High fidelity of internal strand transfer catalyzed by human immunodeficiency virus reverse transcriptase

Resolution of Specific Nucleotide Mismatches by Wild-Type and AZT-Resistant Reverse Transcriptases during HIV-1 Replication

A key contributor to HIV-1 genetic variation is reverse transcriptase errors. Some mutations result because reverse transcriptase (RT) lacks 3' to 5' proofreading exonuclease and can extend mismatches. However, RT also excises terminal nucleotides to a limited extent, and this activity contributes to AZT resistance. Because HIV-1 mismatch resolution has been studied in vitro but only indirectly during replication, we developed a novel system to study mismatched base pair resolution during HIV-1 replication in cultured cells using vectors that force template switching at defined locations. These vectors generated mismatched reverse transcription intermediates, with proviral products diagnostic of mismatch resolution mechanisms. Outcomes for wild-type (WT) RT and an AZT-resistant (AZT(R)) RT containing a thymidine analog mutation set-D67N, K70R, D215F, and K219Q-were compared. AZT(R) RT did not excise terminal nucleotides more frequently than WT, and for the majority of tested mismatches, both WT and AZT(R) RTs extended mismatches in more than 90% of proviruses. However, striking enzyme-specific differences were observed for one mispair, with WT RT preferentially resolving dC-rC pairs either by excising the mismatched base or switching templates prematurely, while AZT(R) RT primarily misaligned the primer strand, causing deletions via dislocation mutagenesis. Overall, the results confirmed HIV-1 RT's high capacity for mismatch extension during virus replication and revealed dramatic differences in aberrant intermediate resolution repertoires between WT and AZT(R) RTs on one mismatched replication intermediate. Correlating mismatch extension frequencies observed here with reported viral mutation rates suggests a complex interplay of nucleotide discrimination and mismatch extension drives HIV-1 mutagenesis.

Primer Extension Reactions for the PCR- based α- complementation Assay

The PCR- based- α- complementation assay is an effective technique to measure the fidelity of polymerases, especially RNA-dependent RNA polymerases (RDRP) and Reverse Transcriptases (RT). It has been successfully employed to determine the fidelity of the poliovirus polymerase 3D-pol (DeStefano, 2010) as well as the human immunodeficiency virus Reverse Transcriptase (HIV RT) (Achuthan et al., 2014). A major advantage of the assay is that since the PCR step is involved, even the low yield of products obtained after two rounds of low yield of RNA synthesis (for RDRP) or reverse transcription (for RT) can be measured using the assay. The assay also mimics the reverse transcription process, since both RNA- and DNA- directed RT synthesis steps are performed. We recently used this assay to show that the HIV RT, at physiologically relevant magnesium concentration, has accuracy in the same range as other reverse transcriptases (Achuthan et al., 2014). Here, we describe in detail how to prepare the inserts using the primer extension reactions. The prepared inserts are then processed further in the PCR- based- α- complementation assay.

Alternative divalent cations (Zn²⁺, Co²⁺, and Mn²⁺) are not mutagenic at conditions optimal for HIV-1 reverse transcriptase activity

BACKGROUND

Fidelity of DNA polymerases can be influenced by cation co-factors. Physiologically, Mg(2+) is used as a co-factor by HIV reverse transcriptase (RT) to perform catalysis; however, alternative cations including Mn(2+), Co(2+), and Zn(2+) can also support catalysis. Although Zn(2+) supports DNA synthesis, it inhibits HIV RT by significantly modifying RT catalysis. Zn(2+) is currently being investigated as a component of novel treatment options against HIV and we wanted to investigate the fidelity of RT with Zn(2+).

METHODS

We used PCR-based and plasmid-based alpha complementation assays as well as steady-state misinsertion and misincorporation assays to examine the fidelity of RT with Mn(2+), Co(2+), and Zn(2+).

RESULTS

The fidelity of DNA synthesis by HIV-1 RT was approximately 2.5 fold greater in Zn(2+) when compared to Mg(2+) at cation conditions optimized for nucleotide catalysis. Consistent with this, RT extended primers with mismatched 3' nucleotides poorly and inserted incorrect nucleotides less efficiently using Zn(2+) than Mg(2+). In agreement with previous literature, we observed that Mn(2+) and Co(2+) dramatically decreased the fidelity of RT at highly elevated concentrations (6 mM). However, surprisingly, the fidelity of HIV RT with Mn(2+) and Co(2+) remained similar to Mg(2+) at lower concentrations that are optimal for catalysis.

CONCLUSION

This study shows that Zn(2+), at optimal extension conditions, increases the fidelity of HIV-1 RT and challenges the notion that alternative cations capable of supporting polymerase catalysis are inherently mutagenic.

Human immunodeficiency virus reverse transcriptase displays dramatically higher fidelity under physiological magnesium conditions in vitro

The fidelity of human immunodeficiency virus (HIV) reverse transcriptase (RT) has been a subject of intensive investigation. The mutation frequencies for the purified enzyme in vitro vary widely but are typically in the 10(-4) range (per nucleotide addition), making the enzyme severalfold less accurate than most polymerases, including other RTs. This has often been cited as a factor in HIV's accelerated generation of genetic diversity. However, cellular experiments suggest that HIV does not have significantly lower fidelity than other retroviruses and shows a mutation frequency in the 10(-5) range. In this report, we reconcile, at least in part, these discrepancies by showing that HIV RT fidelity in vitro is in the same range as cellular results from experiments conducted with physiological (for lymphocytes) concentrations of free Mg(2+) (~0.25 mM) and is comparable to Moloney murine leukemia virus (MuLV) RT fidelity. The physiological conditions produced mutation rates that were 5 to 10 times lower than those obtained under typically employed in vitro conditions optimized for RT activity (5 to 10 mM Mg(2+)). These results were consistent in both commonly used lacZα complementation and steady-state fidelity assays. Interestingly, although HIV RT showed severalfold-lower fidelity under high-Mg(2+) (6 mM) conditions, MuLV RT fidelity was insensitive to Mg(2+). Overall, the results indicate that the fidelity of HIV replication in cells is compatible with findings of experiments carried out in vitro with purified HIV RT, providing more physiological conditions are used.

IMPORTANCE

Human immunodeficiency virus rapidly evolves through the generation and subsequent selection of mutants that can circumvent the immune response and escape drug therapy. This process is fueled, in part, by the presumably highly error-prone HIV polymerase reverse transcriptase (RT). Paradoxically, results of studies examining HIV replication in cells indicate an error frequency that is ~10 times lower than the rate for RT in the test tube, which invokes the possibility of factors that make RT more accurate in cells. This study brings the cellular and test tube results in closer agreement by showing that HIV RT is not more error prone than other RTs and, when assayed under physiological magnesium conditions, has a much lower error rate than in typical assays conducted using conditions optimized for enzyme activity.

Fifteen to twenty percent of HIV substitution mutations are associated with recombination

HIV undergoes high rates of mutation and recombination during reverse transcription, but it is not known whether these events occur independently or are linked mechanistically. Here we used a system of silent marker mutations in HIV and a single round of infection in primary T lymphocytes combined with a high-throughput sequencing and mathematical modeling approach to directly estimate the viral recombination and mutation rates. From >7 million nucleotides (nt) of sequences from HIV infection, we observed 4,801 recombination events and 859 substitution mutations (≈1.51 and 0.12 events per 1,000 nt, respectively). We used experimental controls to account for PCR-induced and transfection-induced recombination and sequencing error. We found that the single-cycle virus-induced mutation rate is 4.6 × 10(-5) mutations per nt after correction. By sorting of our data into recombined and nonrecombined sequences, we found a significantly higher mutation rate in recombined regions (P = 0.003 by Fisher's exact test). We used a permutation approach to eliminate a number of potential confounding factors and confirm that mutation occurs around the site of recombination and is not simply colocated in the genome. By comparing mutation rates in recombined and nonrecombined regions, we found that recombination-associated mutations account for 15 to 20% of all mutations occurring during reverse transcription.

The origin of genetic diversity in HIV-1

One of the hallmarks of HIV infection is the rapid development of a genetically complex population (quasispecies) from an initially limited number of infectious particles. Genetic diversity remains one of the major obstacles to eradication of HIV. The viral quasispecies can respond rapidly to selective pressures, such as that imposed by the immune system and antiretroviral therapy, and frustrates vaccine design efforts. Two unique features of retroviral replication are responsible for the unprecedented variation generated during infection. First, mutations are frequently introduced into the viral genome by the error prone viral reverse transcriptase and through the actions of host cellular factors, such as the APOBEC family of nucleic acid editing enzymes. Second, the HIV reverse transcriptase can utilize both copies of the co-packaged viral genome in a process termed retroviral recombination. When the co-packaged viral genomes are genetically different, retroviral recombination can lead to the shuffling of mutations between viral genomes in the quasispecies. This review outlines the stages of the retroviral life cycle where genetic variation is introduced, focusing on the principal mechanisms of mutation and recombination. Understanding the mechanistic origin of genetic diversity is essential to combating HIV.

Effect of reaction conditions and 3AB on the mutation rate of poliovirus RNA-dependent RNA polymerase in a alpha-complementation assay

The fidelity of poliovirus RNA-dependent RNA polymerase (3Dpol) was determined using a modified alpha-complementation assay. Several polymerases have been analyzed by this approach allowing comparisons to be drawn. Various conditions including high and low MgCl(2), replacing MgCl(2) with MnCl(2), skewed nucleotide pools, and the presence of poliovirus protein 3AB were analyzed. The assay included RNA synthesis by 3Dpol on an RNA template that coded for a region of the alpha peptide of beta-galactosidase (lacZ-alpha). The product of this reaction was used as a template for a second round of 3Dpol synthesis and the resulting RNA was reverse transcribed to DNA by reverse transcriptase. The DNA was amplified by PCR and inserted into a vector used to transform Escherichia coli. The bacteria were screened for beta-galactosidase activity by blue-white phenotype analysis with white or faint blue colonies scored as errors made during synthesis on lacZ-alpha. Although 3AB strongly stimulated 3Dpol synthesis as expected, no change in fidelity was detected. Changes in MgCl(2) also showed little effect. Mutation rates of approximately 9 x 10(-5) (approximately 1 error per 11,000 incorporations) were estimated for these conditions. In contrast, MnCl(2) or skewed nucleotide pools were highly mutagenic resulting in lowered fidelity.

Superinfection of defective human immunodeficiency virus type 1 with different subtypes of wild-type virus efficiently produces infectious variants with the initial viral phenotypes by complementation followed by recombination

Superinfection rates of human immunodeficiency virus type 1 (HIV-1) have increasingly been leading to more variation in HIV-1, as evidenced by the emergence of circulating recombinant forms (CRFs). We recently reported complementation in a persistently replication-defective subtype B-infected cell clone, L-2, by superinfection with CRF15_01B. The L-2 cells continuously produce immature particles due to a one-base insertion at pol protease. Proviruses in the superinfected cells carried both subtypes and produced particles with a mature morphology. In this study, we examined possible recombination following complementation to generate replication-competent variants by using three cell clones prepared from superinfected L-2 cells. The individual clones predominantly expressed the initial subtype B-derived mature Gag proteins. However, the viral particles carried both subtype B with the mutation and wild-type CRF15_01B at pol, suggesting the generation of virions with heterozygous RNAs. Interestingly, with cell-free passages of the progeny, defective particles disappeared, and were replaced with heterogeneous recombinants in the pol region with sequences derived from CRF15_01B that expressed subtype B phenotype. Thus, even a defective form of persistent HIV-1 can become replication-competent through superinfection-mediated complementation followed by recombination. These findings suggest the significance of long-lived infected cells as recipients for superinfection.

Random strand transfer recombination (RSTR) for homology-independent nucleic acid recombination

Random strand transfer recombination (RSTR) relies on the ability of viral reverse transcriptases to undergo homology-independent template switches during DNA synthesis. To facilitate strand transfer events single stranded template DNA was hybridized via noncoding complementary sequence stretches which flank the genes. The resulting bulb shaped heteroduplex was used as template for the reverse transcriptase driven DNA polymerization in which template switch events occur spontaneously. Switches were promoted by induced pausing of the viral polymerase at digoxygenin labeled nucleotides. RSTR of homologous genes was demonstrated by the recombination of two genes encoding (R)-hydroxynitrilase isoenzymes from Prunus amygdolus. For non-homologous RSTR we recombined the genes of the Havea brasiliensis (S)-hydroxynitrilase and the estC gene from Burkholderia gladioli. Base-pairing dependent recombination took place spontaneously and at high frequency between genes with low and high sequence homology.

Analysis of mutations made during active synthesis or extension of mismatched substrates further define the mechanism of HIV-RT mutagenesis

The effect of reverse transcriptase (RT) catalyzed mutations on continued extension of the nascent DNA chain was investigated. A system using the alpha-lac gene of beta-galactosidase as template and two sets of conditions was used. In one, RT was allowed to reassociate with the primer-template after falling off, while in a second RT was sequestered after dissociating. In the first condition, subsequent extension of errors that may have initially caused enzyme dissociation can occur. In the second, such errors would not be extended. Fully extended products were assayed by alpha-complementation to assess mutation frequency. A lower frequency in the latter scenario implies that some errors caused the polymerase to dissociate. Allowing only a single binding event lowered the mutation frequency of the products by about 1/2 suggesting that approximately 1 in 2 errors terminated synthesis. In other experiments, when added to a primer-template with a terminal mismatch at the 3' end, RT dissociated from the template about 50-90% of the time (depending on mismatch type) rather than extending. Running start reactions indicated that extension was more likely if an actively synthesizing RT made the mutation. RT RNase H cleavage analysis showed that 3' mismatches weakened the association of RT with the primer-terminus. Taken together, these results suggest that an actively synthesizing RT enzyme that has just made a mistake is likely bound in a configuration that generally enhances extension of the mistake. This is in contrast to RTs that must bind to then extend mismatches. The importance of these findings with respect to the mechanism of mutagenesis is discussed.

Molecular basis of fidelity of DNA synthesis and nucleotide specificity of retroviral reverse transcriptases

Reverse transcription involves the conversion of viral genomic RNAinto proviral double-stranded DNA that integrates into the host cell genome. Cellular DNA polymerases replicate the integrated viral DNA and RNA polymerase II transcribes the proviral DNA into RNA genomes that are packaged into virions. Although mutations can be introduced at any of these replication steps, reverse transcriptase (RT) errors play a major role in retroviral mutation. This review summarizes our current knowledge on fidelity of reverse transcriptases. Estimates of retroviral mutation rates or fidelity of retroviral RTs are discussed in the context of the different techniques used for this purpose (i.e., retroviral vectors replicated in culture, misinsertion and mispair extension fidelity assay, etc.). In vitro fidelity assays provide information on the RT's accuracy during the elongation reaction of DNA synthesis. In addition, other steps such as initiation of reverse transcription, or strand transfer, and factors including viral proteins such as Vpr [in the case of the human immunodeficiency virus type 1 (HIV-1)] have been shown to influence fidelity. A comprehensive description of the effect of amino acid substitutions on the fidelity of HIV-1 RT is presented. Published data point to certain dNTP-binding residues, as well as to various amino acids involved in interactions with the template or the primer strand, and to residues in the minor groove-binding track as major components of the fidelity center of retroviral RTs. Implications of these studies include the design of novel therapeutic strategies leading to virus extinction, by increasing the viral mutation rate beyond a tolerable threshold.

Structures of complexes formed by HIV-1 reverse transcriptase at a termination site of DNA synthesis

This study presents structural parameters associated with termination of human immunodeficiency virus, type 1 (HIV-1) reverse transcriptase (RT) at Ter2, the major termination site located in the center of the HIV-1 genome. DNA footprinting studies of various elongation complexes formed by RT around wild type and mutant Ter2 sites have revealed two major structural transformations of these complexes when the enzyme gets closer to Ter2. First, the interactions between RT and the DNA duplex are less extended, although the global affinity of the enzyme for this duplex is only decreased by 2-fold. Second, there is an atypical positioning of the RT RNase H domain on the DNA duplex. We interpret our data as indicating that the A(n)T(m) motif located upstream of Ter2 prevents a classical positioning of the enzyme on the double-stranded part of the DNA duplex at some precise positions of elongation downstream of this motif. Instead, novel species of binary and/or ternary complexes, characterized by atypical footprints, are formed. The new rate-limiting step of the reaction, characterized in the preceding paper (Lavigne, M., Polomack, L., and Buc, H. (2001) J. Biol. Chem. 276, 31429-31438), would be a transition leading from these new species to a catalytically competent ternary complex.

Determination of the mutation rate of poliovirus RNA-dependent RNA polymerase

The fidelity of poliovirus RNA-dependent RNA polymerase (3D(pol)) was determined using a system based on the fidelity of synthesis of the alpha-lac gene which codes for a subunit of beta-galactosidase. Synthesis products are screened for mutations by an alpha-complementation assay, in which the protein product from alpha-lac is used in trans to complement beta-galactosidase activity in bacteria that do not express alpha-Lac. Several polymerases have been analyzed by this approach allowing comparisons to be drawn. The assay included RNA synthesis by 3D(pol) on an RNA template that coded for the N-terminal region of alpha-Lac. The product of this reaction was used as a template for a second round of 3D(pol) synthesis and the resulting RNA was reverse transcribed to DNA by MMLV-RT. The DNA was amplified by PCR and inserted into a vector used to transform Escherichia coli. The bacteria were screened for beta-galactosidase activity by blue-white phenotype analysis with white or faint blue colonies scored as errors made during synthesis on alpha-lac. Results showed a mutation rate for 3D(pol) corresponding to approximately 4.5x10(-4) errors per base (one error in approximately 2200 bases). Analysis of mutations showed that base substitutions occurred with greater frequency than deletions and insertions.

In vitro strand transfer from broken RNAs results in mismatch but not frameshift mutations

An in vitro system to compare the fidelity of strand transfers from truncated vs full-length RNAs was constructed. A donor RNA, on which reverse transcriptase (RT)-directed DNA synthesis was initiated, shared homology with an acceptor RNA, to which DNAs initiated on the donor could transfer. All RNAs were derived from the N-terminal portion of the alpha-lac gene. On full-length donors, transfers occurred when DNAs migrated to the acceptor prior to being completed on the donor. On donors that were truncated, most transfers occurred after DNAs reached the end of the donor. Transfer products were amplified by PCR and used to replace the corresponding region in a vector containing the alpha-lac gene. Transformed Escherichia coli were screened for alpha-complementation by blue-white phenotype analysis, with white colonies scored as those with errors in alpha-lac. These errors were derived from RT synthesis and strand transfer. The mutant colony frequency approximately doubled for transfer products derived from truncated donors (0.026+/-0.005 vs. 0.053+/-0.011 (three experiments +/- SD), for full-length vs. truncated, respectively). The increases resulted from additional non-template-directed bases (mostly thymidines) added to the DNAs before transfer. Sequence analysis of DNAs synthesized on truncated donors showed that about 60% had additions (20/34); however, those without additions transferred at a much higher rate than those with. Transfer of the DNAs with additions always resulted in substitutions; no frameshifts were observed. Results are consistent with RT adding nontemplated nucleotides at template termini. Transfer and subsequent extension of these products is severely inhibited relative to products without additions. The potential relevance of these findings to retrovirus replication is discussed.

Effect of distance between homologous sequences and 3' homology on the frequency of retroviral reverse transcriptase template switching

Deletion of direct repeats in retroviral genomes provides an in vivo system for analysis of reverse transcriptase (RT) template switching. The effect of distance between direct repeats on the rate of deletion was determined for 16 murine leukemia virus (MLV)-based vectors containing a 701-bp direct repeat of overlapping fragments of the herpes simplex virus thymidine kinase gene (HTK). The direct repeats were separated by spacer fragments of various lengths (0.1 to 3.5 kb). Southern analysis of infected cells after one replication cycle indicated that all vectors in which the distance between homologous sequences was >1,500 bp deleted at very high rates (>90%). In contrast, vectors containing <1,500 bp between homologous sequences exhibited lower frequencies of deletion (37 to 82%). To analyze the pattern of locations at which RT switched templates, restriction site markers were introduced to divide the downstream direct repeat into five regions. RT switched templates within all five regions of the 701-bp direct repeat and the frequency of template switching was greater within the 5' regions in comparison to the 3' regions. The probability of RT switching templates within the 5' regions doubled when the MLV packaging sequence (Psi) was placed between the 701-bp direct repeats. However, Psi did not increase the rate of template switching for shorter direct repeats. These results indicate that linear distance between homologous sequences increases the rate of template switching and suggest that duplex formation between nascent DNA and homologous template sequences 3' of RT promote template switching.

RNase H requirements for the second strand transfer reaction of human immunodeficiency virus type 1 reverse transcription

Retroviral reverse transcriptase (RT) enzymes are responsible for transcribing viral RNA into double-stranded DNA. An in vitro assay to analyze the second strand transfer event during human immunodeficiency virus type 1 (HIV-1) reverse transcription has been developed. Model substrates were constructed which mimic the viral intermediate found during plus-strand strong-stop synthesis. Utilizing wild-type HIV-1 RT and a mutant E478Q RT, the requirement for RNase H activity in this strand transfer event was analyzed. In the presence of Mg2+, HIV-1 RT was able to fully support the second strand transfer reaction in vitro. However, in the presence of Mg2+, the E478Q RT mutant had no detectable RNase H activity and was unable to support strand transfer. In the presence of Mn2+, the E478Q RT yields the initial endoribonucleolytic cleavage at the penultimate C residue of the tRNA primer yet does not support strand transfer. This suggests that subsequent degradation of the RNA primer by the RNase H domain was required for strand transfer. In reactions in which the E478Q RT was complemented with exogenous RNase H enzymes, strand transfer was supported. Additionally, we have shown that HIV-1 RT is capable of supporting strand transfer with substrates that mimic tRNAHis as well as the authentic tRNA3Lys.