Cleavage of single stranded oligonucleotides by EcoRI restriction ...

Volume 15 Number 2 1987 Volume 15 Number 2 1987

Nucleic Acids Research Nucleic Acids Research

Cleavage of single stranded oligonucleotides by EcoRI restriction endonuclease

Norbert Bischofberger*, Peter G.Ng, Thomas R.Webb and Mark D.Matteucci

Department of Molecular Biology, Genentech, Inc., 460 Point San Bruno Boulevard, South San Francisco, CA 94080, USA Received November 7, 1986; Accepted December 9, 1986

ABSTRACT The 31mer 5'-TCA ACG CTA GM TTC GGA TCC ATC GCT TGG T, the complementary 33mer 5'-CCA AGC GATGGA TCC GM TTC TAG CGT TGA GAT, the 40mer 5'-GGC CAG GAT GGT GM GM TTC GAT CCG GTA CGT AGC TM G, and the complementary 42mer 5'-TAC T1A GCTTACG TAC CGG ATC GM TTC TTC ACC ATC CTG GCC were synthesized and their reactivity towards EcoRI was studied. It was found that the 31mer and the 40mer were cleaved at a comparable rate to the 31mer-33mer hybrid and the 40mer-42mer hybrid, respectively. The rate of cleavage of the 33mer and the 42mer was an order of magnitude lower. To rule out possible intermolecular duplex forT~tion, the 33mer was immobilized on cellulose by ligation and labeled with a P-dCTP using Klenow fragment of E. coli DNA polymerase. EcoRI cleaved this immobilized oligomer into specific fragments.

INTRODUCTION Sequence specific, or type II, restriction endonucleases recognize and cleave double-stranded (ds) DNA within a defined sequence (1). As such, they have emerged as valuable research tools and are indispensable for molecular biology. About eighty restriction enzymes with different sequence specificities are now commercially available; surprisingly little, however, is known about their mode of recognition and cleavage (2). Restriction endonucleases were generally believed to cleave only ds DNA. Cleavage of single-stranded (ss) DNA has been observed (3,4) and recently it has been suggested that ss cleavage is a general behavior of restriction enzymes (5). These studies were done with DNA from natural sources, however, and due to the complexity of the DNA it remains unclear whether restriction endonucleases act indeed on ss DNA or only on transiently formed duplex structures ("canonical structures"). Only a few investigations have used oligonucleotides as simple, well-defined systems (6). In one report MspI was shown to cleave ss oligonucleotides, although the rate of ss cleavage was an order of magnitude lower than ds cleavage (7). The synthesis of oligonucleotides now being a routine procedure (8), 1011 1 R L Press Umited, Oxford, England.

709

Nucleic Acids Research such studies are becoming more commonplace (9). EcoRI is the most extensively studied restriction enzyme (10) and it provides a model system for sequence-specific DNA-protein interactions. It recognizes and cleaves the sequence 5'-GAM TTC-3' and it is believed to act only on ds DNA. From kinetic data using the self-complementary octanucleotide T GAA TTC A it was inferred that the reacting species was actually the duplex TGAATTCA/ACTTMGT (11). In the present paper we report our findings that EcoRI cleaves ss and ds oligonucleotides in solution and also ss oligonucleotides immobilized on solid support. This has consequences on the enzyme's presumed mechanism of action.

MATERIALS AND METHODS Enzymes and Chemicals Oligonucleotides were synthesized by the H-phosphonate method (12) on a Biosearch Model 8600 DNA Synthesizer and were purified by gel electrophoresis. Restriction enzymes, DNA polymerase I (Klenow fragment), T4-DNA ligase and polynucleotide kinase were from New England Biolabs. Oligo dT cellulose was from Sigma. y-32P-dATP and a-32P-dCTP were from Amersham Corporation. Snake venom phosphodiesterase was from Boehringer Mannheim. Restriction Enzyme Digests A mixture containing the 5,_p32 labeled oligomer at a given concentration in the buffer system recommended by the commercial supplier of the enzymes was incubated with a given amount of restriction endonuclease at a certain temperature (volume was 20 to 50 ul). At time intervals aliquots (2 to 10 il) were withdrawn, quenched by heating to 1000C for 2 min, the DNA was ethanol precipitated and analyzed on a 20 percent acrylamide/7M urea sequencing gel. The gel was exposed to photographic film and the darkness of the spots was measured by gel scanning. The oligomers were annealed by incubating an equimolar amount of the complementary oligonucleotides in the restriction enzyme buffer system at 70 C for 30 min followed by 37 C for 2 hours. The success of the reaction was checked by gel electrophoresis under non-denaturing conditions. The oligonucleotide concentrations were calculated from the A260 value. Ligation of the 33mer to oligo dT cellulose (see Fig. 2) A mixture containing oligo dT cellulose (75 mg), the 54mer 5'-TTG ATC TCA ACG CTA GM TTC GGA TCC ATC GCT TGG A18 (0.5 A260 units) and the 710

Nucleic Acids Research kinased 33mer 5'-pCCA AGC GAT GGA TCC GM TTC TCG CGT TCA GAT (0.5 A260 units) in ligation buffer (250 ul, 30 mM Tris.HCL pH 8, 4mM MgCl2, 0.1 mM EDTA, 1 mM DTT; 50 ug/ml BSA) was incubated under annealing conditions by heating to 70°C for 30 min followed by 37 C for 2 hours with occasional vortex mixing. T4 DNA ligase (800 U) and ATP (20 .l of a 10 mM solution) were added and the mixture was incubated at 37 C for 20 hours and was finally washed with buffer solution. To determine the yield of the ligation step, the experiment was repeated using 32P-labeled 33mer. After ligation '15 percent of the total counts were found to remain on the cellulose after repeated washings with water. On incubation of this labeled, immobilized oligomer with snake venom phosphodiesterase, more than 90 percent of the total counts were released from the cellulose. In a control experiment to determine the amount of unspecific adsorption the experiment with the 32P-labeled 33mer was repeated without addition of T4 ligase; no significant radioactivity was incorporated into the cellulose. Labeling of the immobilized oligomer with Klenow fragment A mixture containing the oligomer duplex from the ligation step, dATP (0.5 mM), dCTP (0.5 mM), a-32 dCTP (10 jM; 400 Ci/mmol) and Klenow fragment (15 U) in buffered solution (100 uil) was incubated at ambient temperature with occasional mixing. After 2 hours and again after 4 hours more enzyme (15 U) was added. After 18 hours the polymer was washed with H20, 0.1 N NaOH, H20 and finally with EcoRI buffer. The cellulose had an activity of 1.5 jCi (= 20 percent yield). In a control experiment the procedure was repeated without adding Klenow fragment. No significant radioactivity was incorporated into the cellulose. Incubation of the labeled immobilized oligomer with snake venom phosphodiesterase caused >95 percent of the radioactivity to be released from the cellulose.

RESULTS AND DISCUSSION Interaction of oligomers with EcoRI in solution Figure 1 shows the results of the cleavage of the 31mer 5'-TCA ACG CTA GM TTC GGA TCC ATC GCT TGG T, the 33mer 5'-CCA AGC GAT GGA TCC GAA TTC TAG CGT TGA GAT and the 31mer-33mer hybrid with EcoRI. As can be seen, the ss 31mer and the ds 31mer-33mer hybrid are cleaved with similar efficiency into their specific fragments, whereas the ss 33mer is cleaved at a much lower rate. Cleavage of the 33mer was observed at higher enzyme concentrations,

711

Nucleic Acids Research

0 1 2

c

B

A

30

1

2

3

3-333 :31 Illme r

I

m

ner

_ -.6

*

I

Fig. 1. Cleavage of ss 3lmer (lanes A), ds 31mer-33mer (lanes B) and ss 33mer (lanes C) with EcoRI. 0: after 0 min; 1: after 15 min; 2: after 30 min; 3: after 60 min. The samples containelI the 5'- '2P-labeled oligonucleotides ('-0.5 1AM), EcoRI (1.5 U ul- ) in 20 uil of a buffered solution. The 3lmer is cleaved into a lOmer and cleavage of the 33mer yields a l6mer fragment. however, and was estimated to be -50x slower than the cleavage of the 3lmer at 1 uM oligomer concentration. The unexpected efficiency with which EcoRI cleaved the ss 3lmer led us to investigate this phenomenon with other oligomers and restriction endonucleases. We synthesized the 40mer 51-GGG GAG GAT GGT GAA GAA TTG GAT CGG GTA CGT AGG TAA G and the complementary 42mer 5'-TAC TTA GGT AGG TAG CGGG ATG GAA TTG TTG AGG ATG GTG GGG, which contain multiple restriction enzyme recognition sites. On treatment of the 40mer, the 42mer and the 40mer-42mer hybrid with EcoRI under the same conditions as for the 31mer-33mer, we found that the ss 4Omer and the 40mer-42mer hybrid were again cleaved with similar rates, whereas the rate of cleavage of the ss 42mer was an order of magnitude or so lower (data not shown). Digests with other restriction endonucleases (Sau3AI, AluI, MboII, HpaII, FokI, HphI, DdeI, RsaI and BstNI) showed only efficient cleavage of the ds 40mer-42mer, although ss cleavage might have occurred at a lower rate (lox). The difference in the rate of ss cleavage is remarkable. The ss 3lmer 712

Nucleic Acids Research ®-TTnT-3' + 54 mer

+ 33 mer -5- phosphate

AA15AAGGT.TCG.CTA-CCT.AGGCCTTTAAG-ATC-GCA.ACT.CTA.GTT-5'

O-TTnT

pCCA'AGC-GAT*GGA-TCC.GAA.TTC-TAG-CGT.TGA-GAT-3'

4 T4-DNA ligase

A17A GGT.TCG.CTA-CCT.AGG.CTT.AAG.ATC-GCA-ACT CTA*GTT-5'

-TTnT*CCA-AGC*GAT.GGA.TCC-GAATTC.TAG.CGT-TGA.GAT -3' Kienow fragment

32P_ dCTP

dATP

A 17AGGT.TCG-CTA.CCT-AGG-CTT.AAG.ATC4GCA.ACT CTA.G T T- 5' c-TTnT.CCA-AGC-GAT.GGA.TCC.GAA.TTC-TAGCCGT.TGA.GATCVA(A) -3'

, 0.1 N NaOH

®-TTnT.CCA-AGC.GAT.GGA.TCCGAATTC-TAG-CGT.TGA-GAT.C-A(A)-3' Fig. 2.

Attachment of oligonucleotides to oligo dT cellulose.

and the ss 40mer are both cleaved at a much higher rate than the complementary ss 33mer and ss 42mer, respectively. Within the duplexes 31mer-33mer and 40mer-42mer, however, both individual strands are cleaved at a similar rate. This difference in the rate of ss oligomer cleavage is consistent with the observation that restriction endonucleases cleave different sites in DNA with different rates (13,14). It has been suggested that the bases surrounding the recognition site may play a role in enhancing the binding or catalytic activity of the enzyme. Our results confirm this idea and suggest that purine bases surrounding the EcoRI recognition site enhance the rate of cleavage, as is observed for the 31mer TCA ACG CTA GAA TTC GGA TCC ATC GCT TGG T and the 40mer GGC CAG GAT GGT GAA GAA TTC GAT CCG GTA CGT AGC TM G. The present results showing efficient cleavage of ss oligonucleotides with EcoRI indicate that duplex formation is not a necessary requirement for cleavage to occur. A computer and visual homology search of the 31mer and 713

Nucleic Acids Research

rTrjTC C A AG C GAT

1

2 3.

*

&'^S

GG- A TC C G AA T TC TAG CGT TGA GAT CA (A T~20

Eco RI

- TrtT CCA AGC GAT GGA TCC G

AA TTC TAG CGT TGA GAT CA(A) 19 (20)ner

Fig. 3. Cleavage of oligomers immobilized to cellulose with EcoRI. The oligonucleotides attached to oligo dT cellulose were digested with EcoRI (100 U) for 24 hours. The percent radioactivity released into the supernatant was determined and the nature of the radioactivity in the supernatant was analyzed by ethanol precipitation and gel electrophoresis. Lane 1: immobilized oligoiner was annealed with the complementary 31mer 5'-TCA ACG CTA GAM TTC GGA TCC ATC GCT TGG T prior to digestion with RI; 96 percent of the radioactivity was released from the cellulose. Lane 2: ss oligomer attached to cellulose digest; 45 percent of the radioactivity was released. Lane 3: poly-dT reference standard. In a control experiment without added enzyme less than 1 percent of the radioactivity was released from the support. the 40mer revealed no secondary structure that could create ds GAATTC regions by intramolecular duplex (hairpin) formation. However, intermolecular duplex formation between the GAATTC regions of two ss oligonucleotides could occur, resulting in the observed ss cleavage. Due to the instability of such transiently formed structures, the rate of ss cleavage would be expected to be significantly lower compared to the rate of ds DNA cleavage. Our finding that both the ss 31mer and the ss 40mer are cleaved at a rate similar to the rate of the ds 31mer-33mer and the ds 40mer-42mer cleavage suggests that EcoRI cleavage of ss oligonucleotides is not caused by intermolecular duplex formation. Attachment of Oligonucleotides to Cellulose and Interaction of EcoRI with the Immobilized Oligomers To further exclude the possibility of intermolecular duplex formation of the oligomers in solution, we decided to immobilize the oligomers on solid 714

Nucleic Acids Research support. Since the immobilized oligonucleotides were to be tested as potential substrates for the EcoRI restriction enzyme, we chose to use enzymatic methods to bind the oligomers to the solid support rather than chemical methods (15) to assure accessibility for the enzyme. The method is depicted in Fig. 2: Oligo-dT cellulose was annealed with the 54mer 3'-A18 GGT TCG CTA CCT AGG CTT MG ATC GCA ACT CTA GTT and the kinased 33mer 5'-pCCA AGC GAT GGA TCC GM TTC TCG CGT TCA GAT. The resulting duplex served as a substrate for T4 DNA ligase to link the 33mer to oligo dT cellulose. Radioactive label was introduced with a32P-dCTP using DNA polymerase I (Klenow fragment), resulting in a partially repaired hybrid (vide infra). The resulting labeled duplex DNA was finally denatured by washing the polymer repeatedly with 0.1 N NaOH (see Materials and Methods). The results of the restriction endonuclease treatment of the oligomers immobilized to cellulose are summarized in Fig. 3. As can be seen, EcoRI caused the release of radioactivity from the cellulose from both the ss and the annealed oligomers, the rate of release from the annealed cellulose being higher. In order to distinguish between specific endonucleolytic activity of EcoRI and unspecific nuclease activity, the radioactive DNA released from the polymer was ethanol precipitated and the precipitate analyzed by gel electrophoresis. As shown in Fig. 3, the radioactive materials released from the cellulose were a mixture of 19 and 20mers (as a result of partial Klenow repair), indicating specific EcoRI cleavage. Similar results were obtained with the 33mer ligased to oligo dT cellulose (data not shown). In this model system where oligomers are attached to solid support intermolecular duplex formation between the oligonucleotides is not possible. Our finding that EcoRI cleaves these immobilized oligomers specifically strongly supports the notion that EcoRI acts on single-stranded DNA, and that duplex formation is not necessary for the enzymatic cleavage.

ACKNOWLEDGMENTS We would like to thank Mark Vasser and Parkash Jhurani for assistance with the oligonucleotide synthesis. REFERENCES 1. Wells, R.D., Klein, R.D., Singleton, C.K. (1981) in The Enzymes, Vol. 14, p. 157-191, Academic Press, New York. 2. Chirikjian, J.G. (1981) in Gene Amplification and Analysis, Vol. 1: Restriction Endonucleases, Elsevier.

715

Nucleic Acids Research 3. Horiuchi, K. and Zinder, N.D. (1975) Proc. Natl. Acad. Sci. 72, 2555-2558. 4. Blakesley, R.W., Dodgson, J.B., Nes, I.F. and Wells, R.D. (1977) J. Biol. Chem. 20, 7300-7306. 5. Nishigaki, K., Kaneko, Y., Wakuda, H., Husimi, Y. and Tanaka, T. (1985) Nucl. Acids Res. 13, 5747-5760. 6. Kita, K., Hiraoka, N., Kimizuka, F., Obayashi, A., Kojima, H., Takahashi, H. and Saito, H. (1985) Nucl. Acids Res. 13, 7015-7024. 7. Yoo, O.J. and Agarwal, K.L. (1980) .J. Biol. Chem. 23T, 10559-10562. 8. Caruthers, M.H. (1985) Science 230, 281-285. 9. a) Baumstark, B.R., Roberts, R.T and RajBhandary, U.L. (1979) J. Biol. Chem. 254, 8943-8950; b) Yolov, A.A., Gromova, E.S., Kubareva, E.A., Potapov, V.K., and Shabarova, Z.A. (1985) Nucl. Acids Res. 13, 8969-8981; c) Yolov, A.A., Vinogradova, M.N., Gromova, E.S., Rosenthal, A., Cech, D., Veiko, V.P., Metelev, V.G., Kosykh, V.G., Buryanov, Y.I., Bayer, A.A., Shabarova, Z.A. (1985) Nucl. Acids Res. 13, 8983-8998. 10. A crystal structure of EcoRI has been published: Frederick, C.A., Grable, J., Melia, M., Samudzi, C., Jen-Jacobson, L., Wang, B-C., Greene, P., Boyer, H.W. and Rosenberg, J.M. (1984) Nature 309, 327-331. 11. Greene, P.J., Poonian, M.S., Nussbaum, A.L., Tobias, L., Garfin, D.E., Boyer, H.W., and Goodman, H.M. (1975) J. Mol. Biol. 99, 237-261. 12. a) Froehler, B.C. and Matteucci, M.D. (1986) Tet. Lett. 27, 469-472; b) Froehler, B.C., Ng, P.G., and Matteucci, M.D. (1986) NucIT Acids Res. 14, 5399-5407. 13. Thomas, M. and Davis, R.W. (1975) J. Mol. Biol. 91, 315-328. 14. Farsblom, S., Rigler, R., Ehrenberg, M., Pettersson, U. and Philipson, L. (1976) Nucl. Acids Res. 3, 3255-3269. 15. Goppelt, M., Pingoud, A.; Maass, G., Mayer, H., Koster, H., and Frank, R. (1980) Eur. J. Biochem. 104, 101-107.

716