Human Immunodeficiency Virus Reverse Transcriptase-associated RNase H Activity*

Biochemical characteristics of the RNase H activity associated with immunoaffinity purified human immunodeficiency virus (HIV) reverse transcriptase (RT) were examined. Glycerol gradient centrifugation of HIV RT resulted in a single peak of RNase H, associ- ated with RT activity, with an apparent molecular weight of 110,000. HIV RNase H exhibited a marked substrate preference for poly(dC)*[SH]poly(rG) compared to p~ly(dT).[~H]poly(rA). It did not hydrolyze single-stranded RNA or the DNA component of DNA-RNA hybrids. Products of the HIV RT-associated RNase H reaction consisted primarily of monomers, dimers, and trimers with 3’ OH groups. This reaction was Mg2+ dependent, with greater than 90% of maximum activity at MgClz concentrations between 4 and 12 mM. The optimum KC1 concentration for HIV RNase H was 50 mM, which was lower than that for HIV RT catalyzed polymerization with a poly(rA)*(dT)lo template. The optimum pH for HIV RNase H activity was between 8.0 and 8.5, in contrast to an optimum pH of 7.5 to 8.0 for HIV RT activity. The association of RNase H activity with the p66 component of HIV RT was demonstrated by activity gel analysis. These re- sults indicate that HIV RT has an integral RNase H activity; however, some of its properties are different from those of RNase H associated with other retroviral RT’s, and optimal assay conditions are different than those for HIV RT catalyzed

dition, retroviral RT typically catalyzes hydrolysis of the RNA component of DNA. RNA hybrids, known as RNase H activity (2).
HIV RT is composed of two polypeptides, p66 and p51, which share a common amino terminus (4, 5). p51 is derived from the proteolytic cleavage of p66, with release of a small polypeptide, p15 (6,7). This processing is thought to result in production of a stable p66/p51 dimer, which may be the natural form of HIV RT (8). However, after separation of p66 and p51 by denaturing polyacrylamide gel electrophoresis, p66 can renature, with recovery of DNA-and RNA-dependent DNA polymerase activity, but recovery of p51 catalytic activity is very poor in comparison (3).
Amino acid sequence analysis of retroviral pol genes suggests that polymerase sequences are present within the aminoterminal portion of p66, while the carboxyl terminus is homologous with RNase H from Escherichia coli (9). The association of RNase H activity with HIV RT produced by expression of a recombinant gene in E. coli (lo), and with RT from lysed virions (6), was recently reported. RNase H and RT activity co-purify during ion exchange and DNA affinity chromatography, as well as glycerol gradient centrifugation. In addition, proteolytic processing of p66 results in production of p51 and a small polypeptide (p15) that retains RNase H activity but lacks polymerase activity (6).
There are several possible functions of RNase H in proviral DNA synthesis. These include removal of 5' genome RNA in DNA.RNA after strong stop minus-strand DNA synthesis, removal of the tRNA primer at the 5' end of minus-strand DNA, and a role in generation of the primer for plus-strand DNA synthesis (2). In view of the importance of RNase H in the generation of proviral DNA, this enzyme activity is a possible chemotherapeutic target for inhibition of HIV replication. This article examines some of the biochemical characteristics of the RNase H activity associated with immunoaffinity purified HIV RT.
HIV positive human serum and immunoaffinity purified HIV RT were kindly provided by Biotech Research Laboratories (Rockville, MD). 7.5 ml of lysed virions from HTLV-IIIb-infected H9 cell culture supernatant was applied to a 10-ml affinity column, which consisted of monoclonal antibody specific for HIV p66 and p51 covalently linked to Sepharose (5). The column was washed with 275 ml of phosphate-buffered saline, then with 30 ml of phosphate-buffered saline with 0.5 M NaCl, followed by elution with 40 ml of 0.2 M NHIOH. Fractions of 3 ml were collected during elution, and were neutralized by addition of an equal volume of 0.5 M Tris-HC1, pH 6.8. Fractions which corresponded to the peak of RT activity were pooled, and BSA was added to a final concentration of 100 pglml. Pooled enzyme was dialyzed against 50 mM Tris-HC1, pH 7.5, 2 mM DTT, 20% glycerol, and was stored at -70 "C. General properties of this enzyme preparation have been described (11 . ~2P]poly(rG)-This was prepared in a reaction identical to that described above except that the reaction volume was 50 p1 and contained 0.2 A260 units of poly(dC) and 500 p M [w3'P]GTP (10 Ci/mmol).
DNA. L3H]RNA-The complex was prepared as described above in a 50-p1 reaction which contained 10 pg of heat-denatured calf thymus DNA and 500 pM (2 Ci/mmol) of each rNTP. A portion of the purified DNA. [3H]RNA was heat denatured by incubation at 80 "C for 15 min followed by immediate cooling on ice.
PHIDNA. RNA-This was prepared with the use of MMLV RT in a reaction which contained the same buffers and salts as for E. coli RNA polymerase reactions, and 10 pg of poly(A)+ RNA (prepared from KB cells as described in Ref. 12 poly(rG) unless otherwise indicated). Reactions were initiated with 2-5 pl of enzyme and incubated at 37 "C for 30 min. Reactions were terminated by transfer to wet ice followed by addition of 50 pl of icecold 7% perchloric acid. After 30 min on ice, precipitate was pelleted by centrifugation at 10,000 rpm for 15 min. 75 pl of the supernatant was added to 10 ml of scintillation mixture. One unit is defined as the amount of enzyme which produces 1 nmol of acid-soluble ribonucleotide/h at 37 "C.

Glycerol Gradient Centrifugation
Tris-HC1, pH 7.5, 1 mM DTT, and 200 mM KC1 in a total volume of Continuous 10-30% glycerol gradients which contained 25 mM 4.8 ml were prepared. 60 units of HIV RT in a final volume of 200 pl (10% glycerol) was layered onto the top of the gradient. Identical gradients were layered with solution containing molecular weight standards (20 pg each) in buffer containing 10% glycerol. Ultracentrifugation was performed at 4 "C for 20 h a t 40,000 rpm in a Beckman SW 50.1 rotor. Fractions of approximately 200 pl each were collected from the top of the gradient and assayed for the presence of various enzyme activities as described above. Positions of molecular weight standards were identified after PAGE in the presence of sodium dodecyl sulfate (SDS) as previously described (13), followed by staining with Coomassie Blue (14). Recovery of all enzyme activities was greater than 80%.
Product Analysis 0.05 or 0.5 units of HIV RNase H were incubated with poly(dC) .
[32P]poly(rG) using standard reaction conditions for the time periods indicated in Fig. 2. Two 10-pl portions of each reaction were treated with 10 pl of 7% perchloric acid and processed for determination of acid-soluble radioactivity as described above. One 5-pl portion of each reaction was diluted with 20 pl of H20 and 12.5 p1 of sequencing gel sample buffer (50 mM EDTA and 0.02% bromphenol blue in 88% formamide). 5 pl of an 80% acid-soluble reaction was further treated for 15 min with 0.1 unit of phosphodiesterase I at 37 "C, followed by addition of 44 pl of H 2 0 and 25 p1 of sequencing gel sample buffer. for 5 min and fractionated by electrophoresis on a 15% polyacrylamide-urea DNA sequencing gel as previously described (15), followed by autoradiography with Kodak X-Omat K film.

RNase H Actiuity Gel
A 10% polyacrylamide gel which contained SDS was prepared essentially as described by Laemmli (13), except for the addition of poly(dC).
[32P]poly(rG) (IO6 cpm) prior to polymerization. Enzyme samples for activity recovery were HIV RT (12 units), MMLV RT (200 units), and E. coli RNase H (10 units), each in a final volume of 40 p1 which contained 5% glycerol, 2 mM EDTA, 1% SDS, 50 mM Tris-HCl, pH 7.5,0.02% bromphenol blue, 1 mg/ml BSA, and 0.5 mM 2-mercaptoethanol. These were incubated for 5 min at 37 "C before application to the gel. One lane was reserved for protein standards and another for HIV RT transfer to nitrocellulose. These were mixed with standard sample buffer (13) and boiled before application to the gel. After electrophoresis, the lane with molecular weight standards was removed and stained with Coomassie Blue (14). The lane with HIV RT in standard sample buffer was transferred to nitrocellulose (16), and transferred proteins were stained by incubation with human HIV positive serum, followed by biotinylated anti-human IgG, and avidin D-horseradish peroxidase conjugate, in a blocking buffer which consisted of 5% Carnation non-fat dry milk and 4% IgG-free calf serum in phosphate-buffered saline. Development was performed with a 4-chloro-1-naphthol-based reagent in the presence of HZOZ as recommended by the horseradish peroxidase substrate supplier.
The remainder of the gel, which contained enzymes for activity recovery, was shaken gently at room temperature for 1 h in 2 changes (1 liter each) of 50 mM Tris-HC1, pH 8.0, 2 mM DTT, 20% glycerol, followed by 16 h in 2 changes (1 liter each) of the same buffer plus 50 mM KC1 and 8 mM MgCL The gel was shaken for another 8 h in two changes (I liter each) of 50 mM Tris-HC1, pH 8.0, 50 mM KC1, 2 mM DTT, 8 mM MgC12, in the absence of glycerol, followed by 16 h in the same buffer at 37 "C without shaking. This was followed by gentle shaking for 4 h in 4 (1 liter) changes of cold 5% trichloroacetic acid, 10 mM pyrophosphate. The gel was dried under vacuum and autoradiography was performed with Kodak X-Omat K film.

Glycerol Gradient
Centrifugation-Immunoaffinity purified HIV RT was subjected to ultracentrifugation in a 10-30% glycerol gradient in the presence of 200 mM KC1, and fractions were analyzed for the presence of polymerase and RNase H activities (Fig. 1)    electrophoresis in a polyacrylamide/urea DNA sequencing gel (Fig. 2). Incubation with 0.05 units of R T for 20, 40, or 120 min resulted in the increasing appearance of a series of 32Plabeled oligomers, primarily monomers, dimers, and trimers. 80% of the substrate was acid-soluble at 120 min. When a 10-fold higher concentration of HIV R T was employed, 100% of the substrate was rendered acid soluble, and the products consisted almost exclusively of monomers, dimers, and trimers. Treatment of HIV RNase H reaction products with venom phosphodiesterase resulted in complete conversion to [5'-"P]GMP. Divalent Cation Requirement-The HIV RNase H reaction with poly(dC) .
[3H]poly(rG) as the substrate was absolutely dependent on the presence of divalent cation. MgC12 was preferred to MnC12 (not shown). Greater than 90% of the maximum RNase H activity was obtained a t MgClz concentrations between 4 and 12 mM (Fig. 3). This is in contrast to the MgC12 dependence of HIV R T activity measured with a poly(rA). (dT)lo template, which exhibited a comparatively sharp MgC12 optimum of 6-8 mM, and retained only 54% of maximum activity a t a concentration of 12 mM. Effect of Salt-The effect of KC1 on RNase H activity and on polymerase activity with several different templates was examined (Fig. 4). KC1 was added to various assay mixtures at 25 mM increments to 0.25 M. R T activity with a poly(rA) . RNase H activity exhibited a salt profile different from that of polymerase activity with any of these templates. Maximum activation of 2.1-fold was obtained a t 50 mM KCl, and the presence of KC1 at concentrations greater than 100 mM was strongly inhibitory.
pH Optimum-HIV RT-associated RNase H had an alkaline pH optimum of 8.0-8.5, with either Tris or Hepes buffers (Fig. 5). In contrast, polymerase activity with a poly(rA). (dT),o template had a sharp pH optimum between 7.5 and 8.0, and retained about 60% of maximum activity at pH 8.5.
RNase H Activity Gel-The RNase H substrate poly(dC) .
[32P]poly(rG) was incorporated into a polyacrylamide gel prior to polymerization. Purified HIV R T was denatured and subunits were separated during electrophoresis through this gel. One lane which contained HIV R T was transferred to nitrocellulose and immunostained with a monoclonal antibody against this enzyme (Fig. 6a). The remainder of the gel was washed extensively to allow protein renaturation, followed by incubation a t 37 "C under conditions optimal for detection of HIV RNase H activity. The gel was then dried, and autoradiography performed (Fig. 6b). RNase H activity was associated with the p66 subunit of HIV R T as demonstrated by the absence of radioactive substrate in the portion of the gel occupied by p66. A smaller band of nuclease activity which corresponded to a molecular weight of approximately 200,000 was also noted.

DISCUSSION
Immunoaffinity purified HIV R T possessed RNase H activity as well as DNA-and RNA-dependent DNA polymerase activity. All three activities co-migrated in a glycerol gradient during ultracentrifugation. We were unable to observe any small molecular weight RNase H in glycerol gradient fractions or by activity gel analysis. RNase H activity was an integral part of HIV RT, since it was associated with the p66 subunit. A smaller amount of RNase H activity with an apparent molecular weight of 200,000 was also observed, which raises the interesting possibility that the gag-pol precursor could have RNase H activity; however, we have not ruled out the possibility that the presence of high molecular weight RNase H could be due to enzyme aggregation. Others have observed concurrent production of p51 and a 15,000 RNase H during proteolytic cleavage of a carboxyl-terminal fragment from p66 (6). Several factors may contribute t o the absence of this activity in our preparation. The monoclonal antibody used for purification of HIV RT recognizes a determinant shared by p66 and p51, and would not be expected to recognize the carboxyl-terminal portion of p66. Therefore, p15 should not be present unless it forms a complex with RT, or is produced after purification of RT by the action of a contaminating protease.
Several factors indicate that HIV RT contains distinct catalytic sites for polymerase and RNase H activity. Amino acid sequence analysis demonstrates homology with viral and bacterial polymerases within the amino-terminal portion of p66, and homology with other RNase Hs near the carboxyl terminus (9). Release of an active 15,000 RNase H which lacks polymerase activity during processing of p66 also supports this idea (6). Proteolytic fragments of AMV RT which possess RNase H activity but lack polymerase activity have also been reported (18). In spite of apparently distinct active sites for RNase H and polymerase, removal of the 15,000 fragment results in a 51,000 polypeptide that has greatly reduced RT activity. We previously reported that we were unable to detect polymerase activity associated with p51 by activity gel analysis (3); however, with highly concentrated enzyme, we are now able to detect a small amount of p51associated polymerase activity. This is consistent with a recent report that p51 has 20-80-fold less RT activity than p66/ mg of protein (19).
RNase H activity catalyzed by p15 H exhibits a random mode of action, in contrast with RT-associated RNase H, which is processive (6). The change to a random mode of RNase H action after loss of the polymerase active site could be due to the loss of a nucleic acid binding domain. That this is the case for HIV p15 is supported by a site-directed muta-genesis study which indicates that the template binding site is present within the amino-terminal half of p66 (20).
Our results obtained with immunoaffinity purified HIV RT indicate several differences in optimum assay conditions for polymerase and RNase H activities. RNase H exhibits a broader MgClz optimum, lower KC1 optimum, and prefers a more alkaline pH, than polymerase. For these reasons and those discussed above, it may be possible to discover selective inhibitors of HIV RNase H.
RNase H plays an important role in proviral DNA synthesis catalyzed by RT. Studies in progress may indicate whether or not this enzyme activity will be a useful target for inhibition of HIV replication. In addition, degradation of viral genomic or messenger RNA by RNase H could be an important mechanism of action of antisense oligonucleotides, a chemotherapeutic approach which is under active investigation in several laboratories. This article provides important basic information for investigation of these possibilities.