Sequence and Structural Determinants Required for ... - NCBI - NIH

JOURNAL OF VIROLOGY, Aug. 1996, p. 5288–5296 0022-538X/96/$04.0010 Copyright q 1996, American Society for Microbiology

Vol. 70, No. 8

Sequence and Structural Determinants Required for Priming of Plus-Strand DNA Synthesis by the Human Immunodeficiency Virus Type 1 Polypurine Tract MICHAEL D. POWELL

AND

JUDITH G. LEVIN*

Laboratory of Molecular Genetics, National Institute of Child Health and Human Development, Bethesda, Maryland 20892 Received 13 February 1996/Accepted 3 May 1996

At the 3* end of all retroviral genomes there is a short, highly conserved sequence known as the polypurine tract (PPT), which serves as the primer for plus-strand DNA synthesis. We have identified the determinants for in vitro priming by the human immunodeficiency virus type 1 (HIV-1) PPT. We show that when the PPT is removed and placed into different nucleotide contexts, new priming sites are produced at the precise 3* end of the PPT. In addition, we find that a hybrid consisting of a 15- or 20-nucleotide RNA primer annealed to a 35-nucleotide DNA template is competent for initiation of plus-strand synthesis with HIV-1 reverse transcriptase. Thus, no cis-acting elements appear to be required for priming activity. Changes at the 5* end of the PPT have no effect on primer function, whereas the identity of bases at the 3* end is crucial. A primer containing only the 6 G residues from the 3* end of the wild-type PPT sequence and 9 bases of random sequence at the 5* end functions like a wild-type PPT. A short hybrid having a similar helical structure but a primary sequence different from that of the PPT is cleaved imprecisely, resulting in initiation of synthesis at multiple sites; however, total primer extension is close to the wild-type level. We conclude that helical structure as well as the presence of particular bases at the 3* end of the PPT is essential for PPT function.

synthesis and are determined by the distance between the polymerase and RNase H active sites (1, 14, 16, 18, 19, 28, 30, 33, 36, 40, 51) or whether they are uncoupled from DNA synthesis (11) during reverse transcription in vivo (for a review, see reference 4). In either case, many small RNA fragments will be produced which could theoretically serve as primers for plus-strand synthesis. However, plus-strand initiation occurs selectively at the 39 terminus of a short, highly conserved, purine-rich sequence, immediately upstream of the U3 region, which is known as the polypurine tract (PPT) (for a review, see reference 4). The PPT sequence is absolutely required for virus replication (55). There is also evidence for the existence of multiple plus-strand priming sites upstream of the PPT (for reviews, see references 2 and 4), and in the case of human immunodeficiency virus type 1 (HIV-1) (45, 49, 60) and other lentiviruses (3, 20, 54), there is a second copy of the PPT located in the integrase coding region near the center of the genome. Mutation of the HIV-1 central PPT results in decreased efficiency of replication (6, 27). Interestingly, there do not appear to be any PPT-like sequences in the U3 region of the genome (4). In vitro studies demonstrated that initiation of plus-strand synthesis by the 39 PPT involves selective cleavage of viral RNA to produce the correct 39 end for subsequent extension by RT (13, 24, 47, 48, 53, 62). The PPT primer is removed from plus-strand DNA by specific cleavages at or near the 59 terminus of the PPT and at the 39 RNA-DNA junction (5, 24, 46). Ultimately, the proper processing of the 39 PPT is of great importance since it precisely defines the left end of the linear DNA, which serves as a recognition and cleavage site for viral integrase (58). The focus of the present study was to determine what features of the 39 HIV-1 PPT are important for determining the specific cleavage and extension required for initiation of plusstrand DNA synthesis. Using a simple model system consisting

Retroviruses are plus-strand RNA viruses that replicate through a double-stranded DNA intermediate, which is ultimately integrated into the host genome (59). Synthesis of viral DNA is carried out by the virion-associated reverse transcriptase (RT) in a complex series of steps involving RNA- and DNA-dependent DNA polymerase activities and RNase H activity, which degrades the RNA moiety in an RNA-DNA hybrid (17; for reviews, see references 29 and 59). The conversion of single-stranded viral RNA into doublestranded DNA requires specific priming events for each strand of DNA produced. Minus-strand DNA synthesis is initiated by a cellular tRNA primer, which is selectively encapsidated into virions by RT sequences (31, 34, 39, 50; for a review, see reference 61) contained in the Gag-Pol precursor (34) and is annealed to an 18-nucleotide (nt) primer binding site at the 59 end of the viral RNA genome (59). As the tRNA is extended, 59 template sequences are degraded by the RNase H activity of RT (10). When RT reaches the 59 end of the template, the last 14 to 18 bases are cleaved by a 39-OH-independent (i.e., polymerase-independent) mechanism (16, 18, 40), thereby allowing nascent minus-strand strong-stop DNA to be repositioned at the 39 end of the genome for further elongation (9, 14, 38, 56). This transfer, which is known as the first template switch, is mediated by base pairing of sequences in an area of homology known as the R region, which is present as direct repeats at the 39 and 59 ends of the viral genome (23, 37; for a review, see reference 57). As minus-strand DNA synthesis proceeds, the genomic RNA template is further degraded by RNase H. It is not known whether these cleavages are coordinated with DNA

* Corresponding author. Mailing address: Laboratory of Molecular Genetics, NICHD, Building 6B, Room 216, NIH, Bethesda, MD 20892. Phone: (301) 496-1970. Fax: (301) 496-0243. Electronic mail address: [email protected]. 5288

VOL. 70, 1996

PRIMING OF HIV-1 PLUS-STRAND DNA SYNTHESIS

5289

TABLE 1. RNA oligonucleotides used in this study Sequencea

Designation

215 21 115 Viral sequenceb ............................................................................................................UU UUU AAA AGA AAA GGG GGG ACU GGA AGG GCU AAU Dwn ............................................................................................................................... ACU GGA AGG GCU AAU WT................................................................................................................................. AAA AGA AAA GGG GGG 59 .................................................................................................................................... GCG CGA AAA GGG GGG 39 .................................................................................................................................... AAA AGA AAA GGC CCC GA ................................................................................................................................. GAG AGA GAG AGA GAG CG ................................................................................................................................. AAA AGA AAA GGC GCG GC ................................................................................................................................. AAA AGA AAA GGG CGC 6G’s ............................................................................................................................... UCA UAC CAU GGG GGG a b

Sequences are shown 59339. Underlining indicates positions with sequences different from that of the WT HIV-1 sequence. HIV-1 LAI, nt 8657 to 8691 (60; numbering according to GenBank accession number K02013).

of short RNA-DNA hybrids, we find that bases at the 39 end of the PPT are essential for priming activity. The helical structure of the hybrid may play a role in primer recognition by RT, but it is not the sole determinant of activity. MATERIALS AND METHODS Materials. Restriction enzymes, Taq polymerase, T4 ligase, and T4 DNA polymerase were purchased from Boehringer Mannheim (Indianapolis, Ind.). T4 polynucleotide kinase, RNasin, and RNase ONE, an RNase which degrades RNA without base specificity, were obtained from Promega Biotec (Madison, Wis.). HIV-1 RT was purchased from Worthington Biochemical Corp. (Freehold, N.J.). Radioactively labeled nucleotides [a-32P]dATP (3,000 Ci/mmol), [g-32P]ATP (3,000 Ci/mmol), and [a-35S]dATPaS were purchased from Amersham Life Science Inc. (Arlington Heights, Ill.). RNA oligonucleotides were purchased from Oligos Etc., Inc. (Wilsonville, Oreg.). Sequence ladders were prepared with reagents from a Sequenase kit (U.S. Biochemicals Corp.) according to the manufacturer’s instructions. Preparation of minus-strand single-stranded DNA. Circular minus-strand single-stranded DNA templates (3,352 nt in length), which contain a 185-nt region complementary to the PPT and the surrounding region, were prepared from an HIV-1 LAI clone (pKP8115). This clone was derived from a plasmid containing the HIV-1 LAI 39 long terminal repeat and part of nef (gift of Keith Peden, Center for Biologics Evaluation and Research, Food and Drug Administration) and contains nt 8587 to 8772 (60, numbering according to GenBank accession number K02013) inserted between the EcoRI and XbaI sites of the pGEM 3Zf(1) vector (Promega Biotec). The PPT is located in the center of the retroviral sequence. Single-stranded DNA was produced from this phagemid by coinfection with helper phage M13K07 according to the instructions from Promega Biotec. The isolated single-stranded DNA was purified on a 1% agarose gel, and the final concentration was determined by spectrophotometry. Linear singlestranded DNA templates of the same region were used in some assays and were produced by asymmetric PCR with forward and reverse primers JL14 (nt 8587 to 8609 of LAI) and JL15 (nt 8772 to 8753 of LAI), respectively. Primer JL15 was used in 10-fold excess over primer JL14. The final product was purified by isolation from a 2% agarose gel and quantified by measuring A260. Oligomeric 35-nt minus-strand single-stranded DNA templates were synthesized with a 39-terminal sequence complementary to the 15-nt wild-type (WT) PPT or 15-nt PPT mutant sequences. The remainder of the 35-nt sequence was complementary to the 20 bases immediately downstream of the WT PPT. For example, the WT PPT template had the sequence 59-AGT GAA TTA GCC CTT CCA GTC CCC CCT TTT CTT TT-39. (The underlined region was the same regardless of whether the PPT sequence was WT or mutant.) The downstream DNA template (Dwn) had a sequence whose 39 end started with the complement of the first base downstream of the PPT. DNA oligonucleotides were synthesized by Lofstrand Laboratories (Gaithersburg, Md.) and were purified by electrophoresis on 15% sequencing gels prior to use. Preparation of plus-strand RNA. RNA containing the PPT region (185 nt in length) was produced from pKP8115 by in vitro transcription. One microgram of pKP8115 DNA, linearized at a unique XbaI site, was used as a template in a 20-ml T7 RNA transcription reaction mixture (MEGAscript; Ambion) according to the manufacturer’s recommendations. The RNA transcript was treated with DNase I, sequentially extracted with phenol and phenol-chloroform, and finally precipitated with 2 volumes of absolute ethanol. The final product was resuspended in DEPC-treated water and was quantified by spectrophotometry. All RNA transcripts were stored at 2808C prior to use. In some reactions, 15- or 20-nt RNA oligonucleotides were used in place of T7 RNA transcripts. The RNA oligonucleotides were gel purified by electrophoresis in 15% sequencing gels with diethyl pyrocarbonate (DEPC)-treated water, concentrated by precipitation with absolute ethanol, and were then dried and resuspended in DEPC-treated water. RNA oligonucleotides were quantitated by

spectrophotometry and stored at 2808C prior to use. When indicated, primers were labeled with 32P at their 59 ends as described by Guo et al. (19). In vitro synthesis of plus-strand DNA by HIV-1 RT. Three assays were used to study aspects of in vitro synthesis of plus-strand DNA. In assay I, the start site for initiation of DNA synthesis was determined by performing primer extension on the DNA products. Assays II and III directly measure the 32P-labeled DNA products formed by incorporation of [a-32P]dATP, with preformed linear RNADNA hybrids as substrates. (i) Assay I. One picomole of phagemid single-stranded DNA was annealed to 10 pmol of in vitro-transcribed RNA in 10 mM KCl (total volume, 5 ml) by heating to 908C for 5 min and then slowly cooling to room temperature. The RT reaction mixture, in a volume of 10 ml, contained RT buffer (50 mM Tris-HCl [pH 7.8], 50 mM KCl, 10 mM MgCl2, 6 mM dithiothreitol), 10 U of RNasin, 250 mM each the four deoxynucleoside triphosphates (dNTPs), and 10 pmol of HIV-1 RT. The reaction (final volume, 15 ml) was started by the addition of 5 ml of hybrid to the reaction mixture and then incubation for 40 min at 378C. (A 40-min incubation time was chosen, since control experiments showed that fulllength extension reached a sharp plateau at 40 min [data not shown].) The reaction was terminated by heating to 908C for 5 min. The solution was then adjusted to 0.3 M NaOH and incubated at 658C for 20 min to hydrolyze RNA present in the mixture. After the volume had been adjusted to 45 ml with distilled water, 0.1 volume of 3 M sodium acetate was added. The solution was extracted sequentially with 50 ml of phenol and phenol-chloroform and was then precipitated with 2 volumes of absolute ethanol. The precipitate was washed once with 70% ethanol and dried. To locate the exact position at which plus-strand DNA synthesis was initiated, DNA products were analyzed by primer extension. A 59 32P-labeled primer, JL76 (nt 8746 to 8727 of LAI), which is complementary to sequences at the 39 end of the reverse transcript, was annealed to the reaction products and extended with T4 DNA polymerase. The extension reaction was carried out in a total volume of 10 ml containing buffer supplied by the manufacturer, 250 mM each of the four dNTPs, and 2 U of T4 DNA polymerase and was incubated at 378C for 30 min. The reaction was terminated by the addition of 4 ml of formamide loading buffer (Sequenase kit). Two microliters of each reaction mixture was subjected to electrophoresis in a 6% sequencing gel. A sequencing ladder was generated with primer JL76 (see above) and the appropriate pGEM plasmid DNA. (ii) Assay II. Reactions were performed with linear minus-strand singlestranded DNA (185 nt in length) annealed to a complementary single-stranded RNA. The reaction conditions used were the same as those described above for assay I, except that the products were internally labeled by the addition of 3 pmol of [a-32P]dATP and unlabeled dATP at a final concentration of 50 mM. After incubation at 378C for 40 min, the reactions were terminated by the addition of 4 ml of formamide STOP solution, and 2 ml of each reaction mixture was directly analyzed on a 6% sequencing gel without prior treatment with NaOH. The products were compared with a sequencing ladder generated with primer JL15 (see above). (iii) Assay III. Ten picomoles of either 15- or 20-nt RNA oligonucleotides was annealed to 1 pmol of 35-nt single-stranded DNA templates. The reaction conditions were the same as those described above for substrates containing a long linear RNA-DNA hybrid (assay II), except that incubation was at 378C for 15 min. Two microliters of each reaction mixture was directly analyzed on an 8% sequencing gel. To determine whether any RNA was attached to the plus-strand DNA products, a 5-ml portion of the reaction mixture was terminated with 2 ml of STOP solution and set aside. The remaining 10 ml was heated to 908C for 5 min to inactivate the RNasin present, then treated with 10 U of RNase ONE in the same RT buffer, and incubated for an additional 30 min. The reaction was terminated by the addition of 4 ml of formamide STOP solution. All samples were heated to 908C for 5 min before electrophoresis. Construction of a mutation which changes the five U’s immediately upstream of the PPT. A mutation in which the five U’s immediately upstream of the PPT (Table 1) were changed to an AvaI restriction site (UUUUU3CCGAG) was constructed by first digesting pKP8115 with BglII and EcoRV restriction en-

5290

POWELL AND LEVIN

J. VIROL.

zymes; this removes a 67-nt fragment containing the PPT. This region was then replaced by a pair of complementary 67-nt DNA oligonucleotides containing the PPT and the new 59 upstream sequence. The resulting clones were screened for the presence of the introduced unique AvaI site. The identities of positive clones were confirmed by dideoxy sequencing with [a-35S]dATPaS with the Sequenase kit. Construction of context mutations. pKP8115 was modified to move the PPT into two different upstream nucleotide contexts (see Fig. 2). First, the PPT and surrounding context were removed from the clone by digestion with BglII and EcoRV restriction enzymes. The BglII site was then blunt ended by filling in with T4 DNA polymerase. Subsequent ligation with T4 ligase restored the BglII site. The resulting clone (pMP9447) contained two unique restriction sites, Asp718 at nt 8608 and the restored BglII site at nt 8644, but lacked the PPT and surrounding context. The pMP9447 plasmid was then cut with either Asp718 or BglII, and the resulting ends were blunted with T4 DNA polymerase. The PPT sequence was formed by annealing two complementary DNA oligonucleotides containing either the 15-base PPT sequence or the PPT sequence plus the immediate downstream 5 bases (nt 11 to 15). After annealing, the DNA oligonucleotides were ligated into pMP9447 at either the Asp718 or the BglII site. The resulting clones contained the PPT sequence in two new nucleotide contexts. The sequences were confirmed by dideoxy sequencing (see above). Circular dichroism (CD) spectroscopy on RNA-DNA hybrids. Equimolar mixtures (;1.0 A260) of a 15-nt RNA oligonucleotide and a 15-nt complementary DNA oligonucleotide were annealed by heating to 1008C for 5 min and then by slow cooling to room temperature (approximately 60 min). Monitoring of the hypochromic shift at 260 nm was used to confirm that the RNA and DNA oligonucleotides were present in equimolar ratios. Data were collected every 0.2 nm from 350 to 200 nm on a Jasco 700 spectropolarimeter. In each case, the plotted spectra represent an average of two scans at 208C.

RESULTS Mutation of the five U’s immediately upstream of the PPT. In HIV-1, there is a highly conserved stretch of five U’s immediately 59 of the PPT (35). To determine if this conserved sequence has any functional significance in initiation of plusstrand DNA synthesis, a mutant in which the five U’s (Table 1) were changed to an AvaI restriction site (CCGAG) was constructed. In an RT assay (see assay II above) with 185-nt linear RNA-DNA hybrids containing the WT sequence or the mutation of five U’s, no difference in either the amounts or the lengths of plus-strand DNA products made could be detected (Fig. 1). Comparison of the products with a sequencing ladder showed that initiation occurred at the 11 A residue, immediately following the 21 G residue of the PPT (the residues within the PPT and upstream of the PPT are numbered by negative integers [39359], and residues downstream of the PPT are numbered by positive integers [59339]; Table 1). These results demonstrate that plus-strand priming is not dependent on the presence of the five U’s flanking the 59 end of the PPT. In parallel experiments with a clone (pMP9447) from which 67 nt including the PPT and the surrounding context were removed (see Materials and Methods; Fig. 2), no sites of initiation were observed (data not shown). This finding suggests that none of the other RNA fragments generated by RNase H cleavage of the preformed RNA-DNA hybrid (see schematic diagram at the bottom of Fig. 1) could be used as primers by HIV-1 RT. Effect of context on plus-strand initiation. To further explore the question of whether the surrounding nucleotide context is important for plus-strand initiation, pMP9447 was used for construction of clones in which the PPT, with or without nt 11 to 15, was moved into two new contexts (Fig. 2). One context, at a unique Asp718 restriction site (Asp and Asp15), is 53 nt upstream of the 59 terminus of the PPT; a second context, at a unique BglII restriction site (Bgl and Bgl15), is 18 nt upstream of this locus. Hybrids consisting of runoff RNA transcripts (185 nt) annealed to complementary phagemid single-stranded DNA were used as substrates to determine the start site for initiation of plus-strand synthesis (see assay I above; schematic diagram at the bottom of Fig. 3). In each instance (except that of Asp [see below]), synthesis

FIG. 1. Effect of mutating the conserved five U’s upstream of the PPT on initiation of plus-strand DNA synthesis. The five U’s immediately upstream of the PPT were changed to an AvaI restriction site (CCGAG), as described in Materials and Methods. A 185-nt linear RNA-DNA hybrid was used to synthesize the plus-strand DNA product, which was internally labeled with [a-32P]dATP (see Materials and Methods [assay II]). Lane 1, WT upstream sequence; lane 2, mutant sequence. A schematic diagram of the assay is shown in the lower portion of the figure. Curved lines, RNA fragments generated by nonspecific cleavage of RNA in the hybrid by RNase H; small vertical arrow, the site of specific cleavage at the 39 end of the PPT and primer extension; small dashed lines, the newly synthesized plus-strand DNA; ssRNA and ssDNA, singlestranded RNA and DNA, respectively.

of the plus-strand DNA product was initiated primarily at the 11 A position (Fig. 3), indicating that cleavage occurred at the site normally used in the WT situation, i.e., between the 21 G and 11 A residues (4). Unlike the other clones, the Asp construct has a C residue instead of an A at the 11 position (Fig. 2). However, under the conditions of this experiment, cleavage appeared to be unaffected by substitution of the 11 C. Thus, HIV-1 RT was able to initiate plus-strand synthesis at the 11 residues in both the Asp and the Asp15 substrates (Fig. 3). These results are consistent with the finding by Pullen et al. (42) that the major cleavage site for an HIV PPT substrate with a C at the 11 position is between positions 21 and 11, although some additional cleavages can also be detected. Taken together, these results demonstrate that (i) the PPT sequence alone is sufficient to initiate plus-strand DNA synthesis and (ii) this reaction is not significantly influenced by interactions with the immediate or more distant 59 and 39 nucleotide contexts. Selection of the PPT as the primer for plus-strand DNA synthesis. From the results of the context experiments, it appears that the determinants of plus-strand priming reside within the 15 nt of the PPT. To analyze the contribution of elements within the PPT to priming activity, we developed a simple assay procedure utilizing short RNA-DNA hybrids containing a 15-nt RNA oligonucleotide annealed to a complementary 35-nt single-stranded DNA template (see assay III above). If the 15-nt PPT primer remains covalently attached to

VOL. 70, 1996

FIG. 2. Schematic representation of the context mutants. Portions of the nucleotide sequence upstream and downstream of the PPT in each new context are shown. The PPT sequence is shaded. The lower portion of the figure shows the relative positions of the new PPT loci at Asp718 and BglII sites in clones which lack the region between the EcoRV and BglII sites (see Materials and Methods). In each case, the PPT itself (clones termed Asp or Bgl, as appropriate) or the PPT plus nt 11 to 15 (indicated by a line extended from the box enclosing the PPT and the clones termed Asp15 or Bgl15, as appropriate) was transferred to the new locations.

the extension product, a 35-nt DNA will be detected. Removal of the RNA primer at the RNA-DNA junction yields a 20-nt DNA (schematic diagram in Fig. 4C). The integrity of the primers was monitored in control reactions with T4 DNA polymerase, which was found to extend all of the RNA sequences (Fig. 4A, even-numbered lanes). A 15-nt DNA version of the PPT used as an additional positive control was a very efficient primer and produced roughly the


5291

same amount of 35-nt DNA with both HIV-1 RT (lane 17) and T4 DNA polymerase (lane 18). With the WT PPT oligonucleotide (Table 1) and HIV-1 RT, the predominant product was a 20-nt band, indicating that most of the 15-nt PPT sequence had been removed (Fig. 4A, lane 1). In contrast, with T4 DNA polymerase (lane 2), which does not have RNase H activity under the conditions of the assay (data not shown), all of the radioactivity was in the 35-nt product. However, when RT was added in trans after synthesis of the 35-nt DNA by T4 DNA polymerase, the 20-nt DNA could be seen in the gel (data not shown). These results suggest that removal of the 15-nt PPT is catalyzed by the RNase H activity associated with RT. Treatment of the WT 20-nt product with RNase ONE did not alter the size of the 20-nt DNA (Fig. 4B; compare lanes 1 and 2). This indicates that the 20-nt product consisted solely of DNA and was not simply a premature termination product with intact RNA primer. Similarly, RNase ONE treatment of the 35-nt DNA formed by T4 DNA polymerase generated a 20-nt DNA, presumably after digestion of the 15-nt RNA primer (Fig. 4B; compare lanes 17 and 18). Other studies (15, 24, 32, 43) have shown that only the PPT or closely related PPT-like sequences (46) can function as a primer for plus-strand synthesis. To test the specificity of our assay, we investigated the activity of an RNA-DNA hybrid containing the sequence from nt 11 to 115 (Dwn [Table 1]). The results showed that very little 35-nt product was formed and no 20-nt DNA was detected (Fig. 4A, lane 3). These data are in accord with findings by Randolph and Champoux (43) with a Moloney murine leukemia virus (MuLV) system and show that the present assay reflects the selective recognition of the PPT by HIV-1 RT. Effect of helical structure on recognition of the PPT. The exclusive purine-base composition and the unique run of six G’s in the PPT sequence raised the possibility that the PPT has an unusual structure which might be important for its selection as a primer and could account for the observation that RT is

FIG. 3. Initiation of plus-strand DNA synthesis with constructs containing the PPT at loci upstream of its normal position in the genome. Assay conditions are as described in Materials and Methods (assay I), and a schematic diagram is shown at the bottom of the figure. Note that after extension with RT, the products were digested with 0.5 N NaOH to remove any RNA primer. The 59 boundary of plus-strand DNA (dashed line) was determined by primer extension with T4 DNA polymerase with a 59 32P-end-labeled primer, as described in Materials and Methods. Sequencing ladders are shown for each construct. For a description of the Asp, Asp15, Bgl, and Bgl15 constructs, see the legend to Fig. 2 and text. ssRNA, single-stranded RNA.

5292

POWELL AND LEVIN

J. VIROL.

FIG. 5. CD spectra of 15-nt RNA-DNA hybrids containing the WT PPT (F), Dwn (■), or GA (h) sequences. The data are plotted as εL-εR in units of mol liter21 cm21 versus wavelength. The GA sequence (Table 1) is a reference for R-Pu/D-Py hybrids that contain all purines in the RNA strand (25). Note the negative band at 210 nm, which is characteristic of this class of hybrids (25, 44). The CD spectrum generated by the PPT sequence also has this feature, while the CD spectrum obtained with the Dwn hybrid lacks a prominent 210-nm band.

FIG. 4. Primer activity of 15-nt WT PPT and mutant RNA oligonucleotides hybridized to 35-nt minus-strand DNA templates. A description of the 35-nt template DNAs is given in Materials and Methods. The designations and sequences of the RNA oligonucleotides are listed in Table 1. Plus-strand DNA products were internally labeled with [a-32P]dATP, as described in Materials and Methods (assay III). (A) The 15-nt RNAs were assayed with HIV-1 RT (oddnumbered lanes) or T4 DNA polymerase (even-numbered lanes). The conditions for the control T4 DNA polymerase reactions are described in Materials and Methods. In lanes 17 and 18, a 15-nt DNA primer with the PPT sequence was used as a control. (B) The products of each HIV-1 RT reaction (lanes 1 to 16) were treated with RNase ONE (even-numbered lanes) or were not treated (odd-numbered lanes) to determine if any RNA bases remained attached to the plus-strand DNA product (see Materials and Methods). The identity of the residual RNase One-stable 35-nt product is not known, but it is unlikely to be undigested RNA. A small amount of this product was also seen when the 35-nt DNA synthesized in a T4 DNA polymerase control reaction (CON) (lane 17) was treated with RNase One (lane 18). (C) Schematic diagram of the assay. Note that after synthesis of plus-strand DNA (dashed line), if the RNA primer remains attached to the DNA, a 35-nt product will result; removal of the primer will produce a 20-nt DNA.

unable to extend the Dwn oligonucleotide. To investigate a possible role for helical structure in PPT recognition, we used CD spectroscopy (Fig. 5) to study the solution structure of a 15-nt RNA-DNA hybrid of the PPT. Hung et al. (25) and Ratmeyer et al. (44) have shown that RNA-DNA hybrids having only purines (Pu) in the RNA strand and only pyrimidines (Py) in the DNA strand (R-Pu/D-Py) belong to a different structural class than hybrids with the complementary composition, i.e., R-Py/D-Pu. The CD spectra of PPT-containing hybrids are consistent with the spectra generated by hybrids in the R-Pu/D-Py structural class (Fig. 5 and data not shown). Note the presence of a negative band at 210 nm, which is a characteristic feature of these spectra (Fig. 5). In contrast, a prominent 210-nm band is missing from the CD spectrum of the Dwn RNA-DNA hybrid (Fig. 5), which contains a non-

priming sequence (Fig. 4A, lane 3). These results suggest that helical structure is important for primer activity. To investigate whether helical structure alone determines whether a given sequence can serve as a primer, we selected a 15-nt sequence (alternating GA [Table 1]) from the R-Pu/D-Py class (25) and tested for primer activity with HIV-1 RT. As anticipated, the CD spectrum of a hybrid containing the GA sequence was very similar to that of a PPT-containing hybrid (Fig. 5). The results of the primer extension assay (assay III) showed that the GA sequence could be extended with an efficiency of 80% compared with that of the WT PPT, as determined by densitometric analysis of the extension products produced with HIV-1 RT (Fig. 4A [compare lanes 1 and 9] and B [compare lanes 1 and 7]). However, unlike the WT PPT (Fig. 4A [lane 1] and B [lane 1]), extension of the GA oligonucleotide was initiated from several sites and products 20 to 23 nt in length were generated (Fig. 4A [lane 9] and B [lane 7]). Treatment of the reaction products with RNase ONE eliminated the 23-nt product and resulted in an increase in the amount of 22-nt DNA (Fig. 4B; compare lanes 7 and 8). Taken together, these observations indicate the following. (i) The 15-nt GA primer was cleaved first between positions 21 and 11, 21 and 22, and 22 and 23 and then each fragment was subsequently extended by RT to produce 20-, 21-, and 22-nt DNA products, respectively. (ii) Some of the 22-nt DNA was left with a single RNA base attached, reflecting incomplete primer removal. Cleavage within the WT PPT sequence has also been shown to occur to some extent in other in vitro priming systems (15, 42). Effect of sequence on recognition of the PPT. The finding that the GA sequence could be extended but was not a substrate for precise initiation of plus-strand DNA synthesis indicates that some factor(s) other than helical structure must also play a role in PPT function. To investigate the contribution of nucleotide sequence within the PPT to its activity, we tested a number of different 15-nt mutant PPT sequences (Fig. 4A; Table 1). With a change in the 4 bases at the 59 end of the PPT (nt 212 to 215; designated 59 [Table 1]), the total amount of DNA made was equivalent to that synthesized by WT RT (Fig. 4A; compare lanes 1 and 5); the major DNA product was 20 nt in length. This shows that the primer extension and primer

VOL. 70, 1996

removal activities of HIV-1 RT with the 59 mutant are comparable to the corresponding activities with the WT PPT. In contrast, mutation of the G’s at positions 21 to 24 (designated 39 [Table 1]) caused a striking reduction in primer extension and specific cleavage was not detected (Fig. 4A, lane 7). Mutants with less severe changes at the 39 end of the PPT, e.g., with changes of the G’s to C’s at positions 22 and 24 (CG in Table 1; Fig. 4A, lane 11) or at positions 21 and 23 (GC in Table 1; Fig. 4A, lane 13), could not be extended as efficiently as the WT and, in contrast to the situation with the WT PPT, synthesis was initiated at more than one site (Fig. 4A; compare lanes 11 and 13 with lane 1). The products formed from extension of the 59, CG, and GC oligonucleotides were unchanged by treatment with RNase ONE (Fig. 4B, lanes 4, 10, and 12, respectively). This shows that the products consist entirely of DNA. In the case of the CG and GC oligonucleotides, the results imply that cleavage of the RNA oligonucleotide between positions 21 and 22 must have preceded extension to produce the additional 21-nt product seen in these reactions (Fig. 4A [lanes 11 and 13] and 4B [lanes 9 to 12]). These results demonstrate that bases at the 39 end of the PPT are more important for priming activity than bases at the 59 end. This conclusion is further supported by the unexpected finding that a mutant PPT retaining the six G’s at the 39 end and having a random sequence in nt 27 to 215 (Table 1) was extended and processed in a manner almost indistinguishable from that of the WT PPT (Fig. 4A; compare lane 15 with lane 1). The DNA products in this case were also stable to RNase ONE treatment (Fig. 4B, lanes 13 and 14). Cleavage of 20-nt RNA oligonucleotides. In the assay described above, the enzyme is presented with a 15-nt RNA primer that already possesses the 39 end normally generated by RNase H cleavage between the last G (21) of the PPT and the downstream A (11). To determine whether the WT PPT and mutant RNA oligonucleotides could be properly cleaved at this site, we used the same procedure (assay III) to test the activity of 20-nt RNAs, containing an additional 5 nt (nt 11 to 15) downstream of the WT PPT (Fig. 6 [see schematic diagram in Fig. 6C]). The DNA product synthesized with a 15-nt DNA PPT primer (Fig. 6A, lane 9) serves as a marker for 35-nt DNA. Extension of the 20-nt WT PPT-containing oligonucleotide gave rise to the same product as the 15-nt WT PPT (Fig. 6A; compare lanes 1 and 2). In each case, the major product was a 20-nt DNA. When the reactions were carried out with unlabeled dNTPs and the 15- or 20-nt WT primers labeled with 32 P at their 59 ends, the RNA was recovered as a 15-nt species (compare products in lanes 11 and 12 of Fig. 6B with 15- and 20-nt markers in lanes 10 and 13, respectively). These results indicate that cleavage of the extra 5 bases from the 20-nt RNA must occur first, which is followed by extension and primer removal to produce a 20-nt DNA and a 15-nt RNA (see schematic diagram in Fig. 6C). In addition, the data show that the base-paired region of an HIV-1 20-nt PPT-containing RNA– 35-nt DNA hybrid is sufficient to allow specific cleavage at the junction between the 21 and 11 positions of the RNA. Assays of other 20-nt RNA oligonucleotides also gave results consistent with the data shown in Fig. 4 for the 15-nt RNA primers. An RNA with the 39 mutation (Fig. 6A, lane 4; Table 1) was poorly extended, and the corresponding hybrid was not a substrate for specific cleavage. In contrast, 20-nt RNAs with the 59 and 6Gs mutations (Table 1) gave essentially the same results as those seen with the WT 20-nt primer (Fig. 6A; compare lane 2 with lanes 3 and 8). These findings demonstrate that proper cleavage and processing of the PPT can occur even with a primer whose only resemblance to the au-


5293

FIG. 6. Primer activity of 20-nt RNA oligonucleotides and fate of 59 32Plabeled WT RNA primers. (A) The 20-nt RNAs, containing WT or mutant PPT sequences plus nt 11 to 15, were hybridized to 35-nt minus-strand DNA templates and were assayed for primer extension. The assay conditions were the same as those described in the legend to Fig. 4 (assay III). Products analyzed in lanes 1 to 9 were internally labeled with [a-32P]dATP. The RNA oligonucleotides used (Table 1) were as follows: 15-nt WT PPT (lane 1), 20-nt WT PPT (lane 2), 20-nt 59 (lane 3), 20-nt 39 (lane 4), 20-nt GA (lane 5), 20-nt CG (lane 6), 20-nt GC (lane 7), 20-nt 6G’s (lane 8), and 15-nt DNA primer with PPT sequence (positive control). (B) Parallel reactions were carried out on the WT template with 59 32P-labeled 15-nt (lane 11) or 20-nt (lane 12) WT RNA oligonucleotide. In these reactions, all four dNTPs were unlabeled. The positions to which the unreacted labeled 15- and 20-nt oligonucleotides migrate are illustrated in lanes 10 and 13, respectively. (C) Schematic diagram of the assay. Note that in step two, a specific cleavage precedes extension to produce a 20-nt DNA product.

thentic PPT primer is the presence of the six G’s at positions 21 to 26. The products from the 20-nt oligonucleotide versions of the GA, CG, and GC mutations were the same as those seen in the experiments utilizing the 15-nt RNA primers (compare Fig. 6A, lanes 5, 6, and 7, respectively, with Fig. 4A, lanes 9, 11, and 13, respectively). This suggests that in each case, cleavage could occur within the PPT to produce products 1 or 2 bases longer than that seen with the WT PPT. A summary of the results of multiple experiments, including the data from Fig. 4 and 6, is presented in Fig. 7. The results demonstrate that the 59 end of the PPT is not required for proper extension, cleavage, or primer removal in an in vitro assay for PPT function. In contrast, the six G’s at the 39 end of the PPT are essential. DISCUSSION The proper priming of plus-strand DNA is crucial for retrovirus replication. Since the priming event also serves to define one end of the linear double-stranded DNA intermediate prior to integration (58), precise selection and extension of the correct plus-strand primer are essential. In the present work, we have investigated the determinants of plus-strand priming by the HIV-1 39 PPT and HIV-1 RT in an in vitro system. We have addressed several questions. (i) Is the context surrounding the PPT required for initiation of plus-strand DNA synthesis? (ii) What is the minimum size of a PPT-containing RNA-DNA hybrid which can serve as primer template? (iii) Does the helical structure of the PPT play a role in its selection as a primer? (iv) Which bases within the PPT are required for priming activity? Our results demonstrate that the presence of the six G’s at the 39 end of the PPT represents a critical element in determining whether the PPT can function as a

5294

POWELL AND LEVIN

FIG. 7. Summary of results obtained with WT and mutant RNA oligonucleotides. In each case, reactions with the 15 and 20-nt oligonucleotides gave identical results. Underlined bases, changes in the WT PPT sequence; vertical arrows, locations of cleavage sites; sizes of the arrows, approximate amounts of each cleavage product relative to the amount of 20-nt DNA produced in the reaction with the WT PPT; horizontal arrows, the total amounts of primer extension obtained with each RNA (the larger the arrow is, the greater the total amount of synthesis). Horizontal arrows with an X through them represent total synthesis of less than 5% relative to the amount of plus-strand DNA made with the WT PPT. The estimates of relative activity are based on densitometry measurements and reflect the results of multiple experiments. The shaded box around the 23 G in the GA oligonucleotide represents an imprecise removal of the RNA primer which results in a base of RNA remaining on the DNA product (see Results).

primer in vitro; in contrast, the identities of the bases at the 59 end do not appear to be important. The HIV-1 PPT is a relatively short sequence, consisting of only 15 nt (45, 49, 60). In view of sequence conservation in the PPT region (47), it seemed possible that upstream and/or downstream nucleotides in the viral RNA also contribute to PPT activity. However, when we inserted the 15-nt PPT sequence (either alone or with nt 11 to 15) into new contexts within the viral genome (Fig. 2), we found that in each case, specific cleavage and extension at the PPT were not affected by transfer to the new locus (Fig. 3). Mutation of a conserved run of five U’s flanking the 59 end of the PPT to a random sequence also had no effect on initiation of plus-strand synthesis (Fig. 1). In addition, Huber and Richardson (24) showed that deletion of one of the five U’s does not affect cleavage at the 39 end of the PPT. Taken together, these data demonstrate that the 15-nt PPT alone contains sufficient information to specify the origin of plus-strand DNA synthesis, without any additional requirement for interactions with cis-acting elements. A similar conclusion was reached in studies showing that both the MuLV and HIV-1 RTs can correctly prime plus-strand synthesis with either the HIV-1 or MuLV WT PPTs, despite sequence divergence in bases outside a 20-nt PPT region (41, 42). In vivo studies indicate that plus-strand priming from the central PPT is also independent of the surrounding context. For example, Charneau et al. (6) showed that transfer of the HIV-1 central PPT (together with 14 nt of upstream sequence) to a new site approximately 1 kb downstream of the normal locus created a new plus-strand priming site. The fact that neither the immediate nor the more distant context is needed for plus-strand initiation has made it possible

J. VIROL.

to develop simple assay systems to assess primer activity. In the case of MuLV, which has a 13-nt PPT (22, 52), Randolph and Champoux (43) were able to reproduce steps involved in plusstrand initiation with a 13-, 16-, or 20-nt PPT-containing RNA oligonucleotide annealed to a complementary 190-nt circular DNA. Using the procedures described for assay II (see Materials and Methods), we have made a similar observation with 15-nt HIV-1 PPT-containing RNA oligonucleotides (data not shown). In addition, we have further reduced the sequence requirements for primer recognition, specific cleavage, extension, and primer removal to a 20-nt RNA annealed to a 35-nt DNA template (assay III; Fig. 6). A number of studies have demonstrated that plus-strand DNA synthesis is initiated solely by the PPT or sequences closely related to the PPT (15, 24, 32, 43, 46). In accord with findings with an in vitro MuLV system (43), we observed that an HIV-1 RNA oligonucleotide with downstream nucleotides (Dwn [Table 1]) is not extended by HIV-1 RT (Fig. 4). These results indicate that there must be some mechanism by which RT selects the primer for plus-strand synthesis from among the many RNA fragments generated during minus-strand DNA elongation. One possibility that we considered is that the PPT has an unusual helical structure which is recognized by RT. The helical structure of RNA-DNA hybrids is intermediate between the A form (e.g., as in RNA-RNA duplexes) and B form (e.g., as in DNA-DNA duplexes) (12). Thus, the RNA strand in the hybrid is in an A-like conformation (C39-endo sugars), while the DNA strand sugars have an unexpected intermediate O49-endo conformation; in addition, the minor groove of the hybrid is narrower than that of the A form but wider than that of the B form (12). A number of factors, including nucleotide sequence, can influence the structures of RNA-DNA hybrids. For example, hybrids which have only purines in the RNA strand (R-Pu/D-Py) have similar helical structures and constitute a distinct conformational class (25, 44). On the basis of the results of CD spectroscopy, we showed that the structure of a 15-nt PPT hybrid is consistent with that of other hybrids in the R-Pu/D-Py class (Fig. 5) (25, 44). We also observed that a hybrid designed to mimic the structure of the PPT (Fig. 5) but having a different nucleotide sequence, i.e., an alternating GA sequence (Table 1), is able to function as a fairly efficient primer with HIV-1 RT (Fig. 4). In contrast, the nonpriming Dwn hybrid is in a structural class different from that of the GA and PPT oligonucleotides (Fig. 5), even though the Dwn sequence is relatively purine rich (10 of 15 bases) (Table 1). This suggests that helical structure plays some role in primer recognition. Interestingly, the GA sequence used in the present study is very similar to the PPT sequence present in the Ty3 retrotransposon (59-GAGAGAGAGGAA GA [21]). Analysis of the 6G’s mutant shows that it is possible to disrupt a significant portion of the PPT sequence and yet retain normal PPT activity (Fig. 4 and 6). This finding suggests that any structural determinants for PPT activity may be confined to the 39 end of the PPT sequence. It is known from the crystal structure of HIV-1 RT complexed with double-stranded DNA that the DNA primer template undergoes an A- to B-form transition in the helical structure after a bend of 40 to 458 (28). The DNA present in A form is near the primer terminus, and the bend is distributed over approximately 4 nt from the 39 end of the primer in the vicinity of helix aH in the p66 thumb (28). Since the 39 bases are the most critical for PPT function (Fig. 7), it is possible that proper recognition of the PPT hybrid requires a specific helical structure in the region encompassing the run of G’s at positions 21 to 26. While it appears that helical structure is important for PPT

VOL. 70, 1996

priming, the results of assays with mutant PPTs (Fig. 4 and 6; Table 1) indicated that sequence determinants also have a major role in correct initiation of plus-strand synthesis. Earlier mutational analyses of the MuLV and HIV-1 PPTs showed that single- or double-base changes as well as small insertions or deletions within the PPT did not abolish plus-strand priming (42, 47). However, certain mutations (at nt 27 [MuLV] and at nt 22 or 24 [HIV-1]) were found to significantly reduce the precision of cleavage at the 21/11 junction (42, 47), presumably because of improper positioning of the RNase H domain. In our study, we made large mutations in bases at the 59 and 39 ends of the HIV-1 PPT. We found that changes at the 59 end (Table 1), e.g., a change in 4 bases (nt 212 to 215) or a more drastic change (nt 27 to 215), generated primers whose activities were virtually indistinguishable from those of the WT (Fig. 4, 6, and 7). Double-base changes at nt 21 and 23 (GC [Table 1]) or at nt 22 and 24 (CG [Table 1]) reduced priming efficiency and precision of initiation of plus-strand DNA synthesis (Fig. 4, 6, and 7) but did not totally eliminate priming activity, in agreement with the results described by Pullen et al. (42). In contrast, changing the four G’s at the 39 end of the PPT (nt 21 to 24) to four C’s (39 [Table 1]) resulted in a dramatic loss in priming activity (Fig. 4, 6, and 7), suggesting that the 39 mutant sequence is very poorly recognized by RT. It is interesting to note that within the sequence of the LAI clone of HIV-1 (60), there are three stretches of six G’s in addition to those present in the 39 and central PPTs. One stretch is only 3 bases immediately downstream of the central PPT; the other two are located toward the 59 end of genomic RNA. Our in vitro studies with the 6G’s mutant (Fig. 4, 6, and 7) suggest, that these G-containing sequences have the potential to provide additional sites of plus-strand initiation. However, whether such sites would be functional has not yet been resolved by in vivo studies of HIV-1 replication which address the question of alternate initiation sites. Thus, Miller et al. (34a) reported that unintegrated, linear HIV-1 DNA contains a large number of discontinuous plus strands of heterogeneous size which can be integrated into target DNA in vitro. This observation led to the proposal that multiple initiation sites are used for synthesis of plus-strand DNA and that gap repair may occur after integration (34a; see also reference 2). In contrast, others have shown that initiation of plus-strand DNA synthesis from the central PPT produces a discrete plus-strand discontinuity in unintegrated DNA (7, 26), which ultimately results in a linear molecule with a defined central plus-strand overlap (8). This implies that HIV-1 has only one additional site of plus-strand synthesis (7, 8). In a recent study utilizing spleen necrosis virus vectors, Bowman et al. (2a) reported that plusstrand DNAs initiated at sites other than the 39 PPT are rarely used for strand transfer (second jump), and it was concluded that such initiation events would, therefore, result in nonproductive infection. In summary, our data demonstrate that interaction of HIV-1 RT with the 39 end of the PPT is critical for proper function. These results also indicate that recognition of the PPT by HIV-1 RT depends on both structural and sequence determinants. Since the 6G’s mutant appears to be functionally equivalent to the WT PPT in vitro, it seems likely that any requirements for specific structure and/or sequence are restricted to the (26 to 21) bases of the PPT. In view of the fact that the entire PPT sequence is highly conserved in all retroviruses, the question remains as to what role the 59 bases play in synthesis of plus-strand DNA or, more generally, in virus replication. Determination of the requirements for PPT priming in vivo may help to answer this intriguing question.


5295

ACKNOWLEDGMENTS We thank Keith Peden for a plasmid containing sequences from the 39 end of HIV-1 LAI, Klara Post for construction of the HIV-1 LAI subclone (pKP8115) used in this study, Jane Sayer and Donald Jerina for generously allowing us to use their Jasco 700 spectropolarimeter for determination of CD spectra, and Henri Buc for valuable discussion. We are also grateful to James Champoux, Gloria Fuentes, and Robert Bambara for communicating results prior to publication. This work was supported in part by the National Institutes of Health Intramural AIDS Targeted Antiviral Program. REFERENCES 1. Ben-Artzi, H., E. Zeelon, B. Amit, A. Wortzel, M. Gorecki, and A. Panet. 1993. RNase H activity of reverse transcriptases on substrates derived from the 59 end of retroviral genome. J. Biol. Chem. 268:16465–16471. 2. Boone, L. R., and A. M. Skalka. 1993. Strand displacement synthesis by reverse transcriptase, p. 119–133. In A. M. Skalka and S. P. Goff (ed.), Reverse transcriptase. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 2a.Bowman, E. H., V. K. Pathak, and W.-S. Hu. 1996. Efficient initiation and strand transfer of polypurine tract-primed plus-strand DNA prevent strand transfer of internally initiated plus-strand DNAs. J. Virol. 70:1687–1694. 3. Chakrabarti, L., M. Guyader, M. Alizon, M. D. Daniel, R. C. Desrosiers, P. Tiollais, and P. Sonigo. 1987. Sequence of simian immunodeficiency virus from macaque and its relationship to other human and simian retroviruses. Nature (London) 328:543–547. 4. Champoux, J. J. 1993. Roles of ribonuclease H in reverse transcription, p. 103–117. In A. M. Skalka and S. P. Goff (ed.), Reverse transcriptase. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 5. Champoux, J. J., E. Gilboa, and D. Baltimore. 1984. Mechanism of RNA primer removal by the RNase H activity of avian myeloblastosis virus reverse transcriptase. J. Virol. 49:686–691. 6. Charneau, P., M. Alizon, and F. Clavel. 1992. A second origin of DNA plus-strand synthesis is required for optimal human immunodeficiency virus replication. J. Virol. 66:2814–2820. 7. Charneau, P., and F. Clavel. 1991. A single-stranded gap in human immunodeficiency virus unintegrated linear DNA defined by a central copy of the polypurine tract. J. Virol. 65:2415–2421. 8. Charneau, P., G. Mirambeau, P. Roux, S. Paulous, H. Buc, and F. Clavel. 1994. HIV-1 reverse transcription. A termination step at the center of the genome. J. Mol. Biol. 241:651–662. 9. Cirino, N. M., C. E. Cameron, J. S. Smith, J. W. Rausch, M. J. Roth, S. J. Benkovic, and S. F. J. Le Grice. 1995. Divalent cation modulation of the ribonuclease functions of human immunodeficiency virus reverse transcriptase. Biochemistry 34:9936–9943. 10. Collett, M. S., P. Dierks, J. T. Parsons, and A. J. Faras. 1978. RNase H hydrolysis of the 59 terminus of the avian sarcoma virus genome during reverse transcription. Nature (London) 272:181–184. 11. DeStefano, J. J., R. G. Buiser, L. M. Mallaber, T. W. Myers, R. A. Bambara, and P. J. Fay. 1991. Polymerization and RNase H activities of reverse transcriptases from avian myeloblastosis, human immunodeficiency, and Moloney murine leukemia viruses are functionally uncoupled. J. Biol. Chem. 266:7423–7431. 12. Fedoroff, O. Y., M. Salazar, and B. R. Reid. 1993. Structure of a DNA:RNA hybrid duplex. Why RNase H does not cleave pure RNA. J. Mol. Biol. 233:509–523. 13. Finston, W. I., and J. J. Champoux. 1984. RNA-primed initiation of Moloney murine leukemia virus plus strands by reverse transcriptase in vitro. J. Virol. 51:26–33. 14. Fu, T.-B., and J. Taylor. 1992. When retroviral reverse transcriptases reach the end of their RNA templates. J. Virol. 66:4271–4278. 15. Fuentes, G. M., L. Rodriguez-Rodriguez, P. J. Fay, and R. A. Bambara. 1995. Use of an oligoribonucleotide containing the polypurine tract sequence as a primer by HIV reverse transcriptase. J. Biol. Chem. 270:28169–28176. 16. Furfine, E. S., and J. E. Reardon. 1991. Reverse transcriptase z RNase H from the human immunodeficiency virus. Relationship of the DNA polymerase and RNA hydrolysis activities. J. Biol. Chem. 266:406–412. 17. Gilboa, E., S. W. Mitra, S. Goff, and D. Baltimore. 1979. A detailed model of reverse transcription and tests of crucial aspects. Cell 18:93–100. 18. Gopalakrishnan, V., J. A. Peliska, and S. J. Benkovic. 1992. Human immunodeficiency virus type 1 reverse transcriptase: spatial and temporal relationship between the polymerase and RNase H activities. Proc. Natl. Acad. Sci. USA 89:10763–10767. 19. Guo, J., W. Wu, Z. Y. Yuan, K. Post, R. J. Crouch, and J. G. Levin. 1995. Defects in primer-template binding, processive DNA synthesis, and RNase H activity associated with chimeric reverse transcriptases having the murine leukemia virus polymerase domain joined to Escherichia coli RNase H. Biochemistry 34:5018–5029. 20. Guyader, M., M. Emerman, P. Sonigo, F. Clavel, L. Montagnier, and M. Alizon. 1987. Genome organization and transactivation of the human immu-

5296

POWELL AND LEVIN

nodeficiency virus type 2. Nature (London) 326:662–669. 21. Hansen, L. J., and S. B. Sandmeyer. 1990. Characterization of a transpositionally active Ty3 element and identification of the Ty3 integrase protein. J. Virol. 64:2599–2607. 22. Herr, W. 1984. Nucleotide sequence of AKV murine leukemia virus. J. Virol. 49:471–478. 23. Hu, W.-S., and H. M. Temin. 1990. Retroviral recombination and reverse transcription. Science 250:1227–1233. 24. Huber, H. E., and C. C. Richardson. 1990. Processing of the primer for plus strand DNA synthesis by human immunodeficiency virus 1 reverse transcriptase. J. Biol. Chem. 265:10565–10573. 25. Hung, S.-H., Q. Yu, D. M. Gray, and R. L. Ratliff. 1994. Evidence from CD spectra that d(purine) z r(pyrimidine) and r(purine) z d(pyrimidine) hybrids are in different structural classes. Nucleic Acids Res. 22:4326–4334. 26. Hungnes, O., E. Tjøtta, and B. Grinde. 1991. The plus strand is discontinuous in a subpopulation of unintegrated HIV-1 DNA. Arch. Virol. 116:133– 141. 27. Hungnes, O., E. Tjøtta, and B. Grinde. 1992. Mutations in the central polypurine tract of HIV-1 result in delayed replication. Virology 190:440– 442. 28. Jacobo-Molina, A., J. Ding, R. G. Nanni, A. D. Clark, Jr., X. Lu, C. Tantillo, R. L. Williams, G. Kamer, A. L. Ferris, P. Clark, A. Hizi, S. H. Hughes, and E. Arnold. 1993. Crystal structure of human immunodeficiency virus type 1 reverse transcriptase complexed with double-stranded DNA at 3.0 Å resolution shows bent DNA. Proc. Natl. Acad. Sci. USA 90:6320–6324. 29. Katz, R. A., and A. M. Skalka. 1994. The retroviral enzymes. Annu. Rev. Biochem. 63:133–173. 30. Kohlstaedt, L. A., J. Wang, J. M. Friedman, P. A. Rice, and T. A. Steitz. 1992. Crystal structure at 3.5 Å resolution of HIV-1 reverse transcriptase complexed with an inhibitor. Science 256:1783–1790. 31. Levin, J. G., and J. G. Seidman. 1981. Effect of polymerase mutations on packaging of primer tRNAPro during murine leukemia virus assembly. J. Virol. 38:403–408. 32. Luo, G., L. Sharmeen, and J. Taylor. 1990. Specificities involved in the initiation of retroviral plus-strand DNA. J. Virol. 64:592–597. 33. Luo, G., and J. Taylor. 1990. Template switching by reverse transcriptase during DNA synthesis. J. Virol. 64:4321–4328. 34. Mak, J., M. Jiang, M. A. Wainberg, M.-L. Hammarskjo ¨ld, D. Rekosh, and L. Kleiman. 1994. Role of Pr160gag-pol in mediating the selective incorporation Lys into human immunodeficiency virus type 1 particles. J. Virol. of tRNA 68:2065–2072. 34a.Miller, M. D., B. Wang, and F. D. Bushman. 1995. Human immunodeficiency virus type 1 preintegration complexes containing discontinuous plus strands are competent to integrate in vitro. J. Virol. 69:3938–3944. 35. Myers, G., B. Korber, B. H. Hahn, K.-T. Jeang, J. W. Mellors, F. E. McCutchan, L. E. Henderson, and G. N. Pavlakis. 1995. Human retroviruses and AIDS. A compilation and analysis of nucleic acid and amino acid sequences, p. I-A-376–379. Los Alamos National Laboratory, Los Alamos, N. Mex. 36. Oyama, F., R. Kikuchi, R. J. Crouch, and T. Uchida. 1989. Intrinsic properties of reverse transcriptase in reverse transcription. Associated RNase H is essentially regarded as an endonuclease. J. Biol. Chem. 264:18808–18817. 37. Panganiban, A. T., and D. Fiore. 1988. Ordered interstrand and intrastrand DNA transfer during reverse transcription. Science 241:1064–1069. 38. Peliska, J. A., and S. J. Benkovic. 1992. Mechanism of DNA strand transfer reactions catalyzed by HIV-1 reverse transcriptase. Science 258:1112–1118. 39. Peters, G. G., and J. Hu. 1980. Reverse transcriptase as the major determinant for selective packaging of tRNA’s into avian sarcoma virus particles. J. Virol. 36:692–700. 40. Post, K., J. Guo, E. Kalman, T. Uchida, R. J. Crouch, and J. G. Levin. 1993. A large deletion in the connection subdomain of murine leukemia virus reverse transcriptase or replacement of the RNase H domain with Escherichia coli RNase H results in altered polymerase and RNase H activities. Biochemistry 32:5508–5517. 41. Pullen, K. A., and J. J. Champoux. 1990. Plus-strand origin for human immunodeficiency virus type 1: implications for integration. J. Virol. 64: 6274–6277. 42. Pullen, K. A., A. J. Rattray, and J. J. Champoux. 1993. The sequence

J. VIROL.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52. 53. 54.

55.

56.

57.

58.

59.

60. 61.

62.

features important for plus strand priming by human immunodeficiency virus type 1 reverse transcriptase. J. Biol. Chem. 268:6221–6227. Randolph, C. A., and J. J. Champoux. 1994. The use of DNA and RNA oligonucleotides in hybrid structures with longer polynucleotide chains to probe the structural requirements for Moloney murine leukemia virus plus strand priming. J. Biol. Chem. 269:19207–19215. Ratmeyer, L., R. Vinayak, Y. Y. Zhong, G. Zon, and W. D. Wilson. 1994. Sequence specific thermodynamic and structural properties for DNA z RNA duplexes. Biochemistry 33:5298–5304. Ratner, L., W. Haseltine, R. Patarca, K. J. Livak, B. Starcich, S. F. Josephs, E. R. Doran, J. A. Rafalski, E. A. Whitehorn, K. Baumeister, L. Ivanoff, S. R. Petteway, Jr., M. L. Pearson, J. A. Lautenberger, T. S. Papas, J. Ghrayeb, N. T. Chang, R. C. Gallo, and F. Wong-Staal. 1985. Complete nucleotide sequence of the AIDS virus, HTLV-III. Nature (London) 313:277–284. Rattray, A. J., and J. J. Champoux. 1987. The role of Moloney murine leukemia virus RNase H activity in the formation of plus-strand primers. J. Virol. 61:2843–2851. Rattray, A. J., and J. J. Champoux. 1989. Plus-strand priming by Moloney murine leukemia virus. The sequence features important for cleavage by RNase H. J. Mol. Biol. 208:445–456. Resnick, R., C. A. Omer, and A. J. Faras. 1984. Involvement of retrovirus reverse transcriptase-associated RNase H in the initiation of strong-stop (1) DNA synthesis and the generation of the long terminal repeat. J. Virol. 51:813–821. Sanchez-Pescador, R., M. D. Power, P. J. Barr, K. S. Steimer, M. M. Stempien, S. L. Brown-Shimer, W. W. Gee, A. Renard, A. Randolph, J. A. Levy, D. Dina, and P. A. Luciw. 1985. Nucleotide sequence and expression of an AIDS-associated retrovirus (ARV-2). Science 227:484–492. Sawyer, R. C., and H. Hanafusa. 1979. Comparison of the small RNAs of polymerase-deficient and polymerase-positive Rous sarcoma virus and another species of avian retrovirus. J. Virol. 29:863–871. Schatz, O., J. Mous, and S. F. J. Le Grice. 1990. HIV-1 RT-associated ribonuclease H displays both endonuclease and 39359 exonuclease activity. EMBO J. 9:1171–1176. Shinnick, T. M., R. A. Lerner, and J. G. Sutcliffe. 1981. Nucleotide sequence of Moloney murine leukaemia virus. Nature (London) 293:543–548. Smith, J. K., A. Cywinski, and J. M. Taylor. 1984. Specificity of initiation of plus-strand DNA by Rous sarcoma virus. J. Virol. 52:314–319. Sonigo, P., M. Alizon, K. Staskus, D. Klatzmann, S. Cole, O. Danos, E. Retzel, P. Tiollais, A. Haase, and S. Wain-Hobson. 1985. Nucleotide sequence of the visna lentivirus: relationship to the AIDS virus. Cell 42:369– 382. Sorge, J., and S. H. Hughes. 1982. Polypurine tract adjacent to the U3 region of the Rous sarcoma virus genome provides a cis-acting function. J. Virol. 43:482–488. Tanese, N., A. Telesnitsky, and S. P. Goff. 1991. Abortive reverse transcription by mutants of Moloney murine leukemia virus deficient in the reverse transcriptase-associated RNase H function. J. Virol. 65:4387–4397. Telesnitsky, A., and S. P. Goff. 1993. Strong-stop strand transfer during reverse transcription, p. 49–83. In A. M. Skalka and S. P. Goff (ed.), Reverse transcriptase. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. Varmus, H., and P. Brown. 1989. Retroviruses, p. 53–108. In D. E. Berg and M. M. Howe (ed.), Mobile DNA. American Society for Microbiology, Washington, D.C. Varmus, H., and R. Swanstrom. 1984. Replication of retroviruses, p. 369– 512. In R. Weiss, N. Teich, H. Varmus, and J. Coffin (ed.), Molecular biology of tumor viruses, RNA tumor viruses, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Wain-Hobson, S., P. Sonigo, O. Danos, S. Cole, and M. Alizon. 1985. Nucleotide sequence of the AIDS virus, LAV. Cell 40:9–17. Wilson, S. H., and J. Abbotts. 1992. tRNA in the molecular biology of retroviruses, p. 1–21. In D. L. Hatfield, B. J. Lee, and R. M. Pirtle (ed.), Transfer RNA in protein synthesis. CRC Press, Inc., Boca Raton, Fla. Wo ¨hrl, B. M., and K. Moelling. 1990. Interaction of HIV-1 ribonuclease H with polypurine tract containing RNA-DNA hybrids. Biochemistry 29:10141–10147.