Comparison of DNA Polymerase Activities Between Recombinant Feline Immunodeficiency and Leukemia Virus Reverse Transcriptases

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Virology 335 (2005) 106 – 121 www.elsevier.com/locate/yviro

Comparison of DNA polymerase activities between recombinant feline immunodeficiency and leukemia virus reverse transcriptases

Darwin J. Operario1, Holly M. Reynolds1, Baek KimT

Department of Microbiology and Immunology, School of Medicine and Dentistry, University of Rochester, 601 Elmwood Avenue, Box 672, Rochester, NY 14642, USA

Received 27 November 2004; returned to author for revision 22 December 2004; accepted 11 February 2005

Abstract

In this study, we present enzymatic differences found between recombinant RTs derived from feline leukemia virus and feline immunodeficiency virus. Firstly, FIV RT showed low steady state Km values for dNTPs compared to FeLV RT. Consistent with this, FIV RT synthesized DNA more efficiently than FeLV RT at low dNTP concentrations. We observed similar concentration-dependent activity differences between other lentiviral (HIV-1 and SIV) and non-lentiviral (MuLV and AMV) RTs. Second, FeLV RT showed less efficient misincorporation with biased dNTP pools and mismatch primer extension capabilities, compared to FIV RT. In M13mp2 lacZa forward mutation assays, FeLV RT displayed approximately 11-fold higher fidelity than FIV RT. Finally, FeLV RT was less sensitive to 3TCTP and ddATP than FIV RT. This study represents the comprehensive enzymatic characterization of RTs from a lentivirus and a non-lentivirus retrovirus from the same host species. The data presented here support enzymatic divergences seen among retroviral RTs. D 2005 Elsevier Inc. All rights reserved.

Keywords: FIV; FeLV; Reverse transcriptase

Introduction human and simian T cell leukemia viruses [HTLV and STLV, respectively], murine leukemia virus [MuLV], and Retro-elements, including retroviruses, are characterized feline leukemia virus [FeLV]). by their unique process of replicating single-stranded These two groups of retroviruses also exhibit differences positive sense genomic RNA into double-stranded DNA, in cell type-specific infection. Lentiviruses efficiently infect and have been widely identified in various host species from both non-dividing/terminally-differentiated cells (i.e., mac- yeast (e.g., Ty elements) to humans (e.g., retroviruses). rophage) and dividing cells (i.e., activated T cells), whereas Animal retroviruses have been extensively studied due to most oncoretroviruses studied thus far only productively their pathological roles in various diseases of infected hosts. replicate in dividing cells during natural infections (Gendel- Family Retroviridae is phylogenetically classified into seven man et al., 1986; Lewis and Emerman, 1994; Lewis et al., distinct groups based upon their genomic sequences (Fields 1992; McGuire and Henson, 1971). The exception to this et al., 2001). However, with the exception of non- generalization would be Avian Sarcoma Virus (ASV), which pathogenic spumaviruses, animal retroviruses have been has been shown to infect non-dividing cells (Greger et al., pathologically classified into two major groups: the immu- 2004). However, actively dividing avian cells such as nodeficiency causing lentiviruses (e.g., human, simian, and chicken embryo fibroblasts have been used for the feline immunodeficiency viruses [HIV-1, SIV, and FIV, productive viral replication and culture of this virus (Estis respectively]), and the cancer causing oncoretroviruses (e.g., and Temin, 1979). All retroviruses encode their own DNA polymerase, T Corresponding author. Fax: +1 585 473 9573. called reverse transcriptase (RT). RT catalyzes both RNA- E-mail address: [email protected] (B. Kim). and DNA-dependent DNA polymerization reactions that 1 These authors contributed equally to this work. synthesize the (À) and (+) strands of the proviral DNA,

0042-6822/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.virol.2005.02.010 D.J. Operario et al. / Virology 335 (2005) 106–121 107 respectively, from single-stranded positive sense viral RNAs 1999), all of which have been successfully purified, show during replication. The RTs from HIV-1 (a lentivirus) and higher fidelity than lentiviral RTs. The fidelity difference MuLV (an oncoretrovirus) are the most extensively charac- between lentiviral and oncoretroviral RTs implies that the terized RTs, in terms of enzymology and structure (Skalka active sites of RTs from these two major groups of and Goff, 1993). These two RTs share general enzyme retroviruses may have different molecular architectures. It activities essential for viral replication, such as DNA has also been reported that the lower capability of MuLV RT polymerase and RNase H activities (and the resulting strand to extend mismatched primers contributes to its high transfer action). They also share similar structural organ- replication fidelity (Bakhanashvili and Hizi, 1992). There izations (Georgiadis et al., 1995; Jacobo-Molina et al., 1993; are several lines of evidence that support that the fidelity of Kohlstaedt et al., 1992). Their DNA polymerase domains, retroviral RTs contributes to viral genomic mutagenesis. A consisting of the fingers, palm and thumb sub-domains, lie study by O’Neil et al. (2002) which analyzed mutations in at their N-terminal regions, while the RNase H sub-domain HIV-1 LTRs during a single round of replication using both is located at their C-terminal regions. MuLV RT (¨71 kDa) HIV-1 RT and RNA pol II, suggested that HIV-1 RT, and not is slightly larger than HIV-1 RT (¨66 kDa) mainly due to its host RNA pol II, plays a primary role in the genomic larger fingers sub-domain. In addition, MuLV RT (and mutagenesis of HIV-1. Recent reports demonstrated that Bovine Leukemia Virus [BLV] RT) functions as a monomer alterations in enzymatic fidelity of HIV-1 RT affect viral (Avidan et al., 2002; Perach and Hizi, 1999), whereas HIV-1 genomic mutagenesis during a single round of viral RT acts as a p66/p51 heterodimer (Kohlstaedt et al., 1992). replication (Mansky et al., 2003; Weiss et al., 2004). These Several enzymatic differences between HIV-1 and MuLV biochemical and virological findings suggest that replication RTs have also been observed. Firstly, HIV-1 RT appears to fidelity of retroviral RTs contributes to viral mutagenesis. have lower steady state Km values for dNTP substrates Thirdly, MuLV RT shows less sensitivity to nucleotide RT compared to MuLV RT, although these kinetic differences inhibitors (NRTIs), particularly to NRTIs containing modi- vary significantly depending on the properties and sequences fied L-form sugar moieties such as 3TCTP (Back et al., 1996; of the template-primers used in the study (Chowdhury et al., Cherrington et al., 1996). Structural studies suggest that the 1996; Diamond et al., 2003; Furge and Guengerich, 1997; Shi amino acid sequence variation of residue FX_ within the et al., 2002). This indicates that HIV-1 RT can synthesize conserved YXDD catalytic sites of HIV-1 RT (YMDD) and DNA more efficiently than MuLV RT at low dNTP MuLV RT (YVDD) contribute to their different sensitivities concentrations. In keeping with this, our recently published to 3TCTP (Rezende et al., 1998). Accordingly, clinical data indicate that HIV-1 RT has a higher binding affinity for observations show that HIV-1 isolates containing M184I dNTPs compared to MuLV RT (Skasko et al., 2005). Further and M184V RT mutations display 3TC resistance (Back et data generated by our lab have shown that the dNTP al., 1996; Wainberg et al., 1996). Additionally, a Met-to-Thr concentration in terminally differentiated/non-dividing mutation at the FX_ residue found in mutants of FIV RT was human macrophage (¨40 nM) is 50–100 times lower than shown to also confer resistance to 3TC (Smith et al., 1997). activated human CD4+ T cells (2–5 AM) (Diamond et al., The FX_ residue in the YXDD motif is also important in the 2004) . Because the ability of an RT to efficiently incorporate enzyme fidelity of retroviral RTs. Mutants of MuLV RT dNTPs in different dNTP concentration environments may containing Met or Ala at the YXDD motif, instead of the place limitations on the different types of cells that a wild-type Val, exhibit decreased replication fidelity (Kaushik particular retrovirus may infect, our combined data suggest et al., 2000b), while the Val and Ile mutations in the YXDD that the ability of HIV-1 RT to efficiently synthesize DNA at motif of HIV-1 RT enhance fidelity (Rezende et al., 1998). low dNTP concentrations may contribute to the unique Several oncoretroviruses and lentiviruses have been capability of HIV-1 to infect nondividing cells such as identified from a single host species (e.g., HTLV-1 and macrophages. HIV-1). However, it has been difficult to perform parallel Secondly, HIV-1 and MuLV RTs differ in their fidelity of biochemical characterizations of lentiviral and oncoretro- DNA polymerization. HIV-1 RT has an error rate of 2.5–5 Â viral RTs isolated from a single host species, mainly due to 10À4 mutations per nucleotide copied, a rate 15 times higher difficulties in purifying large amounts of active RT than that of MuLV RT (Bebenek et al., 1989; Preston et al., proteins from the oncoretroviruses in these hosts. A 1988; Roberts et al., 1989; Williams and Loeb, 1992). Other parallel study of this sort would be useful in determining lentiviral RTs, such as SIV RT (Diamond et al., 2001) and if the enzymatic differences found between lentiviral and Equine Infectious Anemia Virus RT (Bakhanashvili and Hizi, oncoretroviral RTs studied thus far are preserved when two 1993), are also unfaithful DNA polymerases with error rates retroviruses from different retroviral groups evolve in the similar to that of HIV-1 RT. In addition, RT proteins of same host species. Moreover, the cell type specificity of oncoretroviruses, such as those from mouse mammary tumor these retroviruses as related to their respective RTs is virus (MMTV, Taube et al., 1998a, 1998b), Moloney murine biochemically relevant. Most oncoretroviral RTs success- leukemia virus (MoMuLV, Roberts et al., 1989), avian fully purified thus far (e.g., RTs of MuLV, AMV, and myeloblastosis virus (AMV, Battula and Loeb, 1974; Roberts MMTV) have been isolated from oncoretroviruses in host et al., 1989), and BLV (Avidan et al., 2002; Perach and Hizi, species where a lentivirus has yet to be identified. While 108 D.J. Operario et al. / Virology 335 (2005) 106–121 recombinant HTLV-1 RT has been previously isolated RT, like MuLV and Friend MLV RTs, has a Val residue at (Owen et al., 1998; Trentin et al., 1998), it has not been X in the YXDD motif, whereas the analogous sequence purified in sufficient amounts to be able to compare to its found in FIV RT is similar to that of the other three HIV-1 counterpart. BLV RT has also been previously lentiviral RTs. In addition, the residues known to be isolated (Avidan et al., 2002; Perach and Hizi, 1999), but important in enzymatic fidelity of HIV-1 and MuLV RTs has not been compared with BIV RT. In this study, we are also observed in FIV and FeLV RTs (see references in have successfully purified active RTs from both FeLV (an Fig. 1A legend). oncoretrovirus, genus gammaretrovirus)andFIV(a FeLV and FIV RT expression plasmids were constructed lentivirus), retroviruses that infect the same host species. by cloning the amplified RT genes into pET28a (Novagen, Results from enzymatic comparison between FeLV and WI), giving the expressed proteins cleavable 6-His tags. FIV RTs support conserved enzymatic divergences The proteins were purified from bacterial overexpression between RTs of lentiviruses and oncoretroviruses. Possible systems using Ni2+ affinity columns (see details in virological contribution of the enzymatic differences Materials and methods). The RT proteins initially eluted between lentiviral and oncoretroviral RTs are discussed. from the Ni2+ columns had considerable contamination as shown in Fig. 1B (Ni). The eluted proteins were therefore treated with Thrombin to cleave the N-terminal 6-His tag, Results and then applied to a fresh Ni2+ column. This second column was used to remove cleaved hexahistidine tags as Cloning and purification of FeLV and FIV RTs well as contaminants that non-specifically bound to the first column. The flow-through fractions containing cleaved RT The RTs used in the work presented below are derived proteins (unbound) were collected for analysis. As shown in from FIV (strain PPR) and FeLV (strain 61E). FIV-PPR Fig. 1B (Thr), Thrombin treatment increased the purity of infects feline macrophage and peripheral blood leukocytes both FeLV and FIV RT proteins. The observed sizes of (Phillips et al., 1990). Infection with FIV causes immuno- FeLVand FIV RT proteins were similar with those of MuLV deficiency disease with a clinical course similar to that and p66 HIV-1 RT proteins, respectively (see Fig. 1B). found with HIV disease in humans, and as such has been Using this purification system, we were able to purify proposed as a small animal model for HIV/AIDS disease in approximately 0.2 mg of FeLV RT and 0.5 mg of FIV RT, humans (Bendinelli et al., 1995; Torten et al., 1991). In each from 1 L culture. Both enzymes were tested for contrast, FeLV-61E induces lymphosarcoma, and is feline T polymerase activity using poly (rA)/oligo (dT) (data not cell-tropic (Overbaugh et al., 1988). Both recombinant shown) as well as heteropolymeric RNA and DNA untagged FIV RT (North et al., 1994) and GST-tagged templates (see below), and both purified RTs were found FIV RT (Amacker et al., 1995) proteins have previously to be highly active. Purified RT proteins were also tested for been purified. While FeLV RT has been previously isolated bacterial RNase contamination by incubating the RTs with a from virions (Rho and Gallo, 1980), no recombinant FeLV 32P-labeled RNA template annealed to DNA primer in the RT has been purified thus far. absence of dNTP, and no significant degradation of RNA FeLV and FIV RT genes (2001 bp and 1677 bp, template was observed, compared to ‘‘No RT control’’ (data respectively) were amplified from molecular clones, pFIV- not shown). The successful purification of these recombi- PPR (Phillips et al., 1990; Talbott et al., 1989) and p61E nant proteins allows us to perform the first comparative (Donahue et al., 1988; Overbaugh et al., 1988), respec- characterization of RT proteins derived from a lentivirus and tively (NIH AIDS Research and Reference Reagent oncoretrovirus isolated from the same host species. Program). The 5V and 3V ends of the coding regions for both RTs were determined by aligning them against the Steady state kinetic analysis of FIV and FeLV RTs Asp residues of the conserved YXDD motif of HIV-1 and MuLV RTs (i.e., Asp185 of HIV-1 RT and Asp224 of In order to compare enzymatic properties of the purified MuLV RT). The amino acid sequences of FeLV and FIV FeLV and FIV RT proteins, we performed steady state RTs near the active site, as well as those sequences at the single nucleotide incorporation assays using an RNA N- and C-terminal ends of the proteins were compared template. For this assay, four different 5V 32P-labeled with those of other oncoretroviral (MuLV and Friend primers, which were individually annealed to a 38-mer MLV) and lentiviral (HIV-1, HIV-2 and SIVmac239) RTs, RNA template, were extended by RTs in the presence of respectively (Fig. 1A). As shown in Fig. 1A, the three each of the four individual dNTPs. Km and kcat values were catalytically essential Asp residues (each indicated as a determined for both RTs. As shown in Fig. 2A, the kcat bold FD_) known to interact with metal ions in HIV-1 RT values of both FIV and FeLV RT proteins were similar. (Huang et al., 1998; Kohlstaedt et al., 1992) and the However, the Km values of FeLV RT for dNTPs were adjacent residues are highly conserved between FeLV RT 16–58 times higher than that of FIV RT. Fig. 2B shows and the three other oncoretroviral RTs, as well as between that FeLV RT maintains its maximum dCTP incorporation FIV RT and the three other lentiviral RTs (Fig. 1A). FeLV rate above 5 AM (lane 4, FeLV RT), while FIV RT is able D.J. Operario et al. / Virology 335 (2005) 106–121 109

Fig. 1. Amino acid sequence comparison and purification of FeLV and FIV RTs. (A) Amino acid sequence comparison of four RTs of lentiviruses (FIV-PPR, HIV-1 HXB2, HIV-2 ST, and SIVmac239) and three RTs of oncoretroviruses (FeLV-61E, MuLV, and Friend MLV) at their fingers domain interacting with template and the palm domain containing three Asp residues that are catalytically important (bold) as well as both N- and C-terminal ends of the proteins. The conserved YXDD domains are underlined. Residue numbers of HIV-1 and MuLV RTs are marked at the first of last residues of each region. Residues known to be important for the fidelity of HIV-1 RT and MuLV RT are marked by ‘‘*’’: D76 (Kim et al., 1996, 1999), R78 (Kim et al., 1999), Y115 (Cases-Gonzalez et al., 2000) V148 (Diamond et al., 2003), and Q151 (Kaushik et al., 1997; Weiss et al., 2002) of HIV-1 RT, F155 (equivalent to F119 in MMTV RT) (Entin-Meer et al., 2002), Q190 (Singh et al., 2000), and V223 (Kaushik et al., 2000a) of MuLV RT. These RT sequences were obtained from Gene Bank: FIV-PPR (GenBank accession no. M36968), HIV-1 HXB2 (GenBank accession no. NC_001802), HIV-2 ST (GenBank accession no. M31113), SIVmac239 (GenBank accession no. M33262), FeLV p61E (GenBank accession no. M18247), MuLV (GenBank accession no. NC_001501) and Friend MLV (GenBank accession no. NC_ 001362). Lentiviral RT overall sequence homology was 22.6% over all four proteins, and 47.3% between HIV-1 and FIV RTs specifically. Oncoretroviral RT overall sequence homology was 74.6% between the three proteins, and 75.3% between MuLV and FeLV RTs specifically. (B) SDS-PAGE (12%) for purified RT proteins. Ni: His-tagged FeLV and FIV RT proteins eluted from Ni2+ column, Thr: RT proteins collected from a second Ni2+ column after the Thrombin treatment. Purified MuLV (M) and HIV-1 RT (H) proteins were used for the size comparison. Molecular weight markers (MW, kDa) are also shown. to retain its maximum dCTP incorporation rate at concen- incorporate dNTPs much more efficiently than FeLV RT at trations as low as 0.25 AM (lane 7, FIV RT), reflecting the low dNTP concentrations. higher Km values of FeLV RT compared to FIV RT. Interestingly, HIV-1 RT has been shown to have lower Km dNTP concentration dependent DNA polymerase activity of values for dNTPs than MuLV RT, even though this Km lentiviral and oncoretroviral RT proteins difference depends on the template/primer (T/P) and type of assay (single vs. multiple nucleotide incorporation) The steady state kinetic studies shown in Fig. 2 suggested employed in the studies (Chowdhury et al., 1996; Diamond that FIV RT utilizes dNTPs more efficiently than FeLV RT at et al., 2001, 2004; Furge and Guengerich, 1997; Shi et al., low dNTP concentrations. We tested whether differences in 2002; Ueno et al., 1995; Woodside and Guengerich, 2002). DNA synthesis efficiency at low dNTP concentrations Yet another lentiviral RT, SIV RT (SIVMNE CL8 RT), also between FIV and FeLV RTs were also true between other has low Km values (¨50 nM, Diamond et al., 2001). These lentiviral and oncoretroviral RTs. For this, we employed data suggest that FIV RT, like HIV and SIV RTs, can HIV-1 and MuLV RTs for comparison. 110 D.J. Operario et al. / Virology 335 (2005) 106–121

Fig. 2. Steady state kinetic analysis of FeLV and FIV RT proteins. (A) Steady state kinetic parameters of FeLV and FIV RTs in incorporation of dNTPs. Fold differences of Km values between FIV and FeLV RTs were shown in parentheses. These kinetic values were obtained from three independent experiments. Standard deviations were <25%. (B) dCTP incorporation by FeLV and FIV RTs. 32P-labeled 17-mer Extend-C primer annealed to a 38-mer RNA template was extended by 10 mM RT proteins for 5 min with decreasing concentrations of dCTP (50, 25, 12.5, 5, 1, 0.5, 0.25, 0.125, 0.06, and 0.03 AM for FeLV RT or 12.5, 5, 1, 0.5, 0.25, 0.125, 0.06, 0.03, 0.015, and 0.008 AM for FIV RT). Reactions were analyzed on a 14% denaturing polyacrylamide-urea gel. E: extended primer, S: unextended primer.

We performed primer extension assays using a 32P-labeled protein that yielded full extension of approximately 50% of 17-mer A-primer (See Materials and Methods) annealed to a primer (see ‘‘F’’ in Fig. 3) with 250 AM dNTPs (for each 38-mer RNA template. Primer extension by the RT proteins dNTP) at 37 -C for 5 min. Note that 250 AM dNTP is much was assayed in the presence of all four dNTPs at four different higher than the Km values of both oncoretroviral RTs and concentrations. For this, we first determined the amount of lentiviral RTs used in this study (Fig. 2A). Next, the primer

Fig. 3. dNTP concentration dependence of primer extension two lentiviral and two oncoretroviral RT proteins. The 32P-labeled 17-mer A-primer bound to the 38-mer RNA template was extended by an equivalent activity of two lentiviral RTs (A), FIV and HIV-1 RTs, and two oncoretroviral RTs (B), FeLV and MuLV RTs, showing approximately 50% primer extension with 25 AM dNTPs (10 nM HIV-1 RT, 40 nM FIV RT, 5 nM MuLV RT, and 60 nM FeLV RT, see the first lane in each RT reaction) at 37 -C for 5 min. The reactions were repeated with three different decreasing concentrations of dNTPs (2.5 AM, 0.25 AM, and 0.05 AM). Reaction conditions and RT concentrations used were described in Materials and methods. Reactions were analyzed on a 14% denaturing polyacrylamide gel. F: fully extended products, S: unextended primer. D.J. Operario et al. / Virology 335 (2005) 106–121 111 extension reactions were performed at decreasing dNTP complementary dNTP is withheld (‘‘stop site’’; indicated concentrations, 2, 0.2, and 0.05 AM, with the same amount of by asterisks ‘‘*’’ in Fig. 4). A higher efficiency of primer RNA-dependent DNA polymerization (RDDP) activity of extension beyond the stop site reflects higher ability to RT proteins used with the 25 AM dNTP reactions. As utilize incorrect dNTPs, or lower fidelity of the RT. Under observed in Fig. 3, the oncoretroviral RTs showed dramatic our reaction conditions, the small amount of potential decreases in polymerization activity in reactions containing contamination by correct dNTP does not significantly low dNTP concentrations, whereas the lentiviral RTs main- contribute to the observed results, as previously shown tained significant RDDP activity (Fig. 3A) even at the low (Diamond et al., 2003). Initially, we determined RT dNTP concentrations (0.2 and 0.05 AM). Our recent studies concentrations that gave approximately 20% (1Â) and also demonstrated that SIVMNE CL8 RT, but not AMV RT, 80% (4Â) of primer extension, as estimated from the showed efficient DNA synthesis at low dNTP concentrations amount of fully extended primer (‘‘F’’ in Fig. 4) following (Diamond et al., 2004) which is consistent with the low Km a 5-min incubation at 37 -C in the presence of all four value of SIVMNE CL8 RT for dNTPs (Diamond et al., 2003). dNTPs (250 AM). This dNTP concentration is well above Therefore, the data presented here with FIVand FeLV RTs are the Km values for all RT proteins tested, therefore allowing consistent with the idea that lentiviral RTs efficiently reactions to proceed at dNTP saturation. When incubated synthesize DNA at low dNTP concentrations compared to with mixtures of only three dNTPs (Fig. 4: minus dGTP or oncoretroviral RTs. minus dCTP), FIV and HIV-1 RTs catalyzed substantial extension past the stop sites on RNA templates (*), Misincorporation assay of FIV and FeLV RTs indicating their low fidelity. However, under the same reaction conditions, FeLV RT behaved similar to MuLV Next, we examined the fidelity of FIV and FeLV RT RT, catalyzing less efficient extension beyond the stop sites proteins using the misincorporation assay. The low-fidelity than the two lentiviral RT proteins (Fig. 4), indicating HIV-1 RT and high-fidelity MuLV RT were also included higher replication fidelity. In addition, FeLV and MuLV for comparison. The misincorporation assay is a primer RTs are similar in their misincorporation fidelities, as extension assay which qualitatively monitors both mis- judged by the amount of the primer extension beyond the insertion and mismatch extension in the presence of biased first stop site. Similar differences were observed in minus dNTP pools containing only three of the four natural dATP and minus dTTP reactions as well as in reactions dNTPs. Assays were performed using 32P-labeled 17-mer with DNA templates (data not shown). We previously A-primer annealed to the 38-mer RNA or DNA template. showed that SIV RT (CL8, Diamond et al., 2001, but not In the absence of one of the dNTPs, primer extension stalls AMV RT Kim et al., 1999), showed efficient primer one base before the template nucleotide for which the extension beyond the stop sites in similar assays.

Fig. 4. Misincorporation assay. The 32P-labeled 17-mer primer (‘‘S’’: A-primer) annealed to 38-mer RNA template was extended by RTs of two lentiviral RTs (A), FIV and HIV-1 RTs, and two oncoretroviral RTs (B), FeLV and MuLV RTs, at 37 -C for 5 min. The extension reactions were performed in the presence of all 4, or only 3 complementary dNTPs (minus dGTP or minus dCTP). The two same RT activities (see Materials and methods) of each RT protein showing approximately 20 (1Â) and 80 (4Â) % of the primer extension with all 4 dNTPs were used in this assay. The reactions were analyzed by 14% polyacrylamide- urea denaturing gel. The sequence of the extended part of the primer is shown. F: fully extended 38-nucleotide-long products, S: unextended primer, ‘‘*’’: stop sites. The sequence of the first 4 nucleotides in the extended part of the primer is shown. 112 D.J. Operario et al. / Virology 335 (2005) 106–121

Extension of mismatched primers mismatched primers (‘‘MP’’ in Fig. 5) were extended with the same amount of DNA polymerase activities (4Â and Creation of a mutation during DNA synthesis involves 1Â) of FIV, FeLV, HIV-1, and MuLV RTs as used in the two sequential steps: (1) incorporation of an incorrect misincorporation assay (see ‘‘All dNTPs’’ in Fig. 4) for 10 dNTP, or misinsertion, and (2) extension past the min. Extension of these mismatched primers in the mismatch generated by the misinsertion. In the misincor- presence of all four dNTPs generated the 38 nucleotide poration assay shown in Fig. 4, extension of the primer long product (FF_ in Fig. 5). As shown in Figs. 5A and B, beyond the stop site measures the capability of the RT the two oncoretroviral RTs showed less efficient extension proteins to both misinsert and to extend the mismatched from both G/T (Fig. 5A) and C/A (Fig. 5B) mismatched primer. Next, we examined the capability of FIV and primers than FIV and HIV-1 RTs. These data suggest that, FeLV RT proteins to perform only the second step of like MuLV RT, FeLV RT has higher mismatch extension mutation synthesis, the extension of a mismatched primer fidelity than FIV RT and HIV-1 RT. terminus. In this assay, we used two different mismatched primers annealed to either RNA or DNA template. One M13mp2 lacZa forward mutation assay T/P contains a G/T mismatch at the 3V end of the primer annealed to an RNA template. The other T/P contains a To verify the fidelity differences observed between C/A mismatch at the 3V end of the primer annealed to a FeLV and FIV RTs in the gel based assays, we measured DNA template. These G/T (1st DNA strand synthesis) and fidelity of FeLV and FIV RTs using the M13mp2 lacZa C/A (2nd DNA strand synthesis) mismatched T/Ps forward mutation assay (Bebenek and Kunkel, 1995). represent replication intermediates encountered during the Three independent gap-filling reactions of FIV and FeLV process of G to A mutation, the predominant lentiviral RTs using the M13 lacZa gapped DNA substrate were mutation (Vartanian et al., 1991). While the G to A performed, each over a 20-min reaction time. The mutant transitions that arise during HIV-1 replication can occur by frequency of each RT was determined as the ratio of the actions of APOBEC3G (Gu and Sundquist, 2003; mutant (pale blue or clear) plaques to total plaques Sheehy et al., 2002; Vartanian et al., 2003), the counted. As shown in Table 1, FeLV RT displayed contribution of misinsertion/misincorporation by HIV-1 approximately 11-fold lower mutant frequency than FIV RT to this type of mutation cannot be ruled out. We tested RT. In fact, the mutation rate of FIV RT is comparable to whether FeLV RT (and MuLV RT) has a reduced that previously observed with HIV-1 RT (Preston et al., capability to extend these two mismatched T/Ps, compared 1988). Previous studies have shown that MuLV RT has 15- with FIV RT (and HIV-1 RT). The G/T and C/A fold higher fidelity than HIV-1 RT in the same M13mp2

Fig. 5. Extension of two mismatched primers by FIV and FeLV RT proteins. The 32P labeled mismatched G/T 16-mer (A) and C/A 19-mer (B) primers (‘‘MP’’) annealed to 38-mer RNA and DNA templates, respectively, were extended by the same two quantities (4Â and 1Â) of RT proteins used in Fig. 4 at 37 -C for 10 min. The extension reactions were performed in the presence of all 4 dNTPs. Reaction conditions and RT concentrations used are described in Materials and methods. The reactions were analyzed by 14% polyacrylamide-urea denaturing gel. F: fully extended 38-nucleotide-long products, MP: unextended mismatched primer. D.J. Operario et al. / Virology 335 (2005) 106–121 113

Table 1 Sensitivity of FeLV and FIV RTs to 3TCTP, AZTTP, and Mutant frequency of FeLV and FIV RT proteins determined by M13mp2 ddATP lacZa forward mutation assay RT Total Mutant Mutant Fold It has been demonstrated that RTs that contain the Val proteins plaquesa plaques frequency difference residue in the conserved YXDD sequence (e.g., MuLV RT) FIV RT 14,220 134 0.00942 X11 display inefficient incorporation of 3TCTP, compared to FeLV RT 15,820 14 0.00088 X1 a RTs that contain the Met residue (e.g., HIV-1 RT) (Gao et Numbers of plaques obtained from three independent gap-filling al., 2000; Rezende et al., 1998). This inefficient 3TCTP reactions. incorporation is also observed in 3TC-resistant HIV-1 RT variants that contain the Val residue in the YXDD sequence lacZa forward mutation assay (Roberts et al., 1989). (Gao et al., 2000; Miller et al., 2002; Sarafianos et al., Therefore, together with the results from the misincorpora- 1999). Since RTs of FeLV and FIV contain Val and Met tion assay (Fig. 4) and mismatch extension assay (Fig. 5), residues, respectively, we compared the sensitivity of FeLV these data demonstrate that FeLV RT has higher fidelity and FIV RTs to 3TCTP, using the primer extension reaction than FIV RT. that is similar with the assay previously used (Klarmann et al., 2000). We also compared the AZTTP and ddATP Mg2+ vs. Mn2+ usage by RT proteins sensitivity of FeLV and FIV RTs. For this, the 17-mer A- primer annealed to the 38-mer RNA template was extended Previous reports have also indicated that HIV-1 RT by RT proteins showing approximately 80% of primer performs optimally in the presence of Mg2+ while Moloney extension with 10 AM dNTPs (see 4Â in Fig. 4). The Murine Sarcoma-Leukemia Virus RT has optimal perform- reactions were then repeated in the presence of mixtures of ance with Mn2+ (Gerard and Grandgenett, 1975; Valverde- 10 AM dNTPs and various concentrations of 3TCTP, Garduno et al., 1998). Using poly(rA)/oligo(dT) as a AZTTP and ddATP. As shown in Fig. 7A, FeLV RT substrate, we tested the activity of FIV and FeLV RTs in showed low sensitivity to 3TCTP, compared to FIV RT. We the presence of two different concentrations (5 and 0.05 see that FIV RT was inhibited by 3TCTP even at the lower mM) of these divalent cations, using HIV and MuLV RTs concentrations of 3TCTP tested. We calculated an IC50 for for comparison. HIV-1 RT and FIV RT show higher activity 3TCTP of approximately 150 AM for FIV RT, consistent with 5 mM Mg2+, but can effectively use Mn2+ at 0.05 mM. with a previous report by Smith et al. (1997). Because FeLV In contrast, MuLV and FeLV RTs show significantly higher RT showed a very low sensitivity to 3TCTP, we were unable 2+ activities in the presence of 0.05 mM Mn , but does not to determine an IC50 even with the highest 3TCTP 2+ efficiently use Mg in comparison (Fig. 6). concentration (400 AM, Fig. 7A). However, the IC50 of

Fig. 6. Mg2+ and Mn2+ dependence of RT Activity. RT proteins were tested for their preference of metal in RT polymerization activity by using a homopolymeric template/primer complex as a substrate in the presence of either 5 mM or 0.05 mM divalent cation (see Materials and methods). The data are presented such that the total cpm values incorporated in the presence of the preferred metal ion (and preferred concentration) is set to 1, with the relative percent activities in the presence of the metal conditions. Each experiment was performed in duplicate. 114 D.J. Operario et al. / Virology 335 (2005) 106–121

AM and 60 AM for FIV RT and FeLV RT, respectively (Fig. 7B). ddATP also effectively inhibited FIV RT. An 80% inhibition was observed even at the lowest concentration tested. We calculated an IC50 of approximately 15 AM for ddATP against FIV RT. While not being inhibited as dramatically as FIV RT, FeLV RT was able to be inhibited by ddATP at the high AM range, with an IC50 of approximately 280 AM(Fig. 7C).

Discussion

This study describes the characterization of active recombinant RT proteins from an oncoretrovirus (FeLV) and lentivirus (FIV) isolated from a single host species. RT proteins from FeLV and FIV were produced from bacterial overexpression systems and further purified by reverse Ni2+ column chromatography after the removal of the hexahistidine tag from both RT proteins. Both purified RT proteins were highly active as shown by their steady state kcat values (Fig. 2A), and by primer extension reactions (see Fig. 4, ‘‘All dNTPs’’).We investigated the enzymatic differences between FeLVand FIV RT proteins and compared them with retroviral RTs previously characterized (e.g., RTs of HIV-1, SIV, MuLV and AMV). This study therefore presents the first extensive biochemical comparison of oncoretroviral and lentiviral RTs isolated from a single host species. RTs of FeLV and FIV displayed clear differences in their steady state kinetics. As shown in Fig. 2A, FeLVand FIV RTs showed similar kcat values, indicating that the enzyme preparations were equally active. These values are also similar with those of MuLV and HIV-1 RTs previously reported (Preston et al., 1988; Roberts et al., 1989). However, FeLV RT showed higher Km values than FIV RT for each of the four dNTPs (Fig. 2A). Differences in Km values for the incoming dNTPs have also been observed between MuLV and HIV-1 RTs as well as in SIV RT in previous kinetic studies (Chowdhury et al., 1996; Diamond et al., 2001; Furge and Guengerich, 1997). Importantly, the lower Km values of FIV RT (and HIV-1 RT) indicate that FIV RT efficiently incorporates dNTP at low dNTP concentrations, compared to FeLV RT (and MuLV RT). A previous study that used Fig. 7. 3TCTP, AZTTP and ddATP sensitivity of FeLV and FIV RTs. The multiple nucleotide incorporation assays (North et al., 1994) 32P-labeled 17-mer A-primer annealed to the 38-mer RNA template was extended by FIV (closed square) and FeLV (open circle) RT proteins reported higher Km values of FIV RT than those observed in showing approximately 75% of the primer extension (see Materials and this study. The kinetic values of DNA polymerases also methods) with dNTPs (10 AM each, no drug) at 37 -C for 5 min. The widely vary depending on the type of assay (i.e., single or extension reactions were repeated in the presence of four dNTPs (10 AM multiple nucleotide incorporation) and nature of substrates each) and increasing concentrations of 3TCTP (A), AZTTP (B) and ddATP used (i.e., templates and template sequences). In addition, the (C). The reactions were analyzed by 14% polyacrylamide-urea denaturing gel. Percent of fully extended products at different concentrations of FIV strain used in this earlier study is different from the strain inhibitors were plotted to determine the drug concentrations showing 50% (FIV-PPR) used in this study. Therefore, to verify differences inhibition (IC50). in the DNA synthesis rate difference between FeLV and FIV RTs at low dNTP concentrations, we examined the multi- round primer extension capabilities of FeLV and FIV RTs at FeLV RT to 3TCTP appears to be at least 10 times higher various dNTP concentrations using a 38-mer heteropoly- than that of FIV RT. Unlike 3TCTP, these two enzymes meric RNA template, using HIV-1, SIVMNE CL8, MuLV, showed similar sensitivity to AZTTP, showing IC50sof42 and AMV RTs for comparison. Indeed, we found that the D.J. Operario et al. / Virology 335 (2005) 106–121 115 lentiviral RTs executed DNA synthesis much more efficiently supported by our recent observation that HIV-1 variants than the oncoretroviral RTs at low dNTP concentrations (i.e., harboring HIV-1 RT mutants which kinetically mimic 0.05 AM, Fig. 3), consistent with their Km differences oncoretroviral RTs in having reduced DNA polymerization observed in Fig. 2. The lowest dNTP concentrations used capability in low dNTP concentration environments, failed to in this experiment, 0.2 and 0.05 AM, are much lower than the infect macrophage even though these mutant HIV-1s Km values of the oncoretroviral RTs, but near the Km values of successfully infected dividing cells such as activated CD4+ the lentiviral RTs. Further experiments employing pre-steady T cells and transformed cell lines containing high dNTP state assays will be necessary to further elucidate mechanistic concentrations (Diamond et al., 2004). elements that kinetically differentiate FIV and FeLV RTs. In three different assays presented in this work, FeLV RT Contributions of the dNTP utilization efficiency differ- showed higher fidelity than FIV RT. The M13mp2 lacZa ence between RTs of these two retrovirus groups on viral forward mutation assay showed that FeLV RT has an 11-fold characteristics have not been clearly discussed. One viro- higher fidelity than FIV RT. In previous studies, RTs of MuLV logical difference between lentiviruses and oncoretroviruses RT and AMV RT showed 12- to 15-fold higher fidelity than is their cell type specificity. Oncoretroviruses productively HIV-1 RT (Roberts et al., 1989). We previously reported that replicate primarily in actively dividing cells (e.g., activated T SIVCL8 RT also has low fidelity (Diamond et al., 2001). This cells), while lentiviruses efficiently infect both dividing cells combined evidence supports previous data indicating that and non-dividing/terminally-differentiated cells (e.g., macro- lentiviral RTs have a lower fidelity than non-lentiviral RTs phage) (Gendelman et al., 1986; McGuire and Henson, studied thus far. It is becoming increasingly apparent that the 1971). For example, MuLV infection requires cell prolifer- fidelity of RTs plays important roles in viral genomic ation (Lewis and Emerman, 1994; Lewis et al., 1992) and mutagenesis. Recent studies using pseudotyped viruses AMV infects dividing macrophages (Durban and Boettiger, demonstrated that RT, but not host RNA polymerase II, is 1981a, 1981b). Recent studies employing Avian Sarcoma responsible for viral genetic mutagenesis (O’Neil et al., Virus (alpharetrovirus) have shown that ASV is capable of 2002). It was also recently reported that pseudotyped HIV-1 mounting an infection of non-dividing cells (Greger et al., containing various RT mutants with high fidelity drove down 2004; Hatziioannou and Goff, 2001; Katz et al., 2002). viral mutation rates during a single round of infection However, these studies employ either a modified version of (Mansky et al., 2003; Weiss et al., 2004). Furthermore, the virus, or use vector systems based on ASV, and not the similar pseudotyped systems containing HTLV showed lower naturally occurring virus. Moreover, most of the studies mutation rates than HIV-1 (Mansky, 2000). We predict that conducted use murine neurons, HeLa, or human fibrosar- differences in the fidelity of the FeLVand FIV RTs observed coma cell lines, none of which are the natural hosts of ASV.In in this study very likely reflects different mutation rates addition, productive replication of this virus requires dividing between these two retroviruses. Further kinetic studies such chicken embryo fibroblast cells (Estis and Temin, 1979). as pre-steady state kinetic analysis will elucidate mechanisms Numerous studies have demonstrated that non-dividing of the fidelity difference between these enzymes. cells have lower dNTP concentrations than dividing cells In looking at the divalent cation usage of our recombi- (Traut, 1994). Recently, we determined that the dNTP nant proteins, we see that the FIV and FeLV RTs show concentration of human macrophage to be approximately similar patterns of metal dependence compared to HIV-1 40 nM, which is roughly 50–100 times lower than that of and MuLV RTs, respectively. The similar requirement of activated T cells (2–5 AM) (Diamond et al., 2004). FIV RT for Mg2+ to HIV-1 RT is consistent with a report by 2+ Interestingly, the Km values for dNTPs of lentiviral and North et al. (1994), while the preference for Mn by FeLV oncoretroviral RTs studied here are near the dNTP concen- is consistent with a previous report by Rho and Gallo tration in macrophage and activated T cells, respectively. (1980). The results presented here are also consistent with Therefore, kinetic comparisons of RTs from these two older reports studying metal usage by RTs (Gerard and retroviral groups (Fig. 2) support the idea that RTs of Grandgenett, 1975; Valverde-Garduno et al., 1998). retroviruses may have evolved to efficiently execute DNA It was previously found that 3TCTP is not an efficient synthesis in the specific host cells that support productive inhibitor for MuLV RT (Cherrington et al., 1996), whereas infection. It is likely that the dNTP concentration level in wild type HIV-1 RT is efficiently inhibited by 3TCTP feline macrophage, a known target for several strains of FIV (Boyer et al., 2002). This 3TCTP sensitivity difference can (Dean et al., 1999), is well below the dNTP concentration that be attributed to the sequence difference at the X residue in can be found in activated feline Tcells although this has yet to the conserved YXDD region (Rezende et al., 1998). RTs of be confirmed. The ability of FIV RT to synthesize DNA at MuLV and HIV-1 have Val and Met residues at their YXDD low dNTP concentrations would certainly be a factor in the sequence, respectively (See Fig. 1A). This possibility was ability of FIV to replicate in macrophage. In contrast, the further supported by the clinical finding that 3TC-resistant higher Km values of the RT of the FeLV 61E strain and the T HIV-1 variants contain the M184V (or M184I) RT mutation cell tropism of this virus imply that the active site of the 61E in the YMDD motif (Back et al., 1996). Structural analysis FeLV RT may have adapted to the high dNTP concentration demonstrates that the Val mutation to the YMDD sequence environment of T cells. This possibility can be further of HIV-1 RT prevents entry of dNTP analogs containing a 116 D.J. Operario et al. / Virology 335 (2005) 106–121 structurally altered sugar moiety (e.g., 3TCTP) to the active effective inhibitor of FIV viral replication to both wild-type site (Gao et al., 1993; Sarafianos et al., 1999). Our results and AZT-resistant strains of FIV. clearly demonstrate that FeLV RT is less sensitive to 3TCTP The data presented in this study with purified FIV and than FIV RT. The presence of Val and Met at their YXDD FeLV RT proteins contribute to validating the enzymatic catalytic sites of FeLV and FIV RTs (Fig. 1A), respectively, differences found thus far between lentiviral and oncoretro- very likely contributes to the difference in inhibitory effects viral RTs currently studied in the literature. These enzymatic of 3TCTP on the DNA polymerase activity of these two RT differences observed between these two groups of retro- proteins. A known FIV RT Met-to-Thr mutation at the viruses appear to be related with their virological differences. analogous position in the YXDD motif confers 3TC More specifically, enzymatic differences between lentiviral resistance (Smith et al., 1997). However, viral sensitivity and non-lentiviral RTs in the Km values, fidelity and drug of FeLV and FIV to 3TC has not been directly compared. In sensitivity may correlate to virological differences in cell type contrast to 3TCTP, we observed that both FIV and FeLV specificity, genomic mutation rate and sensitivity to nucleo- RTs are sensitive to AZTTP even though FeLV RT is side RT inhibitors. Therefore, this observation suggests that slightly less sensitive than FIV RT (Fig. 7B). It was RTs of retroviruses may have evolved to facilitate viral previously demonstrated that MuLV RT is less sensitive to replication within the specific cellular environments that AZTTP than HIV-1 RT (Quan et al., 1998). AZT and 3TC these two groups of retroviruses encounter upon infection. are drugs used in anti-FIV therapy. AZT in particular has Indeed, characterization of more RT proteins from various been studied both in monotherapy and in combination with retroviruses and relating the biochemical findings to viro- other drugs for the treatment of FIV infection both in vivo logical properties of each retrovirus (i.e., cell type specificity and in vitro (Arai et al., 2002; Bisset et al., 2002; North et and drug sensitivity) will shed light on unique evolutionary al., 1989). In vitro studies have shown that AZT/3TC and virological roles of RT. treatment, in combination with abacavir, is effective in preventing FIV replication, similar to what has been seen in HIV therapy (Bisset et al., 2002). A recent study has shown Materials and methods that AZT/3TC combination therapy is a good prophylactic for preventing infection, and that the treatment can either Strains and plasmids prevent or delay seroconversion when administered near the time of infection. The same report shows that the therapy Escherichia coli NM522 (Stratagene, CA) was used can prevent replication of FIV in primary PBMC (Arai et during the construction of plasmids; E. coli BL21 (DE3) al., 2002). In contrast, the treatment of the FeLV infection pLysS (Novagen, WI) was used for over-expression of RT with 3TC has not been reported. However, AZT has been proteins. Hexahistidine-tagged p66/p66 HIV-1 RT (Weiss et used for treating FeLV infection. Inhibition of FeLV al., 2002), p66/p66 SIVMNE CL8 RT (SIV RT (Diamond et replication by AZT was observed in in vitro studies (Mathes al., 2001)), and MuLV RT proteins (Malboeuf et al., 2001) et al., 1992). For in vivo treatments, proper dosage of AZT were purified as described previously. AMV RT was can prevent antigenemia, but not seroconversion, indicating purchased from Stratagene (CA). pFIV-PPR (Phillips et that the treatment is not effective for preventing primary al., 1990; Talbott et al., 1989), which encodes the full-length infection. This seems to be true for both prophylactic genome of FIV-PPR, was obtained from Dr. John Elder treatment as well as AZT treatment post-infection (Mathes through the NIH AIDS Research and Reference Reagent et al., 1992). Program, Division of AIDS, NIAID, NIH. p61E (Donahue In addition, replication inhibition of both FIV and FeLV et al., 1988; Overbaugh et al., 1988), which encodes the full- by ddI was observed in in vitro studies (Meldin et al., 1996; length genome of FeLV-61E, was similarly obtained from Mukherji et al., 1994). The IC50 values of FIV (FIV-PPR) Dr. James Mullins. pET28a (Novagen) was used for over- and FeLV (61E), which are determined using the same expression of FIV and FeLV RT proteins in BL21 (DE3) conditions, are not currently available. This in vitro culture pLysS. DNA oligomers used in this study were obtained data can be used to address the enzymatic inhibition profile from Integrated DNA Technologies (IA). differences observed in this study. In fact, the biochemical IC50 values determined for any inhibitor may be different Construction of FIV and FeLV RT expression plasmids when employing different assays (i.e., single nucleotide addition vs. multi-nucleotide addition), different templates, The N-terminal (N-T) and C-terminal (C-T) sequences of and may also be dependent on the neighboring sequence FIV and FeLV RT proteins were determined by aligning variation (i.e., secondary structure) of the template. Our data positions of the conserved YXDD sequences of HIV-1 and also show that ddATP is an effective inhibitor of FIV RT. MuLV RTs (i.e., Asp185 of HIV-1 RT and Asp224 of MuLV This is consistent with a previous report that showed that RT), respectively, using Jellyfish 1.5 (LabVelocity, CA) to ddATP was an effective inhibitor of FIV RT activity when determine both alignment and sequence similarity. The RT using AX-174 ssDNA as a template (Cronn et al., 1992). proteins were not cloned according to the precise cleavage The same report demonstrates that, in vitro, ddA is an site within the pol polyprotein because this information has D.J. Operario et al. / Virology 335 (2005) 106–121 117 thus far not been published. The FIV RT gene encoding 560 incubation at 95 -C for 3 min. Denatured products were amino acid residues was amplified from pFIV-PPR using N- analyzed by electrophoresis on 14% polyacrylamide-urea Tprimer(5V-AAAAAAAAACATATGCAAATTTCTGA- denaturing gel. TAAGATTCC-3V) and C-T primer (5V-AAAAAAGAGCTC- FIV RT p66/p66 dimer was used in this study, and a GAGTTACCCTTCTATTATCATCATTG-3V). The FeLV RT previous study of FIV RT shows that the p66/p66 gene, encoding 668 amino acid residues, was amplified homodimer does shows comparable activity to the p51/p66 from p61E using N-T primer (5V-AAAAAAAAACATAT- heterodimer (Amacker et al., 1995). GACTCTACAATTAGAAGAG-3V)andC-Tprimer(5V- AAAAAAAAAAAGCTTTTATAAGATGGTTAGTGAT- Primer extension assay with DNA or RNA templates and GATTG-3V). The amplified RT genes were inserted into Misincorporation assay pET28a using NdeI and XhoI (FIV RT) or NdeI and HindIII (FeLV RT), generating pFIVRT and pFeLVRT, respectively. Procedures were modified from those of Preston et al. The cloned RT genes in pET28a are fused to sequences (1988). The template-primer was prepared by annealing a coding for a hexahistidine tag and a Thrombin cleavage site 38mer RNA template (5V-GCUUGGCUGCAGAA- at their N-T ends. The N-T 1.2 kb regions of these cloned UAUUGCUAGCGGGAAUUCGGCGCG-3V,Dharmacon RT genes were sequenced to confirm each RT clone. Research, CO) or 38mer DNA template (5V-GCTTGG CTGCAGAATATTGCTAGCGGGAATTCGGCGC G-3V) Purification of RT proteins to a 17-mer DNA (5V-CGCGCCGAATTCCCGCT-3V,A- primer or 5V-GCCGAATTCCCGCTAGCAATATT-3V The hexahistidine-tagged FIV and FeLV RT proteins Extend-C primer) 32P-labeled at the 5V end using were purified using protocols, reagents, and buffers pro- [g-32P]ATP and T4 polynucleotide kinase (template/primer, vided by the manufacturer (Novagen). The expression of 2.5:1). Reaction mixtures (20 Al) contained 10 nM FIV and FeLV RT proteins was induced at an OD600 = 0.1 template/primer, RT proteins showing the same DNA with 1 mM IPTG for 3 h at 37 -C in BL21 (DE3) pLysS polymerase activities indicated below, 3 or 4 dNTPs harboring pFIVRT or pFeLVRT. The harvested cell pellets (250 AM each dNTP) in multiple extension assays or only were resuspended in 1Â Binding Buffer, and kept at À70 -C dCTP in single nucleotide extensions (see Fig. 2B legend overnight. The frozen cell pellets were thawed on ice for 2 for concentrations), 25 mM Tris–HCl (pH 8.0), 100 mM h, then centrifuged for 15 min (12,000 Â g at 4 -C). RT KCl, 2 mM DTT, 5 mM MgCl2,10AM (dT)20, 0.1 mg/ml proteins were purified from the cell lysates using Ni2+ bovine serum albumin. Approximately 50% of the primer chelation chromatography (Novagen) as previously was fully extended by 40 nM FIV RT and 60 nM FeLV described (Kim, 1997). Purified proteins were dialyzed RT under this reaction condition. Reactions were incubated against 500 mL 1Â Dialysis Buffer (50 mM Tris–HCl pH at 37 -C for 5 min and terminated by 4 Alof40mM 7.5, 1 mM EDTA, 200 mM NaCl, 10% glycerol) at 4 -C for EDTA, 99% formamide. Under this reaction condition the 12 h. The dialyzed proteins were treated with biotinylated level of the fully extended primer was linear to RT Thrombin under the conditions recommended from the concentrations used (i.e., as used in Fig. 3; concentrations manufacturer (Novagen). After removing the biotinylated designated 1Â and 4Â in Figs. 4 and 5: HIV-1 RT 16 nM Thrombin from the digested RT solutions by Stravidine and 4 nM, FIV RT: 64 nM and 16 nM, MuLV RT: 8 nM columns (Novagen), the RT proteins were re-applied to Ni2+ and 2 nM, FeLV RT: 96 nM and 24 nM.). Reaction chelation chromatography, and the flow-through fractions products were immediately denatured by incubating at 95 containing digested RT proteins were pooled for dialysis -C for 3 min and analyzed by electrophoresis on 14% against 1Â Dialysis Buffer with DTT (1 mM) for an polyacrylamide-urea denaturing gels. additional 12 h at 4 -C. Amounts of purified proteins were estimated in the Bradford assay (BioRad, CA) with a bovine Extension of mismatched primers serum albumin standard; most preparations had >90% purity, estimated by visual inspection of Coomassie Blue- To measure the capability of each RT to extend stained SDS-polyacrylamide gels. With this protocol, we mismatched primer, we used two different mismatched were able to purify 0.2–0.5 mg of RT proteins from each 1 primers, G/T and C/A. G/T mismatched primer (5V- L culture. Purified proteins were also assayed for RNase CGCGCCGAATTCCCGT-3V) was annealed to the 38-mer contamination. Reaction mixtures contained purified RT RNA template (see above) with the G/T mismatch at the 3V (FIV RT: 64 nM, FeLV RT: 96 nM), 10 nM template/primer end nucleotide underlined. C/A mismatched primer (5V- (see Primer extension assay below), 25 mM Tris–HCl (pH CGCGCCGAATTCCCGCTAA-3V) was annealed to the 38- 8.0), 100 mM KCl, 2 mM DTT, 5 mM MgCl2,10AM mer DNA template (see above) with the C/A mismatch at (dT)20, 0.1 mg/ml bovine serum albumin, and no dNTP. the 3V end nucleotide underlined. Extension conditions for Reactions were carried out at 37 -C for 5 min, and then these mismatched primers were the same as in the terminated with the addition of 4 Al of 40 mM EDTA, 99% misincorporation assay described above except 10 min formamide. Reaction products were then denatured by incubation and T/Ps. Equal RT activities of RT proteins 118 D.J. Operario et al. / Virology 335 (2005) 106–121 determined by the extension reactions with the matched A- plaques. Five mutant phages in each RT assay were primer (see above) were used for the extension of the two G/ isolated and the gap regions in the mutant phage DNAs T and C/A mismatched primers. The reactions were were sequenced. All mutants analyzed had a mutation in incubated at 37 -C for 10 min. The reactions were also the gapped region. analyzed by 14% denaturing gel. Comparison of RT activities in the presence of either Mg2+ Steady state kinetic analysis of RT proteins or Mn2+

The steady state kinetic assay protocol described by A test of preference for divalent cation was performed Boosalis et al. (1987) was modified to determine DNA using a template-primer complex of homopolymeric polymerization efficiencies (kcat/Km) of RT proteins. Reac- poly(rA) (286–428 nt, Amersham Pharmacia, NJ) tion conditions were the same as those described in the annealed to a 20-nt oligo(dT) primer (Amersham-Pharma- misincorporation assay except for primers and dNTP cia) (template/primer ratio 1:4) as a substrate. Reaction concentration. For the analysis of four individual dNTP mixtures (20 AL) contained approximately 0.1 Ag purified incorporation kinetics, four different primers (A-, G-, C-, RT proteins, 25 mM Tris–HCl (pH 8.0), 100 mM KCl, 2 and T-primers for dATP, dGTP, dCTP and dTTP incorpo- mM DTT, 3H-TTP (10 AM, Amersham-Pharmacia), dTTP ration) annealed to the 38-mer RNA template (see above) (100 AM), and either MgCl2 or MnCl2 at either 0.05 mM were used. The sequence of the A-primer was described or 5 mM final concentration. Reactions were carried out in above in misincorporation assay, and the sequences of the duplicate at 37 -C for 5 min. FIV RT, FeLV RT, HIV-1 RT, G-(18-mer), C-(19-mer) and T-(16-mer) primers are the and MuLV RT were all examined in this manner. The assay same as A-primer except their 3V ends either extended or protocol has been previously described (Kim et al., 1996). shortened from the A-primer based on the template sequence (see Primer extension assay above). The amounts Sensitivity of RT proteins to 3TCTP, AZTTP, and ddATP of RTs and incubation time were adjusted to yield extension of 30–40% of the total labeled primer (200 fmol) at the The assay for sensitivity of RT proteins to these NRTIs highest dNTP concentration (500 AM), which is at least 5 was the primer extension assay described above with the times higher than the Km values. Then, the reactions were T/P (17-mer A-primer and 38-mer RNA template; see repeated with decreasing 8 concentrations of each dNTP. above) and mixtures of natural dNTPs (10 AM) and Products were resolved in 14% polyacrylamide-urea gel and varying concentrations of 3TCTP (Moravek, CA), AZTTP quantitated by phosphor-imaging analysis (Cyclone, Perki- (Moravek), or ddATP (Amersham-Pharmacia). Polymerase nElmer, CA). The percent of primer extension for each activity of RT proteins, which showed ¨80% primer reaction was used to calculate the kcat and Km values by the extension in the absence of NRTIs under the reaction Lineweaver–Burke plot as described by Creighton and condition described in misincorporation assay, was chosen. Goodman (1995). Extension products were analyzed on 14% polyacrylamide- urea denaturing gel. The percent of full extension (38 nt) M13mp2 lacZa forward mutation assay was determined by comparing products extended in the presence of different concentrations of NRTI with product The fidelity for the purified RT proteins was measured produced in the presence of dNTPs only (i.e., no drug). essentially as previously described (Bebenek and Kunkel, 1995; Weiss et al., 2000). M13mp2 DNA containing a 361-nucleotide single-stranded gap was prepared as Acknowledgments specified (Weiss et al., 2000). Gapped M13mp2 DNA (1 This work was supported by research grant, AI49781 (to Ag) was incubated with RTs (0.5 Ag) at 37 -C for 20 min B.K.), and training fellowship AI056668 (to D. J. O.) from under the conditions described above for the misincorpo- the National Institutes of Health. We would also like to ration assays. The gapped DNAs incubated with RTs were thank Drs. Tracy Diamond and Mini Balakrishnan for analyzed by 0.7% agarose for examining the production of critical readings of the manuscript. double strand M13 DNA with filled gaps (Weiss et al., 2000). Three independent gap-filling reactions per RT protein were performed. The extended gapped DNAs were References transformed to MC1016 cells, and the transformed cells were plated to M9 plates containing X-gal (50 Ag/ml) and Amacker, M., Hottiger, M., Hu¨bscher, U., 1995. Feline immunodeficiency IPTG (1 mM) with CSH50 lawn cells. Phenotype of virus reverse transcriptase: expression, functional characterization, and 4000–6000 plaques per gap-filling reaction was examined. reconstitution of 66- and 51-kilodalton subunits. J. Virol. 69 (10), 6273–6279. The mutant frequency was determined as the ratio Arai, M., Earl, D.D., Yamamoto, J.K., 2002. Is AZT/3TC therapy effective between number of mutant plaques (pale blue and/or against FIV infection or immunopathogenesis? Vet. Immunol. Immu- clear) and number of mutant plus wild type (dark blue) nopathol. 85 (3–4), 189–204. D.J. Operario et al. / Virology 335 (2005) 106–121 119

Avidan, O., Meer, M.E., Oz, I., Hizi, A., 2002. The processivity and fidelity Diamond, T.L., Souroullas, G., Weiss, K.K., Lee, K.Y., Bambara, R.A., of DNA synthesis exhibited by the reverse transcriptase of bovine Dewhurst, S., Kim, B., 2003. Mechanistic understanding of an altered leukemia virus. Eur. J. Biochem. 269 (3), 859–867. fidelity simian immunodeficiency virus reverse transcriptase mutation, Back, N.K., Nijhuis, M., Keulen, W., Boucher, C.A., Oude Essink, B.O., V148I, identified in a pig-tailed macaque. J. Biol. Chem. 278 (32), van Kuilenburg, A.B., van Gennip, A.H., Berkhout, B., 1996. Reduced 29913–29924. replication of 3TC-resistant HIV-1 variants in primary cells due to a Diamond, T.L., Roshal, M., Jamburuthugoda, V.K., Reynolds, H.M., processivity defect of the reverse transcriptase enzyme. EMBO J. 15 Merriam, A.R., Lee, K.Y., Balakrishnan, M., Bambara, R.A., Planelles, (15), 4040–4049. V., Dewhurst, S., Kim, B., 2004. Macrophage tropism of HIV-1 depends Bakhanashvili, M., Hizi, A., 1992. Fidelity of the RNA-dependent DNA upon efficient cellular dNTP utilization by reverse transcriptase. J. Biol. synthesis exhibited by the reverse transcriptases of human immunode- Chem. 279 (49), 51545–51553. ficiency virus types 1 and 2 and of murine leukemia virus: mispair Donahue, P.R., Hoover, E.A., Beltz, G.A., Riedel, N., Hirsch, V.M., extension frequencies. Biochemistry 31 (39), 9393–9398. Overbaugh, J., Mullins, J.I., 1988. Strong sequence conservation among Bakhanashvili, M., Hizi, A., 1993. Fidelity of DNA synthesis exhibited in horizontally transmissible, minimally pathogenic feline leukemia vitro by the reverse transcriptase of the lentivirus equine infectious viruses. J. Virol. 62 (3), 722–731. anemia virus. Biochemistry 32 (29), 7559–7567. Durban, E.M., Boettiger, D., 1981a. Differential effects of transforming Battula, N., Loeb, L.A., 1974. The infidelity of avian myeloblastosis virus avian RNA tumor viruses on avian macrophage. Proc. Natl. Acad. Sci. deoxyribonucleic acid polymerase in polynucleotide replication. J. Biol. U.S.A. 78 (6), 3600–3604. Chem. 249 (13), 4086–4093. Durban, E.M., Boettiger, D., 1981b. Replicating, differentiated macro- Bebenek, K., Kunkel, T.A., 1995. Analyzing fidelity of DNA polymerases. phages can serve as in vitro targets for transformation by avian Methods Enzymol. 262, 217–232. myeloblastosis virus. J. Virol. 37 (1), 488–492. Bebenek, K., Abbotts, J., Roberts, J.D., Wilson, S.H., Kunkel, T.A., 1989. Entin-Meer, M., Sevilya, Z., Hizi, A., 2002. The role of phenylalanine- Specificity and mechanism of error-prone replication by human 119 of he reverse transcriptase of mouse mammary tumour virus in immunodeficiency virus-1 reverse transcriptase. J. Biol. Chem. 264 DNA synthesis, ribose selection and drug resistance. Biochem. J. 367, (28), 16948–16956. 381–391. Bendinelli, M., Pistello, M., Lombardi, S., Poli, A., Garzelli, C., Matteucci, Estis, L.F., Temin, H.M., 1979. Suppression of multiplication of avian D., Ceccherini-Nelli, L., Malvaldi, G., Tozzini, F., 1995. Feline sarcoma virus by rapid spread of transformation-defective virus of the immunodeficiency virus: an interesting model for AIDS studies and same group. J. Virol. 31 (2), 389–397. an important cat pathogen. Clin. Microbiol. Rev. 8 (1), 87–112. Fields, B.N., Knipe, D.M., Howley, P.M., Griffin, D.E., Martin, M.A., Bisset, L.R., Lutz, H., Boni, J., Hofmann-Lehmann, R., Luthy, R., Lamb, R.A., Roizman, B., Straus, S.E., 2001. Retroviridae, (4th edR) Scupbach, J., 2002. Combined effect of zidovudine (ZDV), lamivudine Fields Virology, vol. 2. Lippincott Williams and Wilkins, Philadelphia, (3TC) and abacavir (ABC) antiretroviral therapy in suppressing in vitro pp. 1871–2139 (2 vols.). FIV replication. Antiviral Res. 53 (1), 35–45. Furge, L.L., Guengerich, F.P., 1997. Analysis of nucleotide insertion and Boosalis, M.S., Petruska, J., Goodman, M.F., 1987. DNA polymerase extension at 8-oxo-7,8-dihydroguanine by replicative T7 polymerase insertion fidelity. Gel assay for site-specific kinetics. J. Biol. Chem. 262 exo- and human immunodeficiency virus-1 reverse transcriptase using (30), 14689–14696. steady-state and pre-steady-state kinetics. Biochemistry 36 (21), Boyer, P.L., Sarafianos, S.G., Arnold, E., Hughes, S.H., 2002. Nucleoside 6475–6487. analog resistance caused by insertions in the fingers of human Gao, Q., Gu, Z., Parniak, M.A., Cameron, J., Cammack, N., Boucher, C., immunodeficiency virus type 1 reverse transcriptase involves ATP- Wainberg, M.A., 1993. The same mutation that encodes low-level mediated excision. J. Virol. 76 (18), 9143–9151. human immunodeficiency virus type 1 resistance to 2V,3V-dideoxyino- Cases-Gonzalez, C.E., Gutierrez-Rivas, M., Menendez-Arias, L., 2000. sine and 2V,3V-dideoxycytidine confers high-level resistance to the (-) Coupling ribose selection to fidelity of DNA synthesis. The role of enantiomer of 2V,3V-dideoxy-3V-thiacytidine. Antimicrob. Agents Che- Tyr-115 of human immunodeficiency virus type 1 reverse transcriptase. mother. 37 (6), 1390–1392. J. Biol. Chem. 275 (26), 19759–19767. Gao, H.Q., Boyer, P.L., Sarafianos, S.G., Arnold, E., Hughes, S.H., 2000. Cherrington, J.M., Fuller, M.D., Mulato, A.S., Allen, S.J., Kunder, S.C., The role of steric hindrance in 3TC resistance of human immunodefi- Ussery, M.A., Lesnikowski, Z., Schinazi, R.F., Sommadossi, J.P., Chen, ciency virus type-1 reverse transcriptase. J. Mol. Biol. 300 (2), 403–418. M.S., 1996. Comparative kinetic analyses of interaction of inhibitors Gendelman, H.E.N.O., Kennedy-Stoskopf, S., Kennedy, P.G., Ghotbi, Z., with Rauscher murine leukemia virus and human immunodeficiency Clements, J.E., Stanley, J., Pezeshkpour, G., 1986. Tropism of sheep virus reverse transcriptases. Antimicrob. Agents Chemother. 40 (5), lentivirus for monocytes: susceptibility to infection and virus gene 1270–1273. expression increase during maturation of monocytes to macrophages. Chowdhury, K., Kaushik, N., Pandey, V.N., Modak, M.J., 1996. Elucidation J. Virol. 58 (1), 67–74. of the role of Arg 110 of murine leukemia virus reverse transcriptase in Georgiadis, M.M., Jessen, S.M., Ogata, C.M., Telesnitsky, A., Goff, S.P., the catalytic mechanism: biochemical characterization of its mutant Hendrickson, W.A., 1995. Mechanistic implications from the structure enzymes. Biochemistry 35 (51), 16610–16620. of a catalytic fragment of Moloney murine leukemia virus reverse Creighton, S., Goodman, M.F., 1995. Gel kinetic analysis of DNA transcriptase. Structure 3 (9), 879–892. polymerase fidelity in the presence of proofreading using bacteriophage Gerard, G.F., Grandgenett, D.P., 1975. Purification and characterization of T4 polymerase. J. Biol. Chem. 270 (9), 4759–4774. the DNA polymerase and RNase H activities in moloney murine Cronn, R.C., Remington, K.M., Preston, B.D., North, T.W., 1992. sarcoma-leukemia virus. J. Virol. 15 (4), 785–797. Inhibition of reverse transcriptase from feline immunodeficiency virus Greger, J.G., Katz, R.A., Taganov, K., Rall, G.F., Skalka, A.M., 2004. by analogs of 2V-deoxyadenosine-5V-triphosphate. Biochem. Pharmacol. Transduction of terminally differentiated neurons by avian sarcoma 44 (7), 1375–1381. virus. J. Virol. 78 (9), 4902–4906. Dean, G.A., Himathongkham, S., Sparger, E.E., 1999. Differential cell Gu, Y., Sundquist, W., 2003. Good to CU. Nature 424, 21–22. tropism of feline immunodeficiency virus molecular clones in vivo. Hatziioannou, T., Goff, S.P., 2001. Infection of nondividing cells by rous J. Virol. 73 (4), 2596–2603. sarcoman virus. J. Virol. 75 (19), 9526–9531. Diamond, T.L., Kimata, J., Kim, B., 2001. Identification of a simian Huang, H., Chopra, R., Verdine, G.L., Harrison, S.C., 1998. Structure of a immunodeficiency virus reverse transcriptase variant with enhanced covalently trapped catalytic complex of HIV-1 reverse transcriptase: replicational fidelity in the late stage of viral infection. J. Biol. Chem. implications for drug resistance. Science 282 (5394), 1669–1675. 276 (26), 23624–23631. Jacobo-Molina, A., Ding, J., Nanni, R.G., Clark Jr., A.D., Lu, X., Tantillo, 120 D.J. Operario et al. / Virology 335 (2005) 106–121

C., Williams, R.L., Kamer, G., Ferris, A.L., Clark, P., et al., 1993. Mukherji, E., Au, J.L.-S., Mathes, L.E., 1994. Differential antiviral Crystal structure of human immunodeficiency virus type 1 reverse activities and intracellular metabolism of 3V-azido-3V-deoxythymidine transcriptase complexed with double-stranded DNA at 3.0 A resolution and 2V,3-dideoxyinosine in human cells. Antimicrob. Agents Chemo- shows bent DNA. Proc. Natl. Acad. Sci. U.S.A. 90 (13), 6320–6324. ther. 38 (7), 1573–1579. Katz, R.A., Greger, J.G., Darby, K., Boimel, P., Rall, G.F., Skalka, A.M., North, T.W., North, G.L., Pedersen, N.C., 1989. Feline immunodeficiency 2002. Transduction of interphase cells by avian sarcoma virus. J. Virol. virus, a model for reverse transcriptase-targeted chemotherapy for 76 (11), 5422–5434. acquired immune deficiency syndrome. Antimicrob. Agents Chemother. Kaushik, N., Harris, D., Rege, N., Modak, M.J., Yadav, P.N., Pandey, V.N., 33 (6), 915–919. 1997. Role of glutamine-151 of human immunodeficiency virus type-1 North, T.W., Hansen, G.L., Zhu, Y., Griffin, J.A., Shih, C.K., 1994. reverse transcriptase in RNA-directed DNA synthesis. Biochemistry 36 Expression of reverse transcriptase from feline immunodeficiency (47), 14430–14438. virus in Escherichia coli. Antimicrob. Agents Chemother. 38 (2), Kaushik, N., Chowdhury, K., Pandey, V.N., Modak, M.J., 2000a. Valine of 388–391. the YVDD motif of maloney murine leukemia virus reverse tran- O’Neil, P.K., Sun, G., Yu, H., Ron, Y., Dougherty, J.P., Preston, B.D., 2002. scriptase: role in fidelity of DNA synthesis. Biochemistry 39 (17), Mutational analysis of HIV-1 long terminal repeats to explore the 5155–5165. relative contribution of reverse transcriptase and RNA polymerase II to Kaushik, N., Chowdhury, K., Pandey, V.N., Modak, M.J., 2000b. Valine of viral mutagenesis. J. Biol. Chem. 31, 31. the YVDD motif of moloney murine leukemia virus reverse tran- Overbaugh, J., Donahue, P.R., Quackenbush, S.L., Hoover, E.A., scriptase: role in the fidelity of DNA synthesis. Biochemistry 39, Mullins, J.I., 1988. Molecular cloning of a feline leukemia virus 5155–5165. that induces fatal immunodeficiency disease in cats. Science 239, Kim, B., 1997. Genetic selection in Escherichia coli for active human 906–910. immunodeficiency virus reverse transcriptase mutants. Methods 12 (4), Owen, S.M., Lal, R.B., Ikeda, R.A., 1998. Cloning and expression of a 318–324. human T-lyphotropic virus type one protein with reverse transcriptase Kim, B., Hathaway, T.R., Loeb, L.A., 1996. Human immunodeficiency activity. J. Virol. 72 (6), 5279–5284. virus reverse transcriptase. Functional mutants obtained by random Perach, M., Hizi, A., 1999. Catalytic features of the recombinant reverse mutagenesis coupled with genetic selection in Escherichia coli. J. Biol. transcriptase of bovine leukemia virus expressed in bacteria. Virology Chem. 271 (9), 4872–4878. 259 (1), 176–189. Kim, B., Ayran, J.C., Sagar, S.G., Adman, E.T., Fuller, S.M., Tran, N.H., Phillips, T.R., Talbott, R.L., Lamont, C., Muir, S., Lovelace, K., Elder, J.H., Horrigan, J., 1999. New human immunodeficiency virus, type 1 reverse 1990. Comparison of two host cell range variants of feline immuno- transcriptase (HIV-1 RT) mutants with increased fidelity of DNA deficiency virus. J. Virol. 64 (10), 4605–4613. synthesis. Accuracy, template binding, and processivity. J. Biol. Chem. Preston, B.D., Poiesz, B.J., Loeb, L.A., 1988. Fidelity of HIV-1 reverse 274 (39), 27666–27673. transcriptase. Science 242 (4882), 1168–1171. Klarmann, G.J., Smith, R.A., Schinazi, R.F., North, T.W., Preston, B.D., Quan, Y., Liang, C., Inouye, P., Wainberg, M.A., 1998. Enhanced 2000. Site-specific incorporation of nucleoside analogs by HIV-1 impairment of chain elongation by inhibitors of HIV reverse tran- reverse transcriptase and the template grip mutant P157S. Template scriptase in cell-free reactions yielding longer DNA products. Nucleic interactions influence substrate recognition at the polymerase active Acids Res. 26 (24), 5692–5698. site. J. Biol. Chem. 275 (1), 359–366. Rezende, L.F., Drosopoulos, W.C., Prasad, V.R., 1998. The influence of Kohlstaedt, L.A., Wang, J., Friedman, J.M., Rice, P.A., Steitz, T.A., 1992. 3TC resistance mutation M184I on the fidelity and error specificity of Crystal structure at 3.5 A resolution of HIV-1 reverse transcriptase human immunodeficiency virus type 1 reverse transcriptase. Nucleic complexed with an inhibitor. Science 256 (5065), 1783–1790. Acids Res. 26 (12), 3066–3072. Lewis, P.F., Emerman, M., 1994. Passage through mitosis is required for Rho, H.M., Gallo, R.C., 1980. Biochemical and immunological properties oncoretroviruses but not for human immunodeficiency virus. J. Virol. of the DNA polymerase and RNAase H activities of purified feline 68 (1), 510–516. leukemia virus reverse transcriptase. Cancer Lett. 10, 207–221. Lewis, P., Hensel, M., Emerman, M., 1992. Human immunodeficiency Roberts, J.D., Preston, B.D., Johnston, L.A., Soni, A., Loeb, L.A., virus infection of cells arrested in the cell cycle. EMBO J. 11 (8), Kunkel, T.A., 1989. Fidelity of two retroviral reverse transcriptases 3053–3058. during DNA-dependent DNA synthesis in vitro. Mol. Cell. Biol. 9 (2), Malboeuf, C.M., Isaacs, S.J., Tran, N.H., Kim, B., (2001). Thermal effects 469–476. on reverse transcription: improvement of accuracy and processivity in Sarafianos, S.G., Das, K., Clark Jr., A.D., Ding, J., Boyer, P.L., Hughes, cDNA synthesis. Biotechniques. 30 (5), 1074–8, 1080, 1082, passim. S.H., Arnold, E., 1999. Lamivudine (3TC) resistance in HIV-1 reverse Mansky, L.M., 2000. In vivo analysis of human T-cell leukemia virus type 1 transcriptase involves steric hindrance with beta-branched amino acids. reverse transcription accuracy. J. Virol. 74 (20), 9525–9531. Proc. Natl. Acad. Sci. U.S.A. 96 (18), 10027–10032. Mansky, L.M., Le Rouzic, E., Benichou, S., Gajary, L.C., 2003. Influence of Sheehy, A., Gaddis, N., Choi, J., Malim, M., 2002. Isolation of a human reverse transcriptase variants, drugs, and Vpr on human immunodefi- gene that inhibits HIV-1 infection is suppressed by the viral Vif protein. ciency virus type 1 mutant frequencies. J. Virol. 77 (3), 2071–2080. Nature 418, 646–650. Mathes, L.E., Polas, P.J., Hayes, K.A., Swenson, C.L., Johnson, S., Kociba, Shi, Q., Singh, K., Srivastava, A., Kaushik, N., Modak, M.J., 2002. Lysine G.L., 1992. Pre- and postexposure chemoprophylaxis: evidence that 3V- 152 of MuLV reverse transcriptase is required for the integrity of the azido-3V-dideoxythymidine inhibits feline leukemia virus disease by a active site. Biochemistry 41 (50), 14831–14842. drug-induced vaccine response. Antimicrob. Agents Chemother. 36 Singh, K., Kaushik, N., Jin, J., Madhusudanan, M., Modak, M.J., 2000. (12), 2715–2721. Role of Q190 of MuLV RT in ddNTP resistance and fidelity of DNA McGuire, T.C.C.T., Henson, J.B., 1971. Immunofluorescent localization of synthesis: a molecular model of interactions with substrates. Protein equine infectious anemia virus in tissue. Am. J. Pathol. 62, 283–292. Eng. 13 (9), 635–643. Meldin, H.K., Zhu, Y., Remington, K.M., Phillips, T.R., North, T.W., 1996. Skalka, A.M., Goff, S.P., 1993. Reverse transcriptase. Cold spring harbor Selection and characterization of a mutant feline immunodeficiency monograph series. Cold Spring Harbor Labs Press, Cold Spring Harbor. virus resistant to 2V,3V-dideoxycytidine. Antimicrob. Agents Chemother. Skasko, M., Weiss, K.K., Reynolds, H.M., Jamburuthugoda, V., Lee, K., 40 (4), 953–957. Kim, B., 2005. Mechanistic differences in RNA dependent DNA Miller, V., Stark, T., Loeliger, A.E., Lange, J.M., 2002. The impact of the polymerization and fidelity between murine leukemia virus and human M184V substitution in HIV-1 reverse transcriptase on treatment immunodeficiency virus type 1 reverse transcriptases. J. Biol. Chem. response. HIV Med. 3 (2), 135–145. (Electronic publication). D.J. Operario et al. / Virology 335 (2005) 106–121 121

Smith, R.A., Remington, K.M., Lloyd, J.R.M., Schinazi, R.F., North, T.W., of HIV-1 reverse transcriptase motif C: site-directed mutagenesis and 1997. A novel Met-to-Thr mutation in the YMDD motif of reverse metal cation interaction. J. Mol. Evol. 47 (1), 73–80. transcriptase from feline immunodeficiency virus confers resistance to Vartanian, J.P., Meyerhans, A., Asjo, B., Wain-Hobson, S., 1991. Selection, oxathiolane nucleosides. J. Virol. 71 (3), 2357–2362. recombination, and G-A hypermutation of human immunodeficiency Talbott, R.L., Sparger, E.E., Lovelace, K.M., Fitch, W.M., Pedersen, N.C., virus type 1 genomes. J. Virol. 65 (4), 1779–1788. Luciw, P.A., Elder, J.H., 1989. Nucleotide sequence and genome Vartanian, J.P., Sommer, P., Wain-Hobson, S., 2003. Death and the organization of feline immunodeficiency virus. Proc. Natl. Acad. Sci. retrovirus. Trends Mol. Med. 9 (10), 409–413. U.S.A. 86 (15), 5743–5747. Wainberg, M.A., Drosopoulos, W.C., Salomon, H., Hsu, M., Borkow, G., Taube, R., Avidan, O., Bakhanashvili, M., Hizi, A., 1998a. DNA Synthesis Parniak, M., Gu, Z., Song, Q., Manne, J., Islam, S., Castriota, G., exhibited by the reverse transcriptase of mouse mammary tumor virus: Prasad, V.R., 1996. Enhanced fidelity of 3TC-selected mutant HIV-1 processivity and fidelity of misinsertion and mispair extension. Eur. J. reverse transcriptase. Science 271 (5253), 1282–1285. Biochem. 258 (3), 1032–1039. Weiss, K.K., Isaacs, S.J., Tran, N.H., Adman, E.T., Kim, B., 2000. Taube, R., Loya, S., Avidan, O., Perach, M., Hizi, A., 1998b. Reverse Molecular architecture of the mutagenic active site of human transcriptase of mouse mammary tumor virus: expression in immunodeficiency virus type 1 reverse transcriptase: roles of the beta bacteria, purification and biochemical characterization. Biochem. J. 8-alpha E loop in fidelity, processivity, and substrate interactions. 329, 579–587. Biochemistry 39 (35), 10684–10694. Torten, M., Franchini, M., Barlough, J.E., George, J.W., Mozes, E., Lutz, Weiss, K.K., Bambara, R.A., Kim, B., 2002. Mechanistic role of residue H., Pedersen, N.C., 1991. Progressive immune dysfunction in cats Gln151 in error prone DNA synthesis by human immunodeficiency experimentally infected with feline immunodeficiency virus. J. Virol. virus type 1 (HIV-1) reverse transcriptase (RT). Pre-steady state kinetic 65 (5), 2225–2230. study of the Q151N HIV-1 RT mutant with increased fidelity. J. Biol. Traut, T.W., 1994. Physiological concentrations of purines and pyrimidines. Chem. 277 (25), 22662–22669. Mol. Cell. Biochem. 140 (1), 1–22. Weiss, K.K., Chen, R., Skasko, M., Reynolds, H.M., Lee, K., Bambara, Trentin, B., Rebeyrotte, N., Mamoun, R.Z., 1998. Human T-cell R.A., Mansky, L.M., Kim, B., 2004. A role for dNTP binding of human leukemia virus type 1 reverse transcriptase (RT) originates from immunodeficiency virus type 1 reverse transcriptase in viral muta- the pro and pol open reading frames and requires the presence of genesis. Biochemistry 43 (15), 4490–4500. RT-RNase H (RH) and RT-RH-integrase proteins for its activity. Williams, K.J., Loeb, L.A., 1992. Retroviral reverse transcriptases: error J. Virol. 72 (8), 6504–6510. frequencies and mutagenesis. Curr. Top. Microbiol. Immunol. 176, Ueno, T., Shirasaka, T., Mitsuya, H., 1995. Enzymatic characterization of 165–180. human immunodeficiency virus type 1 reverse transcriptase resistant to Woodside, A.M., Guengerich, F.P., 2002. Effect of the O6 substituent multiple 2V,3V-dideoxynucleoside 5V-triphosphates. J. Biol. Chem. 270 on misincorporation kinetics catalyzed by DNA polymerases at (40), 23605–23611. O(6)-methylguanine and O(6)-benzylguanine. Biochemistry 41 (3), Valverde-Garduno, V., Gariglio, P., Gutierrez, L., 1998. Functional analysis 1027–1038.