Nucleoside triphosphate-dependent DNA-binding ... - Europe PMC

1 downloads 0 Views 1004KB Size Report
Jan 20, 1987 - manuscript and Don Blair and Robert Fisher for helpful discussions. We thank Donna Dubel for expert technical assistance and Susan. Kelly for ...

Proc. Natl. Acad. Sci. USA Vol. 84, pp. 3560-3564, June 1987 Biochemistry

Nucleoside triphosphate-dependent DNA-binding properties of mos protein (v-mos gene/mos deletion mutants/ATP-dependent DNA binding)

ARUN SETH*, ESTHER PRIELt, AND GEORGE F. VANDE WOUDE* *Bionetics Research, Inc.-Basic Research Program, National Cancer Institute-Frederick Cancer Research Facility, Frederick, MD 21701; and tBen-Gurion University of the Negev, Beer-Sheva, Israel

Communicated by Howard L. Bachrach, January 20, 1987 (received for review October 27, 1986)

ABSTRACT We have previously shown that the mos gene product, p40f°, produced in Escherichia coli binds ATP and has ATPase activity. In the present study, we investigated the DNA-binding properties of p4011S and two mos deletion mutant proteins. Nitrocellulose blot protein-DNA binding assays showed that p40otm binds DNA in the presence of Mg2+-ATP and certain other nucleoside triphosphates. Ninety percent of the p40'°"-bound DNA is dissociated if the complex is washed in the presence of 1 M NaCl or in the absence of ATP. p40"'0-DNA binding is not observed in the presence of AMP or the nonhydrolyzable ATP analog adenosine 5'-[3, r-methylene]triphosphate; however, in the presence of ADP, p40'° binds DNA at 20% of the level that is observed with ATP. An N-terminal-deletion mutant protein, pl9mOS, has no DNAbinding activity, whereas a C-terminal-deletion mutant protein, p25rnS, does. p25m"' contains the ATP-binding domain, binds DNA in the presence of either ADP or ATP, and shows 5% and 45% binding (relative to that in the presence of ATP) in the presence of AMP and adenosine 5'-[8,y-methylene]triphosphate, respectively. These results suggest that the Nterminal domain of p40IlS is responsible for nucleoside triphosphate-mediated DNA binding. We also observed differential histone-DNA binding in the presence and absence of ATP.

Therefore, we expressed the mos product and several mos deletion mutant proteins in Escherichia coli by using derivatives of high-level-expression vectors pJL6 and pANH-1 (20, 21). The derivative containing the complete mos gene (pA28) directed the synthesis of p40os (22), whereas those vectors lacking either N-terminal (pANHA28) or C-terminal (pA28A8) moieties expressed truncated versions of p4Omos, called pl9moS and p25m0s, respectively (21, 22). The mos protein has strong homology with the ATP-binding domain of the cAMP-dependent bovine protein kinase and members of the src kinase family of oncogene products (23). We have shown that purified p40mOS expressed in E. coli binds ATP and has ATPase activity (22). However, ATPase activity was not observed with either p25m0' or pl9mos (ref. 22 and unpublished results). The ATP-binding activity of the p40mns protein prompted us to investigate whether it, like some other ATP-binding proteins (13, 24), also has nucleic acid-binding properties. Here we show that p40m0s binds both DNA and RNA, but only in the presence of ATP or certain other nucleoside triphosphates. The mutant protein p19m0s, lacking the Nterminal half of the mos product, does not bind DNA, whereas the mutant protein p25mos, lacking the C terminus of the mos product, binds DNA in the presence of all nucleoside phosphates tested.

Protein-DNA interactions have been implicated in the regulation of gene expression and cell growth. Several proteins localized to the nucleus, such as simian virus 40 (SV40) large tumor (T) antigen (1, 2) and the cellular myc (3, 4) and myb (5) protooncogene proteins, as well as several proteins localized to the cytoplasm or cytoplasmic membrane, such as the gag-mil fusion protein (6) and the polyomavirus middlesized T antigen (7), have been shown to have DNA-binding activity (3-10). Although the myc and myb proteins have not been shown to bind to specific DNA sequences (4, 5, 11), the SV40 large T antigen shows both sequence-specific and nonspecific binding to DNA sequences (8-10, 12). Recently, the DNA sequence-specific binding of SV40 large T antigen was shown to be abolished in an ATP-dependent manner by the allosteric effect of purine nucleoside triphosphates (13). However, in another case, the Tn3 transposase has been shown to bind specifically to the Tn3 inverted repeat in the presence of ATP (14). The viral mos (v-mos) gene of Moloney murine sarcoma virus (Mo-MSV) encodes a 37-kDa env-mos fusion protein (15, 16). The mos proteins expressed by Mo-MSV strains tsllO and 124 have been shown to have serine/threonine autophosphorylation activity (17, 18). The v-mos product has been localized to the cytoplasm in acutely infected and transformed cells and is present at extremely low levels (19). The low level of mos product in transformed cells makes the characterization of its biochemical properties difficult.

MATERIALS AND METHODS Materials. ATP, GTP, CTP, UTP, dATP, ADP, AMP, and the nonhydrolyzable ATP analog adenosine 5'-[,B,y-methylene]triphosphate (p[CH2]ppA) were obtained from Boehringer Mannheim. Histones were purchased from Sigma. [a32P]dCTP and [a-32P]dATP (>600 Ci/mmol; 1 Ci = 37 GBq) for nick-translation of DNA probes were purchased from New England Nuclear. Production of mos and mos Deletion Mutant Proteins in E. coli. The mos and the mos deletion mutant proteins were synthesized in E. coli using recombinant plasmids pA28, pA28A&8, and pANHA28, derived from expression vectors pJL6 and pANH-1 as described (21, 22). The proteins were purified by washing the insoluble protein pellet with 1 M NaCl/1% (vol/vol) Triton X-100/1.75 M guanidine hydrochloride. The resulting protein pellet was solubilized in 7 M urea as described (22). The maps and protein sizes of mos and deletion mutants are shown in Fig. la. mos Protein-DNA Binding Assay. We determined mos protein-DNA binding with a nitrocellulose blot protein-DNA binding assay (11, 25). The purified mos and deletion mutant proteins were resolved in a NaDodSO4/12% polyacrylamide gel and were transferred electrophoretically onto nitrocellulose paper (26). The paper was incubated with 1% non-fat dry milk (27) in 0.5 M NaCl/50 mM Tris HCl, pH 8.0, for 1 hr to

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Abbreviations: SV40, simian virus 40; T antigen, tumor antigen; p[CH2]ppA, adenosine 5'-[1B,y-methylene]triphosphate. 3560

Biochemistry: a

Seth et al. 101

Proc. Natl. Acad. Sci. USA 84 (1987) 107

GSGGFGS

Plasmid

3561 Protein Size

(121) mos

pA28

p40

pA28A8

p25

pANHA28

p19

374

1

amos

197 Amos

215

374

b 43 26 18 14 1

2

3

4

FIG. 1. (a) Map of the mos and mos deletion mutant products expressed in E. coli. Plasmid names and the different size proteins they encode indicated. Open bars represent protein sequence and the amino acid positions where the deletion mutants begin and end. These are according to the numbering system of Seth and Vande Woude (22). The position of the putative ATP binding site [lysine (K)-121] and the associated glycine-rich consensus sequence (standard one-letter amino acid abbreviations) are shown. (b) Purification of mos protein and mos deletion mutant proteins expressed in E. coli. The mos proteins were purified as described (22) and resolved in a NaDodSO4/12% polyacrylamnide gel. The proteins were visualized by staining with 0.2% Coomassie blue in 50% trichloroacetic acid and destaining with 40% methanol/5% acetic acid (22). Lanes: 1, histones; 2, p25mSo; 3, pl9mOS; 4, p40m05. The outside lanes show prestained marker polypeptides (molecular masses in kDa at left). are

block nonspecific binding. The paper was washed first with 0.5 M NaCl/50 mM Tris HCI, pH 8.0, and then with binding buffer [20 mM Tris-HCl, pH 8.0/20 mM KCI/10 mM MgCl2/1 mM dithiothreitol/0.5 mM EDTA/15% (vol/vol) glycerol] and then was incubated with nick-translated 32P-labeled DNA (40,000 cpm/ml; specific activity of 2 x 108 cpm/gg) for 16 hr. Excess DNA was removed, and the blots were washed four times with binding buffer containing 0.5% bovine serum albumin for 1 hr. The blots were then air-dried and exposed to x-ray film with intensifying screen for 2-4 hr. Labeling of DNA. Various cloned DNA fragments [e.g., Ava 1-HindI11 c'-mos fragment (22)] were nick-translated using [a--32P~dCTP and [a-32P]dATP as described (28). Unincorporated 32P-labeled nucleotides were removed by Sephadex G-50 gel filtration.

RESULTS p4O""°-DNA Binding Is ATP-Dependent. Preliminary studies using a nitrocellulose filter binding assay showed that p4Ornos and certain mos deletion mutant proteins (Fig. la) bind DNA (data not shown). We performed subsequent analyses using a modified protein-DNA binding assay on nitrocellulose paper to minimize potential nonspecific DNA binding (see Materials and Methods). The purified mos proteins were resolved in a NaDodSO4/12% polyacrylamide gel and then electrophoretically transferred onto nitrocellulose paper in the presence of 20% methanol. This transfer procedure has been shown to remove NaDodSO4 and to aid in renaturation of proteins (29, 30). The nonspecific and low-affinity DNA-binding sites were blocked with 1% non-fat dry milk (27), and the blot was incubated with a nicktranslated 32P-labeled DNA probe. With this method as little as 10-100 ng of protein is sufficient for detection (data not shown).

A typical Coomassie blue-stained polyacrylamide gel electrophoretogram of bacterially expressed, purified mos proteins is shown in Fig. lb. The mos-specific proteins are indicated with an asterisk (p25m°s, lane 2; pl9mos, lane 3; and p4Omos, lane 4). Previously, we showed that p40Ons binds ATP and has ATPase activity (22). We therefore tested whether ATP has an effect on mos protein-DNA binding. The data (Fig. 2a) show that DNA binding occurs with p40Ons only in the presence of ATP. The mos deletion mutant protein pl9rns, lacking the N terminus, shows less than 5% DNA binding, whereas p25r°S, lacking the C terminus, binds DNA as efficiently as p40Oos. We used histones as a positive control for the DNA binding assay. Our analyses showed that histones bind DNA in the presence or absence of ATP. However, we also observed differences in the ability of the different classes of histones to bind DNA in the presence or absence of ATP (Fig. 2a). Thus, histones H2A, H2B, H3, and H4 showed more bNA binding in the presence of ATP, whereas histone Hi showed more DNA binding in the absence of ATP (Fig. 2a). We also determined that 1 M NaCl is sufficient to dissociate all of the bound DNA from histones, p40ms, and the mutant p25m°s, suggesting that the DNA binding is caused by electrostatic interactions (data not shown). We found that p4Omos in the presence of ATP will bind single-stranded DNA 2-3 times better than double-stranded DNA (data not shown). Since the v-mos product has been shown to be a soluble cytoplasmic protein (19), we tested the ability of p4QmOS to bind RNA. The binding experiments were performed on nitrocellulose blots in the presence of 32P-labeled RNA transcribed in vitro from a plasmid containing the SP6 promoter and the human c-mos gene. It was observed that p40mos and the p25mos mutant protein also bound RNA (data not shown). Effect of ATP Concentration on p4O""°-DNA Binding. To determine the optimal concentration of ATP required for DNA binding to p40ms, we compared the DNA binding of

3562

Biochemistry: Seth et al.

Proc. Natl. Acad. Sci. USA 84 (1987)

b 6,400

5,600

0o

p25mos

A

p4QmoS

4,800 z a

a

4,000

a.

n 3,200 1 2 3 4

0

43-

_. _

26-

0.Ea

43-

2,400

1 2 3 4

_

3 mM

L

"O

Wm

=

1,600

26-

3 mM

4 mM

800 p19mos

1

2

3

4

10mM ATP

1

2

3

4

No ATP

ATP (mM)

FIG. 2. (a) Nitrocellulose blot protein-DNA binding assay. Purified p4OmOs and mos deletion mutant proteins were subjected to electrophoresis in a NaDodSO4/12% polyacrylamide gel and transferred to nitrocellulose paper as described (26). The nitrocellulose paper was treateo with 1% non-fat dry milk (27) in 0.5 M NaCl/50 mM Tris HCl, pH 8.0, for 1 hr to block nonspecific binding sites. Then the paper was washed with 0.5 M NaCl/50 mM Tris HCl, pH 8.0, followed by binding buffer and was incubated with nick-translated 32P-labeled DNA. DNA binding was assayed in presence (10 mM) or absence of ATP. Lanes: 1, histones (2 ,ug); 2, p25m05 (1 ,g); 3, pl9r°s (1 ,ug); 4, p40Ons (1 pg). Positions of marker proteins (43 kDa and 26 kDa) run in parallel are indicated. (b) Effect of ATP concentration on p40mos-DNA binding. DNA binding was assayed on nitrocellulose blots with increasing concentrations of ATP (see Inset). The labeled bands were excised for measurement of radioactivity, and the cpm bound was plotted vs. ATP (mM) concentration. (Inset) Autoradiogram of nitrocellulose blots incubated with nick-translated 32P-labeled DNA with 1, 2, 3, or 4 mM ATP. Lanes: 1, histones; 2, p25m°S; 3, pl9mos; 4, p4ornS.

p40mos, p25mos, and pl9mos in the presence of various concentrations (0-10 mM) of ATP (Fig. 2b). With p40mns we observed some DNA binding at 2 mM ATP, and binding increased with increasing ATP concentration through 7 mM, at which point the level of binding plateaued. We observed no significant binding with the pl9mrs mutant protein. On an equimolar basis p25mns appeared to bind more DNA than p40m's at low ATP concentrations. However, both proteins bind DNA to the same extent at higher ATP concentrations. Effect of AMP, ADP, and the Nonhydrolyzable ATP Analog p[CH2]ppA on mos Protein-DNA Binding. To determine if the p40mns ATPase activity (22) is required for binding, we tested the ability of p40mns to bind DNA in the presence of AMP, ADP, or p[CH2]ppA. The binding assay was performed as described above, the labeled bands were cut out, and the radioactivity was determined (Table 1). No p40mos-DNA complexes were observed in the presence of AMP or p[CH2]Table 1. Binding of 32P-labeled DNA to mos proteins in the presence of ATP, ADP, AMP, or p[CH2]ppA DNA bound, cpm Protein ATP AMP ADP p[CH2]ppA 8346 1694 (20%) 0 140 (1.6%) p40MOS 9184 7184 (80%) 470 (5%) 4120 (45%) p25mo$ 690 120 0 0 p19M°S Binding was assayed on a nitrocellulose blot and the labeled bands were excised for Cerenkov counting after autoradiography. DNA binding in the presence of ATP was considered as 100% binding.

ppA, and in the presence of ADP, the binding was only 20% of that observed with ATP. However, p25tm0 bound DNA in the presence of ADP (80%) or ATP and showed approximately 5% and 45% binding in the presence of AMP and p[CH21ppA, respectively. These results do not exclude the possibility that binding of DNA by p40m°0 requires hydrolysis of ATP. The observations that p25m°S lacks ATPase activity (22) and binds DNA in the presence of ADP or p[CH2]ppA indicate that ATP hydrolysis is not required for its DNA binding. Effect of Other Nucleoside Triphosphates on p40°s-IDNA Binding, We studied the influence of dATP, GTP, and UTP on p40mos-DNA binding. At the same molar concentration, less than 10% p40tm0-DNA binding was observed with UTP as compared to ATP (100%). However, the C-terminaldeletion mutant product, p25mrS, showed 30% binding in the presence of UTP (Fig. 3). The binding of DNA to p40m05 in the presence of dATP or GTP (Fig. 3) or CT!P (data not shown) was comparable to the level observed with ATP. The N-terminal-deletion mutant product, pl9mos, showed no significant DNA binding. DISCUSSION Using a nitrocellulose blot protein-DNA binding assay, we have shown that p40mos binds DNA in the presence of nucleoside triphosphates and that this DNA-binding activity is not observed with AMP or the nonhydrolyzable ATP analog p[CH2]ppA. We cannot determine by this assay whether mos protein-DNA binding is sequence-specific or nonspecific. The mutant protein pl9mOs, lacking the N ter-

Biochemistry: Seth et al. GTP

do

1

.,

Proc. Natl. Acad. Sci. USA 84 (1987)

ATP

dATP

_0_M

_0

UTP

.

._

2 3 4

1 2 3 4

1

2 3 4

1

2

3 4

FIG. 3. DNA binding was tested in the presence of 4 mM GTP, ATP, dATP, or UTP. Lanes: 1, histones; 2, p25rOS; 3, pl9OS; 4, p4oOs.

minus of p40os, does not show significant DNA binding in the presence or absence of ATP. In contrast, the deletion mutant protein p25mrs, which lacks the C terminus of p40mOS, shows DNA binding in the presence of ATP and other nucleoside phosphates. Equimolar quantities of p4QnmOS and p25moS bind DNA to the same extent at high ATP concentrations, but DNA binding by p25mrs is greater at lower ATP concentrations (Fig. 2b). Although we have not identified the specific N-terminal amino acid sequences participating in the DNA binding, the p25mOS protein contains the ATP-binding domain, while the pl9mos product does not (22, 23). We cannot exclude the possibility that amino acids upstream of the c-mos sequence present in the p4Omos and p25m`s products [encoded by vector sequence, helper env sequence, and sequence preceding the first ATG in the protooncogene locus (22)] participate in DNA binding. However, preliminary analysis with a p27mrs, a product that has a 143 amino acid internal deletion in the ATP-binding domain (residues 55-197; ref. 22) but possesses the N-terminal amino acids upstream of the c-mos ATG as in p4Omos and p25m0s, shows a marked reduction in DNA binding (data not shown). Previously, it has been shown that the gag-mil fusion protein binds DNA and RNA (6). It is possible that other members of the src tyrosine kinase family possessing the ATP-binding domain (23) also bind DNA. We have shown that binding of DNA to p4QmOS does not occur in the presence of AMP, UTP, or p[CH2]ppA, whereas some binding of DNA to p2510S occurs in the presence of all nucleoside phosphates tested. Thus, we observe 80% DNA binding with ADP, 30% with UTP, and 45% with the nonhydrolyzable ATP analog p[CH21ppA. It is thus possible that p25mos- and p40mos-DNA binding may occur via a conformational change effected by the nucleoside phosphates. We are not sure why p25mos is less specific than p4Omos in its nucleoside phosphate-dependent DNA binding. First, there are only slight differences in the basic amino acid content of p2smOs, p40oOs, and pl9mos, whose calculated isoelectric points are 8.2, 8.1, and 7.7, respectively. Second, ADP and UTP should be able to facilitate DNA binding to p40O4os as well as to the smaller p25mos protein. Even though the phosphate bond angles are different in ATP and p[CHj2ppA (31), it seems unlikely that p[CH2]ppA can influence p25mos conformation but not influence p40mOS. We have demonstrated that, with the p25mOS mutant protein, ATP hydrolysis is not required for DNA binding. However, p4Omos possesses ATPase activity (22), and we cannot exclude that ATP hydrolysis could regulate p40o°s-DNA binding. It is also possible that the nucleoside phosphates affect the structure of the DNA in some way that allows it to bind to the protein, as has been shown for an E. coli replication protein, DnaB (32). The modulation of protein-DNA binding by nucleoside triphosphates has been demonstrated for several proteins (13, 14). We used histones as a control in the nitrocellulose blot

3563

assay and observed that ATP can also modulate their relative extent of DNA binding (Fig. 2a). This was an unexpected result, but nucleoside triphosphates have been shown to influence conformation of histones and unfolding of nucleosomes (33, 34). It has been reported (14) that Tn3 transposase binds only to the Tn3 inverted repeat sequence in the presence of ATP, whereas DNA binding in the absence of ATP is both specific and nonspecific. The mos product is a soluble cytoplasmic protein (19), and as such one would expect its ability to bind RNA to be more important than its DNA-binding properties. The mos product localization experiments were performed with transformed and acutely infected cells (19), and it is conceivable that a small amount of protein present in the nucleus was not detected by these analyses, since the level of mos protein in cells is extremely low (16, 19). It is also possible that the mos protein functions during mitosis [or meiosis, since mos is expressed in gonadal tissue (35)] when the nuclear envelope breaks down. We wish to thank Dennis Watson for critical review of the manuscript and Don Blair and Robert Fisher for helpful discussions. We thank Donna Dubel for expert technical assistance and Susan Kelly for preparation of the manuscript. The research was sponsored by the National Cancer Ipstitute, under Contract NO1-CO-2309 with Bionetics Research, Inc. 1. Kalderon, D., Richardson, W. D., Markham, A. F. & Smith, A. E. (1984) Nature (London) 311, 33-38. 2. Santos, M. & Butel, J. S. (1982) Virology 120, 1-17. 3. Donner, P., Greiser-Wilke, I. & Moelling, K. (1982) Nature (London) 2%, 262-266. 4. Persson, H. & Leder, P. (1984) Science 225, 718-721. 5. Klempnauer, K.-H. & Sippel, A. E. (1986) Mol. Cell. Biol. 6, 62-69. 6. Bunte, T., Greiser-Wilke, I. & Moelling, K. (1983) EMBO J. 2, 1087-1092. 7. Bolen, J. B., Cary, K., Scheller, A., Basilico, C., Israel, M. A. & Prives, C. (1986) J. Virol. 58, 157-164. 8. Carroll, R. B., Hager, L. & Dulbecco, R. (1974) Proc. Natl. Acad. Scj. USA 71, 3754-3757. 9. Tijan, R. (1978) Cell 13, 165-179. 10. Prives, C., Barnet, B., Scheller, A., Khoury, G. & Jay, G. (1982) J. Virol. 43, 73-82. 11. Watt, R. A., Shatzman, A. R. & Rosenberg, M. (1985) Mol. Cell. Biol. 5, 448-456. 12. Prives, C., Beck, Y. & Shure, H. (1980) J. Virol. 33, 689-696. 13. Vogt, B., Vakalopoulou, E. & Fanning, E. (1986) J. Virol. 58, 765-772. 14. Wishart, W. L., Broach, J. R. & Ohtsubo, E. (1985) Nature (London) 314, 556-558. 15. van Beveren, C., van Straaten, F., Galleshaw, J. A. & Verma, I. M. (1981) Cell 27, 97-108. 16. Papkoff, J., Verma, I. M. & Hunter, T. (1982) Cell 29, 417-426. 17. Kloetzer, W. S., Maxwell, S. A. & Arlinghaus, R. B. (1983) Proc. Natl. Acad. Sci. USA 80, 412-416. 18. Maxwell, S. A. & Arlinghaus, R. B. (1985) Virology 143, 321-333. 19. Papkoff, J., Nigg, E. A. & Hunter, T. (1983) Cell 33, 161-172. 20. Lautenberger, J. A., Court, D. & Papas, T. S. (1983) Gene 23, 75-84. 21. Seth, A., Lapis, P., Vande Woude, G. F. & Papas, T. (1986) Gene 42, 49-57, 22. Seth, A. & Vande Woude, G. F. (1985) J. Virol. 56, 144-152. 23. Hunter, T. & Cooper, J. A. (1986) in The Enzymes: Control by Phosphorylation, eds. Boyer, P. D. & Krebs, E. G. (Academic, New York), Vol. 17, pp. 191-246. 24. Clertant, P., Gaudray, P., May, E. & Cuzin, F. (1984) J. Biol. Chem. 259, 15196-15203. 25. Miskimins, W. K., Roberts, M. P., McClelland, A. & Ruddle, F. H. (1985) Proc. Natl. Acad. Sci. USA 82, 6741-6744. 26. Towbin, H., Staehelin, T. & Gordon, J. (1979) Proc. NatI.

3564

Biochemistry: Seth et al.

Acad. Sci. USA 76, 4350-4354. 27. Johnson, D. A., Gautsch, J. W., Sportsman, J. R. & Elder, J. H. (1984) Gene Anal. Tech. 1, 3-8. 28. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY). 29. Shapiro, A. L. & Maizel, J. V. (1969) Anal. Biochem. 29, 505-514. 30. Bers, G. & Garfin, D. (1985) BioTechniques 3, 276-288.

Proc; Natl. Acad. Sci. USA 84 (1987) 31. Castora, F. J. & Kelly, W. G. (1986) Proc. Natl. Acad. Sci. USA 83, 1680-1684. 32. Arai, K. & Kornberg, A. (1981) J. Biol. Chem. 256, 5253-5259. 33. Wickett, R. & Isenberg, I. (1972) Proc, Natl. Acad. Sci. USA 69, 2687-2690. 34. Pentz, M., Vatev, R. & Goldthwait, D. A. (1986) Nucleic Acids Res. 14, 5513-5529. 35. Propst, F. & Vande Woude, G. F. (1985) Nature (London) 315, 516-518.