-t include the 0.59-0.54 region.
Fig. 6 indicates that SV40 large-T is coded for by two non-
Proc. Natl. Acad. Sci. USA 75 (1978)
Biochemistry: Paucha et al. contiguous sequences in SV40 DNA. This model is based on the following findings: (i) as we have demonstrated here, the amino terminus of large-T is coded at 0.65 map units; (ii) the carboxy terminus is coded for at or near 0.17 map units because deletions at 0.21 map units generate a shortened form of T-Ag whereas those at 0.18 map units give essentially wild-type-sized large-T (7); (iii) deletions in SV40 DNA in the region 0.59-0.54 do not affect the mobility of large-T (7, 8); and (iv) there are termination codons in all three phases of SV40 DNA at 0.55 map units (31, 19). The simplest interpretation of all these data is that the region 0.59-0.54 is not represented in the mRNA for large-T. The mRNA must therefore be composed of sequences from two noncontiguous transcripts of SV40 DNA (approximately 0.65-0.59 map units and 0.54-0.17 map units) which at some stage during or after transcription are spliced together. Indeed, direct evidence has recently been obtained showing that there are two major SV40 early mRNAs, both of which are spliced (A. Berk and P. A. Sharp, personal communication). The presence of similar splices in the mRNAs for various viral and eukaryotic proteins has been deduced. In some cases the interruption in the sequence is present in a leader sequence on the mRNA (20-24); in other cases it is present within the sequences coding for protein (25-29). The splicing model for large-T mRNA is also consistent with our finding that the mRNA active in the synthesis of large-T spans the early region, hybridizes equally with SV40 DNA restriction fragments Hae III-E and HindII/II-B, and yet sediments marginally but reproducibly more slowly than the mRNA for small-t (6; and unpublished data). The model also explains our failure to synthesize full-length large-T in vitro in response to SV40 cRNA. Presumably cRNA could not produce full-sized large-T if the latter were the product of a "spliced" mRNA since it is unlikely that splicing could occur during the synthesis of SV40 cRNA with E. coli RNA polymerase. The location of the sequences coding for the amino termini of both large-T and small-t at a site less than 2 map units from the origin of SV40 DNA replication indicates that the greater part, perhaps over 95%, of the early region of SV40 DNA codes for polypeptide. However, the exact details of the amino acid sequences contained in'large-T and small-t have yet to be determined. Of particular interest are the sequences unique to small-t and those coded for by the region around the splice site in large-T mRNA. To establish these requires further sequence studies on both the early proteins and their mRNAs. We thank Drs. Palmiter and Illingworth for advice about pretreatment with citrate synthase, Drs. Fiers and Weissman and their respective colleagues for discussions and sequence information before publication, and Prof. R. R. Porter and Dr. L. E. Mole for access to a Beckman sequencer during breakdown of our own. E.P. was supported by a Fellowship from the Canadian Medical Research Council. We also thank Drs. Kamen and Fried for critical reading of the manuscript and Mrs. C. Conway for its preparation. 1. Kelly, T. K. & Nathans, D. (1977) Adv. Virus Res. 21,85-173.
2169
2. Fried, M. & Griffin, B. E. (1977) Adv. Cancer Res. 24, 67113. 3. Tegtmeyer, P. (1974) Cold Spring Harbor Symp. Quant. Biol. 39, 9-15. 4. Rundell, K., Collins, J. K., Tegtmeyer, P., Ozer, H. L., Lai, C-J. & Nathans, D. (1977) J. Virol. 21, 636-646. 5. Prives, C., Gilboa, E., Revel, M. & Winocour, E. (1977) Proc. Natl. Acad. Sci. USA 74,457-461. 6. Paucha, E., Harvey, R., Smith, R. & Smith, A. E. (1977) INSERM Colloq. 69, 189-198. 7. Crawford, L. V., Cole, C. N., Smith, A. E., Paucha, E., Tegtmeyer, P., Rundell, K. & Berg, P. (1978) Proc. Nati. Acad. Sci. USA 75, 117-121. 8. Shenk, T. E., Carbon, J. & Berg, P. (1976) J. Virol. 18, 664671. 9. Pawson, T., Martin, G. S. & Smith, A. E. (1976) J. Virol. 19, 950-967. 10. Kessler, S. W. (1975) J. Immunol. 115, 1617-1624. 11. Hewick, R. M., Fried, M. & Waterfield, M. D. (1975) Virology
66,408-419.
12. Brauer, A. W., Margolies, M. N. & Haber, E. (1975) Biochemistry
14,3029-3035. 13. Edman, P. & Begg, G. (1967) Eur. J. Biochem. 1, 80-91. 14. Zimmerman, C. L., Appella, E. & Pisano, J. J. (1977) Anal. Biochem. 77, 569-573. 15. Palmiter, R. D. (1978) J. Biol. Chem. 252, 8781-8784. 16. Cole, C. N., Landers, T., Goff, S. P., Manteuil-Brutlag, S. & Berg, P. (1977) J. Virol. 24, 277-294. 17. Contreras, R., Rogiers, R., Van de Voorde, A. & Fiers, W. (1977) Cell 12,529-538. 18. Pan, J., Reddy, V. B., Thimmappaya, B. & Weissman, S. M. (1977) Nucleic Acids Res. 4, 2539-2548. 19. Thimmappaya, B. & Weissman, S. M. (1977) Cell 11, 837843. 20. Berget, S. M., Moore, C. & Sharp, P. A. (1977) Proc. Natl. Acad. Sci. USA 74,3171-3175. 21. Chow, L. T., gelinas, R. E., Broker, T. R. & Roberts, R. J. (1977) Cell 12, 1-8. 22. Aloni, Y., Dhar, R., Laub, O., Horowitz, M. & Khoury, G. (1977) Proc. Natl. Acad. Sci. USA 74,3686-3690. 23. Hsu, M-T. & Ford, J. (1977) Proc. Natl. Acad. Sci. USA 74, 4982-4985. 24. Celma, M. L., Dhar, R., Pan, J. & Weissman, S. M. (1977) Nucleic Acids Res. 4,2549-2559. 25. Brack, C. & Tonegawa, S. (1977) Proc. Natl. Acad. Sci. USA 74, 5652-5656. 26. Jeffreys, A. & Flavell, R. (1977) Cell, 12, 1097-1108. 27. Doel, M. T., Houghton, M., Cook, E. A. & Carey, N. H. (1977) Nucleic Acids Res. 4, 3701-3713. 28. Breathnach, R., Mandel, J. L. & Chambon, P. (1977) Nature 270, 314-318. 29. Tilghman, S. M., Tiemeier, D. C., Seidman, J. G., Peterlin, B. M., Sullivan, M., Maizel, J. V. & Leder, P. (1977) Proc. Natl. Acad. Sci. USA 75,725-729. 30. Subramanian, K. V., Dhar, R. & Weissman, S. M. (1977) J. Biol. Chem. 252, 355-367. 31. Volckaert, G., Van de Voorde, A., & Fiers, W. (1978) Proc. Natl. Acad. Sci. USA 75,2160-2164.
