Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2008 Apr 30;3(4):e2074.
doi: 10.1371/journal.pone.0002074.

Utilization of a Deoxynucleoside Diphosphate Substrate by HIV Reverse Transcriptase

Free PMC article

Utilization of a Deoxynucleoside Diphosphate Substrate by HIV Reverse Transcriptase

Scott J Garforth et al. PLoS One. .
Free PMC article


Background: Deoxynucleoside triphosphates (dNTPs) are the normal substrates for DNA synthesis catalyzed by polymerases such as HIV-1 reverse transcriptase (RT). However, substantial amounts of deoxynucleoside diphosphates (dNDPs) are also present in the cell. Use of dNDPs in HIV-1 DNA synthesis could have significant implications for the efficacy of nucleoside RT inhibitors such as AZT which are first line therapeutics for treatment of HIV infection. Our earlier work on HIV-1 reverse transcriptase (RT) suggested that the interaction between the gamma-phosphate of the incoming dNTP and RT residue K65 in the active site is not essential for dNTP insertion, implying that this polymerase may be able to insert dNDPs in addition to dNTPs.

Methodology/principal findings: We examined the ability of recombinant wild type (wt) and mutant RTs with substitutions at residue K65 to utilize a dNDP substrate in primer extension reactions. We found that wild type HIV-1 RT indeed catalyzes incorporation of dNDP substrates whereas RT with mutations of residue K65 were unable to catalyze this reaction. Wild type HIV-1 RT also catalyzed the reverse reaction, inorganic phosphate-dependent phosphorolysis. Nucleotide-mediated phosphorolytic removal of chain-terminating 3'-terminal nucleoside inhibitors such as AZT forms the basis for HIV-1 resistance to such drugs, and this removal is enhanced by thymidine analog mutations (TAMs). We found that both wt and TAM-containing RTs were able to catalyze Pi-mediated phosphorolysis of 3'-terminal AZT at physiological levels of Pi with an efficacy similar to that for ATP-dependent AZT-excision.

Conclusions: We have identified two new catalytic functions of HIV-1 RT, the use of dNDPs as substrates for DNA synthesis, and the use of Pi as substrate for phosphorolytic removal of primer 3'-terminal nucleotides. The ability to insert dNDPs has been documented for only one other DNA polymerase, the RB69 DNA polymerase and the reverse reaction employing inorganic phosphate has not been documented for any DNA polymerase. Importantly, our results show that Pi-mediated phosphorolysis can contribute to AZT resistance and indicates that factors that influence HIV resistance to AZT are more complex than previously appreciated.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.


Figure 1
Figure 1. Interactions between the dNTP substrate and RT.
A. Representation of the active site highlighting fingers subdomain residues involved in binding of the incoming nucleotide or AZT resistance. B. Detailed view of the interactions between the Lys65 residue, the incoming nucleotide and the catalytic metal ions. Metal ion coordination is depicted with a dashed purple line, hydrogen bonds with an orange line. Figures were generated from the published HIV-1 RT ternary structure (1rtd) using the UCSF Chimera package .
Figure 2
Figure 2. Utilization of a deoxynucleotide diphosphate substrate by HIV-1 RT.
A. 5′ end-labeled primer, annealed to a template strand, was extended in the presence of a range of dADP concentrations by HIV-1 RT. The primer is extended by a single nucleotide, corresponding to the template ‘T’ at the +1 position. Extension reactions contained 10 nM wild-type enzyme, 5 nM primer-template, and were carried out for 10 minutes. Unextended primer is labeled ‘p’, the single nucleotide extension product is ‘p+1’. B. Single nucleotide extension assays of HIV-1 RT or KF exo- (50 nM each) with 10 µM of either dADP or dATP demonstrated no activity in the reaction containing KF exo- and dADP. Lane B, unextended primer. C. Single nucleotide extension assays utilizing dADP, dCDP, dGDP and dTDP incorporation opposite the complementary template nucleotide. Reactions were carried out with 100 µM of the appropriate dNDP or 0.2 µM dNTP, 50 nM enzyme and 5 nM template-primer. Lanes B1 and B2 contained RT and KF exo- respectively, with no nucleotide added. D. The RT preparation does not contain a contaminating nucleotide kinase activity. Nucleotides, which had been pre-incubated in the presence (labeled RT+ADP) or absence of HIV-1 RT, were used in a single nucleotide extension assay with either KF exo- or HIV-1 RT. Lane B: pre-incubation reaction contained no nucleotide. Unextended primer and primer extended by a single nucleotide are indicated ‘p’ and ‘p+1’ respectively.
Figure 3
Figure 3. Mutations at the lysine 65 position abrogate dADP utilization. A.
Single nucleotide extension assays were performed as in Figure 2A, but with enzyme concentrations of 10 nM and 2 nM for, respectively, dADP and dATP utilization. In both groups, the nucleotide concentration was 100 µM. B. Nucleotide binding measured by DEC formation. Reactions contained 20 nM wild-type or K65R mutant RT, and 2 to 250 µM dADP. After challenge with competitor DNA, reactions were separated by PAGE. Position of the complex is indicated. ‘P-T’ represents primer-template and ‘P’ is the unannealed primer. Reaction B contained no dADP.
Figure 4
Figure 4. Pi-dependent phosphorolysis catalyzed by HIV-1 RT.
A. Diagram showing that pyrophosphorolysis and Pi-dependent phosphorolysis are the reverse reaction of polymerization in which the substrate is a nucleotide triphosphate or diphosphate, respectively. B. Time course of phosphorolysis and pyrophosphorolysis reactions catalyzed by HIV-1 RT and KF exo-. Reactions were performed with 5mM PPi or K2HPO4, 10 nM enzyme and 5 nM template-primer. Aliquots of the reaction were stopped after 1, 2 and 5 minutes and analyzed by denaturing PAGE. Lane B contained a ‘no phosphate’ control in which template-primer was incubated with RT for 20 minutes. Full-length, 5′-end labeled primer is indicated ‘p’, phosphorolysis products shortened by a single nucleotide from the 3′ end are indicated ‘p-1’. C. Radiolabeled nucleotide products released from the primer 3′ terminus through a phosphorolysis reaction were separated by PEI-cellulose TLC. HIV-1 RT or KF exo- (5 nM) were incubated with primer template in the presence of 10 mM Pi or 150 µM PPi, and aliquots removed after 2.5, 5 and 15 minutes. Lanes B1 and B2 contained control reactions incubated for 15 minutes with RT and KF exo- respectively, in the absence of added phosphate. The position of TTP and TDP was determined by comparison to unlabeled standards.
Figure 5
Figure 5. Influence of K65 substitutions on Pi-dependent phosphorolysis.
A. Pi-dependent phosphorolysis is inhibited by substitutions of K65. 5′-end labeled primer was incubated with RT (10 nM) and a range of inorganic phosphate concentrations. Lane ‘B’ contained no enzyme, lane 0 contained no Pi.
Figure 6
Figure 6. TAM mutants utilize dADP and perform Pi-dependent phosphorolysis.
A. Single nucleotide insertion assays performed with 100 µM nucleotide and 10 nM or 2 nM each enzyme, for dADP and dATP reactions respectively. ‘p’ indicates unextended primer and ‘p+1’, the single nucleotide extension product. B. Pi-dependent phosphorolysis reactions contained 10 nM each enzyme, and 0, 0.1, 1 or 10 mM Pi. Full length 5′ end labeled primer is indicated ‘p’, primer with a single nucleotide excised from the 3′ end is shown as ‘p-1’. C. Time course reactions for the Pi-dependent phosphorolysis contained 10 nM each enzyme and 10 mM Pi.
Figure 7
Figure 7. Pi-dependent primer unblocking by wild type and TAM mutants.
5′ end labeled primer was blocked with AZT, followed by incubation with a phosphate donor (Pi, ATP or PPi) and dNTPs. Unblocked primer is extended to the end of the template, resulting in the ‘full length’ product. Control lanes C1, C2 and C3 contained no phosphate donor, no dNTPs and no AZT respectively. B. Graphical representation of the results from three independent primer unblocking experiments. Error bars show the standard deviation of the results.

Similar articles

See all similar articles

Cited by 7 articles

See all "Cited by" articles


    1. Kornberg . New York: W.H. Freeman; 1992. DNA Replication.
    1. Bebenek K, Abbotts J, Roberts JD, Wilson SH, Kunkel TA, et al. Specificity and mechanism of error-prone replication by Human Immunodeficiency Virus-1 reverse transcriptase. J Biol Chem. 1989;264:16948–16956. - PubMed
    1. Kunkel TA. DNA replication fidelity. J Biol Chem. 2004;279:16895–16898. - PubMed
    1. Huang H, Chopra R, Verdine GL, Harrison SC. Structure of a covalently trapped catalytic complex of HIV-1 reverse transcriptase: implications for drug resistance. Science. 1998;282:1669–1675. - PubMed
    1. Garforth SJ, Kim TW, Parniak MA, Kool ET, Prasad VR, et al. Site-directed mutagenesis in the fingers subdomain of HIV-1 reverse transcriptase reveals a specific role for the beta3-beta4 hairpin loop in dNTP selection. J Mol Biol. 2007;365:38–49. - PMC - PubMed

Publication types