| Nucleic Acids Research | Pages |
The processivity of DNA synthesis exhibited by drug-resistant variants of human immunodeficiency virus type-1 reverse transcriptase
Introduction
Materials And Methods
Enzymes
DNA polymerization and processivity experiments
Results
DNA synthesis and relative processivity of the DNA-dependent DNA polymerase activity of the HIV-1 RT variants
Processivity and synthesis of DNA on an RNA template by the HIV-1 RT variants
Discussion
Acknowledgements
References
The processivity of DNA synthesis exhibited by drug-resistant variants of human immunodeficiency virus type-1 reverse transcriptase
ABSTRACT
INTRODUCTION
Retroviruses encode the enzyme reverse transcriptase (RT), which plays a key role in the early stages of their life cycle by catalyzing the process of converting the viral single-stranded RNA into double-stranded DNA. The enzyme has two catalytic functions: a DNA polymerase activity, which copies both RNA and DNA templates, and a ribonuclease H (RNase H) activity, that selectively degrades RNA in RNA·DNA hybrids (1-3). Human immunodeficiency viruses type 1 and type 2 (HIV-1 and HIV-2 respectively), the etiological agents of acquired immunodeficiency syndrome (AIDS), represent a distinct subfamily of retroviruses, the lentiviruses (4-6). Many of the anti-HIV drugs used so far for treatment of AIDS patients are inhibitors of HIV RTs (7-8). Yet prolonged exposure to these drugs leads to rapid emergence of drug-resistant virions harboring enzymatically active RT variants with a reduced sensitivity to these inhibitors. This phenomenon is partly due to the high error rates in copying both RNA and DNA observed with HIV RTs (3,9-12). All RTs lack the 3[prime]->5[prime] exonuclease activity necessary for proof-reading during DNA synthesis. Yet the error-proneness of HIV RT is possibly higher than that observed with non-lentiviral RTs. Therefore, it was suggested that this may reflect unique structural and mechanistic features of lentiviral RTs (12).
Most inhibitors of RTs can be grouped into two classes: non-nucleoside inhibitors and nucleoside analogs. The first group, which inhibits HIV-1 RT with no effect on HIV-2 RT, binds to a hydrophobic pocket near the active site of DNA polymerase, modifying amino acids in the drug binding pocket (13-16). Nucleoside analogs which lack a 3[prime]-hydroxyl group mimic the normal dNTPs and, hence, act as chain terminators of DNA synthesis. Resistance to these analogs is quite complex, since most substituted amino acid residues do not cluster around the putative dNTP binding site of the DNA polymerase subdomain and many of the mutations involved in conferring resistance are located in positions that are in contact with the template-primer substrate (15,16). These structural data were confirmed by biochemical analyses of drug-resistant mutants of both HIV-1 and HIV-2 RTs, which suggest that reduced sensitivity to the analogs can be caused by repositioning of the nucleic acid template-primer (17,18).
The relatively low fidelity of DNA synthesis by a variety of DNA polymerases, including lentiviral RTs, was shown in vitro to result primarily from an ability to incorporate the wrong nucleotide at the 3[prime]-end of the nascent DNA strand and to extend the mispaired DNA (3,9-12,19,20). There might be a potential correlation between HIV RT resistance to nucleoside analogs (namely the ability to incorporate or reject an incoming analog) and fidelity of misinsertion and mispair extension. This presumption was supported by recent in vitro studies of several drug-resistant variants of HIV-1 RT, which showed a correlation between drug resistance and fidelity of DNA synthesis. Thus a decrease in the ability to incorporate nucleoside analogs (which is manifested by increased drug resistance) is accompanied by a reduction in the capacity to misincorporate and extend mismatches (which means enhanced fidelity). This phenomenon was reported for a variety of mutants of HIV-1 RT: Leu74Val, resistant to 3[prime]-azido-3[prime]-deoxythymidine triphosphate (AZTTP); Glu89Gly, resistant to ddGTP and ddITP; Met184Leu and Met184Val, both resistant to 2[prime]3[prime]-dideoxy-3-thiacytidine (3TC) (21-26).
There might be a potential association between fidelity of misinsertion and mispair extension and processivity of DNA synthesis. Processivity of a polymerase is directly proportional to the length of nascent polymeric products formed before the enzyme molecules dissociate from the product molecules and rebind other template-primer molecules. It is possible that reduced processivity of a given RT variant, which is coupled with reduced association of the RT with the 3[prime]-end of the primer, can be associated with a decrease in efficiency of 3[prime] misincorporation and 3[prime] mispair extension. Such a correlation was recently shown for two 3TC-resistant HIV-1 RT mutants, Met184Val and Met184Ile, that synthesize DNA with a higher accuracy than wild-type RT and also exhibit reduced processivity of synthesis. This led to the suggestion that there is an inverse correlation between the fidelity and processivity of DNA synthesis by the HIV-1 RT variants (26). Such a connection, if substantiated, would predict that the apparent enhanced fidelity of misinsertion and mispair extension in vitro is the outcome of reduced processivity of a given DNA polymerase. To further explore this potential mechanistic linkage we have studied other drug-resistant mutants of HIV-1 RT for their processivity of DNA synthesis on both RNA and DNA templates. These mutants, Leu74Val, Glu89Gly, Tyr183Phe and Met184Leu, have already been shown in vitro to have enhanced fidelity of DNA synthesis (21,24). The present comparative study indicates that, apart from the Met184 mutant, all the other variants do not support the inverse correlation between fidelity and processivity. If anything, these three mutants have a relative processivity higher than that of wild-type RT.
MATERIALS AND METHODS
Enzymes
The HIV-1 RTs used in this report are recombinant enzymes expressed in DH5[alpha] Escherichia coli and purified from bacterial extracts. The wild-type and mutant version p66/p51 heterodimers of HIV-1 RT, derived from the BH-10 clone of HIV-1, were expressed in bacteria as described in detail previously (27). The mutated genes were obtained by the cassette mutagenesis technique (28) and include six consecutive histidine residues at the N-termini. The proteins were purified as described (29), resulting in p66/p51 heterodimers in which the specified mutation resides in both subunits. The specific DNA polymerase activities of all RTs used were between 3000 and 10 000 U/µg protein determined as described earlier (30). About equal activities of all enzymes were used for the extension and processivity experiments.
DNA polymerization and processivity experiments
We used two sets of template-primer. For DNA-dependent DNA synthesis single-stranded circular [phis]X174am3 DNA primed with a synthetic 15mer oligonucleotide that hybridizes at positions 588-602 of the DNA was employed (31). The primer sequence is 5[prime]-AAAGCGAGGGTATCC-3[prime]. RNA-dependent DNA synthesis was carried out with 16S E.coli rRNA primed with a 16mer synthetic oligonucleotide that hybridizes to nt 2112-2127 of the RNA (32). The primer sequence is 5[prime]-ATTTCACATCTGACTT-3[prime]. The two primers were labeled at their 5[prime]-end with 32P using [[gamma]-32P]ATP and T4 polynucleotide kinase as described (10-12) and were annealed to their templates with a 2-fold molar excess of unlabeled templates over primers. All reactions were performed in 20 mM Tris-HCl, 2 mM DTT, 10 mM MgCl2 and 0.1 mg/ml bovine serum albumin at a final pH of 7.5. The HIV-1 RT variants tested were added to either the [phis]X174am3 DNA-DNA template-primer (at a final concentration of 12.5 µg/ml) or the 16S E.coli rRNA RNA-DNA template-primer (at 50 µg/ml). The enzymes were preincubated with these substrates for 5 min at 30°C. The reactions were divided into two, without or with an excess of a trap of activated unlabeled salmon sperm DNA at a final concentration of 0.6 mg/ml, prepared as described previously (30). The polymerase reactions were initiated immediately afterwards by adding all four dNTPs, each at a final concentration of 20 µM. The reactions, at final volumes of 12.5 µl, were incubated at 37°C for 30 min. All polymerase reactions were stopped by adding an equal volume of formamide dye mix, desaturated at 100°C for 3 min, quickly chilled on ice and analyzed by electrophoresis through 8% polyacrylamide-urea gels as described earlier (10-12). The relative amounts of the extension products were determined by densitometric scanning of the gel autoradiographs and the percentages of total amounts (extended and unextended) of the labeled primers which were extended were calculated.
Figure 1. DNA-dependent DNA polymerase-induced synthesis and processivity exhibited by the different HIV-1 RT variants. Reactions were performed with the different variants of HIV-1 RT using single-stranded [phis]X174am3 DNA annealed to a 15mer 32P-end-labeled primer and the extension products were analyzed by polyacrylamide-urea gel electrophoresis, as described in detail in Materials and Methods. -, DNA extension with no DNA trap; +, extension in the presence of a DNA trap. Molecular size markers were HinfI-cleaved double-stranded dephosphorylated [phis]X174 DNA (Promega) which were 5[prime]-end-labeled with [[gamma]-32P]ATP as described previously (11-13). The arrows indicate the nucleotide lengths of the single-stranded DNA fragments.
RESULTS
DNA synthesis and relative processivity of the DNA-dependent DNA polymerase activity of the HIV-1 RT variants
All mutants (i.e. Leu74Val, Glu89Gly, Tyr183Phe and Met184Leu) and the wild-type HIV-1 RT were examined for their capacity to synthesize DNA using single-stranded [phis]X174am3 DNA as template. We followed extension of the 5[prime]-32P-end-labeled 15mer primer in the absence or presence of an excess of `quencher' unlabeled activated DNA. This trap was added after preincubation of the enzymes with the labeled template-primer and prior to addition of all four unlabeled dNTPs. Thus the polymerase is trapped by the activated DNA after initial dissociation from the template-primer to prevent further extension of the labeled primer by re-associated RT molecules. Consequently, the extent of primer elongation in the presence of trap DNA is directly proportional to relative processivity of each mutant RT (33). The overall pattern of extension without a trap is similar for all HIV-1 RT variants analyzed (Fig. 1). There are several strong pausing sites, with the most apparent ones obtained under the assay conditions employed at the approximate positions of 20-22mer, 40mer, 89-90mer and 120mer, with a few of the longest products extending up to 400mer in length. We have recently shown that the pausing patterns of other RTs (i.e. of HIV-2, mouse mammary tumor virus, murine leukemia virus and avian myeloblastosis virus) are similar to that found for the HIV-1 RT variants (unpublished data). Inspection of the sequence of the template DNA does not reveal any obvious patterns responsible for these pausing positions (31). It is possible, therefore, that secondary structures of the single-stranded DNA contribute to these pausings. It should be noted that it has already been suggested that wild-type HIV-1 RT tends to stop at template oligo(dA) tracks (34). A quantitative analysis of the size distribution of the extended primers (as a percentage of the total amount) is shown in Table 1. The majority of the products generated by all RT variants are <50mer in length. There is a decrease in the amounts of extended primers ranging in length from 51-100 nt, with a smaller fraction longer than 100mer.
The distribution of the products generated after adding a DNA trap follows the same pattern; a decrease in the relative amount of extended primers as a function of length of the products. However, it is apparent that the overall extension is, as expected, substantially lower than that obtained with no trap present. The data summarized in Table 1 present a quantitative measure of relative processivity calculated as follows:
| relative processivity = ExT/ExN |
Processivity and synthesis of DNA on an RNA template by the HIV-1 RT variants
The extent and processivity of DNA synthesis might potentially depend not only on the RT variant studied, but also on the choice of template copied in terms of both sequence and nature (i.e. RNA versus DNA). Therefore, we have also examined DNA synthesized on an RNA template with a sequence different to that of the [phis]X174 DNA analyzed above. The results obtained with 16S E.coli rRNA (already used by us previously for studies of fidelity of DNA synthesis; 10-12) are shown in Figure 2, along with a quantitative analysis in Table 2. It is apparent, as shown above for DNA-dependent synthesis, that the overall pattern of primer extension is quite similar for all RT variants analyzed. Several of the longest products are up to 150mer in length. There are sites of strong pausing of DNA synthesis at the approximate positions corresponding to lengths of the extended primers of [sim]20-22mer, 40mer, 62-68mer, 80mer, 82mer, 100mer and 119mer. Inspection of the sequence of the template RNA copied (32) reveals common sequence patterns for these pausings. All stops occur before GC-rich sequences. It is possible, however, that pausing also results from the secondary structure of the copied RNA. A quantitative analysis, performed as described above for Table 1, shows that overall elongation of the 16mer primer is significantly lower than the fraction of primers extended in DNA-dependent synthesis (Table 2). This is true for the products of RNA-dependent synthesis generated both in the absence and in the presence of a DNA trap. The calculated relative processivity values are also lower than those in Table 1. Nevertheless, there is a similarity between the comparative processivity results obtained with a DNA template and those shown in Table 2 for an RNA template. Thus the Met184Leu mutant has a relative processivity that is about half that of wild-type RT. The other three mutants show relative processivities significantly higher than that of wild-type HIV-1 RT, with the highest value calculated for the Tyr183Phe mutant ([sim]2.3-fold higher than wild-type RT).
Table 1.
| Product length (nt) | Wild-type RT | Leu74Val RT | Glu89Gly RT | Tyr183Phe RT | Met184Leu RT | |||||
| No trap | With trap | No trap | With trap | No trap | With trap | No trap | With trap | No trap | With trap | |
| 16-50 | 34.7 ± 2.3 | 9.2 ± 0 | 30.9 ± 1.5 | 16.5 ± 1.1 | 28.0 ± 1.2 | 22 ± 0.6 | 34.0 ± 0.9 | 35.4 ± 1.0 | 35.8 ± 1.1 | 9.0 ± 0.7 |
| 51-100 | 28.4 ± 2.4 | 18.5 ± 0.8 | 21.4 ± 1.0 | 17.6 ± 1.3 | 26.7 ± 0.9 | 22.5 ± 1.1 | 24.9 ± 0.1 | 13.8 ± 0.7 | 15.1 ± 2.9 | 2.5 ± 1.3 |
| 101-200 | 12.8 ± 0.8 | 6.6 ± 1.1 | 15.4 ± 0.6 | 5.4 ± 0.7 | 17.2 ± 0.5 | 4.9 ± 0.8 | 16.1 ± 0.5 | 0.4 ± 0.1 | 0 | 0 |
| >201 | 3.9 ± 0.3 | 0.5 ± 0.1 | 9.7 ± 0 | 0.8 ± 0.1 | 10.0 ± 0.2 | 0 | 0 | 0 | 0 | 0 |
| Overall extension | 79.8 ± 0.6 | 34.8 ± 0.8 | 77.4 ± 1.1 | 40.3 ± 0.4 | 81.9 ± 1.4 | 49.4 ± 1.3 | 75.0 ± 0.5 | 49.6 ± 0.4 | 50.9 ± 1.8 | 11.5 ± 2 |
| Relative processivity | 43.6 ± 1.0 | 52.1 ± 0.9 | 60.3 ± 1.9 | 66.1 ± 0.7 | 22.6 ± 4 | |||||
Table 2.
| Product length (nt) | Wild-type RT | Leu74Val RT | Glu89Gly RT | Tyr183Phe RT | Met184Leu RT | |||||
| No trap | With trap | No trap | With trap | No trap | With trap | No trap | With trap | No trap | With trap | |
| 17-50 | 11.7 ± 1.3 | 2.6 ± 0.2 | 11.3 ± 2.2 | 3.8 ± 0.4 | 9.7 ± 1 | 4.5 ± 0.8 | 18.8 ± 1.6 | 10.2 ± 1.6 | 12.6 ± 0.4 | 1.1 ± 0 |
| 51-120 | 4.5 ± 0.7 | 0 | 5.0 ± 0.3 | 0 | 6.3 ± 0.2 | 0 | 8.5 ± 0.2 | 0 | 1.4 ± 0.1 | 0 |
| Overall extension | 16.1 ± 0.6 | 2.6 ± 0.2 | 16.3 ± 1.9 | 3.8 ± 0.4 | 16.0 ± 0.8 | 4.5 ± 0.8 | 27.3 ± 1.8 | 10.2 ± 1.6 | 14.0 ± 0.3 | 1.1 ± 0 |
| Relative processivity | 16.1 ± 1.4 | 23.3 ± 3.6 | 28.1 ± 5.2 | 37.4 ± 6.4 | 7.9 ± 0.17 | |||||
DISCUSSION
Many enzymes are processive, i.e. they can attach to polymeric substrates or templates and perform a sequence of polymerization without intervening dissociation. Thus total processivity of synthesis of either DNA or RNA is one in which the entire DNA or RNA template strand is copied as a consequence of only one polymerase binding event. The main biological advantage of a processive polymerization process is the efficiency of polymerization, especially when the molecular ratio of substrate template-primer to enzyme is high. It could also be speculated that the higher the processivity of a polymerase the higher the chances of misincorporation of a wrong nucleotide and, under conditions of no 3[prime]->5[prime] exonuclease proof-reading (which is the case for reverse transcriptases), the probability of extending these mismatches is also higher. We have tested this hypothesis, recently raised for mutants of Met184 of HIV-1 RT resistant to 3TC and having enhanced fidelity of DNA synthesis (26), by conducting a comparative analysis of other analog-resistant mutants of HIV-1 RT (already shown to also have a fidelity of DNA synthesis higher than that of wild-type RT; 21,24,25). The Met184Leu mutant behaves similarly to the other mutants of Met184 studied previously, Met184Val, Met184Thr and Met184Ile (26,35). All mutants of Met184, including Met184Leu, are resistant to 3TC and have enhanced fidelity of DNA synthesis relative to wild-type HIV-1 RT, as determined by a variety of methods (21-23,26). Similar to the other Met184 mutants studied, the Met184Leu mutant shows reduced relative processivity of DNA synthesis on both DNA and RNA templates. However, the other three drug-resistant mutants tested in the present study, Leu74Val, Glu89Gly and Tyr183Phe, exhibit a relative processivity of DNA synthesis that is not lower but is even higher than that of wild-type HIV-1 RT. Therefore, there is no apparent general mechanistic linkage between accuracy and processivity of DNA synthesis for these three mutants.
Figure 2. RNA-dependent DNA synthesis and processivity catalyzed by the different HIV-1 RT variants. The reactions were performed with 16S E.coli rRNA annealed to a 16mer end-labeled primer as described in Material and Methods. Analysis of the reaction products was carried out as described in the legend to Figure 1. The molecular size markers were also as described in Figure 1. Crystal structure studies of HIV-1 RT show that the conserved Tyr183-Met184-Asp185-Asp186 sequence is part of the [beta]9-[beta]10 turn in the DNA polymerase active site in the `palm' subdomain, therefore they have a direct effect on dNTP binding (15). The C[beta] atom of Met184 interacts directly with the amide group of Asp185 (36). Hence, mutating residue 184 can affect incoming dNTP in such a way that the 3TC analog is less favorably incorporated. This could happen, for example, if the side chain of residue 184 directly sensed the unnatural sugar moiety of the analog. On the other hand, it might be that this mutation indirectly affects incorporation of 3TC by conformational changes in the active site. These changes could simultaneously produce increased resistance and fidelity and decreased processivity. In the light of the present results it is interesting to note that modifying one residue adjacent to Met184 (mutant Tyr183Phe) affects the processivity of DNA synthesis in a manner contrary to that observed with residue 184. It is highly likely that these differences are mainly due to differences in the spatial interactions of residues 183 and 184. In contrast to residues 183 and 184, residues 74 and 89 are located far from the polymerase active site or the putative dNTP binding site in HIV-1 RT, in positions where they interact directly with the nucleic acid template-primer (13,15). Both structural and biochemical data, assembled for both HIV-1 and HIV-2 RTs, support the suggestion that resistance to nucleoside analogs due to mutations at positions 74, 89 and 215 is caused by repositioning of the template-primer (15,17,18). As stated, the apparent inverse relationship between fidelity and relative processivity found for the Met184Leu variant of HIV-1 RT does not apply to the Leu74Val, Glu89Gly and Tyr183Phe mutants. It is not yet clear what the mechanistic reasons for these differences are. Certainly, they may reflect different molecular modes of fidelity of misinsertion and elongation as well as processivity, possibly due to differences in proximity of the mutated residues to the dNTP binding site. This matter needs to be further investigated, both in vitro, in the context of pure RT and in association with other viral proteins and nucleic acids, as well as in vivo in HIV-infected cells. For example, it will be interesting to study proviral DNA synthesis by virions containing the different mutated RT. Furthermore, detailed catalytic studies of the different HIV-1 RT variants should provide an insight into the mechanistic interplay between the parameters of the catalytic `triad' involved in DNA polymerization: drug resistance, accuracy and processivity.
ACKNOWLEDGEMENTS
The plasmids used for expression of the different HIV-1 RT variants were kindly provided by Drs S.H.Hughes and P.L.Boyer. We also thank Dr S.Loya for critical reading of the manuscript.
REFERENCES
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 24 Mar 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Hizi The Reverse Transcriptase of the Tf1 Retrotransposon Has a Specific Novel Activity for Generating the RNA Self-Primer That Is Functional in cDNA Synthesis J. Virol., November 1, 2008; 82(21): 10906 - 10910. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Held, J. D. Kissel, D. Saran, D. Michalowski, and D. H. Burke Differential Susceptibility of HIV-1 Reverse Transcriptase to Inhibition by RNA Aptamers in Enzymatic Reactions Monitoring Specific Steps during Genome Replication J. Biol. Chem., September 1, 2006; 281(35): 25712 - 25722. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. H. Conlan, M. J. Stanger, K. Ichiyanagi, and M. Belfort Localization, mobility and fidelity of retrotransposed Group II introns in rRNA genes Nucleic Acids Res., September 16, 2005; 33(16): 5262 - 5270. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. K. Halvas, E. S. Svarovskaia, and V. K. Pathak Development of an In Vivo Assay To Identify Structural Determinants in Murine Leukemia Virus Reverse Transcriptase Important for Fidelity J. Virol., January 1, 2000; 74(1): 312 - 319. [Abstract] [Full Text]
|
|
|
|
|
|
P. Gerondelis, R. H. Archer, C. Palaniappan, R. C. Reichman, P. J. Fay, R. A. Bambara, and L. M. Demeter The P236L Delavirdine-Resistant Human Immunodeficiency Virus Type 1 Mutant Is Replication Defective and Demonstrates Alterations in both RNA 5'-End- and DNA 3'-End-Directed RNase H Activities J. Virol., July 1, 1999; 73(7): 5803 - 5813. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||