HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kim, B. Articles by Horrigan, J. Search for Related Content PubMed PubMed Citation Articles by Kim, B. Articles by Horrigan, J. Social Bookmarking                      What's this?

J Biol Chem, Vol. 274, Issue 39, 27666-27673, September 24, 1999

New Human Immunodeficiency Virus, Type 1 Reverse Transcriptase (HIV-1 RT) Mutants with Increased Fidelity of DNA Synthesis
ACCURACY, TEMPLATE BINDING, AND PROCESSIVITY^*

Baek Kim^§, Jennifer C. Ayran^¶, Sarah G. Sagar^¶, Elinor T. Adman, Shannon M. Fuller, Nancy H. Tran, and Jeffrey Horrigan

From the  Department of Microbiology and Immunology, University of Rochester, Rochester, New York 14642, the ^¶ Departments of Pathology, and  Biological Structure, University of Washington, Seattle, Washington 98195

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Infidelity of DNA synthesis by human immunodeficiency virus, type 1 reverse transcriptase (HIV-1 RT) is a presumptive determinant of HIV-1 hypervariability and is incompletely understood at the mechanistic and structural levels. Amino acid substitution at only three residues, including Asp-76 (Kim, B., Hathaway, T. R., and Loeb, L. A. (1996) Biochemistry 37, 5831-5839), is known to increase fidelity. We report here that substitution at Arg-78 can also increase accuracy. Mutant R78A RT showed reduced primer extension in misincorporation assays lacking a complementary dNTP and exhibited a 9-fold decrease in mutation frequency in the M13mp2 lacZ forward mutation assay. Previous structural studies indicate that Arg-78 and Asp-76 lie in a region that interacts with template nucleotides. Interestingly, R78A RT exhibited 6- to 8-fold decreases in binding affinity (K_d) for RNA and DNA templates relative to wild type RT. In contrast, D76V RT, which also increases fidelity (Kim et al., 1996), showed a 6- to 7-fold increased affinity. The processivity of R78A RT on both RNA and DNA templates was substantially reduced relative to wild type RT, whereas the processivity of D76V RT was increased. We discuss relationships of fidelity, template binding, and processivity in these and other HIV RT mutants.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Genomic hypervariability is a defense device that allows human immunodeficiency virus, type 1 (HIV-1)¹ to escape from selection pressure such as that imposed by the host immune system or drug therapies. Hypervariation of HIV-1 genomes can be achieved by at least three unique viral replication processes. First, mutations can be synthesized by the virally encoded DNA polymerase, HIV-1 reverse transcriptase (RT) (1, 2). HIV-1 RT is the most error prone of all known DNA polymerases (3-7). Second, mutations produced by HIV-1 RT can be amplified by recombination, which has been observed in virions containing heterogeneous diploid genomes (8). It has been suggested that the occurrence of homologous recombination during viral replication is related to the processivity and strand transfer activity of HIV-1 RT (9, 10). Finally, constant, massive replication amplifies the complexity of the viral population (11).

The structural basis of HIV-1 RT infidelity remains to be fully elucidated. Lack of a 3'-5' proofreading exonuclease activity is a presumptive factor (4), but it cannot be the only determinant, because other retroviral RTs that lack a proofreading exonuclease, such as murine leukemia virus RT and AMV RT, have 10- to 18-fold higher fidelity (3). Our knowledge of the structure-function relationships governing the accuracy of other DNA polymerases rests heavily on the analysis of mutants (12, 13); the classic mutants of Escherichia coli DNA polymerase I are important examples (14). Currently available mutants of HIV-1 RT are mostly drug-resistant variants isolated from patients and cell culture systems. Interestingly, it has been reported that drug resistance mutations at Met-184 (M184V, M184I) and Glu-89 (E89G) confer increased fidelity (15-18). Biochemical studies suggest that these mutations affect accuracy by altering interactions with either the incoming dNTP (M184V and M184I) or with the penultimate nucleotide of the double-stranded template (E89G). Notably, these drug-resistant mutations permit continued viral evolution and escape from selective pressure, suggesting that the changes in fidelity may be rather moderate. Other interactions of HIV-1 RT with its substrates also affect accuracy. Thus, mutations (e.g. G262A and W266A) in the thumb subdomain that interact with the primer have been shown to affect the frequency of frameshift mutations, presumably because of template-primer slippage; these mutations also alter processivity (19, 20).

To better understand the structural and mechanistic basis of HIV-1 RT infidelity, we have undertaken a systematic approach to the study of mutants. To obtain novel fidelity mutations, we employ both genetic and biochemical methods (21). We initially select for catalytically active mutants among a large, randomly mutated RT population by using a genetic complementation system (22, 23). Mutants with altered accuracy are then identified by screening in a primer extension assay. In this way, we obtain numerous catalytically robust fidelity mutants for detailed biochemical, kinetic, and structural analysis. Other amino acid substitutions of interest, suggested by these genetically selected variants, can then be synthesized by site-directed mutagenesis. Substitutions that confer increased accuracy are of particular relevance because they identify amino acid side chains that are responsible for mutation synthesis by the wild type enzyme. Among genetically selected variants with increased DNA synthetic accuracy (we call them "hi-fi" mutants), we have previously described D76V, which exhibited a 9-fold increase in fidelity in a lacZ forward mutation assay, and up to 14-fold increases in the accuracy of nucleotide insertion. Structural studies indicate that Asp-76 interacts with the template nucleotide that base pairs with the incoming dNTP (21, 24). Presumably, in wild type RT, interaction of Asp-76 with the template may promote mispairing with the incoming nucleotide, whereas other residues, such as valine, reduce or abolish this mutagenic interaction. This presumption is consonant with the findings for Glu-89, which also interacts with the template. Glu-89 contacts the template nucleotide that pairs with the penultimate primer nucleotide (18), and mutation, at this site too, can increase fidelity. The mutant E89G exhibits enhanced fidelity (18), higher processivity, and a decreased dissociation constant from a RT·template·primer complex (25). Taken together, these observations are consistent with our premise, based on the D76V mutation, that amino acid residues involved in interactions with the template can play an important role in HIV-1 RT infidelity (21).

Our previous structural modeling (21), undertaken in connection with the D76V mutant, suggested that Arg-78 interacts with the template nucleotide that base pairs with the nucleotide at the 3' primer terminus. In view of this putative interaction of Arg-78 with the template and the possible effect on fidelity, we decided to generate and examine the accuracy of mutants with amino substitutions at this position. Recently, Huang et al. (25) reported the structure of a catalytic complex of HIV RT with a DNA template-primer and a dNTP. Although generally similar to our model, the x-ray structure shows that Arg-78 interacts directly with Asp-76 and indirectly with the template nucleotide base paired to the incoming dNTP. This template nucleotide lies between the nucleotide that interacts with both Asp-76 and base pairs with the incoming dNTP and the nucleotide that interacts with both Glu-89 and base pairs to the penultimate primer nucleotide. We report here new HIV-1 RT variants with mutations at Arg-78 that exhibit increased fidelity and altered interactions with DNA and RNA templates. We interpret available evidence to indicate that interactions of Arg-78 with both Asp-76 and the template nucleotide base pairing with the incoming dNTP play important roles in HIV-1 RT infidelity. We suggest that interactions of these two residues with the template nucleotide may promote the formation of a mutagenic DNA polymerization active site, thereby contributing to a high rate of mutation during viral replication.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Strains and Plasmids-- E. coli NM522 (Stratagene) was used for construction of plasmids, and BL21 (Novagen) was used for over-expression of HIV-1 RT. pBK33 is pHis/NdeI (from Dr. A. Hizi) encoding full-length wild type p66 HIV-1 RT fused at the N terminus to six histidine residues (22). pBK44 is a pET28a (Novagen) derivative expressing wild type p51 HIV-1-1 RT fused at the N terminus to six histidine residues. pBK77 and pBK78 are pHis/NdeI constructs for expression of His-tagged p66 of wild type SIV RT that was derived from pSIV_MNE170 (from Dr. J. T. Kimata, Southwest Foundation for Biomedical Research, TX) and the R78A SIV RT mutant, respectively.

Construction of Arg-78 Mutants of HIV-1 and SIV RTs-- Amino acid substitutions at Arg-78 of HIV-1 RT were generated by using PCR-based site-directed mutagenesis with Arg-78 primers (5'-CTTATTGAGCTCTCTGAANNNTACTAATTTTCTCCATTTAG-3') that contained 100% random nucleotides (designated "N") at the positions encoding residue 78. The products of PCR amplification with Arg-78 primer and SD-RT primer (5'-GAAGATCTAAGCTTAGGAGGTTGTCCCATATGCCCATTAGTCC-3'), which hybridizes to the N-terminal sequence of HIV-1 RT and the NdeI site, were inserted into pBK33 for p66 RT after digestion with BsrGI and SacI. A 126-base pair region between BsrGI and SacI of the plasmids expressing Asp-76 mutants was sequenced to confirm mutations at position 78. After sequencing the 126-base pair fragment inserts to verify the mutations at position 78 of RT, the 126-base pair fragments prepared from pBK33 of individual mutation constructs were cloned to pBK44 for p51 RT mutant proteins. The plasmids expressing His-tagged Arg-78 SIV RT mutants were constructed by PCR-based site-directed mutagenesis using two pieces of PCR products. One is the PCR product encoding the N-T part of SIV RT and an Arg-78 mutation that was amplified with SD-SIV RT (5'-TTTTTTAAGCTTGAAGGAGATATACATATGCCCATAGCTAAGGTAGAGCC-3') and an Arg-78 reverse primer (5'-TTT TTTGAGCTCNNNAAAATCTATCAGTATTCTCC-3') containing 100% randomized sequences at position Arg-78 (designated N) and a SacI site near position 79 of SIV RT. The other is the PCR product encoding the C-T part of the SIV RT that was amplified with a SIV-SacI forward primer (5'-TTTTTTGAGCTCAATAAGGTCACTCAG GAC-3') containing a SacI site at position 79 and a SIV C-T primer (5'-TTTTTTGGATCCGTCGACTTAAACTTGTCTAATCCCTTG-3'). After digestion of N-T PCR by NdeI/SacI and C-T PCR by SacI/BamHI, these two digested DNAs were ligated to pHis/NdeI-digested NdeI and BamHI. For the expression plasmid for wild type SIV RT, the wild type SIV RT gene amplified from pSIV_MNE170 by using SD SIV RT and SIV C-T primers (see the SIV C-T primer sequence above) was cloned to the pHis/NdeI plasmid after digestion with NdeI and BamHI. All SIV RT genes in these constructed expression plasmids were sequenced by ABI system.

Purification of HIV-1 RT and SIV RT-- Procedures for purification of the hexahistidine-tagged p66 and p51 monomers of HIV-1 RT by Ni²⁺ chelation chromatography have been described; equi-molar amounts of separately purified p66 and p51 of HIV-1 RT were mixed prior to dialysis to form p66/p51 heterodimers (22, 23). Reconstructed p66/p51 heterodimers of wild type and Asp-76 mutant proteins were used for the fidelity assays described below. Amounts of purified monomers were estimated in the Bradford assay (Bio-Rad) with a bovine serum albumin standard; all preparations were of >95% purity, estimated by visual inspection of Coomassie Blue-stained SDS-polyacrylamide gels. DNA- and RNA-dependent DNA polymerase activities were determined as described (22); the assays employed a gapped salmon sperm DNA template-primer, poly(rA)/oligo(dT) (Amersham Pharmacia Biotech), or an RNA template encoding the HIV-1 RT gene. Preparation of the RNA template is described below (see primer extension assay). We previously found that a 6 His tag of purified HIV-1 RT proteins does not affect DNA polymerase activity and sensitivity to a nucleotide analog (22). We also found both p66/p51 heterodimers and p66/p66 homodimers of wild type and mutant HIV-1 RT proteins showed the same level of misinsertion in four different types of the primer extension reactions described below. The SIV RT p66 fragment was purified by following the purification protocol established for HIV-1 RT. With this protocol, we were able to purify 2 mg of SIV RT proteins with greater than 95% purity from a 1-liter culture.

Primer Extension Assay with DNA or RNA Templates-- Procedures (22) were modified from those of Preston et al. (7). The deoxyoligonucleotide template-primer was prepared by annealing a 63-mer (5'-TAATACGACTCACTATAGGGAGGAAGCTTGGCTGCAGAATATTGCTAGCGGGAATTCGGCGCG-3') to a 17-mer (5'-CGCGCCGAATTCCCGCT-3') ³²P-labeled at the 5'-end with T4 polynucleotide kinase (template:primer, 2.5:1). Assay mixtures (20 µl) contained 10 nM template-primer, 5-50 nM HIV-1 RT or SIV RT as specified in figure legends, 3 or 4 dNTPs (250 µM each), 25 mM Tris-HCl (pH 8.0), 100 mM KCl, 2 mM dithiothreitol, 5 mM MgCl₂, and 0.1 mg/ml bovine serum albumin. Reactions were incubated at 37 °C for 5 min and terminated by the addition of 5 µl of 40 mM EDTA, 99% formamide. Reaction products were immediately denatured by incubating at 95 °C for 3 min and analyzed by electrophoresis in 14% urea-polyacrylamide gels. The 1.6-kilonucleotide RNA template used in this assay contains the coding sequence of the HIV-1 RT gene and was synthesized by T7 RNA polymerase-catalyzed transcription of pBK8 (22), which encodes the HIV-1 RT gene between the HindIII and EcoRI sites of pBluescript (Stratagene). The transcribed RNA was purified by ethanol precipitation and annealed to a ³²P-labeled 21-mer (3305 RT primer (22)) complementary to nucleotides encoding amino acids 239-246 of HIV-1 RT. Reaction conditions were otherwise the same as described above. Extension products were resolved in 10% denaturing gels.

M13mp2 lacZ Forward Mutation Assay-- The mutation frequencies for wild type and mutant HIV-1 RT were measured essentially as described previously (27). M13mp2 DNA containing a 361-nucleotide single-stranded gap was prepared as specified (27). Gapped DNA (1 µg) was incubated with RT (200 nM) at 37 °C for 10 min under the conditions described below for the misinsertion assays. The extended gapped DNAs by RT proteins were analyzed by 0.7% agarose for checking the production of double strand M13 DNA with filled gaps (27). The extended gapped DNAs were transformed to MC1016 cells, and the transformed cells were plated to M9 plates containing 5-bromo-4-chloro-3-indolyl -D-galactopyranoside (X-gal) and isopropyl-1-thio--D-galactopyranoside with CSH50 lawn cells. The mutation frequency was determined as the ratio of mutant (pale blue and white) plaques to mutant plus wild type (dark blue) plaques as described (27). The lacZa genes encoded in the M13 phage DNA prepared from mutant plaques were sequenced as described (27).

Measurment of K_d to DNA Template-- Binding constants (K_d) of HIV-1 RT proteins to the 63-mer DNA template annealed to the 17-mer primer, and HIV-1 RNA template encoding the 1.7 kilonucleotide RT gene annealed to the 3305 template were measured as described (28, 29). HIV-1 RT proteins (20 nM) were preincubated with various concentrations of the template·primer complexes (T·P) used in the misinsertion assay (1, 2.5, 5, 10, 15, 20, 30, 40, 60, 80, and 100 nM) for 3 min at 37 °C. Each of these T·P concentrations contains 1 nM ³²P-labeled template-primer. The extension reactions (total 20 µl) were initiated by adding a mixture of 0.5 mM dNTPs with 5 mM MgCl₂ (final concentration) and a poly(rA)/oligo(dT) trap (4 µg/20µl), and after a 3-min incubation at 37 °C, the reactions were terminated by adding 4 µl of stop solution (90% formamide, 5 mM EDTA, 0.1% xylene cyanol, and 0.1% bromphenol blue). The terminated reactions were immediately incubated at 95 °C for 5 min. Under this reaction condition, RT extended the primers only once during the incubation. To confirm this single round of extension, two control reactions for each T·P concentration were performed. First, RT proteins were added to the mixture of the T·P of each concentration, poly(rA)/oligo(dT) trap and dNTPs, and the reactions continued for 3 min at 37 °C. In this condition the primer was not able to be extended because all RT molecules were successfully trapped by excess molar concentrations (4 µg/20 µl) of a poly(rA)/oligo(dT) trap. Second, the reactions without the trap showed the larger amount of the primer extended because of multiple rounds of replication. These two control experiments guaranteed the single round of extension reaction under our reaction condition. The extension reactions at different T·P concentrations were analyzed by 14 and 10% urea-polyacrylamide gel electrophoresis for DNA and RNA templates, respectively. The amounts of the extended and unextended primers at each T·P concentration were determined by phosphoimager analysis of the gel. The data were fitted to the equation described previously (28, 29; see Eq. 1 below), and the K_d values for wild type, R78A, and D76V RT proteins were determined.

(Eq. 1)

In Equation 1, E  T·P, K_d, E_t, and T·P indicate productive RT-template concentration, equilibrium dissociation constant for RT-template binding, total enzyme concentration, and total template·primer concentration, respectively.

Processivity Assay-- Reaction condition for measuring processivity also requires a single round of primer extension. The reaction condition for processivity assay was modified from the one used for K_d measurement described above. Two concentrations of three RT proteins, wild type (10 and 20 nM), D76V (2.5 and 5 nM), and R78A (40 and 80 nM) HIV-1 RT proteins were used for RNA and DNA templates. RT proteins were preincubated with either 10 nM poly(rA) (average size, 260 nt, Amersham Pharmacia Biotech) annealed to the ³²P-labeled 20-mer oligo(dT) or the M13mp18 single strand DNA (+) annealed to the ³²P-labeled 24-mer forward primer (Stratagene) for 3 min at 37 °C. The extension reactions were initiated by adding the trap mixture containing dNTPs (0.5 mM with final concentration 5 mM MgCl₂), poly(rA)/oligo(dT) (4 µg/20 µl), and heparin (10 µg/20 µl). The extension reactions were terminated by 2 µl of stop solution after a 3-min incubation at 37 °C. Under this condition, the primer was extended only once during the incubation as confirmed by the same two control reactions with or without the trap (see Fig. 4). The terminated processivity reaction and control reactions were analyzed by 10% urea-polyacrylamide gel electrophoresis after a 3-min heat inactivation.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Construction and Purification of Arg-78 Mutants of HIV-1 RT-- Eleven different amino acid substitutions at Arg-78 of HIV-1 RT were generated in both the p66 and p51 subunits by using PCR-based site-directed mutagenesis. The hexahistidine-tagged, wild type and mutant monomers were overexpressed in E. coli and purified by Ni²⁺ affinity chromatography. Heterodimers, formed by dialyzing equimolar quantities of separately purified p66 and p51 monomers, were used for the work described below. Specific activities of 8 of the 11 Arg-78 mutants (R78A, R78S, R78L, R78I, R78T, R78V, R78C, and R78G), measured on activated calf thymus DNA, ranged from 25 to 35% of the wild type. The remaining three mutants (R78Y, R78E, and R78P) exhibited much lower specific activities ranging from 5 to 10% of the wild type. In accord, specific activities of the Arg-78 mutants measured by using poly(rA) template annealed to oligo(dT) ranged at similar levels of the DNA-dependent DNA polymerase activities.

Misinsertion Assay of R78A HIV-1 RT Mutants with DNA and RNA Templates-- We first employed misinsertion assays to compare the misincorporation capability of wild type and R78A mutant HIV-1 RTs. This assay is a primer extension assay that monitors misinsertion by utilizing a synthetic DNA or RNA template-primer and biased dNTP pools containing only three kinds of dNTPs. A gel-displaying extension of the 17-mer primer annealed to a 63-mer DNA template by two quantities (4× and 1×) of wild type, and R78A HIV-1 heterodimer RT is shown in Fig. 1A. In this assay, the higher efficiency of primer extension will reflect the lower fidelity of the HIV-1 RT protein assayed. The specific activity of the purified R78A mutant, as determined with heterogeneous DNA and RNA templates, was approximately 30% of that of wild type RT. As shown in Fig. 1A, loaded with reactions containing all four dNTPs, the wild type and R78A mutant heterodimers catalyzed approximately the same amount of synthesis leading to fully extended primers. When incubated with mixtures of only 3 dNTPs (Fig. 1, B, dTTP; C, dCTP; D, dGTP; E, dATP), wild type HIV-1 RT catalyzed substantial extension past nucleotides for which a complementary dNTP was deleted (see the sites with an asterisk in Fig. 1) indicating utilization of incorrect nucleotides. On the other hand, the R78A mutant catalyzed less synthesis and shorter extension than the wild type enzyme in reactions with three dNTPs, indicative of higher replicational fidelity (Fig. 1, B-E). Interestingly, the R78A RT showed a larger reduction of the primer extension in the reactions with TTP (B) and dCTP (C) than in the reactions with dGTP (D) and dATP (E).

View larger version (83K): [in this window] [in a new window]   Fig. 1.   Misinsertion assay of wild type and R78A HIV-1 RT mutant proteins with DNA template. The 32P-labeled 17-mer primer (arrow) annealed to a 63-mer DNA template (5 nM) was extended by two different concentrations (4× and 1×) of wild type (40 and 10 nM) and R78A (75 and 25 nM) HIV-1 RT proteins at 37 °C for 5 min as seen in the extension reactions with all four dNTPs (A). The extension reactions were also performed in the presence of only 3 complementary dNTPs. B, minus TTP; C, minus dCTP; D, minus dGTP; E, minus dATP. The extension reactions were analyzed by 14% denaturing gel electrophoresis. The DNA sequence of the first 26 nucleotides of the extended part of the primer is shown in A. The sites with an asterisk indicate the stop sites where the deleted dNTPs would be incorporated in the reactions with only three dNTPs. In this assay, the higher efficiency of elongation of terminated primer with only three nucleotides will reflect the lower fidelity of the HIV-1 RT protein assayed. WT, wild type.

Next, we performed the misinsertion assay with an RNA template encoding the positive strand of the HIV-1 RT gene sequence. In this reaction, like the first round of replication of HIV-1, RT proteins produce a negative strand of viral DNA encoding the RT regions where most of the RT drug-resistant mutations have been identified. We also employed AMV RT as a high fidelity control RT. In the presence of all dNTPs, wild type HIV-1 RT, R78A HIV-1 RT, and AMV RT proteins extended the primer up to 250 nt. Two equal quantities of RNA-dependent DNA polymerase activity of three RT proteins showing the extension of the same 10 and 20% of total primer were determined by the reactions with all dNTPs (Fig. 2A). As shown in Fig. 2, B TTP; C, dGTP; D, dCTP; and E, dATP, AMV RT showed less extended primer than wild type HIV-1 RT in the presence of only three dNTPs, confirming that AMV RT has a higher replicational accuracy than wild type HIV-1 RT (3). AMV RT showed a larger reduction of the primer extension in the reactions with TTP (B) and dATP (E) than the reactions with dGTP (C) and dCTP (D). As seen in Fig. 2, B-E, the R78A mutant HIV-1 RT showed an even lower level of extended primer than AMV RT in the presence of three dNTPs. This result indicates that R78A HIV-1 RT could have enhanced fidelity. Interestingly, unlike AMV RT, R78A showed a larger reduction of the primer extension in the reactions with dGTP (C) as well as the reactions with TTP (B) and dATP (E).

View larger version (77K): [in this window] [in a new window]   Fig. 2.   Misinsertion assay of wild type HIV-1 RT, R78A HIV-1 RT mutant, and AMV RT with RNA template. The 32P-labeled 21-mer 3305 primer (arrow) annealed to a 1.7 kilonucleotide long RNA template encoding a positive strand of HIV-1 RT sequence was extended by two different concentrations (2× and 1×) of wild type HIV-1 RT (10 and 5 nM), R78A (30 and 15 nM), HIV-1 RT and AMV RT (0.1 and 0.05 units) showing equal RNA-dependent DNA polymerase activities at 37 °C for 5 min as seen in the extension reactions with all 4 dNTPs (A). The extension reactions were also performed in the presence of only three complementary dNTPs. B, minus TTP; C, minus dGTP; D, minus dCTP; E, minus dATP. The extension reactions were analyzed by 10% denaturing gel electrophoresis. The first 14 nucleotide HIV-1 sequences of the extended part of the primer are shown in A. The sites with an asterisk indicate the stop sites where the deleted dNTPs would be incorporated in the reactions with only three dNTPs. WT, wild type.

Like the R78A mutant, the Arg-78 HIV-1 RT mutants (R78S, R78L, R78I, R78T, R78V, R78C, and R78G) with high activity showed reduced levels of primer extension in the presence of only three dNTPs (data not shown). However, we could not determine the fidelity of some mutants (R78Y, R78E, and R78P) because of their largely reduced DNA polymerase activities.

LacZ Forward Mutation Assay of R78A HIV-1 RT Mutants-- This method has been used to determine mutation frequencies of DNA polymerases (27). Mutations generated when DNA polymerase copies the gapped region of the lacZ gene in M13mp2 can be scored by the number of plaques with altered color phenotype (pale blue or clear) in a specific indicator strain. As seen in Table I, the R78A HIV-1 RT mutant showed an 8.9-fold reduced rate of mutagenesis than the wild type HIV-1 RT; this mutation frequency of R78A was similar with that of AMV RT (3).

We have sequenced the entire 361-base pair lacZ gene from 23 mutant colonies obtained from copying reactions with wild type RT and 18 mutant colonies obtained with R78A RT. All colonies contained single mutations within the lacZ target, i.e. 21 base substitutions and 2 frameshifts for WT, and 15 base substitutions and 3 frameshifts for R78A. These data establish that all of the sequenced colonies were authentic mutants substantiating the large decrease in mutation frequency observed for R78A.

Determination of K_d of Two Hi-Fi HIV-1 RT Proteins with a DNA Template-- Our structural model suggested that the Asp-76 and Arg-78 residues interact with the template nucleotide. This model was basically confirmed by a structural study by Huang et al. (25) with the exception that the Arg-78 residue interacts with the Asp-76 residue and indirectly with the template nucleotide base pairing with the incoming dNTP. We tested whether mutations in these Asp-76 and Arg-78 residues affect the binding affinity of HIV-1 RT to DNA and RNA templates. We measured the K_d values of wild type, D76V, and R78A HIV-1 RT proteins to two templates: 1) a 63-mer DNA template annealed to a 17-mer primer (Fig. 1) and 2) a HIV-1 RT RNA annealed to the 3305 primer (Fig. 2). The D76V HIV-1 RT mutant is a hi-fi RT mutant that we previously characterized (22). The K_d value, indicative of binding affinity (28, 29), for each protein was calculated as described by Carroll et al. (28). As seen in Table II, the D76V mutant RT showed 6.7- and 5.6-fold lower K_d values than wild type RT with DNA and RNA templates, respectively. Conversely, the R78A mutant showed 7.6- and 6.0-fold higher K_d values than wild type with DNA and RNA templates. Tighter binding (lower K_d) of the D76V RT to the template might be because of the removal of the negatively charged Asp-76 residue, which interacts with the phosphate backbone of the template nucleotide base pairing to the incoming dNTP. Low binding (higher K_d) of the R78A HIV-1 RT mutant to the template might also be because of the removal of the positively charged Arg-78 residue, which interacts with the negative charge of Asp-76 that interacts with the ribose ring of the template nucleotide base pairing with the incoming dNTP (this study and 25).

Processivity of Asp-76 and Arg-78 HIV-1 RT Mutants with DNA and RNA Templates-- Because template binding affinity of these two mutants, R78A and D78V, was altered, we tested these hi-fi HIV-1 RT mutants for altered processivity. We determined the processivity of the wild type D76V and R78A HIV-1 RT mutant protein by using two different templates: poly(rA) annealed to the 20-mer oligo(dT) as a RNA template (Fig. 3A) and single strand M13 DNA annealed to the 20-mer forward primer (Fig. 3B). For these experiments we used equal quantities of RT activities for each of the three enzyme reactions; enzyme activity was standardized in terms of measured activity during a single round of the primer extension assay. As seen in the lanes marked "Processivity" (Fig. 3, A and B), under our reaction conditions wild type HIV-1 RT and D76V with poly(rA) template and M13 single-stranded DNA template extended the primer up to 160 and 80 nt, respectively. However, D76V RT showed a higher portion of largely extended primers (~80-160 nt for RNA template reactions and 40-80 nt in DNA template reactions) in the total extended primer than wild type RT. Average sizes of the total extended primers by wild type RT and D76V were approximately 90 and 110 nt in poly(rA) template and 25 and 35 nt in DNA template, suggesting D76V shows slightly higher processivity. However, R78A showed largely decreased average sizes of the primer in both RNA (60 nt) and DNA (10 nt). Therefore, mutations at the Arg-78 and Asp-76 residues of HIV-1 RT altered enzyme processivity. Our assay showed that the processivity of wild type RT with poly(rA) and M13 DNA templates was lower than previously reported (30). This discrepancy could be because of the higher KCl concentration (100 mM) used in our reaction conditions, which may reduce the processivity of HIV-1 RT as described by Huber et al. (30). Fig. 3A also shows two control experiments that confirmed the single round extension reaction during the processivity measurement. We used a trap to inhibit the binding of the free RT molecules to the template·primer. This trap contained heparin and excess molar poly(rA)/oligo(dT). The +Trap reactions of Fig. 3A show the control reactions that contained the poly(rA)/oligo(dT) and heparin trap before RT proteins were added; this completely blocked RT binding to the labeled template·primer. However, as seen in the Trap reactions of Fig. 3A, because of multiple rounds of primer extension in the absence of the trap, the primers were extended more completely than in the single round extension reactions. These two control experiments confirmed that the processivity of the three RT proteins was measured during a single round of extension. Both + and Trap controls were also performed with the DNA template, and the same results confirming the single round of extension were observed (data not shown).

View larger version (128K): [in this window] [in a new window]   Fig. 3.   Processivity of HIV-1 RT proteins with DNA and RNA templates. Poly(rA) annealed to a 20-mer oligo(dT) (A, RNA template) and M13mp18 single strand DNA annealed to a 20-mer forward primer (B, DNA template) were used in this assay. DV and RA indicate D76V and R78A HIV-1 RT mutant proteins, respectively. Processivity: RNA (A) or DNA T·P (B) was first preincubated with RT proteins (wild type (WT), D76V, R78A), and then the extension reactions were initiated by adding trap mixture containing dNTPs, poly(rA)/oligo(dT), and heparin (see "Experimental Procedures"). This condition allowed only a single round of primer extension by RT. +Trap, RNA T·P was first premixed with the trap mixture, and the extension reactions were initiated by adding RT proteins. This condition blocked RT binding to the labeled T·P. Trap, RNA T·P was premixed with RT proteins, and then the extension reactions were initiated by only dNTPs (without poly(rA)/oligo(dT) and heparin). This condition allowed multiple rounds of primer extension by RT proteins. All reactions were analyzed by 10% denaturing gel electrophoresis. WT, wild type.

Misincorporation of Wild Type and R78A SIV RT Proteins-- The Arg-78 residue is conserved between HIV-1 RT and SIV RT. We tested whether the R78A mutation affects replicational accuracy of SIV RT by using the misinsertion assay. As seen in Fig. 4, extension reactions with all four dNTPs (Fig. 4A) or only three dNTPs (Fig. 4, B and C) were performed with two different protein concentrations (4× and 1×) of two SIV RT proteins. As seen in Fig. 4, B (TTP) and C (dCTP), the SIV RT R78A mutant protein showed much less extended primer in the reactions with only three dNTPs than wild type SIV RT, suggesting that the R78A SIV RT mutant has a reduced misincorporation capability as observed for the HIV-1 in RT R78A and D78V mutants. We also performed misinsertion assays with RNA template encoding the HIV-1 sequence, and in this assay, the R78A SIV RT mutant also showed reduced misincorporation capability compared with the wild type SIV RT protein (data not shown). These data suggested that the SIV RT R78A mutant, like its HIV-1 counterpart, is a hi-fi mutant.

View larger version (102K): [in this window] [in a new window]   Fig. 4.   Misinsertion assay with wild type and R78A SIV RT mutants. The 32P-labeled 17-mer primer (arrow) annealed to a 63-mer DNA template (5 nM) was extended by two different concentrations (4× and 1×) of wild type and R78A SIV RT proteins at 37 °C for 5 min (see A for the reactions with all four dNTPs). The extension reactions were performed in the presence of all four (A) or only three complementary dNTPs. B, minus TTP; C, minus dCTP. The extension reactions were analyzed by 14% denaturing gel electrophoresis. WT, wild type; C, unextended primer.

Modeling Interactions of Arg-78 and Asp-76 Residues with Template Nucleotides-- Fig. 5 shows the structure for the region of HIV-1 RT interacting with the template (25). We previously thought that HIV-1 RT Asp-76 interacted with the template nucleotide base pairing to the incoming dNTPs and that Arg-78 directly interacted with the 5'-phosphate of the template nucleotide base pairing to the 3'-end of the primer. Our model had been constructed from 1hin.pdb (31) and 1hmi.pdb (32) with the former being a structure of an inhibited HIV-RT without bound nucleotides and the latter of HIV-RT with bound DNA but containing only coordinates for C atoms and phosphate backbone atoms. In this model, both Asp-76 and Arg-78 extended toward the template; Asp-76 extended toward the nucleotide that would bind to the incoming dNTP, and Arg-78 extended toward the adjacent nucleotide in the template. The recent 3.2 Å structure (25) of HIV-RT with template, primer, and an incoming bound nucleotide shows that in fact Asp-76 and Arg-78 are hydrogen bonded to each other, and Asp-76 forms close contacts with the ribose ring of the nucleotide bound to the incoming nucleotide. It is entirely possible that the interactions we proposed earlier could happen at a stage prior to the state represented by this most recent x-ray structure.

View larger version (121K): [in this window] [in a new window]   Fig. 5.   Structure of the interactions of residues Arg-78 and Asp-76 of HIV-1 RT with template nucleotides. This figure shows the local structure of HIV-1 RT protein complexed with template, primer, and incoming nucleotides (25) using coordinate set 1rtd.pdb from the Protein Data Bank (38). The single-stranded part of the template is shown in yellow, the double-stranded part of the template annealed to the primer (purple) is light blue, and the nucleotide base paired to the incoming nucleotide is blue-green. The finger domain containing Arg-78 (blue) and Asp-76 (red) is shown as a dark gray space and line representation, and the palm domain containing Glu-89 and Met-184 is a light gray space representation. The side chains of Arg-78 and Asp-76 face toward each other. Asp-76 contacts the ribose ring of the template nucleotide base pairing to the incoming dNTPs, whereas Arg-78 interacts with Asp-76 and indirectly to that template nucleotide. This figure was made using the programs Molscript (39) and Raster3D (33).

Our site-directed mutagenesis studies on Asp-76 (22) and Arg-78 residues (this study) suggest that the charges in these two residues are important in their roles in fidelity and interactions (K_d and processivity) with the template. Two other positions of HIV-1 RT, Met-184 and Glu-89, showing both drug resistance and increased fidelity, are also shown in this picture. Glu-89 interacts with the penultimate template nucleotide, whereas Met-184 at the conserved YMDD catalytic site is near to the dNTP binding site. Because of the difference in the structural locations of these fidelity mutations, it is plausible that mutations at the residues interacting with the template nucleotides (Asp-76, Arg-78, and Glu-89) might affect the replicational accuracy differently from mutations at Met-184 of HIV-1 RT (see "Discussion").

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Analysis of HIV-1 RT mutants with altered replicational accuracy is a key element in pinpointing molecular mechanisms of inaccurate DNA synthesis by HIV-1 RT. It is likely that HIV-1 RT mutants with largely altered fidelity might not be selected from infected hosts if RT is responsible for viral variation. Large alterations of HIV-1 RT fidelity could restrain the capability of HIV-1 to evolve and escape. Therefore, to obtain HIV-1 RT mutants with altered fidelity, we have employed both genetic and biochemical approaches. First, active HIV-1 RT mutants were isolated using a bacterial genetic system that selects for active enzymes generated from a randomly mutated pool of active HIV-1 RT mutants. Second, a misinsertion assay measuring the mutation synthesis capability of HIV-1 RT proteins was used to identify novel HIV-1 RT mutants with altered fidelity. These selected enzyme mutants were then further analyzed for their fidelity by using both the M13mp2 lacZ forward mutation assay and kinetic measurements for the misincorporation rate (f_mis). Previously we isolated an active D76V HIV-1 RT mutant with enhanced accuracy (22). This D76V HIV-1 RT mutant showed approximately 9-fold higher fidelity than wild type HIV-1 RT, and our earlier modeling studies suggested that the Asp-76 residue interacts with the template nucleotide base pairing with the incoming dNTP (Fig. 5). Based on this information, we hypothesized that, in wild type HIV-1 RT, the interaction of the Asp-76 residue with this template nucleotide occasionally interferes with base pairing of the template nucleotide to the incoming dNTP, resulting in frequent mutation synthesis. This model also suggested that the Arg-78 residue directly interacts with 5'-phosphate of the template nucleotide base pairing the primer nucleotide at the 3'-end of the primer, which is the same as the 3'-phosphate of the template nucleotide both base pairing to the incoming dNTP and interacting with the Asp-76 residue.

In this study, we tested whether alterations in the proposed interaction between the Arg-78 residue of HIV-1 RT and the template nucleotide base pairing to the 3'-end primer nucleotide also affect replicational accuracy. As seen in misinsertion assays (Figs. 1 and 2) and the M13mp2 lacZ forward mutation assay (Table I), the R78A HIV-1 RT mutant protein showed 8.9 times higher fidelity, confirming our prediction that the Arg-78 residue of HIV-1 RT is important in enzymatic infidelity. To confirm the interactions of both the Arg-78 residue and the Asp-76 residue with the template nucleotides, K_d values of three RT proteins, wild type, R78A, and D76V, with DNA and RNA templates were determined. The D76V mutant protein showed lower K_d than the wild type enzyme, whereas the R78A protein showed a much higher K_d than wild type for both templates. Tighter binding of D76V to the template might be because of the removal of the negatively charged Asp-76 residue that in turn allows Arg-78 to interact with the phosphate backbone of the template nucleotide. In contrast a previously described (E89G) hi-fi mutant affecting the interaction of the penultimate nucleotide of the double strand part of the template showed the same level of K_d for RNA templates as did wild type RT (26). Instead, the E89G mutant showed reduced k_off compared with wild type. It is therefore likely that the Asp-76 and Glu-89 residues may have different roles in template binding because they interact with different template nucleotides. In contrast to the D76V mutant, the lower binding affinity of R78A to the template might result from the removal of the positive charge of the Arg-78 residue that interacts with the negative charge of Asp-76, which interacts with the template nucleotide.

The R78A hi-fi mutant showed largely decreased processivity in both RNA and DNA templates, suggesting that removal of the interaction between the Arg-78 and Asp-76 residues influences the ability to extend the primer as well as initial binding (K_d) to the template. Many primer grip HIV-1 RT mutants have shown both a decreased processivity and an increased frameshift mutation rate but not a point mutation rate (5, 19-21). It was proposed that the increased frameshift mutation rate of these primer grip mutants might result from the increased mutation synthesis mediated by a T·P slippage mechanism. It is plausible that the decreased processivity by the R78A mutation might not affect T·P slippage-mediated mutation synthesis. In other words, the R78A mutation and primer grip mutations may decrease processivity by different mechanisms, as suggested by the structural models of HIV-1 RT (20, 21). M184V and M184I 3TC resistant mutant RT proteins with increased fidelity also have processivity defects (34), and it was proposed that the imbalance of dNTP pools be a factor for the processivity defect of Met-184 drug-resistant mutants (34, 35).

The potential physiologic relevance of the R78A substitution can be inferred from our misincorporation data and the fact that the GA transitions in the plus strand are believed to be the most common and important mutation synthesized by HIV in vivo (36, 37). These transitions can arise from G:T mispairs formed and extended during synthesis on an RNA template or C:A mispairs formed and extended during synthesis on a DNA template. The R78A mutant exhibited reduced misincorporation in the "minus C" reactions with an RNA template (Fig. 2) and the "minus G" reaction with the DNA template (Fig. 1), suggesting that it may catalyze fewer of the mutations believed to be most crucial to HIV physiology.

Presumably, wild type HIV-1 RT evolved to use both residues Asp-76 and Arg-78 of HIV-1 RT to create an active site structure that favors occasional mutation synthesis. We reason that Arg-78 has been retained during HIV-1 RT evolution because it supports both catalytic activity and viral mutability. It is apparent that HIV-1 has evolved to be hypermutable, presumably because hypermutability promotes evasion of antiviral pressures. If an amino acid affects both catalytic activity and mutability, as does Arg-78, the outcome of natural selection must, perforce, reflect the contribution of that amino acid to both phenotypes. The misinsertion assays with SIV RT proteins suggest that SIV, like HIV-1 RT, might also employ the Arg-78 residue to create a mutagenic DNA polymerase.

Many residues of HIV-1 RT interacting with the template near the DNA polymerase active site of HIV-1 RT have been identified (25). They are the Asp-76, Arg-78, Lys-154, Gly-152, and Glu-89 residues. Among these residues, mutations in Asp-76, Arg-78, and Glu-89 affect both accuracy and interactions with the templates. We recently tested the fidelity and processivity of the HIV-1 RT K154 mutants, and we found that, unlike the Asp-76, R78A, and Glu-89 mutants, Lys-154 mutants affect only processivity and not fidelity, suggesting that not all residues involved in template interactions contribute to HIV-1 RT infidelity (data not shown). Therefore, three RT residues of HIV-1 RT (Asp-76, Arg-78, and Glu-89), which are involved in both replicational accuracy and processivity, provide HIV-1 RT with a mechanism that allows the virus to sustain efficient mutation synthesis required for viral evolution and escape.

	ACKNOWLEDGEMENTS

We thank Drs. Lawrence A. Loeb, Robert A. Bambara, Ann Blank, and Stephen Dewhurst and Peter Gerondelis for critical reading of the manuscript.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant R29-GM55500 (to B. K.). Structure modeling was supported by NIEHS National Institutes of Health Grant P30-ES07033 (to E. T. A.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ To whom correspondence should be addressed: Box 672, 601 Elmwood Ave., Dept. of Microbiology and Immunology, University of Rochester Medical Center, Rochester, NY 14642. Tel.: (716) 275-6916; Fax: (716) 473-9573; E-mail: baek_kim@urmc.rochester.edu.

	ABBREVIATIONS

The abbreviations used are: HIV-1, human immunodeficiency virus, type 1; RT, reverse transcriptase; AMV, avian myeloblastosis virus; PCR, polymerase chain reaction; T·P, template·primer complex; p66, 66-kDa RT polypeptide; p51, 51-kDa RT polypeptide; nt, nucleotide(s); SIV, Simian immunodeficiency virus.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				C. A. Howell, C. M. Kondratick, and M. T. Washington Substitution of a residue contacting the triphosphate moiety of the incoming nucleotide increases the fidelity of yeast DNA polymerase {zeta} Nucleic Acids Res., March 1, 2008; 36(5): 1731 - 1740. [Abstract] [Full Text] [PDF]

				C. Ferrer-Orta, A. Arias, R. Perez-Luque, C. Escarmis, E. Domingo, and N. Verdaguer Sequential structures provide insights into the fidelity of RNA replication PNAS, May 29, 2007; 104(22): 9463 - 9468. [Abstract] [Full Text] [PDF]