MINIREVIEW. Primer trnas for Reverse Transcription

Transcription

1 JOURNAL OF VIROLOGY, Nov. 1997, p Vol. 71, No X/97/$ Copyright 1997, American Society for Microbiology MINIREVIEW Primer trnas for Reverse Transcription JOHNSON MAK 1 AND LAWRENCE KLEIMAN 2 * AIDS Pathogenesis Research Unit, Macfarlane Burnet Centre for Medical Research, Fairfield, Victoria, Australia 3078, 1 and Lady Davis Institute for Medical Research and McGill AIDS Center, Jewish General Hospital, and Departments of Medicine and Microbiology and Immunology, McGill University, Montreal, Quebec, Canada H3T 1E2 2 INTRODUCTION Both retroviruses and long terminal repeat (LTR) retrotransposons use cellular trnas as primers for reverse transcription during their replication cycles. In retroviruses, primer trna is selectively packaged into the virion, where it is placed onto the primer binding site (PBS) of the viral RNA genome and used to prime the reverse transcriptase (RT)-catalyzed synthesis of minus-strand cdna. Studies of how these processes are carried out in different retroviral groups have revealed both similarities and differences. Although less extensively studied, a comparison of processes occurring in LTR retrotransposons with similar processes occurring in retroviruses is also informative and is included herein. For an excellent summary of earlier work on retroviral primer trnas, the reader is referred to the review by Waters and Mullin (83). Later reviews on this topic include those by Litvak et al. (48), Marquet et al. (54), and Leis et al. (41), while reviews providing information on retrotransposon primer trnas include those by Voytas and Boeke (78) and Sandmeyer and Menees (68). Retroviruses can be divided into three major subfamilies, oncoviruses, lentiviruses, and spumaviruses, while retrotransposons can be placed into two categories named after the prototypic Drosophila elements copia and gypsy (68). The replication cycles of retroviruses and LTR retrotransposons show strong similarities, as shown in Fig. 1. Both retroviruses and retrotransposons code for Gag and Gag-Pol proteins. Retroviral Gag proteins contain sequences for matrix (MA), capsid (CA), and nucleocapsid (NC) proteins, while retrotransposon Gag proteins contain sequences for CA and, sometimes, NC proteins (68). Both retroviral and retrotransposon pol genes code for protease (PR), RT, and integrase (IN). copia- and gypsy-like elements can be distinguished from each other by major differences in RT sequences (89), and while in copia-like elements the IN gene precedes the RT-RNase H gene, in gypsy-like elements the RT-RNase H sequence precedes the IN sequence (18, 73), as in retroviruses. Many types of retroviruses have an extracellular stage in their life cycle, and this is associated with the presence of envelope protein (Env). The replication cycle of most retrotransposon elements produces an intracellular virus-like particle (VLP) which contains no Env protein. However, some gypsy elements do have an extracellular stage, and this is associated with the presence of Env protein (36, 60). Some of the similarities and differences among retroviruses and retrotransposons relevant to this review are summarized in Table 1. * Corresponding author. Mailing address: Lady Davis Institute for Medical Research, Jewish General Hospital, 3755 Cote Ste-Catherine Rd., Montreal, Quebec, Canada H3T 1E2. The full-length RNA transcribed from proviral or retrotransposon elements codes for Gag and Gag-Pol precursor proteins, which in the cytoplasm assemble into particles that package the full-length mrna, as well as low-molecularweight trna. In both particle types, the Gag and Gag-Pol precursor proteins are cleaved by a viral protease to the final mature proteins. The reverse transcription of the packaged RNA in both retroviruses and retrotransposons is primed by a cellular trna, and the resulting double-stranded cdna is integrated into the host cell DNA by viral- or retrotransposonencoded IN. THE PRIMER trnas In both retroviruses and retrotransposons, cdna synthesis is initiated from the PBS near the 5 end of the packaged RNA. The primer trnas used in retroviruses and retrotransposons are listed in Table 1 and in many cases have been identified only by the PBS sequence. trna Trp is the primer for all members of the avian sarcoma virus (ASV)-avian leukosis virus (ALV) group examined to date (17, 21, 63, 70, 83, 84). There are three major trna Lys isoacceptors in mammalian cells (65). trna Lys 1,2, representing two trna Lys isoacceptors differing by 1 bp in the anticodon stem, is the primer trna for several retroviruses, including Mason-Pfizer monkey virus and human foamy virus (HFV) (41). trna Lys 3 serves as the primer for mouse mammary tumor virus (62, 82) and the lentiviruses, such as equine infectious anemia virus, feline immunodeficiency virus, simian immunodeficiency virus, and human immunodeficiency virus types 1 and 2 (HIV-1 and HIV-2) (41). trna Pro is the common primer for Moloney murine leukemia virus (Mo-MuLV) (20, 61, 77), but murine retroviruses have also been reported to make use of several alternate trnas as primers. For example, a recombinant MuLV that replicates with wild-type kinetics, but uses trna Gln 1 as a primer, was isolated by Colicelli and Goff (10), and endogenous murine retroviruses that have PBSs complementary to trna Lys 3, trna Phe, and trna Gly have also been reported (10). The specificity of interaction in vitro between RT and the cognate primer trna appears to be greater for HIV (6, 67), Rous sarcoma virus (RSV), and avian myeloblastosis virus (AMV) (11, 23, 59) than for MuLV (58). Retrotransposons and retroviruses use different primer trnas for reverse transcription, and for some retrotransposons only a trna fragment is used as the primer (68, 78). In the yeast Saccharomyces cerevisiae, there are five known Ty retrotransposons: Ty1 (copia), Ty2 (copia), Ty3 (gypsy), Ty4 (copia), and Ty5 (copia). The first three have primer binding sites, complementary to the 3 terminus of trna Met i, of 10, 10, and 8 bases in length, respectively. While this is short compared to 18-base retroviral PBSs, it has been reported that in 8087

2 8088 MINIREVIEW J. VIROL. Downloaded from FIG. 1. Replication cycle of retrovirus and retrotransposon elements. Most VLPs tend to remain intracellular, and most retroviruses have an extracellular stage. The structure of the VLP is not as well characterized as that of retroviruses but probably also presents immature and mature forms which differ from each other in their state of precursor protein proteolysis. HIV-1 a deletion of 12 of the 18 nucleotides at the 3 end of the PBS still allows sufficient complementarity for efficient reverse transcription (80). The PBS in Ty4 is complementary to the 3 18 bases of trna Asn. The primer binding site in the Ty5 element is complementary to a 13-base internal fragment (which includes the anticodon loop) of trna i Met (78). For the Drosophila copia element, a 39-nucleotide fragment from the 5 region of Drosophila trna i Met is used as the primer for copia minus-strand cdna synthesis by RT (35). The Drosophila gypsy and gypsy-like elements whose sequences are known include 17.6, 297, tom, and gypsy. For the first three of these, the PBS is complementary to the 3 18 bases of trna Ser, while the primer binding site of the gypsy element is complementary to the 3 11 bases of trna Lys. VIRAL PROTEIN-tRNA INTERACTIONS INVOLVED IN THE SELECTIVE PACKAGING AND GENOMIC PLACEMENT OF PRIMER trnas Primer trna is selectively packaged during the assembly of retroviruses or VLPs. In general, the term selective packaging of trna refers to the increase in the percentage of the lowmolecular-weight RNA population representing primer trna going from the cytoplasm to the virus. For example, in AMV, the relative concentration of trna Trp changes from 1.4 to 32% (83). In HIV-1 produced from COS-7 cells transfected with HIV-1 proviral DNA, both primer trna Lys Lys 3 and trna 1,2 are selectively packaged, and the relative concentration of trna Lys changes from 5 to 6% to 50 to 60% (52). In AKR- MuLV, selective packaging of primer trna Pro is less dramatic, going from a relative cytoplasmic concentration of low-molecular-weight RNA of 5 to 6% to 12 to 24% (83). trna Lys 3 and trna Lys 1,2 are packaged into HIV-1 with equal efficiencies since the trna Lys 3 /trna Lys 1,2 ratio in the virus reflects the cytoplasmic ratio, even when the cytoplasmic ratio is altered (23). In HIV-1, estimates of 20 molecules of trna Lys /virion (i.e., per 2 molecules of genomic RNA) have been reported (23), but these numbers must remain estimates because the viral population studied may not be homogeneous for packaging of either genomic RNA or trna Lys. As with retroviruses, the primer trna Met i used in the Ty1 retrotransposon replication is also selectively packaged into the VLPs, showing a 10- to 40-fold increase in relative concentration (9, 64). In all of these cases, there are no data on whether the absolute concentration of primer trna (trna/volume) changes in going from the cytoplasm into the virus or VLP. Reverse transcriptase sequences. Genomic RNA packaging is not required for the incorporation of primer trna in ASV (64), MuLV (44), or HIV-1 (30, 52). Several studies indicate that RT sequences in retroviruses are involved in primer trna packaging. In vitro studies using enzymatic and chemical probes (5, 6, 69, 88) or cross-linking (56) have shown that the on June 19, 2018 by guest

3 VOL. 71, 1997 MINIREVIEW 8089 Retrovirus or retrotransposon TABLE 1. Similarities and differences among selected retroviruses and retrotransposons Presence or absence of: Gag Gag-Pol a Env Gag Cys-His box b Primer trna Length of primer binding site (bases) ASLV Trp 18 MuLV Pro 18 HIV-1 Lys-3 18 HFV Lys-1, 2 18 S. cerevisiae transposons Ty1 (copia) Met-i 10 Ty2 (copia) Met-i 10 Ty3 (gypsy) Met-i 8 Ty4 (copia) Asn 18 Ty5 (copia) Met-i fragment 13 Drosophila transposons gypsy Lys (gypsy) Ser (gypsy) Ser 18 tom (gypsy) Ser (gypsy) Arg 18 copia Met-i fragment 18 a In HFV, Pol is made from a singly spliced mrna and is produced separately from Gag (91). b In HFV, the presence of an NC sequence in the Gag protein is determined from it position toward the carboxy end of Gag (as in lentiviruses and oncoviruses) and by the ability of this region to bind RNA in vitro, through glycine- and arginine-rich regions (89). Retrotransposon sequences are defined as coding for the NC protein primarily because of the presence of a Cys-His box near the carboxy terminus of the Gag sequences, a position similar to that found for the NC of lentiviruses and oncoviruses. The actual function of retrotransposon Gag carboxy-terminal sequences, with or without a Cys-His box, is not known, nor has the actual protein binding to VLP RNA been evidenced. anticodon, T C, and D loops of free trna 3 Lys interact with mature RT. In vivo studies have also indicated the importance of RT sequences in primer trna packaging. Selective packaging of primer trna does not occur in RT-negative ASV (63) or RSV (71), in a mutant Mo-MuLV which lacks 40% of the RT sequence from the C terminus (42), or in an RT-negative HIV-1 isolate (52). It is, however, not likely that the mature RT itself is involved in primer trna packaging since viral assembly occurs with larger precursor protein molecules which are not cleaved until budding. Since the presence of RT is required for primer trna packaging, this implies that the Gag-Pol precursor is involved in this process since it contains the RT sequences. Investigations into the packaging of trna Lys into HIV have supported this hypothesis. It has been shown that HIV particles composed of only Pr55 gag do not selectively package trna Lys while particles composed of both Pr55 gag and Pr160 gag-pol do (52). Inhibition of HIV-1 protease also does not prevent the select viral packaging of primer trna 3 Lys (52). The results of an in vitro cross-linking study of a synthetic trna 3 Lys HIV-1 RT complex support a possible interaction between the trna 3 Lys and the thumb and connection subdomains of RT (56). In the in vitro study, no evidence for interaction of trna 3 Lys was found for proteolytic RT fragments containing primarily the finger or palm domains. On the other hand, a carboxy-terminal p66 fragment (amino acid [aa] residues 357 to 560) containing both the carboxy half of the connection domain and the RNase H domain interacted with the 5 terminus of trna 3 Lys, while fragments from both p66 and p51 (representing aa residues 230 to 356) containing both the thumb and the amino portion of the connection domain were found to interact with U 36 in the anticodon of trna 3 Lys. Attempts were also made to map a trna Lys binding site within the HIV-1 RT sequences found in Pr160 gag-pol by measuring trna 3 Lys packaging in viruses produced from COS cells transfected with HIV proviral DNA mutated in various RT subdomains. While it was found that mutations specifically in the thumb and connection subdomains of RT resulted in decreased trna Lys 3 packaging, these mutations also inhibited the packaging of Pr160 gag-pol (53) and therefore did not allow the identification of a trna Lys 3 binding site in RT. Because mature RT interacts with primer trna during the initiation of reverse transcription, RT sequences may also be involved in facilitating genomic placement of primer trna. It has been observed that the trna Lys 3 acceptor stem (which contains the sequences complementary to the PBS) is digested by RNase A only in the presence of retroviral RT, suggesting a partial destabilization of this region by the RT (69). Such destabilization might be important for the genomic placement of primer trna Lys 3 in vivo. In ASV (63) and HIV-1 (13, 52), the absence of RT protein and RT activity is correlated with the absence of both selective primer trna packaging and genomic placement of the trna, but it is difficult to determine from this data if the lack of primer trna genomic placement is due to the absence of RT or the absence of selective primer trna packaging. Additionally, it should be noted that while the absence of RT protein in MuLV also inhibits select primer trna packaging, wild-type levels of primer trna are still placed onto the viral genome; i.e., RT may not be involved in primer trna placement in MuLV (42, 43). If RT protein does facilitate annealing of primer trna to the PBS in avian and human retroviruses, it probably does so as part of the Gag-Pol protein, since genomic placement of primer trna can occur in the absence of precursor processing in a PR-negative virion (11, 26, 76). Exceptions to the observation that primer trna packaging in retroviruses is dependent on Gag-Pol packaging and independent of genomic RNA packaging may exist. For example, in HFV, which uses trna Lys 1,2 as a primer, there is no Gag-Pol precursor; the Pol protein is synthesized from a separate spliced mrna (90). In oncoviruses and lentiviruses, the Gag- Pol precursor is believed to be carried into the virus via interactions with the Gag precursor (59, 72, 74), and it may be that in HFV a Gag-Pol complex, analogous to the Gag Gag-Pol complex in other viruses, is formed which can also carry the trna Lys 1,2 into the virion. However, no studies have been done

4 8090 MINIREVIEW J. VIROL. to determine if primer trna Lys 1,2 packaging in HFV is selective or is dependent on the Pol precursor. It has also been reported that trna Met i packaging in yeast Ty1 VLPs is severely diminished by mutations in either the PBS or regions downstream of the PBS which are complementary to sequences in the T C Met and D arms (87). This report implies that primer trna i packaging depends on genomic RNA packaging, in contrast to observations in retroviruses. NC sequences. NC sequences have also been reported to play an important role in retroviral processes such as primer trna annealing, genomic RNA packaging and dimerization, and minus-strand strong-stop cdna strand transfer (12, 19). There is also evidence that in HIV-1 the NC sequences in Pr55 gag may contribute to the trna Lys binding site associated with viral packaging (22). HIV-1 NC (NCp7) contains two subdomains known either as Cys-His boxes (because they contain the CCHC motif, Cys-X2-Cys-X4-His-X4-Cys) or Zn 2 fingers (because of their ability to bind Zn 2 ). The positions of the two Cys-His boxes in HIV-1 define other subdomains of NCp7. From the N to the C terminus, these may be termed the N-terminal subdomain, box 1, the linker subdomain, box 2, and the C-terminal domain (7). The NCp7 mutant K14-T50 has both Cys-His boxes and the linker region between them deleted. COS cells transfected with this mutant proviral DNA produce Pr55 gag particles which package Pr160 gag-pol but not trna Lys. trna Lys packaging is rescued by cotransfection with a plasmid coding for wild-type Pr55 gag (and not with a plasmid coding for wild-type Pr160 gag-pol ), indicating a possible role of NC sequences in the Gag precursor in forming the trna Lys binding site, possibly as part of a Pr55 gag -Pr160 gag-pol complex (22). Lys HIV-1 NC protein also facilitates the annealing of trna 3 to in vitro-transcribed genomic RNA sequences (16), probably by unwinding the secondary structure of trna Lys 3 in vitro (34). Similar observations have been made in RSV and Mo-MuLV; NC promotes the annealing of trna Trp and trna Pro to the PBSs of the RSV and Mo-MuLV genomic RNA, respectively (64). The promotion of trna Lys 3 placement by NC in vitro occurs in an NC concentration-dependent manner (46), and a recent study indicates that the NC-tRNA Lys 3 interaction is nonspecific, i.e., it results from the general ability of NC to interact with RNA (55). For both Mo-MuLV (15) and HIV-1 (16, 39) it has been reported that the in vitro annealing of primer trna is independent of the presence of the Cys-His boxes but depends on the presence of the amino acids flanking the first, or only, Cys-His box. These observations have recently been confirmed in vivo for HIV-1 (25). Whether nucleocapsid plays such an important role in spumaviruses or in retrotransposons is unclear. All retrovirus NC proteins contain one or two Cys-His boxes, with the exception of the NCs of HFV and simian foamy virus, which do not contain Cys-His box (66). The presence of NC sequences in the carboxy region of HFV Gag protein is indicated by the ability of this region to bind to RNA in vitro through glycine- and arginine-rich regions (91). Many retrotransposon Gag proteins synthesized for VLP production also do not contain this metal finger motif. As for the Gag proteins synthesized by the five Ty transposons, those of Ty1 and Ty2 do not have a Cys-His box while those of Ty3 to Ty5 do. None of the four Drosophila gypsy elements (17.6, 297, tom, and gypsy) has this metal finger motif. The function of retrotransposon Gag carboxy-terminal sequences, with or without a Cys-His box, is not known. SIGNALS ON PRIMER trna ASSOCIATED WITH ITS PACKAGING, GENOMIC PLACEMENT, AND PRIMING FUNCTION The signals on the primer trna which target it for viral packaging are not known. In vitro studies on the interaction of purified trna Lys 3 with mature HIV RT have indicated that an interaction between the anticodon of the trna and RT occurs (5, 6, 69, 88). In HIV-1, mutations in the trna Lys 3 anticodon (SUU changed to CUA, where S mcm 5 S 2 U [5-methoxycarbonylmethyl-2-thiouridine]) prevent the mutant trna from acting as a primer in vivo (24) but do not prevent viral packaging of this trna (23, 24). Similarly, trna Lys Lys 3 and trna 1,2 are packaged into HIV-1 with the same efficiency (23) but have different anticodons (SUU and CUU, respectively). The insensitivity of trna Lys packaging to the anticodon sequence should not be surprising since the interaction of the trna with mature RT may be very different from the interaction with the proteins involved in trna Lys packaging, Pr160 gag-pol and Pr55 gag. Although no common property shared by primer trnas has been associated with their ability to act as primers, one case has been reported in which a specific modification of trna Trp was found to influence its ability to be packaged into AMV (33). Two major trna Trp species have been identified in avian cells, which differ by a methylation of nucleotide 7 in the amino acid acceptor stem: G or m2 G. Only the nonmethylated species is packaged and used as the primer for avian retrovirus reverse transcription. The methylation of G to m2 G may thus prevent the recognition of trna Trp by a retroviral protein(s) responsible for the selective packaging of the primer trna Trp. Boeke and his colleagues have utilized the powerful genetic capabilities of the yeast S. cerevisiae to analyze the effect of mutations in trna Met i on its ability to participate in the Ty1 retrotransposon replication cycle (32). A strain of S. cerevisiae was created that did not contain the four trna Met i genes of this yeast. Translation was allowed through the introduction of a plasmid containing a mutant trna Met i gene capable of producing a trna Met i which could participate in translation but not in retrotransposon replication. The ability of a genetically marked Ty1 transposon (also coded for by this plasmid) to transpose was examined when a second plasmid bearing a mutant trna Met i gene was introduced. Using this system, they found that single mutations in the acceptor stem region complementary to the PBS, and mutations in the T C and D arms of trna Met i, inhibited transposition while mutations in the anticodon arm did not. Unfortunately, the steps in the retrotransposon replication cycle which are inhibited by these mutations are not yet known. Since in the cell trna is found in complexes with proteins associated with translation, one of these cellular proteins could play a role in facilitating an interaction between the primer trna and the viral or VLP proteins. In eukaryotic translation, trna is part of a channeled cycle (75) and aminoacyl-trnas are directly transferred from aminoacyl-trna synthetase to the elongation factor eef1, which carries trnas to the ribosomes. After peptide bond formation, the deacylated trnas are transferred back, via eef1, to their cognate synthetases for recharging with amino acids. The interaction between a viral precursor protein and primer trna might occur at the polysome, where both components are concentrated, and the interaction could be initially with a trna binding protein such as eef1 or aminoacyl-trna synthetase. It has been reported that while all detectable trna Lys in an HIV-1-infected cell is acylated, all detectable trna Lys in HIV-1 is deacylated (23), a state required for trna Lys to function as a primer for RT. It is possible that a trna Lys molecule is bound to a partially

5 VOL. 71, 1997 MINIREVIEW 8091 FIG. 2. Proposed regions of base pairing between trna 3 Lys and the HIV-1 genome. This figure is modified from one of Arts et al. (3). In addition to the PBS interaction, other regions in the viral genomic RNA may interact with the trna. These include the A-rich regions upstream (28, 30) and downstream (38) of the PBS interacting with the anticodon loop (shown by the two arrows), as well as an interaction of the T C loop with a U5 region upstream of the PBS (1, 2). synthesized precursor protein immediately after donating its lysine to the growing peptide chain on the polysome, but there is currently no evidence to indicate whether trna Lys is deacylated before or after its entry into the virion. INTERACTION BETWEEN THE PRIMER trna AND GENOMIC RNA SEQUENCES, AND INITIATION OF REVERSE TRANSCRIPTION Primer trna is found in the virus both in a free state and bound to the viral genome (63). Although the incorporation of most of the primer trna molecules into the virus occurs independently of genomic RNA packaging (32, 36, 46), it is not known whether the actual primer trna molecule bound to the viral RNA genome is placed onto the PBS before or after packaging. Some of the postulated interactions between primer trna and the region around the PBS are shown for the trna 3 Lys and HIV-1 genomic RNA in Fig. 2 (modified from reference 3). The 3 -terminal nucleotides of the primer trna are complementary to the nucleotides comprising the PBS, and studies in HIV-1 (80), MuLV-based retroviral vectors (49), and the Ty1 retrotransposon (32) have shown that a minimum degree of complementarity must be maintained to achieve efficient reverse transcription. Studies with both HIV-1 and ALV have shown that the PBS sequence is not in itself sufficient to determine primer trna identity. Transfection of proviral DNAs with PBS sequences that are complementary to trnas other than the natural primer trna yielded virus with initially slow replication kinetics in studies using either HIV-1 (14, 45, 81) or ALV (85). However, the mutant genomes eventually revert back to their respective wild-type PBS sequences and the revertant virus then grows at wild-type rates. The reversion to the wild-type FIG. 3. Reverse transcription of retroviral genomic RNA into doublestranded proviral DNA. This figure is modified with permission from the authors of reference 77a. Step 1: annealing of primer trna to the PBS, and synthesis of minus-strand strong-stop cdna, with the resulting degradation of R and U5 RNA by the RNase H activity of the reverse transcriptase. Step 2: the first strand transfer, in which minus-strand strong-stop cdna is annealed to the 5 terminus of the genomic RNA via R-R hybridization. Steps 3 and 4: further synthesis of minus-strand cdna, during which the genomic RNA is further degraded by RNase H. A small piece of RNA, the polypurine tract (PPT), remains undegraded and serves as the primer for plus-strand strong-stop cdna (step 5). Step 5: termination of plus-strand strong-stop cdna synthesis 18 nucleotides into the primer trna, thereby generating a new PBS sequence; the trna is released from the minus-strand cdna. Step 6: the second strand transfer, in which plus-strand strong-stop cdna is annealed to the 3 terminus of minus-strand cdna via PBS-PBS hybridization. Step 7: completion of synthesis of doublestranded proviral DNA. PBS sequence probably results from the reverse transcription of the 3 18 nucleotides of the natural primer trna which occurs during the synthesis of plus-strand strong-stop cdna, as shown in step 5 of Fig. 3. After the 3 -terminal nucleotides of the trna primer are copied, strand transfer of this strongstop DNA to the 5 terminus of the minus-strand DNA (step 6) results in annealing of the newly generated PBS sequences to the complementary PBS sequences in the minus-strand DNA. This model, which predicts that the trna 3 -terminal sequences determine the sequence of the newly generated PBS, is used to explain how PBS sequences made complementary to a nonprimer retroviral trna eventually revert back to sequences complementary to a more favored natural primer trna. The method of generating a new PBS may, however, be different in some retrotransposons. Using the engineered yeast strain (32) described above, a trna i Met mutated at position 7 in the 3 -terminal sequence complementary to the PBS was used as the sole primer trna for Ty1 reverse transcription. The PBS retained the wild-type sequence; i.e., the mutant trna i Met did not appear to contribute to the generation of the new PBS (40). This implies that the new PBS is inherited from the old PBS sequence rather than the primer trna. This could be accomplished as follows: if the first strand transfer in step 2 of Fig. 3 went to another copy of undegraded genomic RNA,

6 8092 MINIREVIEW J. VIROL. then the resulting minus-strand cdna synthesized would contain R -U5 at its 3 end. In step 5, plus-strand synthesis would terminate before trna sequences are copied, and the second strand transfer seen in step 6 would be to full-length minusstrand cdna; i.e., pairing would be between R-U5 and R - U5. Further plus-strand cdna synthesis would generate a new PBS sequence from PBS in the minus strand. This mechanism of reverse transcription is different from that shown in Fig. 3; i.e., the first strand transfer in step 2 would be intermolecular rather than intramolecular, and plus-strand strongstop cdna would terminate before copying trna sequences. Although there is no direct evidence for the presence of two copies of genomic RNA in the VLPs, as in retroviruses, some genetic evidence indicates that there is more than one fulllength RNA molecule per particle; i.e., data indicating that the equivalent of retroviral recombination happens at a very high frequency have been reported (8). There is also evidence in Ty1 of some plus-strand strong-stop cdna terminating before using trna sequences as templates (86). The conversion of an altered PBS to one complementary to the natural primer trna, as seen in HIV-1 and ALV, may not occur in some MuLV systems. For example, Lund et al. (50) showed that a replication-defective retroviral vector derived from MuLV could use various cellular trnas as primers to replicate with equal efficiencies. This vector contained the LTRs, PBS, and some leader sequence of MuLV, as well as a neomycin gene, but could not code for viral proteins, which were provided by an RNA packaging cell line. The resulting VLPs could undergo one round of infection, and resulting G418-resistant cells were used as evidence of the occurrence of reverse transcription of the infecting RNA genome. It was found that vectors with PBSs complementary to different trnas could produce as many G418-resistant cells as one with the natural PBS complementary to trna Pro. The MuLVbased retroviral vector has a greater ability to use alternate primer trnas than do the ALV and HIV-1 retroviruses, and this could reflect the finding noted earlier that the restriction on the type of primer trna used in MuLVs may not be as stringent as for HIV or avian retroviruses. However, a potential problem when using this retroviral vector system to study primer trna packaging and initiation of reverse transcription is that while trnas other than trna Pro can be used as primers with similar efficiencies, it is not clear whether the retroviral vector genome using the natural primer trna Pro is itself being efficiently reverse transcribed in this artificial system compared with wild-type MuLV. Also, because this system uses only one round of infection, it may not be sensitive to subtle differences arising when using an alternate trna instead of the natural primer trna. The reversion of altered PBSs in HIV and ALV to the natural wild-type PBS suggests that there are factors other than just the PBS which influence the choice of the primer trna used. It has been reported that regions in the genomic RNA other than the PBS interact with the primer trna, and these interactions are shown in Fig. 2 for the trna 3 Lys -HIV genomic RNA complex. The existence of an interaction of USUU in the anticodon loop of trna 3 Lys with the A-rich loop in this U5 stem A-rich loop was first reported by Ehresmann and colleagues (29), who demonstrated this interaction through a combination of chemical and enzymatic probing and computer modeling (27). The anticodon loop has also been proposed to interact with another A-rich region downstream of the PBS (37). Both interactions are shown by arrows in Fig. 2. Also shown is a proposed interaction between the T C loop and U5 sequences which were first proposed for primer trna Trp based on studies of its interaction with the genomes of ASV and ALV (1, 2). Although all lentiviruses examined (HIV-1, HIV-2, simian immunodeficiency virus, feline immunodeficiency virus, and equine infectious anemia virus) use trna 3 Lys as a primer (41), only HIV-1 contains an A-rich loop in the U5 stem-loop. In vitro, reverse transcription using homologous RT and genomic RNA for each of these lentiviruses can be resolved into initiation and elongation phases, and U5 stem-loop structures may be involved in the transition between these two stages. The initiation phase is manifest by an initial buildup of 1- to 12-base DNA extensions of the trna 3 Lys, depending on the source of viral RNA; i.e., the length of these extensions appears to be correlated with a stem-loop structure upstream of the PBS (4). In the case of HIV-1, the in vitro extensions are 1 to 6 bases in length. Evidence for a discrete initiation stage of HIV-1 reverse transcription has also been found in vivo. In HIV-1 produced from either transfected COS-7 cells or a variety of stably infected cell lines (H9, CEM-SS, U937, and PLB), the trna 3 Lys annealed to the PBS appears to exist in two forms: unextended trna 3 Lys and trna 3 Lys extended by the first two nucleotides added during reverse transcription, dcmp and TMP (26). As shown in Fig. 2, the 2-base extension would extend to the first GC base pair in the stem. A U5 stem-loop structure immediately upstream of the PBS in RSV has been reported (9). Mutations disrupting this structure inhibit reverse transcription in vivo. In vitro experiments indicate that the disruption does not reduce trna Trp placement on the PBS but does reduce the initiation of reverse transcription. Also, a deletion in the 3 part of the U5 region near the PBS in Mo-MuLV causes an inhibition of viral cdna synthesis in vivo, but in vitro experiments indicate that neither primer trna placement nor initiation of reverse transcription was impaired (57). No evidence for the existence of short DNA extensions in avian or murine retroviruses has been reported. At the other extreme, it has been estimated that 10 to 15% of mature-hfv particles contain double-stranded cdnas approaching full length (90), an indication of the occurrence of reverse transcription prior to HFV infection of target cells. Deletion of the four consecutive A s in the U5 stem-loop of HIV-1 does not affect trna 3 Lys genomic placement, either in vitro (4, 28) or in vivo (25, 47). Neither is the early pausing of reverse transcription diminished in vitro (4) or in vivo (25) at low RT-to-primer/template (1:2) ratios, probably because the stem structure is still maintained. When the stem structure is destabilized by using a 9-base preextended trna 3 Lys as the primer, pausing disappears (4). In vivo, HIV-1 containing a deletion of the 4 A s shows decreased reverse transcription and slower replication kinetics, which increase over time after the reversion of two G s to A s in the region of the stem-loop structure (47). The initiation and elongation phases of reverse transcription have been associated with low and high processivity of RT, respectively (38), and in vivo, destabilization of the A-rich loop seems to result in decreased elongation due to lower RT processivity (25). The ability of HIV RT to interact with the trna 3 Lys anticodon bound to the A-rich loop may allow the enzyme to change its conformation after a short cdna extension, but why such an interaction is required only in HIV-1 is unknown. Interestingly, the A-rich loop can play a role in determining the identity of the primer trna used in HIV-1. Morrow s group had found that infection of cells with HIV-1 containing a PBS complementary to trna His rather than trna 3 Lys results in an initial low rate of virus replication followed at later stages by rates of virus production approaching those of the wild-type virus. The increased viral replication is correlated with rever-

7 VOL. 71, 1997 MINIREVIEW 8093 sion of the trna His PBS to a PBS complementary to trna 3 Lys, implying the eventual use of trna 3 Lys again as the primer (81). However, when the A-rich loop as well was altered so that it was complementary to the trna His anticodon, the trna His PBS became stabilized in the viral population (79) and wildtype viral replication rates occurred (although some other mutations which may be further required for this stabilization also arose in the region near the A-rich loop). Similarly, altering both the PBS and the U5 stem A-rich loop in HIV-1 genomic RNA so that they were complementary to the acceptor stem and anticodon loop of trna i Met, respectively, resulted in the stable use of trna i Met as a primer (31). These results clearly indicate a role for the A-rich loop in allowing a trna other than trna 3 Lys to be used as a primer for reverse transcription in HIV-1. Since the deletion of the A-rich loop does not alter genomic placement of primer trna 3 Lys, it seems likely that altering the loop so that it interacts with the new primer trna determined by the PBS sequences probably allows a more highly processive elongation phase of reverse transcription to proceed. From this interpretation, it might be predicted that altering the PBS in a virus not containing such an anticodon stem-loop interaction would be enough to create a new primer trna. This has not turned out to be the case for ALV, in which altered PBSs tend to revert to the native trna Trp PBS; this may be due to other required interactions between trna Trp and the genomic RNA, such as the interaction shown to occur between the T C loop and U5 region in ASV and ALV (1, 2). UNANSWERED QUESTIONS The use of trnas as primers for reverse transcription may reflect an ancient function in the RNA world. This is discussed in the genomic tag hypothesis of Maizels and Weiner (51), which proposes that in the ancient world, when the major replicating macromolecule may have been RNA, trna-like structures were tags at the 3 termini of the RNA genomes, serving as primers for the synthesis of complementary RNA strands. It is not clear today why only a limited number of such primer trnas have been identified. Do they have a common property associated with their ability to act as primers, a property even shared with 5 and internal trna i Met fragments used as primers in some retrotransposons, or is the number of trnas used as primers limited only by the evolution of other components of the priming apparatus to operate most efficiently with select trnas? How are the trnas selectively packaged? In HIV-1, both Gag and Gag-Pol precursors seem to be involved in carrying primer trna into the virus, independently of genomic RNA packaging. In HFV, however, Pol is made separately from Gag. Are Gag Gag-Pol and Gag-Pol complexes formed in HIV-1 and HFV, respectively, and do they perform similar functions? What is the role of genomic RNA in primer trna packaging in Ty1, for which mutations in the packaged RNA seem to affect primer trna packaging? What signals target the trna for packaging? Are there targeting signals within either the trna s primary sequence or its molecular conformation, or do viral proteins initially interact with trna-binding proteins such as the aminoacyl-trna synthetase? Could this mean that only trnas participating in protein synthesis are capable of interacting with viral proteins, and does this interaction occur at the polysome during the synthesis of Gag and Gag-Pol? The study of primer trna packaging is a study in viral assembly, and the mechanism by which a primer trna-viral protein complex is carried to the site of viral assembly is not known. In fact, it remains unclear whether selective packaging of primer trna is required for its efficient genomic placement. In MuLV, disruption of selective packaging of trna Pro in an RT-negative virion does not affect trna Pro genomic placement. On the other hand, RT-negative HIV-1 and ALV virions show neither selective incorporation nor genomic placement of primer trna, but inefficient genomic placement could be due to the absence of RT sequences rather than the absence of selective packaging. In HIV-1, altering both the PBS and the A-rich loop so that they are complementary to the 3 -terminal nucleotides and anticodon loop, respectively, of trna His results in the stable use of trna His as a primer, without evidence of an increase in its relative concentration in the virion. If selective packaging of primer trna is not required, why does this packaging occur in most retroviral and retrotransposon systems examined? Are nucleocapsid sequences required for genomic placement of primer trna? The answer appears to be yes in HIV-1 as well as in avian and murine retroviruses, but what about HFV and retrotransposons, which do not have any Cys- His box? In oncoviruses and lentiviruses, the Cys-His box is a landmark for nucleocapsid sequences, but it is the amino acids flanking these boxes, rather than the boxes themselves, which seem to be responsible for annealing primer trna to the PBS. Are there sequences near the carboxy termini of HFV or retrotranspon Gag proteins which perform the same function in these systems as nucleocapsid sequences in other retroviruses? Clearly, further studies are required to answer these questions, and the comparison of results from several different retroviral and retrotransposon systems may eventually reveal simple underlying principles behind the apparent diversity of primer trna processing and function in these systems. REFERENCES 1. Aiyar, A., D. Cobrinik, Z. Ge, H.-J. Kung, and J. Leis Interaction between retroviral U5 RNA and the T C loop of the trna Trp primer is required for efficient initiation of reverse transcription. J. Virol. 66: Aiyar, A., Z. Ge, and J. Leis A specific orientation of RNA secondary structures is required for initiation of reverse transcription. J. Virol. 68: Arts, E. J., M. Ghosh, P. S. Jacques, B. Ehresmann, and S. F. J. LeGrice Restoration of trna Lys 3 -primed ( ) strand DNA synthesis to an HIV-1 reverse transcriptase mutant with extended trnas. J. Biol. Chem. 271: Arts, E. J., S. R. Stetor, X. Li, J. W. Rausch, K. J. Howard, B. Ehresmann, T. W. North, B. M. Wohrl, R. S. Goody, M. A. Wainberg, and S. F. J. LeGrice Initiation of ( ) strand DNA synthesis from the trna Lys 3 on lentiviral RNAs: implications of specific HIV-1 RNA-tRNA Lys 3 interactions inhibiting primer utilization by retroviral reverse transcriptases. Proc. Natl. Acad. Sci. USA 93: Barat, C., S. F. J. Le Grice, and J.-L. Darlix Interaction of HIV-1 reverse transcriptase with a synthetic form of its replication primer, trna Lys 3. Nucleic Acids Res. 19: Barat, C., V. Lullien, O. Schatz, G. Keith, and J.-L. Darlix HIV-1 reverse transcriptase specifically interacts with the anticodon domain of its cognate primer trna. EMBO J. 8: Berkowitz, R., J. Fisher, and S. P. Goff RNA packaging. Morphogenesis and maturation of retroviruses. Curr. Top. Microbiol. Immunol. 214: Boeke, J. D., C. A. Styles, and G. R. Fink Saccharomyces cerevisiae SPT3 gene is required for transposition and transpositional recombination of chromosomal Ty elements. Mol. Cell. Biol. 6: Cobrinik, D., L. Soskey, and J. Leis A retroviral RNA secondary structure required for efficient initiation of reverse transcription. J. Virol. 62: Colicelli, J., and S. P. Goff Isolation of a recombinant murine leukemia virus utilizing a new primer trna. J. Virol. 57: Crawford, S., and S. P. Goff A deletion mutation in the 5 part of the pol gene of Moloney murine leukemia virus blocks proteolytic processing of the gag and pol polyproteins. J. Virol. 53: Darlix, J.-L., M. Lapadat-Tapolsky, H. de Rocquigny, and B. P. Roques First glimpses at structure-functure relationships of the nucleocapsid protein of retroviruses. J. Mol. Biol. 254:

8 8094 MINIREVIEW J. VIROL. 13. Das, A. T., and B. Berkhout Efficient extension of a misaligned trna primer during replication of the HIV-1 retrovirus. Nucleic Acids Res. 23: Das, A. T., B. Klaver, and B. Berkhout Reduced replication of human immunodeficiency virus type 1 mutants that use reverse transcription primers other than the natural trna 3 Lys. J. Virol. 69: de Rocquigny, H., D. Ficheux, C. Gabus, B. Allain, M. C. Fournie-Zaluski, J.-L. Darlix, and B. P. Roques Two short basic sequences surrounding the zinc finger of nucleocapsid protein NCp10 of Moloney murine leukemia virus are critical for RNA annealing activity. Nucleic Acids Res. 21: de Rocquigny, H., C. Gabus, A. Vincent, M.-C. Fournie-Zaluski, B. Roques, and J.-L. Darlix Viral RNA annealing activities of human immunodeficiency virus type 1 nucleocapsid protein require only peptide domains outside the zinc fingers. Proc. Natl. Acad. Sci. USA 89: Faras, A. J., and N. A. Dibble RNA-directed DNA synthesis by the DNA polymerase of Rous sarcoma virus: structural and functional identification of 4S primer RNA in uninfected cells. Proc. Natl. Acad. Sci. USA 72: Flavell, A. J Ty 1 copia group retrotransposons and the evolution of retroelements in the eukaryotes. Genetica 86: Guo, J., L. E. Henderson, J. Bess, B. Kane, and J. G. Levin Human immunodeficiency virus type 1 nucleocapsid protein promotes efficient strand transfer and specific viral DNA synthesis by inhibiting TAR-dependent self-priming from minus-strand strong-stop DNA. J. Virol. 71: Harada, F., G. G. Peters, and J. E. Dahlberg The primer trna for Moloney murine leukemia virus DNA synthesis: nucleotide sequence and aminoacylation of trna Pro. J. Biol. Chem. 254: Harada, F., R. C. Sawyer, and J. E. Dahlberg A primer RNA for initiation of in vitro Rous sarcoma virus DNA synthesis: nucleotide sequence and amino acid acceptor activity. J. Biol. Chem. 250: Huang, Y., A. Khorchid, J. Wang, M. A. Parniak, J.-L. Darlix, M. A. Wainberg, and L. Kleiman Effect of mutations in nucleocapsid protein (NCp7) upon Pr160 gag-pol and trna Lys incorporation into human immunodeficiency virus type 1. J. Virol. 71: Huang, Y., J. Mak, Q. Cao, Z. Li, M. A. Wainberg, and L. Kleiman Incorporation of excess wild-type and mutant trna 3 Lys into human immunodeficiency virus type 1. J. Virol. 68: Huang, Y., A. Shalom, Z. Li, J. Wang, J. Mak, M. A. Wainberg, and L. Kleiman Effects of modifying the trna 3 Lys anticodon on the initiation of human immunodeficiency virus type 1 reverse transcription. J. Virol. 70: Huang, Y., J. Wang, X. Li, J. L. Darlix, M. A. Wainberg, and L. Kleiman. The role of nucleocapsid and U5 stem/a-rich loop sequences in trna 3 Lys genomic placement and initiation of reverse transcription in HIV-1. Submitted for publication. 26. Huang, Y., J. Wang, A. Shalom, Z. Li, A. Khorchid, M. A. Wainberg, and L. Kleiman Primer trna 3 Lys on the viral genome exists in unextended and two-base extended forms within mature human immunodeficiency virus type 1. J. Virol. 71: Isel, C., C. Ehresmann, G. Keith, B. Ehresmann, and R. Marquet Initiation of reverse transcription of HIV-1: secondary structure of the HIV-1 RNA/tRNA 3 Lys (template/primer) complex. J. Mol. Biol. 247: Isel, C., J. Lanchy, S. F. J. LeGrice, C. Ehresmann, B. Ehresmann, and R. Marquet Specific initiation and switch to elongation of human immunodeficiency virus type 1 reverse transcription require the post-translational modifications of primer trna 3 Lys. EMBO J. 15: Isel, C., R. Marquet, G. Keith, C. Ehresmann, and B. Ehresmann Modified nucleotides of trna 3 Lys modulate primer/template loop-loop interaction in the initiation complex of HIV-1 reverse transcription. J. Biol. Chem. 268: Jiang, M., J. Mak, A. Ladha, E. Cohen, M. Klein, B. Rovinski, and L. Kleiman Identification of trnas incorporated into wild-type and mutant human immunodeficiency virus type 1. J. Virol. 67: Kang, S.-M., Z. Zhang, and C. D. Morrow Identification of a sequence within U5 required for human immunodeficiency virus type 1 to stably maintain a primer binding site complementary to trna Met. J. Virol. 71: Keeney, J. B., K. B. Chapman, V. Lauermann, D. F. Voytas, S. U. Åstrom, U. von Pawel-Rammingen, A. Byström, and J. D. Boeke Multiple molecular determinants for retrotransposition in a primer trna. Mol. Cell. Biol. 15: Keith, G., and T. Heyman Heterogeneities in vertebrate trnas Trp. Avian retroviruses package only as a primer the trna Trp lacking modified m 2 G in position 7. Nucleic Acids Res. 18: Khan, R., and D. P. Giedroc Recombinant human immunodeficiency virus type 1 nucleocapsid (NCp7) protein unwinds trna. J. Biol. Chem. 267: Kikuchi, Y., Y. Ando, and T. Shiba Unusual priming mechanism of RNA-directed DNA synthesis in copia retrovirus-like particles of Drosophila. Nature 323: Kim, A., C. Terzian, P. Santamaria, A. Pelisson, N. Prudhomme, and A. Bucheton Retroviruses in invertebrates: the gypsy retrotransposon is apparently an infectious retrovirus of Drosophila meloanogaster. Proc. Natl. Acad. Sci. USA 91: Kohlstaedt, L. A., and T. A. Steitz Reverse transcriptase of human immunodeficiency virus can use either human trna 3 Lys or Escherichia coli trna 2 Gln as a primer in an in vitro primer-utilization assay. Proc. Natl. Acad. Sci. USA 89: Lanchy, J., C. Ehresmann, S. F. J. LeGrice, B. Ehresmann, and R. Marquet Binding and kinetic properties of HIV-1 reverse transcriptase markedly differ during initiation and elongation of reverse transcription. EMBO J. 15: Lapadat-Tapolsky, M., C. Pernelle, C. Borie, and J.-L. Darlix Analysis of the nucleic acid annealing activities of nucleocapsid protein from HIV-1. Nucleic Acids Res. 23: Lauermann, V., and J. D. Boeke The primer trna sequence is not inherited during Ty1 retrotransposition. Proc. Natl. Acad. Sci. USA 91: Leis, J., A. Aiyar, and D. Cobrinik Regulation of initiation of reverse transcription of retroviruses, p In A. M. Skalka and S. P. Goff (ed.), Reverse transcriptase, vol. 1. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 42. Levin, J. G., S. C. Hu, A. Rein, L. I. Messer, and B. I. Gerwin Murine leukemia virus mutant with a frameshift in the reverse transcriptase coding region: implications for pol gene structure. J. Virol. 51: Levin, J. G., and J. G. Seidman Effect of polymerase mutations on packaging of primer trna Pro during murine leukemia virus assembly. J. Virol. 38: Levin, J. G., and J. G. Seidman Selective packaging of host trna s by murine leukemia virus particles does not require genomic RNA. J. Virol. 29: Li, X., J. Mak, E. J. Arts, Z. Gu, L. Kleiman, M. A. Wainberg, and M. A. Parniak Effects of alternations of primer-binding site sequences on human immunodeficiency virus type 1 replication. J. Virol. 68: Li, X., Y. Quan, E. J. Arts, Z. Li, B. D. Preston, H. de Rocquigny, B. P. Roques, J.-L. Darlix, L. Kleiman, M. A. Parniak, and M. A. Wainberg Human immunodeficiency virus type 1 nucleocapsid protein (NCp7) directs specific initiation of minus-strand DNA synthesis primed by human trna 3 Lys in vitro: studies of viral RNA molecules mutated in regions that flank the primer binding site. J. Virology. 70: Liang, C., X. Li, L. Rong, P. Inouye, Y. Quan, L. Kleiman, and M. A. Wainberg The importance of the A-rich loop in human immunodeficiency virus type 1 reverse transcription and infectivity. J. Virol. 71: Litvak, S., L. Sarih-Cottin, M. Fournier, M. Andreola, and L. Tarrago- Litvak Priming of HIV replication by trna 3 Lys : role of reverse transcription. Trends Biochem. Sci. 19: Lund, A. H., M. Duch, J. Lovmand, P. Jørgensen, and F. S. Pedersen Complementation of a primer binding site-impaired murine leukemia virusderived retroviral vector by a genetically engineered trna-like primer. J. Virol. 71: Lund, A. H., M. Duch, J. Lovmand, P. Jørgensen, and F. S. Pedersen Mutated primer binding sites interacting with different trnas allow efficient murine leukemia virus replication. J. Virol. 67: Maizels, N., and A. M. Weiner The genomic tag hypothesis: modern viruses as molecular fossils of ancient strategies for genomic replication, p In R. F. Gesteland and J. F. Atkins (ed.), The RNA world. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 52. Mak, J., M. Jiang, M. A. Wainberg, M.-L. Hammarskjöld, D. Rekosh, and L. Kleiman Role of Pr160 gag-pol in mediating the selective incorporation of trna Lys into human immunodeficiency virus type 1 particles. J. Virol. 68: Mak, J., A. Khorchid, Q. Cao, Y. Huang, I. Lowy, M. A. Parniak, V. R. Prasad, M. A. Wainberg, and L. Kleiman Effects of mutations in Pr160 gag-pol upon trna 3 Lys and Pr160 gag-pol incorporation into HIV-1. J. Mol. Biol. 265: Marquet, R., C. Isel, C. Ehresmann, and B. Ehresmann trna as primer of reverse transcriptases. Biochimie 77: Mély, Y., H. De Rocquigny, M. Sorinas-Jimeno, G. Keith, B. P. Roques, R. Marquet, and D. Gérard Binding of the HIV-1 nucleocapsid protein to the primer trna 3 Lys, in vitro, is essentially not specific. J. Biol. Chem. 270: Mishima, Y., and J. A. Steitz Site-specific crosslinking of 4-thiouridine-modified human trna 3 Lys to reverse transcriptase from human immunodeficiency virus type 1. EMBO J. 14: Murphy, J. E., and S. P. Goff Construction and analysis of deletion mutations in the U5 region of Moloney murine leukemia virus: effects on RNA packaging and reverse transcription. J. Virol. 63: Panet, A., and H. Berliner Binding of trna to reverse transcriptase of RNA tumor viruses. J. Virol. 26: Park, J., and C. D. Morrow The nonmyristylated Pr160 gag-pol polyprotein of human immunodeficiency virus type 1 interacts with Pr55 gag and is