icon-folder.gif   Conference Reports for NATAP  
 
  15th International HIV Drug Resistance Workshop
June 13-17, 2006
Sitges, Spain
Back grey_arrow_rt.gif
 
 
 
Protease Mutations Outside of Protease Gene Found in GAG Gene
 
 
  Reported by Jules Levin
15th Intl HIV Drug Resistance Workshop
June13-17, 2006, Sitges, Spain
 
Six studies at this meeting & reported below about finding protease inhibitor mutations in the HIV- GAG gene. The two key questions are: 1- can these mutations be present without the standard protease mutations found in the protease gene in previously untreated patients who fail HAART & can the GAG gene mutations lead to viral failure, and 2- in treatment-experienced patients who develop standard PI mutations, does the presence of the GAG mutations amount to additional resistance, and are they clinically significant & relevant to resistance.
 
Substitutions within Gag but outside the cleavage
sites can cause protease inhibitor resistance

 
M Nijhuis, NM van Maarseveen, D de Jong, P
Schipper, IW Goedegebuure and CAB Boucher
 
Department of Virology, University Medical Center Utrecht,
Utrecht, the Netherlands
 
BACKGROUND: Recently, we identified that HIV-1 Gag cleavage site changes can cause PI resistance in vitro. Interestingly, not all in vitro observed PI resistance could be explained by these cleavage site mutations, implying that yet another region in the viral genome contributes to PI resistance. In this study we have investigated which other viral region(s) add to the observed PI resistant phenotype.
 
METHODS: HIV-1 HXB2 was cultured in SupT1 cells in the presence of increasing concentrations (20nM-240nM) of the PI RO033-4649. At the highest drug concentration, viral RNA was isolated and full genomic sequencing was performed. Part of the viral genome (MA-CA) was cloned in a wild type reference strain and phenotypic drug susceptibility was analysed. Subsequently, we investigated the effect of these Gag mutations on overall proteolytic processing. To do so, 293T cells were transfected with either a wild type or a mutant MA-CA construct in the presence or absence of RO033-4649. Viral lysates were analysed by quantitative Western blotting (Li-Cor) using antisera against MA or CA.
 
RESULTS:
 
The in vitro selection experiments resulted in the selection of a virus population 4 to 8-fold resistant to all PI in the absence of a single mutation in the viral protease.
 
Part of the resistance (2 to 5-fold) could be explained by Gag cleavage site changes.

 
Full genomic sequencing analysis demonstrated that multiple mutations were selected in the N-terminal part of Gag (MA: E40K; L75R; K113E and CA: M200I; A224A/V).
 
Interestingly, these changes when introduced in a wild type HXB2 backbone confer a 2.7-fold increase in resistance to RO033-4649 and thereby explain the full protease resistant phenotype as was initially observed in vitro.
 
Comparable levels of resistance were observed to other protease inhibitors, such as lopinavir (3-fold) and tipranavir (2.5-fold). Quantitative Western blot analysis demonstrated that these non-cleavage site substitutions contribute to PI resistance by enhancing the overall Gag processing by wild type protease.
 
CONCLUSION: In this study we have demonstrated that substitutions within Gag but outside of the cleavage site contribute directly to PI resistance by enhancing the overall Gag processing by wild type protease.
 
The pathway leading to TMC114 resistance is different for TMC114 compared with other protease inhibitors
 
S De Meyer, H Azijn, E Fransen, I De Baere, M Van Ginderen, B Maes and M-P de Bethune Tibotec BVBA, Mechelen, Belgium
 
BACKGROUND: TMC114 (darunavir), a novel HIV protease inhibitor (PI), is active against wild-type (WT) and PI-resistant HIV-1. In-vitro selection experiments from WT HIV-1 were conducted in order to characterize the pathway leading to resistance to TMC114.
 
METHODS: HIV-1-infected MT4 cells were exposed to increasing concentrations of TMC114 and other PIs, to select for viruses able to replicate at high PI concentrations. Phenotypes and genotypes of selected HIV populations were determined by the Antivirogram® and virco®TYPE assays, respectively. Site-directed mutants (SDM) were constructed using the Medigenomix proprietary technology.
 
RESULTS:
 
Data already presented has shown that 75 passages (260 days) of HIV-1/IIIB in the presence of increasing concentrations of up to 200nM TMC114 resulted in virus populations with the protease mutations R41T and K70E. These viruses showed resistance to TMC114 (fold change in EC50 [FC] ~10), but replicated poorly. Moreover, SDM strains with these two mutations were not resistant to TMC114.
 
The study has now been extended to 327 passages (1,155 days), and selected viruses have shown TMC114 FC >10. Virus populations did not grow at TMC114 concentrations exceeding 350nM. The H69Q and V77I mutations also accumulated in the protease.
 
Eight mutations in the gag gene, both inside and outside the cleavage sites, were also observed.
 
Recombinant viruses encompassing the protease and reverse transcriptase of the selected viruses remained susceptible to TMC114 (FC<1). Recombinant viruses encompassing gag and protease of the selected viruses had TMC114 FC values between 1 and 10 (further SDM strains with the mutations in the gag gene are currently being constructed).
 
Experiments conducted with other WT HIV-1 strains produced comparable results. In contrast, experiments conducted with other PIs resulted in a more rapid selection of virus populations able to grow at high micromolar PI concentrations; these viruses contained typical PI resistance-associated mutations.
 
CONCLUSION: Selection of TMC114-resistant HIV-1 from WT strains is slower and more difficult than for other PIs. Characterization of the selected viruses showed that resistance to TMC114 occurs through a different pathway compared with other PIs.
 
HIV p7/p1 and p1/p6-gag cleavage site mutations are associated with specific PR mutations and PI resistance profiles.
 
J Verheyen1, E Litau1, T Sing 2, U Schuldenzucker 3, M Däumer1, M Oette 4, T Lengauer 2, D Hoffmann3, H Pfister 1 and R Kaiser 1 1University of Cologne, Cologne, Germany 2Max Planck Institute for Informatics, Saabrücken, Germany 3Caesar, Bonn, Germany 4University Clinic Düsseldorf, Germany
 
BACKGROUND: Protease (PR) and Cleavage site (CS) mutations occur in vitro and in vivo under the selective pressure of Protease-Inhibitors (PI). CS-mutations could compensate for the otherwise compromised viral fitness of viruses with PR-mutations and may also affect directly resistance to PI by improved processing of precursor proteins. However, an approach of analysing CS- and PR-mutations with regard to covariation and PI resistance profiles is still missing.
 
METHODS AND PATIENTS: We analysed two CS (p1/p7, p1/p6-gag) and the pol-gene of 529 HIV-1 subtype B isolates. The collective consists of 287 therapy- naive (TN) patients and 242 therapy-experienced (TE) patients with at least one primary PR mutation (30N,33F,46I/L,54V/A/M/L,82A/F/ST,84V,90M). Predicted RF factors for PI were calculated with geno2pheno (www.genafor.org). Statistical significance was determined using Fisher´s exact test or t-test (*P<0.05, **P<0.01), and covariation was assessed using the phi correlation coefficient with correction for multiple testing using Benjamin-Hochberg method at a false discovery rate of 5%. Cross-sectional data were analysed for evolutionary pathways with the mtreemix package.
 
RESULTS:
Significant accumulation of CS mutations in TE isolates was observed at both CS (p7p1:430K*, 431V**, 436R**, 437V**; p1/p6- gag: 449F**/V**/H*, 451T**, 452S**, 453L**/A**).
 
The detection of CS-mutations (TE=73% versus. TN=16%) in TE isolates was moreover associated with an increased number of primary PR-mutations (2.7±1.3 versus. 1.8±1.1), even more pronounced in viruses (n=59) with at least one mutation at each CS (3.1±1.3). Significantly positive covariation was found for mutations 431V-24I/54V/46I/46L, 449F-30N, 453L-84V/77I and significantly negative covariation for mutations 431V-77I and 453L-82A.
 
In our cross-sectional data for 431V 47% of the observed mutational patterns could be explained as an ordered accumulation of mutations along a pathway 82A-54V-431V, followed by either 24I or 46L. Additional TE viruses with 431V had significantly higher predicted RF for the PI lopinavir**/nelfinavir**/indinavir**/amprenavir**.
 
Similar results were obtained for 453L and saquinavir**/nelfinavir**/amprenavir* RF as well as 449F and nelfinavir**/amprenavir* RF.
 
CONCLUSION: Certain CS mutations are strongly associated with different PR mutations and resistance profiles, but seem to be barely first line mutations during the evolution of PI resistance. Nevertheless, CS mutations could influence the genetic barrier for the evolution of specific PI resistance and should therefore be considered in HIV genotypic resistance tests.
 
Amino acid changes in gag that develop in vivo during protease inhibitor therapy have a dominant effect on viral replication and sensitivity to protease inhibitors ex vivo
 
S Koch1, R Coman1, I. Munoz 1, BM Dunn1, JW Sleasman2 and MM Goodenow1 1University of Florida College of Medicine, Gainesville, FL, USA 2University of South Florida, All Children's Hospital, St. Petersburg, FL, USA
 
BACKGROUND: Optimal HIV-1 protease [PR] activity requires Gag and Pol regions from the same allele. Regions of HIV-1 gag between p2 and p6/p6Pol develop genetic diversity in HIV-1 infected individuals who fail to suppress virus replication by combination protease inhibitor [PI] therapy. The goal of this study was to elucidate the functional relationship between drug-related changes in Gag and PR that emerge in PI-treated individuals.
 
METHODS: Gag-pol variants that developed in vivo following prolonged PI therapy were evaluated. Sitedirected mutagenesis combined pretherapy p2/NC cleavage site, NC, p6 and p6Pol sequences with genotypic and phenotypic drug-resistant or -sensitive PR alleles. A panel of fourteen replication competent recombinant viruses was generated and tested in peripheral blood mononuclear cells for replicative kinetics in the absence of PI and for susceptibility to indinavir or ritonavir by determining IC50.
 
RESULTS:
Post-therapy Gag/PR evolved in vivo with 8 amino acid differences in PR and 10 substitutions in p2/NC cleavage site, NC, p6 and p6Pol. Recombinant viruses constructed with post-therapy gag/pr alleles had severely impaired replicative capacity in the absence of PIs that was related to kinetics of Gag polyprotein processing, but 6- to 7-fold reduced susceptibility to either ritonavir or indinavir, relative to pretherapy gag/pr alleles.
 
Reversion of either post-therapy p2/NC cleavage site or NC to pretherapy amino acids failed to improve replication, but dramatically increased sensitivity to PIs.
 
Combining the pretherapy cleavage site and NC amino acids in the same genome increased replicative ability, but produced no change in susceptibility to indinavir and a modest increase in susceptibility to ritonavir.
 
The p2/NC and NC substitutions combined with pretherapy PI-sensitive PR could modulate replicative capacity, independent of changing PR drug susceptibility.
 
Changes in p6/p6Pol increased replication without modulating drug susceptibility.
 
CONCLUSION: Positions in Gag exert a dominant effect on PR processing of Gag polyproteins, virus replication, and susceptibility to drug. Replicative capacity of viruses can be independent of susceptibility to PIs, while genotypic and phenotypic drug-resistant PR can exhibit susceptibility to PIs by changes in NC. The functional linkage between Gag and PR provides targets for novel therapeutics to inhibit drug-resistant viruses.
 
Identification of novel sites of substrate co-evolution with resistant variants of HIV-1 protease
 
M Kolli1, C Chappey2 N Parkin2 and CA Schiffer1 1Department of Biochemistry and Molecular Pharmacology, University of Massachusetts Medical School, Worcester, MA, USA 2Monogram Biosciences Inc., South San Francisco, CA, USA
 
BACKGROUND: Under selective pressure of the competitive protease inhibitors, HIV-1 protease and occasionally its substrates evolve to cause drug resistance. Co-evolution of Gag and Gag-Pol substrate cleavage sites may occur as a compensatory mechanism to any alterations in protease activity. However, studies have shown that not all substrates co-evolve. We have hypothesized that cleavage sites co-evolve only when they extend beyond the consensus 'substrate envelope' to contact residues that mutate in response to protease inhibitors. In this study we focus on changes in the cleavage sites sequences that co-evolve with particular drug-resistant protease variants.
 
METHODS: HIV-1 protease (PR), reverse transcriptase (RT, amino acids 1-305) and Gag (amino acids 418-500) sequences obtained from samples submitted for routine resistance testing (Monogram Biosciences) were analysed for correlated changes in protease and its substrates using Chi-square analysis. The phi correlation coefficients were also calculated to classify correlations as positive or negative. Modelling studies were performed using PyMOL and MIDAS software to investigate structural changes that may occur as a result of these correlated mutations.
 
RESULTS:
We observed three novel sites of co-evolution within the cleavage sites that we analysed. Correlations were observed between: 1). V82A in protease and the PRRT cleavage site; 2). I50V in protease and PR-RT cleavage site and; 3). I50L in protease and NC-p1 CS.
 
We also confirmed earlier reported co-evolution of the p1-p6 cleavage site with either D30N/N88D or I50V in protease. Exact patterns of substrate co-evolution varied depending on which protease mutation occurs. We are currently analysing the structural basis for this specificity.
 
CONCLUSION: Co-evolution of cleavage site substrates in the Gag and GagPol polyproteins with resistance mutations in HIV-1 protease appears to be more widespread than previously reported. These observations suggest further avenues of experimentation to more fully understand the potential impact of Gag and Gag-Pol cleavage sites mutations on drug susceptibility and protease inhibitor resistant virus fitness.
 
Absence of protease mutations upon virological failure of first-line fosamprenavir/ritonavir is not explained by the alternative selection of p7/p1 or p1/p6 Gag cleavage site mutations
 
M Ait-Khaled1, F Xu2, M Tisdale2, S MacManus2, PJ Yates2, RC Elston3 and W Snowden1 1GlaxoSmithKline, Infectious Diseases-Medicines Development Centre, Greenford, UK 2GlaxoSmithKline, Department of Virology, Stevenage, UK 3Boehringer-Ingelheim, Department of Virology, Canada
 
BACKGROUND: Virological failure (VF) of first line PI/r, including first-line fosamprenavir/ritonavir (FPV/r) is associated with absence of protease (PRO) mutations and reduced incidence of nucleoside backbone mutations in RT. Data suggest a role for CS in primary PI resistance: in vitro passage with Ro-033- 4649 selected I437V in p7/p1 cleavage site (CS); which confers 3-8 fold change (FC) in susceptibility to PIs without mutations in PRO [IHDRW 2004, Abst. 36]. Increased FC is explained by increased affinity of CS substrate to PRO, thereby competing more efficiently with PIs [IHDRW 2005, Abst. 107].
 
METHODS: PRO, p7/p1 CS, and p1/p6 CS genotypic sequences and phenotypes (Monogram Biosciences, GENESEQ and PHENOSENSE) of Wk48 VFs in the FPV/r arm of SOLO (FPV/r QD + ABC/3TC BID, n=322) were examined for changes from baseline at week 12 and/or beyond (first sample of two consecutive VL _ 1000 copies/ml and/or last therapy time point if continuing a failing regimen).
 
RESULTS:
For all 32 VFs, no 'treatment-selected' mutations in PRO were detected. Baseline and on-treatment CS data was available for 29/32 VFs., Apart from subject 1168 with S451N/S at wk12, no mutations were selected in first failures (n=29) or last therapy time-point where patients with ongoing replication continued to receive FPV/r (n=11/29; weeks of failing regimen: median 8; range 4-40).
 
One viral species had I437V at baseline and Week 40, 8/29 viral species had mixtures (R429G/R, P439S/P, S440F/S, H441N/H or Y/H, L449/P/L, S451N/S, and P453T/P or L/P) always present at both baseline and VF or solely baseline in one case and solely VF in another (subject 1168, above). A four aa insertion 451S-QSQR-452R was detected at baseline and VF for one subject.
 
1/32 isolates had baseline PRO mutations and displayed increased FC to most PIs, all other VFs showed no change in FC to any PIs.
 
CONCLUSION: Absence of PRO mutations in firstline FPV/r VFs is not explained by the alternative selection of p7/p1 and p1/p6 CS mutations, even after prolonged replication in the presence of FPV/r. The balance between patients' adherence and the potency and pharmacokinetic characteristics of boosted PIs remains a plausible explanation for the absence of PRO mutations in VFs of PI/r-regimen.