Comparative biochemical analysis of HIV-1 subtype B and C integrase enzymes
© Bar-Magen et al. 2009
Received: 16 June 2009
Accepted: 11 November 2009
Published: 11 November 2009
Skip to main content
© Bar-Magen et al. 2009
Received: 16 June 2009
Accepted: 11 November 2009
Published: 11 November 2009
Integrase inhibitors are currently being incorporated into highly active antiretroviral therapy (HAART). Due to high HIV variability, integrase inhibitor efficacy must be evaluated against a range of integrase enzymes from different subtypes.
This study compares the enzymatic activities of HIV-1 integrase from subtypes B and C as well as susceptibility to various integrase inhibitors in vitro. The catalytic activities of both enzymes were analyzed in regard to each of 3' processing and strand transfer activities both in the presence and absence of the integrase inhibitors raltegravir (RAL), elvitegravir (EVG), and MK-2048.
Our results show that integrase function is similar with enzymes of either subtype and that the various integrase strand transfer inhibitors (INSTIs) that were employed possessed similar inhibitory activity against both enzymes.
This suggests that the use of integrase inhibitors against HIV-1 subtype C will result in comparable outcomes to those obtained against subtype B infections.
Integration of viral cDNA into the host genome is one of the definitive features of retroviral replication. Integration is mediated by the HIV pol- encoded integrase enzyme. Recently, integrase inhibitors have been added to the arsenal of antiviral drugs used in therapy. RAL (Merck) was the first integrase inhibitor to be approved by the US Food and Drug Administration (FDA) after clinical trials demonstrated that this drug promoted a rapid and sustained antiretroviral effect . EVG (GS-9137, Gilead), another integrase inhibitor, is currently in phase III clinical trials . Other integrase inhibitors, such as MK-2048 (Merck), are still in pre-clinical development.
Integrase inhibitors are active against both B- and non-B subtypes in therapy [3, 4]. Subtype C variants are responsible for approximately 50% of global infections, mostly in Sub-Saharan Africa and India . It is therefore important to determine whether the integrase enzymes of different HIV-1 subtypes behave in a parallel manner to one another and whether they respond similarly to the use of integrase inhibitors of HIV-1 replication.
After viral entry and reverse transcription, reverse-transcribed double-stranded blunt-ended DNA is incorporated into the host cell genome through two catalytic activities mediated by integrase: 3' end processing and strand transfer [6, 7]. During 3' end processing, a dinucleotide adjacent to the conserved 3' terminal CA is excised from the 3' end of the recently reverse transcribed HIV-1 DNA genome, generating 3' hydroxyl ends. During the strand transfer reaction, both newly generated 3' ends are covalently linked to target DNA in a concerted fashion via a one-step transesterification reaction . In vitro, integrase can also catalyze two additional reactions: disintegration and specific internal endonucleolytic cleavage [9, 10].
Variability between different HIV-1 integrases at an amino acid level is low, ≈ 8-12%. However, sites of amino acid differences between subtypes are often close to resistance-related amino acids. We were therefore interested in analyzing whether such minor differences might be important in differential acquisition of INSTI resistance mutations in a subtype-specific manner . Furthermore, natural polymorphisms in non-B integrase proteins might alter INSTI binding or activity [12, 13]. An in silico comparison of subtype B and CRF A/G integrase predicted that polymorphisms within subtypes might affect structure and substrate binding characteristics of IN enzymes . In this study, we compared the enzymatic activities of subtype B and C recombinant integrases in the context of inhibition by RAL, EVG, and the novel INSTI MK-2048.
RAL and EVG are INSTIs with high specific activity against strand transfer [18, 19]. MK-2048 is a prototype second-generation INSTI with a resistance profile that is distinct from RAL and EVG [20, 21]. These three drugs have been reported to be approximately 100-fold less specific for the inhibition of 3' processing activity compared to strand transfer [18, 22, 23].
IC50 values for RAL, EVG and MK-2048 for subtype B and subtype C integrase in Mn2+-based enzymatic assays.
3' Processing IC50 a
Strand Transfer IC50 a
IC50 values for RAL, EVG and MK-2048 for subtype B and subtype C integrase in Mg2+-based enzymatic assays.
Strand Transfer IC50
Subtype B a
Subtype C a
EC50 values for RAL, MK-2048 and EVG for subtype B and C viruses cultured in cord blood mononuclear cells.
EC50 Values (μM)
0.0243 ± 0.0025
0.0072 ± 0.0021
0.0056 ± 0.0022
0.0010 ± 0.0002
0.0001 ± 0.00007
0.0301 ± 0.0054
0.0148 ± 0.0032
0.0082 ± 0.0013
0.0027 ± 0.0009
0.0009 ± 0.0003
0.0012 ± 0.0005
0.0003 ± 0.0001
0.0018 ± 0.0006
EC 50 Values (μM)
0.0046 ± 0.0002
0.0011 ± 0.0001
0.0087 ± 0.0007
0.0015 ± 0.0001
0.0008 ± 0.0006
0.0206 ± 0.0090
0.0033 ± 0.0020
0.0022 ± 0.0002
0.0067 ± 0.0009
0.0022 ± 0.0007
0.0008 ± 0.00004
0.0015 ± 0.0004
0.0007 ± 0.0005
0.0001 ± 0.00001
Most HIV-1 patients are infected with non-B subtypes, most commonly subtype C , and subtype-specific differences in the development of drug resistance have been reported . Therefore, it is important to understand the activity of enzymes of different subtypes. In our study, subtype B and C integrase enzymes were evaluated; and the activity of integrase inhibitors against them were compared, since a role for polymorphisms and structure-function differences between subtypes in drug resistance has been demonstrated [11, 12, 23].
Strand transfer inhibitors have been shown as efficient inhibitors of integration amongst a wide range of retroviruses . In silico observations suggest that subtype-specific differences in regard to key amino acids in integrase, including those close to the catalytic site, may pose an effect on the binding of RAL [13, 27, 28]. Therefore, subtype-specific variations in DNA-binding domains could also affect the affinity of RAL for integrase. In vitro, subtype B and C recombinant proteins retain similar enzymatic capacities in the absence of drug (Figures 2, 3, 4), with comparable strand transfer, 3' processing and disintegration activities, as measured by time course experiments. We also show that RAL and EVG had similar effects against both subtype B and C integrase enzymes, regardless of whether Mg2+ or Mn2+ was used as a cation (Tables 1 and 2). In addition to the foregoing, we have evaluated the IC50 values of RAL, EVG and MK-2048 in cell-based assays using clinical isolates of viruses of either subtype B or subtype C origin (Table 3). No significant differences were observed between subtypes in regard to drug susceptibility. These findings are consistent with recent results on similarities vis-à-vis biochemical activity and susceptibility to antiretroviral drugs of reverse transcriptase enzymes derived from HIV-1 subtypes B and C .
Our results provide biochemical and tissue culture evidence that integrase enzymes from HIV-1 subtypes B and C are inhibited by each of RAL, EVG and MK-2048 to a similar extent. These findings are supportive of the use of these inhibitors in patients infected with subtype C virus.
Oligonucleotides were purchased from Invitrogen and then PAGE purified. A 21-mer duplex was formed by annealing INT1 (5'-TGTGGAAAATCTCTAGCAGT-3') and INT2 (5'-ACTGCTAGAGATTTTCCACA-3'). The 35-mer duplex was produced by annealing T35 (5'-ACTATACCAGACAATAATTGTCTGGCCTGTACCGT-3') and SK70 (5'-ACGGTACAGGCCAGACAATTATTGTCTGGTATAGT-3'). The disintegration primer (5'-TGCTAGTTCTAGCAGGCCCTTGGGCCGGCGGCGCTTGCGCC-3') was heated to 95°C and slowly cooled to achieve its secondary structure .
RAL and MK-2048 were obtained from Merck Pharmaceuticals, Inc. EVG was obtained from Gilead Biosciences.
Subtype C integrase was PCR amplified from the molecular clone pINDIE-C1 (accession number: AB023804) and subcloned into the bacterial expression vector pET15B, replacing the subtype B integrase ORF kindly obtained from Dr. Robert Craigie, NIH.
The QuickChange II (Stratagene) site-directed mutagenesis kit was utilized for the introduction of the solubility mutations F185H and C280S into subtype C integrase. The primers utilized for this mutagenesis were INC-FF185H (5'-GCAGTATT CAT TCACAATCATAAAAGAAAAGGGGGG-3'), INC-RF185H (5'-CCCCCCTTTTCTTTT ATG ATTGTGAATGAATACTGC-3'), INC-FC280S (5'-GCAGGTGCTGAT TCT GTGGCAGGTAGACAG-3') and INC-RC280S (5'-CTGTCTACCTGCCAC AGA ATCAGCACCTGC-3').
Wild type and mutant His-tagged integrase proteins were expressed in Escherichia coli BL21(DE3) and purified under non-denaturing conditions. Bacterial cultures were grown at 37°C. When cultures of BL21 achieved an optical density of 0.5 at 600 nm, protein expression was induced by the addition of isopropyl-β-D-thiogalactopyranoside (IPTG) to a final concentration of 1 mM. The cultures were incubated for 3 hours at 37°C, centrifuged (5000 rpm for 12 min), and frozen at -80°C. The pellets were resuspended in lysis buffer (20 mM Hepes pH 7.5, 100 mM NaCl, 2 mM β-ME, protease inhibitors) and lysed by sonication. The lysates were centrifuged (12 500 rpm for 30 min), the supernatants discarded, and the pellets resuspended in binding buffer (1 M NaCl, 20 mM imidazole, 20 mM Hepes pH 7.5, 2 mM β-ME, 100 μM ZnCl2, protease inhibitors). Following centrifugation at 12500 rpm for 30 min, the supernatants were incubated with nickel-nitrioltriacetic acid (Ni-NTA) agarose beads (Qiagen) for 1 hour at 4°C with mild agitation. Proteins were purified utilizing propylene columns (Qiagen). His-tagged integrase protein was then eluted, utilizing a gradient of increasing imidazole concentration (0-2 M) in elution buffer (1 M NaCl, 20 mM Hepes pH7.5, 10% glycerol, 2 mM β-ME, 100 μM ZnCl2). The eluates were analyzed on 10% SDS-polyacrylamide gels with Coomassie staining (Sigma-Aldrich). Proteins were dialyzed overnight at 4°C against 1 M NaCl, 200 mM Hepes pH 7.5, 100 μM ZnCl2, 10% glycerol and 2 mM DTT in dialysis cassettes (10,000 MWCO, ThermoScientific). The samples were aliquoted and fast frozen at -80°C. Protein concentration was measured by Bradford assay utilizing the Bradford Protein Assay kit (Bio-Rad Laboratories).
The His tags were removed from purified recombinant subtype B and C integrases using the Thrombin CleanCleave kit (Sigma) according to manufacturers' instructions. Proteins were dialyzed overnight at 4°C against 1 M NaCl, 200 mM Hepes pH 7.5, 100 μM ZnCl2, 10% glycerol and 2 mM DTT in dialysis cassettes (10,000 MWCO, ThermoScientific).
Oligonucleotide-based assays were performed to measure integrase enzymatic activities. All oligonucleotide probes were gel purified and phenol-chloroform extracted. INT1 oligonucleotide was radiolabeled using the T4 Polynucleotide Kinase kit (Ambion) with [γ-32P] ATP (Perkin Elmer). Unincorporated nucleotides were discarded utilizing 'NucAway columns' (Ambion). A double stranded oligonucleotide substrate was obtained by mixing equal concentrations of INT1 and INT2, heating to 95°C and stepwise cooling to 37°C in 100 mM NaCl. The oligonucleotides T35 and SK70 were annealed to form a second double stranded oligonucleotide. INT1/2 mimicked the HIV-1 U5 long terminal repeat (LTR) and acted as a substrate for 3' processing. T35/SK70 acted as a site for integration of the processed INT1/2 fragment and mimicked host DNA.
Concentrations of enzymes used in the following reactions were optimized in a series of preliminary experiments. Integrase reactions were performed in a buffer containing 20 mM Hepes (pH 7.5), 30 mM NaCl, 1 mM dithiothreitol, 125 μM ZnCl2 and 0.125 pmol dsDNA substrate (INT1/2) in a final volume of 10 μl. Recombinant integrase (3.1 μM), and varying concentrations of integrase inhibitors or water/DMSO as control, were mixed and preincubated at 37°C for 15 min. 3' processing reactions were initiated by addition of 7.5 mM of MnCl2 and incubated at 37°C for 3 min unless otherwise indicated. For strand transfer reactions, 1.25 pmol of dsDNA template (T35/SK70) were added and the reaction mixture was further incubated at 37°C for 1 hour unless otherwise indicated. Reactions were stopped by adding 5 times the volume of gel loading dye (formamide containing 1% SDS, 0.25% bromophenol blue, and 0.25% xylene cyanol) and heating to 95°C. Reaction products were separated on 6% acrylamide, 7 M urea sequencing gels. Gels were dried and exposed utilizing phosphorimager screens (GE Healthcare) and scanned in a Molecular Dynamics Typhoon Phosphorimager (GE Healthcare). Product analysis and quantification were conducted using ImageQuant and GraphPad Prism 4.0 software. Quantification of standard error of the mean (SEM) was performed with GraphPad Prism 4.0 software.
A microtiter plate assay was utilized to evaluate MgCl2-mediated strand transfer as previously described [24, 31]. Briefly, biotinylated oligonucleotides mimicking LTR donor DNA (5'-biotin-ACCCTTTTAGTCAGTGTGGAAAATCTCTAGCAGT and 5'-ACTGCTAGAGATTTTCCACACTGACTAAAAG) were immobilized onto black-colour Reacti-Bind Streptavidin-coated plates (ThermoFisher). 313 nM recombinant enzyme was bound to donor DNA on the plates in the presence of 25 mM MnCl2 and the plates were then washed to remove excess unbound enzyme. 3'-FITC labelled dsDNA (5'-TGACCAAGGGCTAATTCACT-FITC-3' and 3'-FITC-ACTGGTTCCCGATTAAGTGA-5'), used as a reaction target, was added to the wells in 25 mM Hepes (pH 7.8), 25 mM NaCl, 2.5 mM MgCl2, and 50 μg/mL BSA [24, 32] and the plate was incubated at 37°C for 1 hour. Covalently linked target DNA was detected through use of an anti-FITC antibody conjugated to alkaline phosphatase (Roche) and a chemiluminescence substrate (CSPD Sapphire II, Applied Biosystems). Integrase inhibitors were added at increasing concentrations shortly before the addition of target DNA. Strand transfer was evaluated by chemiluminescence.
Unpaired two-tailed t-tests were used to examine statistical significance in subtype B versus subtype C integrase enzymatic assays using GraphPad Prism 4.0 software.
Recombinant viruses (subtype B) (pNL4-3 and IIIb) and viruses obtained from either our primary HIV infection cohort or from long-term infected patients (subtype B and C) were amplified as previously described [33, 34]. Drug susceptibility was measured in cell culture-based phenotypic assays using cord blood mononuclear cells to determine the extent of in vivo HIV replication blockage by integrase inhibitors. 50% drug effective concentrations (EC50s) were determined for each of RAL, EVG and MK-2048 by monitoring the production of p24 antigen, as previously elaborated .
We thank Dr. Robert Craigie for the pET15B IN subtype B expression vector, Dr. Yudong Quan and Dr. Amnon Hizi for helpful discussions, Ms. Estrella Moyal and Ms. Bonnie Spira for assistance with digital artwork and Ms. Emily I. McDonough for technical support. This work was supported by a grant from the Canadian Institutes of Health Research (CIHR). RDS is the recipient of a post-doctoral fellowship from the CIHR Canadian HIV Trials Network and the Canadian Foundation for AIDS Research (CANFAR). DAD is the recipient of a pre-doctoral fellowship from the Fonds de la recherche en santé du Québec (FRSQ). We also thank Dr. Daria Hazuda of Merck Inc. for incisive comments as well as for providing drugs and Dr. Lei Zhang of Merck-Frosst for supplying drugs and for research support.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.