Strand invasion by oligonucleotide-nuclease conjugates - American

Strand Invasion by Oligonucleotide-Nuclease Conjugates. David R. Corey,* Debbie Munoz-Medellin, and Alan Huang. Howard Hughes Medical Institute and th...
0 downloads 0 Views 3MB Size
Bioconjugate Chem. 1995, 6,93-100

93

Strand Invasion by Oligonucleotide-Nuclease Conjugates David R. Corey,* Debbie Munoz-Medellin, and Alan Huang Howard Hughes Medical Institute and the Department of Pharmacology, University of Texas Southwestern Medical Center a t Dallas, 5323 Harry Hines Blvd., Dallas, Texas 75235. Received September 22, 1994@

Conjugates consisting of staphylococcalnuclease crosslinked to oligonucleotides hybridize to supercoiled duplex DNA by Watson-Crick base-pairing. Here we describe this strand invasion. Minity cleavage by these conjugates provides a probe for the local topology of the DNA duplex and is most efficient a t a target DNA sequence known to form a cruciform. Additional supercoiling of the substrate DNA increases selective cleavage a t other sequences. Hybridization of the conjugate to duplex DNA is temperature dependent and is stable over time. M i n i t y cleavage is not substantially inhibited by a 200-fold excess of the analogous unmodified oligonucleotide, demonstrating that hybridization of the unmodified oligonucleotide must be less favored and that the nuclease is involved in substrate binding. Surprisingly, affinity cleavage is also not effectively inhibited by complementary oligonucleotides unless they contain a n extended 5’-sequence capable of separate interactions with the nuclease domain of the conjugate. These results suggest that the oligonucleotide-nuclease conjugate prefers to hybridize to target sequences which will allow interactions with both the oligonucleotide and the nuclease domains. M i n i t y cleavage by oligonucleotide-nuclease conjugates provides general insights for the design of oligonucleotides and their conjugates for strand invasion and affords a convenient competition assay for their hybridization.

INTRODUCTION

The development of strategies for the Watson-Crick recognition of sequences within duplex DNA would allow greater control over recombination, replication, repair, and transcription. Unfortunately, target sequences within duplex DNA are already base-paired, rendering them relatively inaccessible to recognition by complementary oligonucleotides. To avoid the need to disrupt preexisting base-pairing, triple helix formation (Moser and Dervan, 1987; Jayasena and Johnston, 1993a; Colocci et al., 1993; Francois et al., 1989) is being optimized to achieve sequence recognition through hybridization in the major groove of the DNA helix. Recent advances have extended stable hybridization into the physiological range (Koh and Dervan, 1992; Young et al., 19911, but sequence recognition by this technique is currently restricted to targets which are primarily polypurine-polypyrimidine. Other strategies attempt to avoid sequence limitations through the employment of Watson-Crick base-pairing. Peptide nucleic acids can spontaneously hybridize to duplex DNA by a combination of Watson-Crick basepairing and triple helix formation (Neilsen et al., 1991; Peffer et al., 1993; Hanvey et al., 1992; Neilsen et al., 1993; Demidov et al., 1993) but strand invasion currently has the same sequence limitations as for triple helix formation by oligodeoxyribonucleotides. More general recognition is offered by the ability of RNA probes to form R-loops (Chen et al., 1993) and the use of RecA protein to promote strand exchange (Cheng et al., 1988; Ferrin and Camerini-Otero, 1991; Jayasena and Johnston, 1993b; Rao et al., 1993). These strategies permit sequence recognition to occur but require the use of exogenous proteins, stringent conditions, or altered basepairing. To develop a simpler and potentially complementary approach to sequence recognition we have *To whom all correspondence should be addressed. Tel.: (214)648-5096.Fax: (214)648-5095.E-mail: Corey@howie. swmed.edu. Abstract published in Advance ACS Abstracts, December 15, 1994. @

examined the direct incorporation of oligodeoxyribonucleotides and their conjugates into duplex DNA by strand invasion. Single-stranded DNA will spontaneously hybridize to complementary regions within supercoiled duplex DNA to form D-loops (Beattie et al., 1977; Holloman et al., 1975). In contrast to the hybridization of complementary single strands, which is driven by the increase in enthalpy during base-pairing, D-loop formation is entropically driven by the removal of superhelical turns as the invading single strand is incorporated into the duplex. The first step is the transient denaturation of several base-pairs. The subsequent rate-limiting step is the initial nucleation in which a short helix is formed between the single-stranded DNA and the denatured region. If conditions of salt and temperature are such that base-pairing persists, more pairing will occur until either the invading strand is completely incorporated or until enough superhelical turns are removed to eliminate the driving force for incorporation. Radding and coworkers demonstrated that supercoiled DNA would spontaneously incorporate relatively long complementary single strands (Beattie et al., 1977; Holloman et al., 1975). Most recently, oligonucleotide-staphylococcal nuclease conjugates were shown to hybridize with supercoiled DNA via Watson-Crick base-pairing (Corey et al., 1989a). Here we characterize the role of the nuclease domain of an oligonucleotide-nuclease conjugate in allowing hybridization to target sites within duplex DNA to occur. We also utilize the inhibition of affinity cleavage to develop a competition assay for the hybridization of other oligonucleotide motifs. EXPERIMENTAL METHODS

Oligonucleotide Synthesis. Underivatized controlled pore glass was purchased from CPG (Fairfield, NJ) and was derivatized with l-O-(4,4’-dimethoxytrityl)3,3’-thiopropanol (Pei et al., 1990) as described. The thiolated controlled pore glass was used to synthesize 3’thiolated oligonucleotides on a n Applied Biosystems 451 DNA synthesizer (Foster City, CA). Oligonucleotides

1043-1802/95/2906-0093$09.00/0 0 1995 American Chemical Society

Corey et al.

94 Bioconjugafe Chem., Vol. 6,No. 1, 1995

containing a 5'-acridine were obtained from Appligene (Pleasanton, CA). The thiolated oligonucleotides were reduced by treatment with 20 mM dithiothreitol (DTT) overnight a t 37 "C in 1 mM EDTA, 10 mM Tris-HC1, pH 8.0. Most of the DTTwas removed by extraction with water-saturated n-butanol, and the reduced oligonucleotide was desalted on a Bio-Spin 6 column (BioRad, Hercules, CA). The reduced oligonucleotide was added to a n equal volume of 10 mM 2,2'-dithiodipyridine (Aldrich, Milwaukee, WI) in acetonitrile, and the mixture was incubated a t room temperature for 30 min. The solution was extracted with diethyl ether (six times) to remove unreacted 2,2'-dithiodipyridine, and the 3' S thiopyridyl oligonucleotide was desalted on a Bio-Spin 6 column. The presence of the thiopyridyl group was confirmed by treatment with DTT and subsequent monitoring of release of thiopyridyl anion a t 342 nm ( E = 7060). The concentrations of oligonucleotides were determined assuming a ratio of l OD per 20 pg/mL for nonhairpin oligonucleotides and 1 OD per 30 pglmL for hairpin oligonucleotides. Melting temperatures with oligonucleotide dissolved in 5 mM bis-Tris-HC1, pH 6.5, 25 mM NaCl were obtained using a Hewlett-Packard (Wilmington, DE) 8452 diode array spectrophotometer using a thermostated cell holder connected to a n adjustable temperature bath. Plasmid Preparation. Plasmid pUC19 (YanischPerron et al., 1985) (New England Biolabs, Beverly MA) was prepared using strains HBlOl or JMlOl by the alkaline lysis method followed by chromatography using columns obtained from Qiagen (Chatsworth, CA) or by CsCl gradient centrifugation. CsCl of ultrapure grade was obtained from Gibco-BRL. The concentration of plasmid DNA was determined by UV absorbance at 260 nm assuming a ratio of 1 OD per 50 pg/mL. Topoisomerase I was either obtained from Gibco-BRL or prepared from wheat germ as described (Dynan e t al., 1981). Plasmid with increased supercoiling was obtained as described (Keller, 1975) using topoisomerase I and varied concentrations of ethidium bromide. The superhelical density of pUC19 was evaluated by agarose gel electrophoresis of the topoisomers as described (Keller, 1975). DNA Affinity Cleavage. DNA cleavage was essentially performed as described (Pei et al., 1990; Corey et al., 1989a,b)using oligonucleotide-nuclease conjugates that were synthesized via disulfide exchange between the 3' thiolated oligonucleotides and staphylococcal nuclease containing an introduced surface cysteine (K116C). 'lasmid pDCl encoding staphylococcal nuclease containing the K116C mutation was provided by Dr. Peter G. Schultz (UC Berkeley). K116C staphylococcal nuclease was expressed behind a lac promoter and an ompA signal sequence within plasmid pDC1, a pONFl derivative (Takahara e t al., 1985), and was isolated as a mixture of monomer and disulfide-linked dimer. The enzyme was completely reduced to monomer by treatment with 50 mM DTT for 8 h a t 37 "C in 10 mM Tris-HC1, pH 8.0. Monomeric enzyme was separated from DTT by Mono S cation exchange chromatography (Pharmacia) in 50 mM NaHEPES, pH 7.5, 1mM EGTA, and a gradient of 0.01.0 M NaC1. The reduced staphylococcal nuclease was mixed with 3'-S-(thiopyridyl)oligonucleotide, and the coupling was monitored at 342 nm. The conjugate was purified by Mono Q anion exchange chromatography (Pharmacia) in 20 mM Tris-HC1, pH 8.0, 1 mM EGTA, and a gradient of 0.0-1.0 M NaCl. The collected oligonucleotide-nuclease conjugate fractions were concentrated to 200 pL and desalted using BioSpin 6 spin columns (Biorad). Oligonucleotide-nuclease conjugates

were annealed to substrate for greater than 30 s at 2537 "C in 25 mM NaC1, 5 mM bis Tris-HC1, pH 6.5. The mixtures were chilled on ice, and the cleavage reactions were initiated by addition of 2.5 mM CaClz followed by termination after 1-5 s by the addition of 5 mM EGTA. The DNA was then treated with restriction enzyme to generate discrete products of identifiable size which were analyzed by 1%agarose gel electrophoresis. Synthesis of Peptide-Oligonucleotide Coqjugates. Peptides were synthesized on either a Symphony Multiplex synthesizer (Rainin, Emeryville, CA) or a n Applied Biosystems Model 430A peptide synthesizer using Fmoc chemistry. The presence of predominantly one species was confirmed by reversed-phase HPLC (Rainin) using a C-18 Microsorb 5 pm 300 A column (Rainin) and 0.1% trifluoroacetic acid in doubly distilled water (buffer A) and a gradient of 0-100% of buffer B (0.08% trifluoroacetic acid in 95:5 acetonitri1e:doubly distilled water). The peptides were characterized by mass spectral analysis acquired by a VG (Altrincham, England) 30-250 quadrapole mass spectrometer using a standard VG electrospray source. The calculated and observed molecular weights were 1175.5 and 1175 for CAGAAKKACAAKK, 1135 and 1135.4 for CAAKKAAKKAAKK, and 1172.4 and 1173 for CGGSPRKSPRK. Two to four mg of dry peptide were weighed out, dissolved in 10 mM Tris-HC1, pH 8.0 buffer, and incubated for 8-12 h at 37 "C with 10 mM DTT to reduce all material to monomeric form. The reduced peptide was purified by reversed-phase HPLC using the column and gradient described above, neutralized with 115 volume 100 mM Tris pH 10.2, and added to a 1 mL quartz cuvette containing the 3'-(S-thiopyridyl)oligonucleotide. The reaction was monitored a t 342 nM utilizing a HewlettPackard 8452 diode array spectrophotometer. Enough peptide and oligonucleotide were used to ensure a distinct peak a t 342 nm ('0.05 OD342) upon completion of the reaction. This material was then purified by anion exchange chromatography with a Mono Q 515 column (Pharmacia) utilizing 20 mM Tris-HC1, pH 8.0, and a gradient of 0.0-1.0 M NaCl. The attachment of the positively charged peptides caused the conjugates to migrate significantly faster than the parent oligonucleotide. The purified conjugate was concentrated to 200 pL and was desalted using BioSpin 6 columns (BioRad). The absorbance maximum of the conjugates was a t 260 nm, as would be expected of a conjugate containing DNA. Treatment of the conjugate by DTT regenerated the free peptide and free oligonucleotide as monitored by HPLC. RESULTS

Selective Cleavage as a Probe of Plasmid Topology. Oligonucleotide-nuclease conjugates were synthe-

sized by coupling 3'-thiolated oligonucleotides to a staphylococcal nuclease variant (K116C) (Pei et al., 1990) containing an introduced surface cysteine. The ability of the conjugates to hybridize to duplex DNA was evaluated by affinity cleavage of supercoiled plasmid DNA (Corey et al., 1989a). Plasmid DNA was prepared by cesium chloride gradient centrifugation or by chromatography and was not denatured prior to addition of the oligonucleotide-nuclease conjugate. Incubation of the oligonucleotide-nuclease conjugates with plasmid, followed by activation of the staphylococcal nuclease domain by CaZf,yielded specific cleavage. Linearized DNA was not cleaved selectively a t any site by the oligonucleotide-nuclease conjugates. Selective cleavage was most efficient at target sequences which contained inverted repeats which have the potential to form cruciform structures (Lilley et al., 1980; Panayotatos and

Strand Invasion by Oligonucleotide-Nuclease Adducts

Bioconjugate Chem., Vol. 6, No. 1, 1995 95

5'

5:5 '

5' puc19 Template 3'

5'

pUC19 Template

35: (XI) -3-' (w)3-' (Xm 3 ' - w G ' I G G A (xIvl3 ' -T"@atTGGA-

Figure 1. Oligonucleotides and oligonucleotide conjugates used in this study. The hybridization region within pUC19 was within bases 1529- 1583. Self-complementary sequences within linear oligonucleotides or within the separate arms of the hairpin oligonucleotides are shown in bold. (Top) oligonucleotides and oligonucleotide conjugates. (Bottom) hairpin and related oligonucleotides. Semicircular connections represent tetracaidine linker regions.

Wells, 1981; Del Olmo and Perez-Ortin, 1993), but also occurred at sites without evident structure (Corey et al., 1989). Hydrolysis of pUC19 by the underivatized nuclease also yielded cleavage at an inverted repeat a t bases 1542- 1567, but in contrast to oligonucleotide-direct cleavage, hydrolysis was accompanied by substantial cleavage a t other sites and required longer incubations with calcium or higher nuclease concentration (Corey et al., 1989a). Addition of free oligonucleotide in conjunction with free nuclease yielded cleavage which was similar to that produced by nuclease alone. The introduction of additional negative supercoiling into the pUC19 plasmid by treatment with topoisomerase I in conjunction with ethidium bromide (Keller, 1975) resulted in efficient cleavage at sites which were not inverted repeats (Figure 2, lane 5), suggesting that the extent of supercoiling influenced strand invasion by the oligonucleotide-nuclease conjugate. Temperature Dependence of Selective Cleavage. The oligonucleotide-nuclease conjugates may have hybridized to substrate duplex DNA stably or they may have hybridized in a readily reversible manner. Knowledge of the stability of hybridization has important implications for the mechanism of strand invasion and its eventual experimental application. To determine whether the conjugates were stably bound we examined the effect of annealing temperature on selective cleavage. Plasmid pUC19 was mixed with oligonucleotide-nuclease conjugate I (Figure 1)which was directed to a n inverted repeat within pUC19. Incubations to permit annealing were performed a t temperatures ranging from 0 to 37 "C for 10 min, after which the mixtures were cooled to 0 "C. Affinity cleavage was then used to probe whether the incubation conditions had allowed hybridization to occur and whether it was stable over time after cooling. No affinity cleavage was observed after annealing a t less than 23 "C (Figure 3 (top), lanes 3 and 4). Cleavage was observed after annealing at 24-26 "C (lanes 5-7) and reached a maximum a t temperatures above 28 "C (Figure 3 (top), lanes 8 and 9). The observation of affinity cleavage after cooling to 0 "C indicated that stable hybridization of the oligonucleotide-nuclease conjugate was occurring, since reversible

1 2 3 4 5

-

3675 23219291 3 7 3 v 1264 702-

Figure 2. Effect of additional supercoiling within pUC 19 on affinity cleavage by a n oligonucleotide-nuclease conjugate directed to a target site which does not contain an inverted repeat. Lane 1: molecular weight markers. Lane 2: uncut pUC 19. Lanes 3-5: affinity cleavage of pUC19 treated with 0, 2.0, or 6.0 mM ethidium bromide and topoisomerase I as described in the Experimental Methods to obtain superhelical densities (a) of -0.056 f 0.14, -0.068 f 0.14, and -0.12 f 0.2. The treated pUC 19 was incubated with an oligonucleotide-nuclease conjugate containing a n oligonucleotide with sequence 5'GGGGn'CCGCGCACATn'CCCCG-3' directed to bases 25832605 of pUC19, and the products from affinity cleavage, if any, were digested with BsaI. BsaI has a single recognition site in pUC19 at 1766, and the predicted fragment sizes of RsaIl oligonucleotide-nuclease cleavage were -840 and -1850.

hybridization would have yielded unbound oligonucleotide-nuclease conjugate which would not have been able to reanneal at temperatures less than 24 "C. As noted above, the temperature dependence of selective cleavage has implications for the application of strand invasion by oligonucleotide-nuclease conjugates, and we sought to determine if intramolecular interactions within conjugate I might contribute to the observed temperature sensitivity. The oligonucleotide domain of the conjugate I is self-complementary (5'-GGATC- - GATCC-3', Figure l ) , so the observed temperature dependence may have been partially due to intramolecular structure within the oligonucleotide domain of the conjugate. To lessen the potential for intramolecular basepairing we synthesized oligonucleotide-nuclease conjugate I1 which was designed to possess a 14-base

Corey et al.

96 Bioconjugate Chem., Vol. 6, No. 1, 1995

12345678

1 2 3 4 5 6 7 0 9

l-

I

1

.

-

z

1

r-r-

I II*

Figure 4. Effect of excess oligonucleotide on DNA cleavage by the analogous oligonucleotide-nuclease conjugate. pUC19 (0.05

pM)was mixed with varying amounts of oligonucleotide I11 for 10 min at 37 "C. Oligonucleotide-nuclease conjugate I (0.07 pM)

1 2 3 4 5 6 I

was added to the mixture and annealed to plasmid for an additional 10 min a t 37 "C prior to cooling to 0 "C and initiation of cleavage. The products were treated with BamHI to generate discrete fragments. Lane 1: uncut pUC19. Lane 2: pUC 19 linearized by HindIII. Lane 3: BgZ I digest of pUC 19 to yield marker fragments of 1118 and 1568 base pairs. Lanes 4-8: digestion of pUC 19 by I in the presence of 0, 1.75, 3.5, 7, and 14 pM 111.

One explanation for the failure of added oligonucleotide

to reduce specific cleavage is that the oligonucleotide may

Figure 3. Effect of temperature on cleavage of pUC19 by oligonucleotide-nuclease conjugates I or 11. pUC 19 (0.05p M ) was mixed with oligonucleotide-nuclease conjugates I or I1 (0.07 pM) a t 22 "C and annealed at various temperatures for 10 min prior to cooling to 0 "C and initiation of cleavage. After cleavage the products were digested with BamHI. BamHI has a single recognition site within pUC 19 a t 417, and the predicted fragments of cleavage by BamHI and I or I1 were -1140 and -1540. (Top) Lane 1: linearized pUC19. Lane 2: pUC19 treated with BgZ I to yield marker fragments of 1118 and 1568 base pairs. Lanes 3-9 show the effects of annealing I at 22.5, 23, 24, 25, 26, 28, and, 37 "C. (Bottom) Lanes 1-6: effect of annealing I1 a t 14, 16, 18, 20, 22, and 24 "C.

oligonucleotide domain which was similar to I but lacked five nucleotides at the 5' terminus. Conjugate I1 exhibited maximal affinity cleavage a t 18 "C (Figure 3(bottom), lane 3), 8-12 "C lower than I (Figure 3 (top)),suggesting that the lack of self-complementarity contributes to a n altered profile of incubation temperature versus activity. Effect on Affinity Cleavage of the Addition of the Analogous Unmodified Oligonucleotide. By design, the nuclease domain was responsible for affinity cleavage by the conjugate. Could it also effect the stability of hybridization? To determine the impact, if any, of the attached nuclease on strand invasion, we examined the effect on aMinity cleavage by oligonucleotide-nuclease conjugate I from the addition an excess of a n unmodified oligonucleotide of the same sequence 111. Previous studies had shown that the addition of unmodified oligonucleotide efficiently inhibited the selective hydrolysis of single-stranded DNA or RNA substrates by the analogous oligonucleotide-nuclease conjugates (Corey and Schultz, 1987; Zuckermann and Schultz, 1989) by blocking the target annealing site. For duplex substrate DNA, however, we observed that up to a 200-fold excess of I11 did not substantially inhibit affinity cleavage by I (Figure 4, lanes 4-8). We observed this result even when I11 was preincubated with pUC19 for 8 h at 37 "C prior to the addition of oligonucleotide-nuclease conjugate I.

anneal and create a D-loop structure that is readily cleaved by unbound oligonucleotide-nuclease conjugate. To account for this possibility, oligonucleotide I11 (0-200fold excess) was added to plasmid and annealed at 37 "C prior to cooling to 0 "C and addition of oligonucleotide conjugate I. In spite of the opportunity for I11 to anneal and the presence of unbound I, no cleavage of plasmid was observed (results not shown). Effect on Affinity Cleavage of the Addition of Oligonucleotide-Peptide Conjugates. The stabilization of stand invasion by the attachment of small molecules to oligonucleotides rather than by the attachment of staphylococcal nuclease would afford conjugates with the potential for more general utility. Staphylococcal nuclease contains 22 lysines on its surface and has a net positive charge of +12 (Loll and Lattman, 1989). To examine whether the stabilization conferred by staphylococcal nuclease could be replicated by attachment of simple positively-charged synthetic peptides we synthesized conjugates IV and V, consisting of an oligonucleotide linked via disulfide bond formation to peptides containing either four (CAGAAKKAGAAKK IV)or six V ),C -( lysine residues. The dilysine repeats were chosen because they also appear on the surface of staphylococcal nuclease at residue pairs 516, 48/49, 63/64, 71/72, and 133/134 (Tucker et al., 1978). Similar peptides have been shown to bind in an a-helical conformation to duplex DNA (Johnson et al., 1994). The oligonucleotide-peptide conjugates did not inhibit selective cleavage by the oligonucleotide-nuclease conjugate I (Figure 5, lanes 6 and 7) when present at a 50-fold excess. Higher concentrations of the oligonucleotidepeptide conjugates began to inhibit selective cleavage, but inhibition was indistinguishable from that produced by similar concentrations of the free peptide (data not shown). A conjugate (VI)consisting of an oligonucleotide coupled to a peptide (CGGGSPKKSPKK) containing a sequence known to bind independently to DNA at A/Trich sites (Churchill and Suzuki, 1989) also failed to inhibit the analogous oligonucleotide-nuclease conjugate I (Figure 5, lane 5). These results suggest that the stabilization of strand invasion may depend on structural features of staphylococcal nuclease which cannot be readily mimicked by simple cationic peptides.

Strand Invasion by Oligonucleotide-Nuclease Adducts

Bioconjugafe Chem., Vol. 6, No. 1, 1995 97

12345678

Figure 5. Effect of peptide- and acridine-linked oligonucleotides on affinity cleavage of pUC19 (0.05 mM) by oligonucleotide nuclease conjugate I(0.07 mM). Lane 1: uncut pUC 19. Lane 2: pUC 19 linearized with RamHI. Lane 3: pUC 19 treated with BgZI. Lane 4: cleavage by oligonucleotide-nuclease conjugate I alone. Lane 5: effect of addition of 3.5 pM oligonucleotide-peptide conjugate VI. Lane 6: effect of addition of 3.5 pM oligonucleotide-peptide conjugate TV. Lane 7: effect of addition of 3.5 pM oligonucleotide-peptide conjugate V. Lane 8: effect of addition of 14/ t M oligonucleotide-acridine conjugate

12345678

VII. Effect on Affinity Cleavage of the Addition of Oligonucleotide-Acridine Conjugates. Acridine is able to intercalate between DNA bases, and its attachment to oligonucleotides had previously been shown to stabilize hybridization to single-stranded DNA (Sun et al., 1989). To assay if the linkage of acridine to a n oligonucleotide could increase the stability of hybridization sufficiently to block the annealing of the oligonucleotide-nuclease conjugate we obtained oligonucleotide VII, which had been derivatized at the 5' terminus with acridine. A 200-fold excess of oligonucleotide-acridine conjugate VI1 was preincubated with pUC19 prior to addition of oligonucleotide-nuclease conjugate I. After activation of the nuclease and subsequent analysis no inhibition of affinity cleavage was observed (Figure 5, lane 8). Effect on Affinity Cleavage of the Addition of Hairpin and ComplementaryOligonucleotides. We examined the ability of hairpin oligonucleotides to inhibit affinity cleavage. Knowledge of the cause of any inhibition would allow insight into the relative contributions to strand exchange of nuclease-DNA interactions and DNA-DNA base-pairing. Hairpin oligonucleotides are a promising motif for hybridization to duplex DNA, since they can base-pair to both strands of the target sequence and because their enhanced stability to nuclease degradation makes them better candidates for use in complex in vitro or i n vivo systems (Yoshizawa et al., 1994). Hairpin oligonucleotides might inhibit affinity cleavage by directly annealing to the target site on plasmid DNA and blocking subsequent annealing of the oligonucleotide-nuclease conjugate. Alternatively, the oligonucleotide-nuclease conjugate may take advantage of nuclease-DNA interactions to stabilize annealing of the conjugate to the complementary strand of a hairpin, thus preventing it from recognizing its plasmid target. Hairpin oligonucleotides were designed to hybridize to both strands of the target sequence (Figure 1,sequences XI and XII) and block subsequent hybridization of oligonucleotide-nuclease conjugate I by one of the mechanisms outlined above. Under the buffer and salt conditions used for the assay the measured melting temperature of hairpin XI was 72 "C. We incubated hairpin XI with duplex DNA at 37 "C and then added oligonucleotide-nuclease I. The incubation was continued for a n additional 10 min at 37 "C after which the nuclease was activated. We observed that selective hydrolysis was almost completely inhibited at a 50-fold excess of (XI)

Figure 6. (Top) effect of the addition of increasing amounts of hairpin oligonucleotide XI on the affinity cleavage of pUC 19 (0.05 pM) by oligonucleotide-nuclease conjugate I (0.07 pM). Lanes 1 and 2: no hairpin added. Lanes 3-8: 0.07, 1.4, 2.1, 2.8,3.5, and 4.9 pM hairpin oligonucleotide XI added. (Bottom) effect of the addition of various oligonucleotides on the affinity cleavage pUC 19 (0.05 pM) by oligonucleotide nuclease I (0.07 pM). Lane 1: no oligonucleotide added. Lane 2: 2.8 p M (40fold excess) oligonucleotide XI added. Lane 3: 14 pM (200-fold excess) oligonucleotide VI11 added. Lane 4: 14 p M (200-fold excess) oligonucleotide IX. Lane 5: 14 pM (200-fold excess) hairpin oligonucleotide complementary to bases 425-450 within pUC19. Lane 6: 2.8 pM (40-fold excess) oligonucleotide XII. Lane 7: 14 pM (200-fold excess) oligonucleotide X. Lane 8: 2.8 pM (40-fold excess) of oligonucleotide XIII.

(Figure 6 (top), lane 6, and (bottom), lane 2). Partial hairpin (XII)inhibited affinity cleavage similarly (Figure 6 (bottom), lane 6). No inhibition was observed if oligonucleotide-nuclease conjugate I was allowed to hybridize to DNA prior to addition of hairpin XI, indicating that the addition of the hairpin could not cause displacement of a n annealed oligonucleotide-nuclease conjugate. A hairpin oligonucleotide directed to another site within pUC19 did not inhibit selective cleavage by I to any extent (Figure 6 (bottom), lane 5), suggesting that inhibition was due to Watson-Crick base-pairing by XI or XII. Inhibition by the hairpin XI may have been due to its structure or to the sequence of its individual arms. A 200-fold excess of oligonucleotides VI11 and IX,which separately make up the two arms of hairpin XI (Figure l ) , were preincubated with plasmid either singly or in combination. Neither substantially inhibited selective cleavage by conjugate I (Figure 6 (bottom), lanes 3 and 4). The incomplete inhibition after preincubation with a 200-fold excess of oligonucleotide M (Figure 6 (bottom), lane 4) was especially noteworthy because IX was complementary to the oligonucleotide-nuclease I and would have been expected to base-pair with it and thus directly prevent hybridization to the duplex. However, when analogous oligonucleotides with extended 5' sequences were examined (XIII (Figure 6 (bottom), lane 8) (Scheme 1, iv) and XN (results not shown)), we noted an inhibition of selective cleavage similar to that caused by hairpin XI. Hairpin X, which possessed a similar capacity to base-pair to pUC19 as partial hairpin XII,

98

SioconjugafeChem., Vol. 6,No. 1, 1995

Corey et al.

Scheme 1. Ability of Oligonucleotide-Nuclease Conjugate I to Hybridize to pUC19 in the Presence of Oligonucleotides 111, XI, E,XIV, and X Oligonucleotide Adduct

Potential complementarity between Oligonucleotide and Oligonucleotide-NucleaseAdduct

Oligonucleotide

Inhibition of Affinity 'leavage

NO

NONE 5 ' --l.crIcA= A

YES

NO

5'-T,

iv. 5

G-3

YES

5"'P

A

NO

but could make fewer base-pairs with I, did not inhibit selective cleavage (Figure 6 (bottom), lane 7) (Scheme 1, VI.

Scheme 1summarizes the data obtained by the inhibition of affinity cleavage and our view of the base-pairing by the oligonucleotide-nuclease conjugate. (i) An oligonucleotide I11 of the analogous sequence to the oligonucleotide-nuclease conjugate does not efficiently block hybridization by the conjugate (Figure 4). (ii) Hairpin oligonucleotides containing a complementary sequence (XI or XII) can incorporate an oligonucleotide nuclease and prevent it from hybridizing to its target sequence within pUC19 (Figure 6 (top), (bottom, lanes 2 and 6)). I11 Complementary oligonucleotides (E) cannot block hybridization by the oligonucleotide-nuclease (Figure 6 (bottom),lane 4) unless (iv) they contain an extended 5'tail capable of additional interactions with the nuclease domain (Figure 6 (bottom), lane 8 ) (XIII) or (XIV). Finally, (v) a hairpin oligonucleotide (X) which is only partially complementary to conjugate I does not inhibit affinity cleavage (Figure 6 (bottom), lane 7) even though it contains the same potential to base-pair with pUC19 as inhibitory hairpin XII. DISCUSSION

Hybridization of oligonucleotides to duplex DNA by Watson-Crick base-pairing potentially offers the most versatile route to sequence recognition. Such recognition would facilitate the control of proteins involved in recombination, replication, and transcription. This ability might have considerable implications for in vitro and in vivo functional studies. However, before functional studies can be attempted, more information is required concerning the impact on strand invasion of the topology of the target and the impact of modifications to the invading oligonucleotide. For proficient strand invasion optimized oligonucleotides will need to be targeted to optimal sequences.

Oligonucleotide-nuclease conjugates represent a valuable tool for obtaining this information because they can hybridize within duplex DNA and because their hybridization can be evaluated by afinity cleavage. Hybridization of oligonucleotide-nuclease conjugates is limited to supercoiled DNA and is most facile at sites containing inverted repeats or to template with relatively high supercoiling. The need for the target to be supercoiled is a restriction relative to recognition by triple helix formation, which allows hybridization to target sequences within relaxed DNA. However, DNA can be supercoiled in vivo, particularly in regions which are transcriptionally active, and certain sequence motifs are known to be particularly susceptible to strand dissociation (Benham, 1993; Huang and Kowalski, 1993). Therefore, these sequences may be the most promising targets for the extension of this strategy to the targeting of oligonucleotides and their derivatives to duplex DNA for in vitro or in vivo enzymological studies. The oligonucleotide-nuclease conjugates appear to hybridize stably to duplex DNA since the conjugate remains hybridized and capable of affinity cleavage after cooling to 0 "C, a temperature which is too low to permit the initiation of hybridization (Figure 3 (top and bottom)). The temperature sensitivity of affinity cleavage can be altered by modulating the self-complementarity of the oligonucleotide domain of the conjugate. By contrast, unmodified oligonucleotides must not be able to hybridize as stably, since oligonucleotide I11 did not block hybridization by the analogous oligonucleotide-nuclease conjugate I when present in 200-fold excess. This result was also observed if the underivatized oligonucleotide was added first and allowed to incubate with the duplex target for 8 h prior to addition of the conjugate. These observations implicate the attached nuclease in the promotion of stable strand invasion. Staphylococcal nuclease, a 149 amino acid protein, possesses a net positive charge of +12, and this dense positive charge on the nuclease surface may be responsible for shifting

Strand Invasion by Oligonucleotide-Nuclease Adducts

the equilibrium between free and DNA-associated conjugate to favor strand invasion. How might the surface charge of staphylococcal nuclease promote the initiation and stabilization of strand invasion? At least two interrelated mechanisms may contribute. Non-sequence-specific electrostatic interactions between the positively charged surface of the nuclease and the phosphodiester backbone of the duplex could bring the attached oligonucleotide into close proximity to the duplex, increasing its local concentration near the complementary target site and facilitating strand invasion by enhancing the probability of the initiation of base-pairing. The nuclease might also stabilize local disruptions of the DNA helix a t the target site, which might facilitate the initiation of base-pairing by the oligonucleotide or stabilize the base-pairing once it occurs. The promotion and stabilization of duplex unwinding would be consistent with the physiological function of staphylococcal nuclease since its active site can only accommodate nucleotides that are not basepaired (Loll and Lattman, 1989). A recent report that proteins such as bovine serum albumin and transcription factor IIIa are able to promote strand exchange between synthetic duplexes and single-stranded M13 DNA supports the notion that proteins without known roles in genetic recombination can facilitate hybridization (Kmiec and Holloman, 1994). While surface charge may play a role in hybridization, the failure of peptide-oligonucleotide conjugates (Figure 5) to stably anneal indicates that this role cannot yet be readily reproduced by unstructured collections of positively charged residues. The three-dimensional framework of the enzyme may act as a scaffold to orient positively charged residues so as to optimize interactions between the phosphate backbone and the DNA duplex. The eventual successful design of peptides or small molecules which can mimic the stabilizing function of staphylococcal nuclease may require a rigid orientation of charge and remains an attractive goal. Additional evidence for the ability of the attached nuclease domain to affect hybridization of the conjugate stemmed from the observation that staphylococcal nuclease was able to hybridize within a 25 base-pair hairpin oligonucleotide during annealing a t 37 "C(Figure 6 (top), lanes 3-8, and (bottom), lane 2). This hybridization occurs in spite of the hairpin possessing a measured melting temperature of 72 "C and in spite of the entropic favorability of intramolecular reclosure of the hairpin. Similarly, the oligonucleotide-nuclease preferred to hybridize to its plasmid target sequence rather than to preannealed complementary oligonucleotides, unless the complementary oligonucleotides possessed a n extended single-stranded region a t its 5'-terminus (Scheme 1).This is a particularly remarkable result since it suggests that preexisting hybridization can be broken in order to yield a final hybridized complex which allows both base-pairing and interactions between the nuclease and the target DNA. These results are significant because they reinforce the suggestion that the nuclease has a dual role involving both substrate recognition and substrate cleavage. In conclusion, the affhity cleavage of duplex DNA by oligonucleotide-nuclease conjugates affords insights into the potential of oligonucleotides to recognize sequences within duplex DNA by strand invasion. Stable sequence recognition requires the nuclease domain of the conjugate as well as the oligonucleotide, and a detailed understanding of how the nuclease accomplishes this may lead to the design of conjugates which are more amenable t o in vitro and in vivo applications. Any polynucleotide derivative able to block cleavage by the analogous oligo-

Sioconjugate Chem., Vol. 6,No. 1, 1995 99

nucleotide-nuclease conjugate would have favorable hybridization properties and would merit further study. The discovery of these conjugates will be aided by the utilization of the inhibition of affinity cleavage by the oligonucleotide motifs for the rapid evaluation of novel strategies of strand invasion. ACKNOWLEDGMENT

We wish to thank Elana Varnum for skilled technical assistance. This work was supported by a grant from the Welch Foundation (1-1244). D.R.C. is a n Assistant Investigator with the Howard Hughes Medical Institute. LITERATURE CITED Beattie, K L., Wiegand, R. C., and Radding, C. M. (1977)Uptake of homologous single-stranded fragments by superhelical DNA Characterization of the reaction J . Mol. Biol. 116,783803. Benham, C. J. (1993) Sites of predicted stress-induced DNA duplex destabilization occur preferentially at regulatory loci. Proc. Natl. Acad. Sci. U.S.A. 90, 2999-3003. Chen, C. B., Gorin, M. B., and Sigman, D. S. (1993) Sequencespecific scission of DNA by the chemical nuclease activity of 1,lO-phenanthroline-copper (I) targeted by RNA. Proc. Natl. Acad. Sci. U.S.A. 90, 4206-4210. Cheng, S., Van Houton, B., Gamper, H. B., Sancar, A., and Hearst, J. E. (1988)Use of Psoralen-modified oligonucleotides to trap three stranded RecA-DNA complexes by ABC excinuclease. J. Biol. Chem. 263, 15110-15117. Churchill, M. A., and Suzuki, M. (1989) SPKK motifs prefer to bind to DNA at A/T rich sites. EMBO J. 8, 4189-4195. Colocci, N., Distefano, M. D., and Dervan, P. B. (1993) Cooperative Oligonucleotide-directed triple helix formation at adjacent DNA sites. J. Am. Chem. SOC.115, 4468-4473. Corey, D. R., and Schultz, P. G. (1987) Generation of a Hybrid Sequence-Specific Single-Stranded Deoxyribonuclease. Science 238,1401-1403. Corey, D. R., Pei, D., and Schultz, P. G. (1989a) The sequenceselective hydrolysis of duplex DNA by an oligonucleotidedirected nuclease. J. Am. Chem. SOC.111, 8523. Corey, D. R., Pei, D., Schultz, P. G. (198913)The generation of a catalytic oligonucleotide-directed nuclease. Biochemistry 28, 8277-8286. Del Olmo, M., and Perez-Ortin, J. E. (1993) A natural A/T-rich sequence from the yeast F B P l gene exists as a cruciform in Escherichia coli cells. Plasmid 29, 222-232. Demidov, V., Frank-Kamenetskii, M. D., Egholm, M., Buchardt, O., and Neilson, P. E. (1993) Sequence selective cleavage of double stranded DNA cleavage by peptide nucleic acid (PNA) targeting using S1 nuclease. Nucl. Acids Res. 21,2103-2107 Dynan, W. S., Jendrisak, J. J., Hager, D. A., and Burgess, R. R. (1981)Purification and characterization of wheat germ DNA topoisomerase I (nicking-closing enzyme). J . Biol. Chem. 256, 5860-5865. Ferrin, L. J., and Camerini-Otero, D. R. (1991) Selective cleavage of human DNA RecA assisted restriction endonuclease (RARE) cleavage. Science 254, 1494-1497. Francois, J-C., Saison-Behormas, T., Thuong, N. T., and Helene, C. (1989) Inhibition of restriction endonuclease cleavage via triple helix formation by homopyrimidine oligonucleotides. Biochemistry 28, 9617-9619. Hanvey, J. C., Peffer, N. J., Bisi, J. E., Thomson, S. A., Cadilla, R., Josey, J. A., Ricca, D. J.,Hassman, C F., Bonham, M. A., Au, K. G., Carter, S. G., Bruckenstein, D. A., Boyd, A. L., Noble, S. A., and Babiss, L. E. (1992) Anti-sense and antigene properties of peptide nucleic acids. Science 258, 14811485. Holloman, W. K., Wiegand, R., Hoessli, C., and Radding, C. M. (1975) Uptake of homologous single-stranded fragments by superhelical DNA: A possible mechanism for initiation of genetic recombination. Proc. Natl. Acad. Sci. U S A . 72,23942398. Huang, R.-Y., and Kowalski, D. (1993) A DNA unwinding element and an ARS consensus comprise a replication origin within a yeast chromosome EMBO J . 12,4521-4531.

Corey et al.

100 Sioconjugate Chem., Vol, 6, No. 1, 1995

Jayasena, S. D., and Johnston, B. H. (1993a) Sequence limitations of triple helix formation by alternate strand recognition. Biochemistry 32, 2800-2807. Jayasena, V. K.,and Johnston, B. H. (1993b) Complementstabilized D-1oop:RecA-catalyzedstable pairing of linear DNA molecules at Internal sites. J . Mol. Biol. 230, 1015-1024. Johnson, N. P., Lindstrom, J., Baase, W. A., and von Hippel, P. H. (1994) Double-stranded DNA templates can induce ahelical conformation in peptides containing lysine and alanine: Functional implications for leucine zipper and helixloop-helix transcription factors. Proc. Natl. Acad. Sci. U.S.A. 91,4840-4844. Keller, W. (1975) Determination of the number of superhelical turns in simian virus 40 DNA by gel electrophoresis. Proc. Natl. Acad. Sci. U.S.A. 72, 4876-4880. Kmiec, E., B., and Holloman, W. K. (1994)DNA strand exchange promoted in the absence of homologouspairing. J . Biol. Chem. 269, 10163-10168. Koh, J.-S., and Dervan, P. B. (1992) Design of a nonnatural deoxyribonucleoside for recognition of GC base pairs by oligonucleotide-directedtriple helix formation. J . Am. Chem. SOC. 114, 1470-1478. Lilley, D. M. (1980) The inverted repeat as a recognisable structural feature in supercoiled DNA molecules. Proc. Natl. Acad. Sci. U.S.A. 77, 6468-6472. Loll, P. J., and Lattman, E. E. (1989) The crystal structure of the ternary complex of staphylococcalnuclease, Ca2+.and the inhibitor pdTp, refined at 1.65 A. Prot. Struct. Func. Gen. 5, 183-201. Moser, H. E.,and Dervan, P. B. (1987) Sequence Specific Cleavage of double strand DNA by triple helix formation. Science 238, 645-650. Neilsen, P. E.,Egholm, M., Berg, R. H., and Buchardt, 0. (1991) Sequence-SelectiveRecognition by Strand Displacement with a Thymine-Substituted Polyamide. Science 254, 1497-254. Neilsen, P. E., Egholm, M., Berg, P. H., and Buchardt, 0. (1993) Sequence specific inhibition of DNA restriction enzyme cleavage by PNA Nucl. Acids Res. 21, 197-200. Panayotatos, N., and Wells, R. D. (1981) Sequence specific inhibition of DNA restriction enzyme cleavage by PNA. Cruciform structures in supercoiled DNA. Nature 289,466470.

Pei, D., Corey, D. R., and Schultz, P. G (1990) Site specific cleavage of duplex DNA by a semisynthetic nuclease via triple helix formation. Proc. Natl. Acad. Sci. U.S.A. 87, 9858. Peffer, N. J., Hanvey, J. C., Bisi, J. E., Thomson,S. A., Hassman, C. F., Noble. S. A., and Babiss, L. E. (1993) Strand invasion of duplex DNA by peptide nucleic acid oligomers.Proc. Natl. Acad. Sci. U.S.A. 90, 10648-10652. Rao, B. J., Chiu, S. K., and Radding, C . M. (1993) Homologous recognition and ttriplex formation promoted by RecA protein between duplex oligonucleotides and single-stranded DNA. J . Mol. Biol. 229, 328-343. Sun, J.-S., Francois, J.-C., Montenay-Garestier, T., SaisonBehmoaras, T., b i g , V., Thuong, N. T., and Helene, C. (1989) Sequence-specificintercalating agents: intercalation at specific sequences on duplex DNA via major groove recognition by oligonucleotide-intercalator conjugates. Proc. Nut. Acad. Sci. U.S.A. 86, 9198-9202. Takahara, M., Hibler, D. W., Barr, P. J.,Gerlt, J. A., and Inouye, M. (1985) The ompA signal peptide directed secretion of staphylococcal nuclease by Escherichia coli. J . Biol. Chem. 260, 2670-2674. Tucker, P. W., Hazen, E. E., and Cotton, F. A. (1978) Staphylococcal nuclease reviewed: A prototypic study in contemporary enzymology. 1. Isolation; physical and enzymatic properties. Mol. Cell. Biochem. 22, 67-77. Yanisch-Perron, C., Viera, J., and Messing, J. (1985) Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33,103119. Yoshizawa, S., Ueda, T., Ishido, Y., Miura, K., Watanabe, K., and Hirao, I. (1994)Nuclease resistance of an extraordinarily thermostable mini-hairpin DNA fragment, d(GCGAACG)and its application to in vitro protein synthesis Nucl. Acids. Res. 22,2217-2221. Young, S. L., Krawczyk, S. H., and Matteucci, M. D. (1991) Triple helix formation inhibits transcription elongation in vitro. Proc. Natl. Acad. Sci. U.S.A. 10023-10026. Zuckermann, R. N., and Schultz, P. G. (1989) Site-selective cleavage of structured RNA by a staphylococcal nuclease-DNA hybrid. Proc. Natl. Acad. Sci. U.S.A. 86, 1766-1770. BC940090V