Sioconiugate Chem. 1994, 5,327-332
327
DNA-Linked RNase H for Site-Selective Cleavage of RNA? Yohtaro Uchiyama,tlg Hideo Inoue,ll Eiko Ohtsuka,’tll Chieko Nakai,l Shigenori Kanaya,l Yoshio Ueno,tys and Morio Ikeharal*l Faculty of Pharmaceutical Sciences, Science University of Tokyo, Tokyo 162, Japan, Research Institute for Biosciences, Science University of Tokyo, Noda 278, Japan, Faculty of Pharmaceutical Sciences, Hokkaido University, Sapporo 060, Japan, and Protein Engineering Research Institute, Suita, Osaka 565, Japan. Received December 15,1993’
A DNA-linked RNase H (Hybrid Enz-1) (Kanaya e t al. (1992) J.Biol. Chem. 267,8492-8498), in which dGTCATCTCC was attached to E. coli RNase H via a covalent linker of 21 A, was altered to improve the site-specific RNA cleavage by increasing the linker length. The sizes of the linkers on these hybrid enzymes (Hybrid Enz-2, -3, and -4) differed by 3 A, the axial rise of the DNA/RNA hybrid, to give 18-, 24-, and 27-1\ lengths. The conjugate with a size of 27 A was able to cleave a synthetic 22mer RNA (5’-rAAGAUGUCUACGGAGAUGACCA-3’), containing the complementary Smer RNA sequence (underlined), a t one position, A16-Ul7. The kinetic parameters of Hybrid Enz-1, -2, -3, and -4 were examined using a Smer RNA target. The results showed that longer linkers produced higher K,, kcat, and kcatlKmvalues, and the kcaJKmvalue of the conjugate with the 27-A linker reached 83% of that of the wild-type RNase H. Hybrid Enz-4 was found to be useful as an RNA restriction endonuclease.
To cleave RNA site-specifically, investigators have reported several methods. These include chemical cleavage, with compounds such as Fe(I1)-bleomycin (Carter et al., 1990),and enzymatic cleavage with ribozymes (Symons, 1992), staphylococcal nuclease-DNA conjugate (Zuckermann et al., 1988; Zuckermann and Schultz, 19891, and RNase A-DNA conjugate (Zuckermann and Schultz, 1988). In the approaches using RNase H,’ short oligodeoxynucleotides or modified oligonucleotides, such as 2‘modified oligonucleotides (Hayase et al., 1990; Monia et al., 1993) and mixed-phosphate-backbone oligodeoxynucleotides (Agrawal et al., 1990), were used as the complementary strand. RNase H, which participates in DNA replication and repair (Crouch and Derksen, 1982; Crouch, 1990), endonucleolytically degrades the RNA moiety of a DNA/RNA hybrid, in the presence of Mg2+or Mn2+,to produce 5’phosphates a t the hydrolysis sites (Miller et al., 1973; Berkower et al., 1973). The enzyme consists of a single polypeptide chain with 155 amino acid residues (Kanaya and Crouch, 1983),and the sequence is similar to the RNase H domains of the reverse transcriptases from retroviruses, including human immunodeficiency virus (Johnson et al., 1986; Doolittle, 1989). Site-directed mutagenesis experiments and computer analysis of the homology of the amino acid sequence between RNase H and the C-terminal domains of the reverse transcriptases from many species suggest that the residues of the active site are Asp-10, + This work was supported in part by a Grant-in-Aidfrom the Ministry of Education, Science and Culture, Japan. * Author to whom correspondence should be addressed. Tel: +81-11-706-4979;FAX: +81-11-706-4989. t Faculty of Pharmaceutical Sciences, Science University of Tokyo. Research Institute for Biosciences. 11 Faculty of Pharmaceutical Sciences, Hokkaido University. I Protein Engineering Research Institute. Abstractpublished in Advance ACS Abstracts, June 1,1994. 1 Abbreviations: RNase H, ribonuclease H; EDTA, ethylenediaminetetraacetic acid; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; kDa, kilodaltons.
Glu-48, and Asp-70. The three-dimensional structure of RNase H was determined by X-ray crystallographic studies (Katayanagi et al., 1990; Yang e t al., 1990) and showed that Mg2+binds close to the three acidic amino acids within the active site. RNase H has a slight specificity for the base sequence of the hybrid. For cleavage of RNA a t single site, a DNAlinked RNase H (Hybrid Enz-1) (Kanaya et al., 1992) has been constructed using the mutant RNase H (C135/RNase H) (Kanaya et al., 1990) in which all three Cys residues were replaced by Ala, and Glu-135 was replaced by Cys. The Smer DNA (dGTCATCTCC) was linked with a 21-A covalent linker including maleimide a t the 5’-terminus of the DNA and then attached to Cys-135 of the mutant RNase H. In the conjugate, the DNA portion plays the role of cleavage site recognition and the RNase H portion catalyzes the hydrolysis of the RNA that hybridizes with the DNA portion. In the presence of the Smer DNA, the wild-type enzyme and C13WRNase H cleaved the Smer RNA (rGGAGAUGAC)a t the A5-U6 and U6-G7 positions and the 22mer RNA (5’dAGAUGUCUACGGAGAUGACCA-3’) containing the complementary Smer RNA sequence (underlined) a t the A16-Ul7 and U17-Gl8 positions (corresponding to the Smer RNA). On the other hand, Hybrid Enz-1 with the Smer DNA attached via a 21-A linker cleaved the complementary Smer RNA a t a single site (A5-U6) and cleaved the 22mer RNA a t two positions (A14-Gl5 and A16-Ul7). Cleavage by Hybrid Enz-1 a t A14-Gl5, the site the nonconjugated enzymes did not cleave, was thought to be caused by the shorter linker length. In this study, the linker length was altered by 3 A, which is equal to the axial rise of a DNA/RNA hybrid (Saenger, W., 1984), and three hybrid enzymes with linkers of different lengths (18, 24, and 27 A) were constructed. Hybrid Enz-4 (Figure l),whose linker size was 6 8, (the axial rise of two bases) longer than that of Hybrid Enz-1, was found to cleave the 22mer RNA almost site-selectivity a t A16-U17. The effect of the linker length on the activity was examined by measuring the kinetic parameters ( K , and Kcat values) for the cleavage of the Smer RNA.
i o 4 3 - 7 a o 2 1 9 4 1 2 9 0 5 - 0 3 ~ ~ ~ 0 4 . 5 o 10~ 1994 American Chemical Society
Uchiyama et al.
328 Bioconjugate Chem., Vol. 5, No. 4, 1994
(CH2)5
I c=o
0 5' II NH-(CH&-O-P-O-dGTCATCTCC
I
3'
I
0-
Figure 1. Model of Hybrid Enz-4. @-Strandsand a-helices are
represented by arrows and cylinders,respectively. The three acidic residues required for activity are shown by solid circles with residue numbers. MATERIALS AND METHODS
Materials. Aminomodifier I1 (N-Fmoc-Ol-(dimethoxytrity1)-02-[(cyanoethoxy)(diisopropylamino)phosphinyl]3-amino-1,2-propanediol) was purchased from Clontech Co., Ltd., and aminolink 2 (N-(trifluoroacety1)-O-[methoxy(diisopropylamino)phosphinyll-6-aminohexan-l-ol) was from Applied Biosystems Inc. N-[(t-Maleimidocaproy1)oxylsuccinimide (EMCS) and N- [(r-maleimidobutyry1)oxylsuccinimide (GMBS) were purchased from Dojindo Laboratories. Crotalus durissus phosphodiesterase was purchased from Boehringer Mannheim. The Smer RNA (rGGAGAUGAC) and the Smer DNA (dGTCATCTCC) have been described previously (Kanaya et al., 1992). Synthesis of the 22mer RNA (rAAGAUGUCUACGGAGAUGACCA). Oligonucleotides were synthesized on an Applied Biosystems 394 DNA/RNA Synthesizer using the standard phosphoramidite method (Caruthers, 1985) with commercially available reagents (MilliGen Biosearch for RNA). The products were deprotected and purified to give single peaks on reversed-phase and anion-exchange HPLC. Reversed-phase HPLC was carried out on an Inertsil ODS-2 column (4.6 mm X 250 mm) from GL Sciences Inc. For the 22mer RNA, elution was performed with a linear gradient of acetonitrile in 0.1 M triethylammonium acetate buffer, pH 7.0, from 9.0 to 14% (v/v) over 25 min at a flow rate of 1.0 mL/min a t 50 "C. The retention time was 14.7 min. Anion-exchange HPLC was carried out on a DEAE 2SW column (4.6 mm X 250 mm) from Tosoh Co., Ltd. Elution was performed with a linear gradient from 0.4 to 0.9 M ammonium formate in 20% acetonitrile for 25 min at a flow rate of 1.0 mL/min a t 50 "C. The retention time was 20.2 min. Labeling of the 5'-Ends of the Smer and 22mer RNAs. The 5'-ends of the Smer and 22mer RNAs were 32P-labeled with T4 polynucleotide kinase from E. coli strain A19 and [Y-~~PIATP and were purified on a NENSORB 20 column (DuPont). Preparation of Hybrid Enz-1, -2, -3, and -4. Hybrid Enz-1 (dS-ClSB/RNase H) with a 21-Alinker was prepared as described (Kanaya et al., 1992). Hybrid Enz-2 with an 18-A linker was prepared by the same procedure used for Hybrid Enz-1, except GMBS and AminoModifier I1 were used. Hybrid Enz-3 (24 A) and -4 (27 A) were prepared
using GMBS plus aminolink 2 and EMCS plus aminolink 2, respectively. The 5'-end amino group was introduced into the Smer DNA (dGTCATCTCC) in the last step of the DNA synthesis. After the protecting groups were removed, the product was then allowed to react with the bifunctional reagent in dimethylformamide in the presence of 50 mM phosphate buffer, pH 7.5, a t room temperature for 1 h. The product was purified by gel filtration on Sephadex G-25 in water. The purity was checked by reversed-phase HPLC. Elution was performed with a linear gradient of acetonitrile in 0.1 M triethylammonium acetate buffer, pH 7.0, from 0 to 38% (v/v) over 25 min at a flow rate of 1.0 mL/min at 50 "C. The retention times of the linkerDNAs were 21.5 min for HybridEnz-l,l9.6 min for Hybrid Enz-2,21.2 min for Hybrid Enz-3, and 23.1 min for Hybrid Enz-4. The coupling reaction of the mutant RNase H, the C135/ RNase H (10 nmol), and the linker-9mer DNA (10 nmol) was carried out in 100 pL of Tris-HC1 buffer, pH 7.0, at room temperature for 1h. The resultant hybrid enzymes were purified by cation-exchange HPLC on an Asahipack ES 502C column (7.5 mm X 100 mm) from Asahi Chemical Industries Co., Ltd. The column was equilibrated with 20 mM sodium phosphate buffer, pH 6.5, and elution was performed by a linear gradient from 0.0 to 0.3 M Nazi304 in 20 mM sodium phosphate buffer, pH 6.5, over 30 min. The retention times were 22.8 min for Hybrid Enz-l,23.2 min for Hybrid Enz-2, 22.8 min for Hybrid Enz-3, and 22.4 min for Hybrid Enz-4. Assay for Hybrid Enz-1, -2, -3, and -4. The substrate for the wild-type and C135/RNase H was either a Smer RNA/9mer DNA or a 22mer RNA/9mer DNA hybrid duplex, and that of the hybrid enzymes was either a singlestranded Smer RNA or 22mer RNA. The substrate (10 pmol) was hydrolyzed with enzymes (0.02 pmol) a t 30 "C for 15 min in 10 pL of 10 mM TrisHCl buffer, pH 8.0,containing 10mM MgC12,50 mM NaCl, 1 mM 2-mercaptoethanol, and 0.01 % bovine serum albumin. The reaction was stopped by the addition of 20 pL of loading buffer (10 M urea, 50 mM EDTA), and the hydrolysates were fractionated on polyacrylamide sequencing gels (19:1acrylamide/bis(acrylamide)) with 7 M urea (0.3 mm X 40 cm). A 20% gel was used for the Smer, and a 15% gel was used for the 22mer. They were identified by comparing the degraded products with those of a 32P-5'end-labeled Smer RNA treated with snake venom phosphodiesterase (Jay et al., 1974). The amount of each hydrolysate was directly quantitated by measuring the radioactivity with a Fujix BAlOOO bioimage analyzer. Determination of the Kinetic Parameters of Hybrid Enz-1, -2, -3, and -4 with Smer RNA. The hydrolysis was carried out a t 30 "C in the same buffer solution as described. The substrate concentrations were varied from 0.2 to 20.0 pM, and the hydrolysate concentration was
