Search

The heterodimeric human immunodeficiency virus type 1 reverse transcriptase (HIV-1 RT) is composed of p66 and p51 subunits, p66 being the catalytic subunit. Our earlier investigation on the role of p51 in the catalytic process has shown that the p51 subunit facilitates the loading of the p66 subunit onto the template primer (TP). We had postulated that the beta7-beta8 loop of the p51 subunit may be involved in opening the polymerase cleft of p66 for DNA binding [Pandey, V. N., et al. (1996) Biochemistry 35, 2168]. We report here that deletion or alanine substitution of four residues of the beta7-beta8 loop results in severe impairment of the polymerase function of the heterodimeric enzyme. The enzyme activity was restored to the wild-type levels when the mutant p66 subunit was dimerized with the wild-type p51, suggesting that the intact beta7-beta8 loop in the p51 subunit is indispensable for the catalytic function of p66. Further, the template primer binding ability of the enzyme was significantly reduced upon deletion or alanine substitution in the beta7-beta8 loop. Interestingly, the loss of the TP binding ability of the mutant p66 was restored upon dimerization with wild-type p51. Examination of the glycerol gradient ultracentrifugation analysis revealed that while the wild-type HIV-1 RT sediments as a dimeric protein, the mutant enzymes carrying deletion or alanine substitution in both the subunits sediment predominantly as monomeric proteins, suggesting their inability to form stable dimers. In contrast, mutant p66 dimerized with wild-type p51 (p66delta/p51WT and p66Ala/p51WT) sedimented at the dimeric position. Taken together, these results clearly implicate the importance of the beta7-beta8 loop of p51 in the formation of stable functional heterodimers.

Tat, an essential human immunodeficiency virus type 1 protein interacts with the transactivation response element (TAR) and stimulates transcription from the viral long-terminal repeat (LTR). Blockage of Tat-TAR interaction halts viral transcription and hence replication. We have found that polyamide nucleic acid (PNA), targeted to the TAR sequences of viral RNA genome is able to prevent Tat-TAR interaction by efficient sequestration of the TAR. Anti-TAR PNA competes for TAR and prevents Tat-mediated stimulation of HIV-1 LTR transcription in vitro but has no influence on the basal level of transcription in the absence of Tat. Using a reporter gene construct pHIV LTR-CAT and pCMV-Tat in cell culture, we have further shown that anti-TAR PNA is able to block Tat-mediated transactivation of HIV-1 LTR transcription in vivo as judged by the extent of LTR driven CAT gene expression in the absence and presence of anti-TAR PNA. Supplementation of 100 nM of anti-TAR PNA into the culture medium further enhances the suppression of transactivation. Nonspecific scrambled PNA had no influence on Tat-TAR interaction and LTR-driven CAT gene expression in cell culture. These results suggest that PNA targeted to the TAR sequence of the viral genome may be a potential inhibitor of HIV-1 gene expression.

The YXDD motif is highly conserved in the reverse transcriptase family. The variable X residue is occupied by valine and methionine in MuLV RT and HIV-1 RT, respectively. Previous studies have shown that Tyr 222, the Y residue of the YXDD motif in MuLV RT, constitutes a major component of the fidelity center of the enzyme [Kaushik, N., Singh, K., Alluru, I., and Modak, M. J. (1999) Biochemistry 38, 2617-2627]. In this work, we present evidence that reverse transcriptases containing valine in the "X" position of the YXDD motif generally catalyze DNA synthesis with greater fidelity than those containing methionine or alanine. In the MuLV RT system, the two mutants V223M and V223A exhibited an overall reduced fidelity of DNA synthesis, specifically for RNA-templated reactions. Further analysis revealed that these mutants exhibit a higher efficiency of misinsertion on MS2 RNA than the wild-type enzyme for every mispair tested. However, unlike HIV-1 RT, the insensitivity of the wild-type MuLV RT to all four ddNTPs remained unchanged by mutation of V223 to Met or Ala. A 3D molecular model of the ternary complex of MuLV RT, template primer, and dNTP suggests that Val 223 along with its neighboring Tyr 222 stabilizes the substrate binding pocket via hydrophobic interactions with the dNTP substrate and template-primer.

A natural mutation at codon 151 (Gln --> Met; Q151M) of HIV-1 RT has been shown to confer resistance to the virus against dideoxy nucleoside analogues [Shirasaka, T., et al. (1995) Proc. Natl. Acad. Sci. U.S.A. 92, 2398], suggesting that Gln 151 may be involved in conferring sensitivity to nucleoside analogues. To understand its functional implication, we generated two mutant derivatives of this residue (Q151M and Q151N) and examined their sensitivities to ddNTPs and their ability to discriminate against rNTPs versus dNTP substrates on natural U5-PBS HIV-1 RNA template. We found that Q151M was highly discriminatory against all four ddNTPs but was able to incorporate rNTPs as efficiently as the wild type enzyme. In contrast, the Q151N mutant was only moderately resistant to ddNTPs but exhibited a higher level of discrimination against rNTPs. The fidelity of misinsertion was found to be highest for the Q151N mutant followed by Q151M and the wild type enzyme. These results point toward the importance of the amino acid side chain at position 151 in influencing the ability of the enzyme in recognition and discrimination against the sugar moieties of nucleotide substrates.

Tyr183 is a constituent of the highly conserved YXDD motif common to all retroviral reverse transcriptases. The two aspartates in this motif are the crucial members of the catalytic carboxylate triad while residue X, which in the case of HIV-1 RT is Met184, is implicated in dNTP substrate recognition and fidelity of DNA synthesis. In an attempt to understand the function of Tyr183 in the catalytic mechanism, we generated mutants of this residue (Y183F and Y183A) and subjected them to in-depth analysis. The efficiency of reverse transcription of natural U5-PBS HIV-1 RNA template was severely impaired by both the conservative and nonconservative substitutions. The major defect identified was at the level of dNTP binding as determined by a 20-80-fold increase in the Km for the dNTP substrate on both homopolymeric and heteropolymeric RNA and DNA templates. A significant reduction in processivity of DNA synthesis by these mutants was also noted. However, the fidelity of DNA synthesis by the Y183F and Y183A mutants was increased significantly compared to the wild-type enzyme. Interestingly, the reduction in the polymerase activity due to single substitution of Tyr to Phe in the YMDD motif is compensated by a second substitution of Met to Val in the same motif, herein referred to as the FVDD. The loss of dNTP binding as well as decreased processivity of DNA synthesis exhibited by the Y183F mutant was also compensated by mutation at the second site. Curiously, the double mutant did not exhibit any synergistic effect in regard to fidelity of DNA synthesis as might be expected since both the single mutations (Y183F, M184V) exhibited enhanced fidelity compared to the wild-type enzyme. These data implicate Tyr183 and Met184 as important constituents of the dNTP-binding pocket. We propose a model which suggests that subtle structural changes due to mutation in the flexible beta9-beta10 loop region at the active site of the molecule influence the enzyme activity and substrate recognition.

Human immunodeficiency virus type 1 (HIV-1) reverse transcriptase (RT) is a dimeric enzyme consisting of p66 and p51 subunits. The functional role of the p51 subunit remains elusive since all the catalytic functions appear to be executed through the p66 subunit. We report here that the p51 subunit, in addition to providing structural support to the p66 subunit, may be involved in facilitating the loading of the p66 subunit on to the template-primer (TP). This possibility is supported by following observations: (i) Upon binding to the TP, the p51 subunit can be dissociated by acetonitrile treatment and the template-primer-bound p66 monomer alone is capable of catalyzing DNA synthesis. (ii) Photo-cross-linking of template-primer to HIV-1 RT is abolished by dissociation of the p51 subunit prior to the TP binding but remains unaffected after the TP binding step. (iii) The p66-TP covalent complex selectively generated by UV irradiation and separated by gel electrophoresis can incorporate a single nucleotide in situ upon its renaturation in the gel. (iv) Treatment of HIV-1 RT with (tert-butyldimethylsilyl)spiroaminooxathioledioside (TSAO), an inhibitor that specifically binds to the beta7 beta8 loop of p51, destabilizes the heterodimeric enzyme, resulting in the subsequent loss of DNA binding.

A peptide nucleic acid (PNA) oligomer, an analogue of DNA, was examined for its ability to function as a primer or a template to support DNA synthesis catalyzed by DNA polymerases. The analogue, (PNA)19-TpG-OH, comprised of 19 bases in the form of PNA followed by a dinucleotide (TpG-OH) with a single phosphate and a free 3'OH terminus, was recognized as a bona fide primer by 2 reverse transcriptases and also by the Klenow fragment of E. coli DNA polymerase I. The 21-mer PNA chimera is extended on both RNA and DNA templates by all three polymerases. The specificity of binding of the PNA chimeric primer/DNA template at the template-primer binding site of the enzyme was shown by its photo-cross-linking ability to the enzyme which could be effectively competed out by another TP but not by template or primer alone. Furthermore, the chimeric TP-enzyme covalent complex was found to be catalytically active as judged by its ability to incorporate one nucleotide onto the 3'OH terminus of the immobilized primer. PNA sequences were also recognized as template when annealed with a DNA primer. These observations are in variance with the conventional suggestion that the phosphate backbone in the duplex region is essential for recognition and binding by DNA polymerases. The efficient extension of (PNA)19-TpG-OH suggests that the diameter of the duplex region of template primer rather than the phosphate backbone may be sufficient for recognition by DNA polymerases.

We report here that polyamide nucleic acid (PNA) as well as a polyamide nucleic acid-DNA chimera complementary to the primer binding site of the HIV-1 genome can completely block priming by tRNA3Lys and consequently the in vitro initiation of reverse transcription by HIV-1 RT. Conventional heating and cooling is not required for annealing PNA analogs to the complementary nucleotide sequence as effective blockage of reverse transcription results from their invasion in the duplex region of preprimed U5-PBS HIV-1 RNA template-primer and was seen even at ambient temperature. Further, the extension of the initiated nascent (-) strand DNA can also be blocked by inclusion of another PNA, targeted to upstream sequences in the U5 region of the viral RNA. Interestingly, a PNA chimera having only two DNA nucleotides annealed with the U5-PBS RNA is recognized as a bonafide primer by HIV-1 RT, as the 3'OH end of the chimeric molecule is extended by the enzyme in the presence of dNTPs. A significant observation was that RNA/PNA or RNA/(PNA-DNA) hybrids were entirely resistant to the RNase H activity of HIV-1 RT. Furthermore, PNA invasion into the RNA/DNA hybrid completely prevented the cleavage of the RNA strand, suggesting that the RNase H activity of HIV-1 RT which was required in reverse transcription may also be inhibited by the PNA oligomer. These observations suggest that oligomeric PNAs targeted to various critical regions of the viral genome are likely to have strong therapeutic potential for interrupting multiple steps involved in the replication of HIV-1 and warrant serious investigation especially in the area of an effective delivery system.

Glutamine-151 of HIV-1 RT has been shown to be a catalytically important residue through the characterization of its mutant phenotype Glu151Ala (Sarafianos et al., 1995a). To further understand the role of this residue, we have extended this analysis to include polymerization on natural RNA template in addition to DNA template. We find that Q151A mutant exhibited a severe reduction in the polymerase activity without any significant effect on the affinity for dNTP substrate. Unlike DNA-directed reactions, the rate-limiting step for RNA-directed reactions does not appear to be either at the dNTP binding step or the chemical step. Analysis of the products formed on natural heteromeric HIV-genomic RNA template annealed with an 18-mer DNA primer with a sequence complementary to the primer binding site (PBS) has shown that addition of nucleotides is nonlinear with time since the enzyme appears to stall on the RNA template following the incorporation of the first nucleotide. The Q151A mutant was found to be nearly devoid of pyrophosphorolytic activity on a RNA-PBS template-primer. Similar properties have been previously reported for a mutant of R72 (R72A) of HIV-1 RT (Sarafianos et al., 1995b). However, R72 was implicated in stabilizing the transition state ternary complex before and after the phosphodiester bond formation (Kaushik et al., 1996; Sarafianos et al., 1995b). Our results with Q151A suggest that the side chain of Q151 may help stabilize the side chain of R72, and the loss of pyrophosphorolysis activity observed with the Q151 mutant may be the indirect manifestation of this stabilizing effect on R72. These observations point to the functional interdependence of residues Q151 and R72 in the polymerase function of the enzyme. An analysis of the 3D model structure of HIV-1 RT bound to DNA-DNA and RNA-DNA template-primer reveals that the guanidine hydrogen of R72 seems to stabilize Q151 by hydrogen bonding with its amide oxygen. A systematic conformational search of the side chain of Q151 also suggests a stable orientation where its specific interaction with the base of the RNA template may aid in stabilizing it.

Based on the projected three-dimensional equivalence of conserved amino acids in the catalytic domains of DNA polymerases, we propose Arg 110 of MuLV RT to be an important participant in the catalytic mechanism of MuLV RT. In order to obtain evidence to support this proposition and to assess the functional importance of Arg 110, we carried out site directed mutagenesis of Arg 110 and replaced it with Lys, Ala, and Glu. The mutant enzymes were characterized with respect to their kinetic parameters, ability to bind template-primers, and the mode of DNA synthesis. All the three substitutions at 110 position resulted in severe loss of polymerase activity without any significant effect on the RNase H function. In spite of an approximately 1000-fold reduction in kcat of polymerase activity with three mutant enzymes, no significant reduction in the affinities for either template-primer or dNTP substrates was apparent. Mutant enzymes also did not exhibit significant sulfur elemental effect, implying that the chemical step, i.e., phosphodiester bond formation, was not defective. Examination of the mode of DNA synthesis by the mutant enzymes indicated a shift from processive to the distributive mode of synthesis. The mutants of R110 also displayed significant loss of pyrophosphorolysis activity. Furthermore, the time course of primer extension with mutant enzymes indicated severe reduction in the rates of addition of the first nucleotide and even further reduction in the addition of the second nucleotide. These results suggest that the rate limiting step for the mutant enzymes may be before and after the phosphodiester bond formation. Based on these results, we propose that Arg 110 of MuLV RT participates in the conformational change steps prior to and after the chemical step of polymerase reaction.

In order to clarify the role(s) of the individual member of the carboxylate triad in the catalytic mechanism of human immunodeficiency virus type 1 (HIV-1) reverse transcriptase, we carried out site-directed mutagenesis of D185, D186, and D110, followed by the extensive characterization of the properties of the individual mutant enzymes. We find that all three residues participate at or prior to the chemical step of bond formation. The incorporation pattern seen with phosphorothioate analogs of dNTP on both RNA-DNA and DNA-DNA template-primers indicated that D186 may be the residue that coordinates with the alpha-phosphate group of dNTP in the transition-state ternary complex. Further support for the role assigned to D186 was obtained by examination of the ability of the individual carboxylate mutants to catalyze the reverse of the polymerase reaction (pyrophosphorolysis). Mutants of D185 exhibited near-normal pyrophosphorolysis activity, while those of D186 were completely devoid of this activity. Thus, D185 appears to participate only in the forward reaction, probably required for the generation of nucleophile by interacting with the 3'-OH of the primer terminus, while D186 seems to be involved in both the forward and the reverse reactions, presumably by participating in the pentavalent intermediate transition state. Lack of any elemental effects during polymerization with mutant enzymes of residue D110, together with their inability to catalyze pyrophosphorolysis, suggest its probable participation in the metal-coordinated binding to the beta-gamma-phosphate of dNTP or PPi in the forward and reverse reactions, respectively. A molecular model of the ternary complex based on these results is also presented.

In order to identify functionally important residues in the O and O1 helices of Escherichia coli DNA polymerase I, we mutated 9 residues of this region to alanine. The alanine substitutions result in moderate to severe effects on the polymerase activity of the individual mutant enzymes. Severe loss of activity is associated with R754A, K758A, F762A, and Y766A. However, the loss of polymerase activity with different template primers exhibited a rather unique pattern implying differential participation of the individual residue in the synthesis directed by poly(rA), poly(dA), and poly(dC) templates. The ability of all mutants to form E-DNA binary complex was found to be unaffected with the exception of Y766A and F771A, where significant reduction in the cross-linking of both the template and the primer strand was noted. Most interestingly, the catalytic activity of all inactive mutant enzymes, with the exception of K758A, could be restored by substituting Mn2+ in place of Mg2+ as a divalent cation. Based on these results and associated changes in the kinetic parameters and other properties of the individual mutant enzyme, we conclude the following: (a) Tyr 766 and Phe 771 are either involved in the binding of template-primer or are in the vicinity of the DNA binding track. (b) Residues Arg 754, Lys 758, Phe 762, and Tyr 766 appear to be required for the binding of Mg.dTTP, while only Arg 754 and Lys 758 are utilized in the polymerization of Mn.dTTP. (c) In the polymerization of dGTP, only Lys 758 appears essential regardless of the type of divalent cation. (d) Phe 762 participates only in the binding of Mg.dTTP. Finally, (e) based on the analysis of the time course of nucleotide incorporation, processivity, and pyrophosphorolysis reaction, we suggest that Lys 758 is probably involved in a conformational change of the ternary complexes preceding and following the chemical step. In summary, our results suggest that the formation of the dNTP binding pocket is a dynamic process which requires the participation of different residues depending on the type of dNTP and the divalent cation.

Methionine 184 of HIV-1 RT is a constituent of the catalytically crucial and highly conserved YXDD motif in the reverse transcriptase class of enzymes. We investigated the role of this residue by substituting it with Ala and Val by site-directed mutagenesis followed by extensive characterization of the two mutant enzymes. The kinetic parameters governing DNA synthesis directed by RNA and DNA templates indicated that both M184A and M184V mutants are catalytically as efficient as the wild type enzyme. Photoaffinity labeling of both the mutant and the wild type enzyme exhibited an identical affinity for RNA-DNA and DNA-DNA template primers. We further demonstrate that M-->V substitution at 184 position significantly increases the fidelity of DNA synthesis while M-->A substitution results in a highly error-prone enzyme without having compromised its efficiency of DNA synthesis. The M184V mutant exhibited a 25-45-fold increase in mismatch selectivity (ratio of k(cat)/K(m) of correct versus incorrect nucleotides) as compared to the WT enzyme. This pattern of error-prone synthesis is also confirmed by examining the abilities of the enzyme-(template-primer) covalent complexes to incorporate correct versus incorrect nucleotide onto the immobilized template-primer. The nature of error-prone synthesis by the M184A mutant shows an increase in both the mismatch synthesis and extension of the mismatched primer termini. Using a three-dimensional molecular model of the ternary complex of HIV-1 RT, template-primer, and dNTP, we observe that the strategic location of M184 may allow it to interact with the sugar moiety of either the primer nucleotide or the dNTP substrate.

In order to define the role of Gln151 in the polymerase function of HIV-1 RT, we carried out site-directed mutagenesis of this residue by substituting it with a conserved (Q151N) and a nonconserved residue (Q151A). Q151N exhibited properties analogous to those of the wild-type enzyme, while Q151A has severely impaired polymerase activity. The Q151A mutant exhibited a 15-100-fold reduction in kcat with RNA [poly(rC) and poly(rA)] templates, while only a 5-fold reduction could be seen with the DNA [poly(dC)] template. Most interestingly, the affinity of the Q151A mutant for dNTP substrate remained unchanged with RNA templates, but a significant increase in Km was noted with the DNA template. The binding affinity of Q151A for DNA remained unchanged, as judged by photoaffinity cross-linking. However, unlike the wild-type enzyme, the Q151A mutant failed to catalyze the nucleotidyl transferase reaction onto the primer terminus of the covalently immobilized template-primer. The enzyme showed profoundly altered divalent cation preference from Mg2+ to Mn2+. These results strongly implicate Q151 of HIV-1 RT in the substrate dNTP binding function and possibly in the following chemical (catalytic) step. The effects of the mutation seem to be through Q151 of the p66 catalytic subunit, as p66WTt/p51Q151A retains the wild-type kinetic constants and nucleotidyl transferase activity. In contrast, p66Q151A/p51WT is indistinguishable from Q151A (mutated in both subunits). A model of the ternary complex (enzyme-template-primer and dNTP) has been used to infer the possible mode by which Q151 may interact with the base moiety of the substrate as well as with Arg72, a residue present within the active site of HIV-1 RT.

In order to determine the role of Tyr 766 of Escherichia coli DNA polymerase I in the catalysis of DNA synthesis, we investigated the properties of a Tyr 766-->Ser (Y766S) mutant of the Klenow fragment of E. coli DNA polymerase I. We found that the rates of incorporation of only dTTP but not the other dNTP substrates were affected in the reactions catalyzed by the mutant enzyme, when homopolymeric template-primers were used. The mutant enzyme exhibited a reduced rate of synthesis only with poly(rA)- or poly(dA)-directed reactions. Examination of the ability of the mutant and the wild-type enzymes to bind to dGTP and dTTP, as judged by UV-mediated cross-linking, indicated nearly identical binding efficiencies of both nucleotides. However, the ability of the mutant enzyme to bind to poly(rA).(dT)15 and poly(dA).(dT)15 was found to be significantly reduced as compared to the binding to heteropolymeric DNA. In order to further define the nature of template-mediated restriction on the catalytic activity of the mutant enzyme, its ability to copy DNA templates containing a stretch of AAAAA and ACACA sequences was compared. The results show that DNA synthesis catalyzed by the mutant enzyme is significantly retarded when it encounters the AAAAA region of the template but not the ACACA region. Product analysis of the reaction directed by the two template-primers showed that the mutant enzyme stalls/terminates synthesis upon encountering an AAAAA sequence in the template.(ABSTRACT TRUNCATED AT 250 WORDS)

We report here that a high level of V(D)J recombination activity is tightly associated with high-salt-resistant nuclear matrix isolated from thymus glands from 2- to 3-week-old rats. The soluble nuclear fractions either were devoid of or contained a very low level of recombinase activity. This is the first time that the process mimicking V(D)J recombination has been achieved in an in vitro system. The matrix-bound V(D)J recombinase activity was further found to be lymphoid specific, detectable only during early stages of development. These observations suggest that in vitro recombination of V(D)J segments of genes encoding antigen-binding proteins could be a matrix-bound process and that the nuclear matrix may be an important intranuclear domain for the functional organization of the V(D)J recombinase system.

Nuclear matrix prepared from 2-3 week old rat thymuses contains tightly bound TdT activity which has been quantitatively solubilized with nonionic detergent and sonication. TdT is contained in a discrete complex with a sedimentation value of 23 S. The complex is retained on an anti-TdT antibody column and contains DNA ligase and 3'-5' exonuclease activities as well as DNA and several other proteins but is devoid of replicative DNA polymerases. Such a type of multienzyme complex is absent from the nuclear extracts of thymus prepared from older rats and also from liver and spleen extracts of young and old rats.

Using the technique of ultraviolet-mediated cross-linking of substrate deoxynucleoside triphosphates (dNTPs) to their acceptor site [Abraham, K. I., & Modak, M. J. (1984) Biochemistry 23, 1176-1182], we have labeled the Klenow fragment of Escherichia coli DNA polymerase I (Pol I) with [alpha-32P]dTTP. Covalent cross-linking of [alpha-32P]dTTP to the Klenow fragment is shown to be at the substrate binding site by the following criteria: (a) the cross-linking reaction requires dTTP in its metal chelate form; (b) dTTP is readily competed out by other dNTPs as well as by substrate binding site directed reagents; (c) labeling with dTTP occurs at a single site as judged by peptide mapping. Under optimal conditions, a modification of approximately 20% of the enzyme was achieved. Following tryptic digestion of the [alpha-32P]dTTP-labeled Klenow fragment, reverse-phase high-performance liquid chromatography demonstrated that 80% of the radioactivity was contained within a single peptide. The amino acid composition and sequence of this peptide identified it as the peptide spanning amino acid residues 876-890 in the primary sequence of E. coli Pol I. Chymotrypsin and Staphylococcus aureus V8 protease digestion of the labeled tryptic peptide in each case yielded a single smaller fragment that was radioactive. Amino acid analysis and sequencing of these smaller peptides further narrowed the dTTP cross-linking site to within the region spanning residues 876-883. We concluded that histidine-881 is the primary attachment site for dTTP in E. coli DNA Pol I, since during amino acid sequencing analysis of all three radioactive peptides loss of the histidine residue at the expected cycle is observed.