HIV-1 Protease and reverse-transcriptase mutations: correlations with antiretroviral therapy in subtype B isolates and implications for drug-resistance surveillance

Machine learning aided multiscale modelling of the HIV-1 infection in the presence of NRTI therapy

Human Immunodeficiency Virus (HIV) is one of the most common chronic infectious diseases in humans. Extending the expected lifetime of patients depends on the use of optimal antiretroviral therapies. Emergence of the drug-resistant strains can reduce the effectiveness of treatments and lead to Acquired Immunodeficiency Syndrome (AIDS), even with antiretroviral therapy. Investigating the genotype-phenotype relationship is a crucial process for optimizing the therapy protocols of the patients. Here, a mathematical modelling framework is proposed to address the impact of existing mutations, timing of initiation, and adherence levels of nucleotide reverse transcriptase inhibitors (NRTIs) on the evolutionary dynamics of the virus strains. For the first time, the existing Stanford HIV drug resistance data have been combined with a multi-strain within-host ordinary differential equation (ODE) model to track the dynamics of the most common NRTI-resistant strains. Overall, the D4T-3TC, D4T-AZT and TDF-D4T drug combinations have been shown to provide higher success rates in preventing treatment failure and further drug resistance. The results are in line with the genotype-phenotype data and pharmacokinetic parameters of the NRTI inhibitors. Moreover, we show that the undetectable mutant strains at the diagnosis have a significant effect on the success/failure rates of the NRTI treatments. Predictions on undetectable strains through our multi-strain within-host model yielded the possible role of viral evolution on the treatment outcomes. It has been recognized that the improvement of multi-scale models can contribute to the understanding of the evolutionary dynamics, and treatment options, and potentially increase the reliability of genotype-phenotype models.

Analysis of mutational history of multidrug-resistant genotypes with a mutagenetic tree model

Human immunodeficiency virus (HIV) can develop resistance to all antiretroviral drugs. Multidrug resistance, however, is a rare event in modern HIV treatment, but can be life-threatening, particular in patients with very long therapy histories and in areas with limited access to novel drugs. To understand the evolution of multidrug resistance, we analyzed the EuResist database to uncover the accumulation of mutations over time. We hypothesize that the accumulation of resistance mutations is not acquired simultaneously and randomly across viral genotypes but rather tends to follow a predetermined order. The knowledge of this order might help to elucidate potential mechanisms of multidrug resistance. Our evolutionary model shows an almost monotonic increase of resistance with each acquired mutation, including less well-known nucleoside reverse transcriptase (RT) inhibitor-related mutations like K223Q, L228H, and Q242H. Mutations within the integrase (IN) (T97A, E138A/K G140S, Q148H, N155H) indicate high probability of multidrug resistance. Hence, these IN mutations also tend to be observed together with mutations in the protease (PR) and RT. We followed up with an analysis of the mutation-specific error rates of our model given the data. We identified several mutations with unusual rates (PR: M41L, L33F, IN: G140S). This could imply the existence of previously unknown virus variants in the viral quasispecies. In conclusion, our bioinformatics model supports the analysis and understanding of multidrug resistance.

The evolution of the HIV-1 protease folding stability

The evolution of structural proteins is generally constrained by the folding stability. However, little is known about the particular capacity of viral proteins to accommodate mutations that can potentially affect the protein stability and, in general, the evolution of the protein stability over time. As an illustrative model case, here, we investigated the evolution of the stability of the human immunodeficiency virus (HIV-1) protease (PR), which is a common HIV-1 drug target, under diverse evolutionary scenarios that include (1) intra-host virus evolution in a cohort of seventy-five patients sampled over time, (2) intra-host virus evolution sampled before and after specific PR-based treatments, and (3) inter-host evolution considering extant and ancestral (reconstructed) PR sequences from diverse HIV-1 subtypes. We also investigated the specific influence of currently known HIV-1 PR resistance mutations on the PR folding stability. We found that the HIV-1 PR stability fluctuated over time within a constant and wide range in any studied evolutionary scenario, accommodating multiple mutations that partially affected the stability while maintaining activity. We did not identify relationships between change of PR stability and diverse clinical parameters such as viral load, CD4⁺ T-cell counts, and a surrogate of time from infection. Counterintuitively, we predicted that nearly half of the studied HIV-1 PR resistance mutations do not significantly decrease stability, which, together with compensatory mutations, would allow the protein to adapt without requiring dramatic stability changes. We conclude that the HIV-1 PR presents a wide structural plasticity to acquire molecular adaptations without affecting the overall evolution of stability.

Bayesian variable selection using Knockoffs with applications to genomics

Given the costliness of HIV drug therapy research, it is important not only to maximize true positive rate (TPR) by identifying which genetic markers are related to drug resistance, but also to minimize false discovery rate (FDR) by reducing the number of incorrect markers unrelated to drug resistance. In this study, we propose a multiple testing procedure that unifies key concepts in computational statistics, namely Model-free Knockoffs, Bayesian variable selection, and the local false discovery rate. We develop an algorithm that utilizes the augmented data-Knockoff matrix and implement Bayesian Lasso. We then identify signals using test statistics based on Markov Chain Monte Carlo outputs and local false discovery rate. We test our proposed methods against non-bayesian methods such as Benjamini-Hochberg (BHq) and Lasso regression in terms TPR and FDR. Using numerical studies, we show the proposed method yields lower FDR compared to BHq and Lasso for certain cases, such as for low and equi-dimensional cases. We also discuss an application to an HIV-1 data set, which aims to be applied analyzing genetic markers linked to drug resistant HIV in the Philippines in future work.

Adjusting the Benjamini-Hochberg method for controlling the false discovery rate in knockoff-assisted variable selection

We consider the knockoff-based multiple testing setup of Barber & Candés (2015) for variable selection in multiple regression. The method of Benjamini & Hochberg (1995) and an adaptive version of it are adjusted to this setup, transforming them to valid p-value based, false discovery rate controlling methods that do not rely on specifying the correlation structure of the explanatory 15 variables. Simulations and real data applications show that our proposed methods are powerful competitors of the false discovery rate controlling method in Barber & Candés (2015).

Cantilever-centric mechanism of cooperative non-active site mutations in HIV protease: Implications for flap dynamics

The HIV-1 protease is an important drug target in antiretroviral therapy due to the crucial role it plays in viral maturation. A greater understanding of the dynamics of the protease as a result of drug-induced mutations has been successfully elucidated using computational models in the past. We performed induced-fit docking studies and molecular dynamics simulations on the wild-type South African HIV-1 subtype C protease and two non-active site mutation-containing protease variants; HP3 PR and HP4 PR. The HP3 PR contained the I13V, I62V, and V77I mutations while HP4 PR contained the same mutations with the addition of the L33F mutation. The simulations were initiated in a cubic cell universe containing explicit solvent, with the protease variants beginning in the fully closed conformation. The trajectory for each simulation totalled 50 ns. The results indicate that the mutations increase the dynamics of the flap, hinge, fulcrum and cantilever regions when compared to the wild-type protease while in complex with protease inhibitors. Specifically, these mutations result in the protease favouring the semi-open conformation when in complex with inhibitors. Moreover, the HP4 PR adopted curled flap tip conformers which coordinated several water molecules into the active site in a manner that may reduce inhibitor binding affinity. The mutations affected the thermodynamic landscape of inhibitor binding as there were fewer observable chemical contacts between the mutated variants and saquinavir, atazanavir and darunavir. These data help to elucidate the biophysical basis for the selection of cooperative non-active site mutations by the HI virus.

HIV-1C and HIV-1B Tat protein polymorphism in Southern Brazil

The transactivator of transcription (Tat) is a key HIV regulatory protein. We aimed to identify the frequency of key polymorphisms in HIV-1C compared with HIV-1B Tat protein, chiefly in the cysteine-, arginine-, and glutamine-rich domains and identify novel point mutations in HIV-1B and C sequences from Southern Brazil. This study was the first to investigate the genetic diversity and point mutations within HIV-1 Tat C in a Brazilian cohort. This was an observational, cross-sectional study, which included sequences of HIV-1B (n = 20) and HIV-1C (n = 21) from Southern Brazil. Additionally, 344 HIV-1C sequences were obtained from the Los Alamos database: 29 from Brazil and 315 from Africa, Asia, and Europe. The frequency of C31S substitution on HIV-1 Tat C in Brazil was 82% vs. 10% in the HIV-1B group (p < 0.0001). The frequency of the R57S substitution among the HIV-1C sequences from Brazil was 74% vs. 20% in HIV-1B (p = 0.004), and that of substitution Q63E in HIV-1C was 80% and 20% in HIV-1B (p < 0.0001). The mutation P60Q was more frequent in HIV-1B than in HIV-1C (55% and 6.12%, respectively, p < 0.0001)). Novel point mutations in the HIV-1C and B Tat functional domains were described. The frequency of C31S and other key point mutations in HIV-1 Tat C in Brazil were similar to those described in Africa, although lower than those in India. The Tat-B and C sequences found in Southern Brazil are consistent with biological differences and have potential implications for HIV-1 subtype pathogenesis.

The role of drug resistance in poor viral suppression in rural South Africa: findings from a population-based study

BACKGROUND

Understanding factors driving virological failure, including the contribution of HIV drug resistance mutations (DRM), is critical to ensuring HIV treatment remains effective. We examine the contribution of drug resistance mutations for low viral suppression in HIV-positive participants in a population-based sero-prevalence survey in rural South Africa.

METHODS

We conducted HIV drug resistance genotyping and ART analyte testing on dried blood spots (DBS) from HIV-positive adults participating in a 2014 survey in North West Province. Among those with virologic failure (> 5000 copies/mL), we describe frequency of DRM to protease inhibitors (PI), nucleoside reverse transcriptase inhibitors (NRTI), and non-nucleoside reverse transcriptase inhibitors (NNRTI), report association of resistance with antiretroviral therapy (ART) status, and assess resistance to first and second line therapy. Analyses are weighted to account for sampling design.

RESULTS

Overall 170 DBS samples were assayed for viral load and ART analytes; 78.4% of men and 50.0% of women had evidence of virologic failure and were assessed for drug resistance, with successful sequencing of 76/107 samples. We found ≥1 DRM in 22% of participants; 47% were from samples with detectable analyte (efavirenz, nevirapine or lopinavir). Of those with DRM and detectable analyte, 60% showed high-level resistance and reduced predicted virologic response to ≥1 NRTI/NNRTI typically used in first and second-line regimens.

CONCLUSIONS

DRM and predicted reduced susceptibility to first and second-line regimens were common among adults with ART exposure in a rural South African population-based sample. Results underscore the importance of ongoing virologic monitoring, regimen optimization and adherence counseling to optimize durable virologic suppression.

Genetic variation and function of the HIV-1 Tat protein

Human immunodeficiency virus type 1 (HIV-1) encodes a transactivator of transcription (Tat) protein, which has several functions that promote viral replication, pathogenesis, and disease. Amino acid variation within Tat has been observed to alter the functional properties of Tat and, depending on the HIV-1 subtype, may produce Tat phenotypes differing from viruses' representative of each subtype and commonly used in in vivo and in vitro experimentation. The molecular properties of Tat allow for distinctive functional activities to be determined such as the subcellular localization and other intracellular and extracellular functional aspects of this important viral protein influenced by variation within the Tat sequence. Once Tat has been transported into the nucleus and becomes engaged in transactivation of the long terminal repeat (LTR), various Tat variants may differ in their capacity to activate viral transcription. Post-translational modification patterns based on these amino acid variations may alter interactions between Tat and host factors, which may positively or negatively affect this process. In addition, the ability of HIV-1 to utilize or not utilize the transactivation response (TAR) element within the LTR, based on genetic variation and cellular phenotype, adds a layer of complexity to the processes that govern Tat-mediated proviral DNA-driven transcription and replication. In contrast, cytoplasmic or extracellular localization of Tat may cause pathogenic effects in the form of altered cell activation, apoptosis, or neurotoxicity. Tat variants have been shown to differentially induce these processes, which may have implications for long-term HIV-1-infected patient care in the antiretroviral therapy era. Future studies concerning genetic variation of Tat with respect to function should focus on variants derived from HIV-1-infected individuals to efficiently guide Tat-targeted therapies and elucidate mechanisms of pathogenesis within the global patient population.

Polymorphisms and Mutational Covariation Associated with Death in a Prospective Cohort of HIV/AIDS Patients Receiving Long-Term ART in China

Background

HIV drug resistance is associated with faster clinical progression of AIDS. However, the effect of significant polymorphisms and mutational covariation on mortality among HIV/AIDS patients receiving long-term antiretroviral therapy (ART), have rarely been studied.

Methods

In this prospective cohort study from December 2003 to December 2014, we present a new computational modelling approach based on bioinformatics-based models and several statistical methods to elucidate the molecular mechanisms involved in the acquisition of polymorphisms and mutations on death in HIV/AIDS patients receiving long-term ART in China.

Results

This study involved 654 ART-treated patients, who had been followed for 5473.4 person-years, a median of 9.8 years, and 178 died (25.2%, 3.3/100 person-years). The first regimens included AZT/d4T + NVP+ ddI (78.9%) or AZT/d4T + NVP+ 3TC (20.0%). We calculated an individual Ka/Ks value for each specific amino acid mutation. Result showed that 20 polymorphisms (E6D, Q18H, E35D, S37N, T39A, K43E, S68N, L74I, I93L, K103N, V106A, E169D, Y181C, G190A, Q197K, T200V, T200E, T215I, E224D and P225H) were strongly associated with AIDS related deaths. Among them, 7 polymorphisms (L74I, K103N, V106A, Y181C, G190A, T215I and P225H) were known to be drug resistance mutations, 7 polymorphisms (E6D, E35D, S37N, I93L, E169D, T200V and T200E were considered to be potential drug resistance mutations, and 6 polymorphisms (T39A, K43E, S68N, Q197K, T200V and E224D) were newly found to have an association with drug resistance mutations, which formed a complex network of relationships.

Conclusions

Some polymorphisms and mutational covariation may be the important influencing factors in the failure of treatment. Understanding these mechanisms is essential for the development of new therapies, designing optimal drug combinations, and determining effective clinical management of individual patients.

Molecular mechanisms and design principles for promiscuous inhibitors to avoid drug resistance: lessons learned from HIV-1 protease inhibition

Molecular recognition is central to biology and ranges from highly selective to broadly promiscuous. The ability to modulate specificity at will is particularly important for drug development, and discovery of mechanisms contributing to binding specificity is crucial for our basic understanding of biology and for applications in health care. In this study, we used computational molecular design to create a large dataset of diverse small molecules with a range of binding specificities. We then performed structural, energetic, and statistical analysis on the dataset to study molecular mechanisms of achieving specificity goals. The work was done in the context of HIV‐1 protease inhibition and the molecular designs targeted a panel of wild‐type and drug‐resistant mutant HIV‐1 protease structures. The analysis focused on mechanisms for promiscuous binding to bind robustly even to resistance mutants. Broadly binding inhibitors tended to be smaller in size, more flexible in chemical structure, and more hydrophobic in nature compared to highly selective ones. Furthermore, structural and energetic analyses illustrated mechanisms by which flexible inhibitors achieved binding; we found ligand conformational adaptation near mutation sites and structural plasticity in targets through torsional flips of asymmetric functional groups to form alternative, compensatory packing interactions or hydrogen bonds. As no inhibitor bound to all variants, we designed small cocktails of inhibitors to do so and discovered that they often jointly covered the target set through mechanistic complementarity. Furthermore, using structural plasticity observed in experiments, and potentially in simulations, is suggested to be a viable means of designing adaptive inhibitors that are promiscuous binders. Proteins 2015; 83:351–372. © 2014 Wiley Periodicals, Inc.

Impact of Y181C and/or H221Y mutation patterns of HIV-1 reverse transcriptase on phenotypic resistance to available non-nucleoside and nucleoside inhibitors in China

BACKGROUND

The aim of this study was to investigate the role of K101Q, Y181C and H221Y emerging in HIV-1 reverse transcriptase with different mutations patterns in phenotypic susceptibility to currently available NNRTIs (nevirapine NVP, efavirenz EFV) and NRTIs (zidovudine AZT, lamivudine 3TC, stavudine d4T) in China.

METHODS

Phenotype testing of currently available NNRTIs (NVP, EFV) and NRTIs (AZT, 3TC, d4T) was performed on TZM-b1 cells using recombined virus strains. P ≤ 0.05 was defined significant considering the change of 50% inhibitory drug concentration (IC50) compared with the reference, while P ≤ 0.01 was considered to be statistically significant considering multiple comparisons.

RESULTS

Triple-mutation K101Q/Y181C/H221Y and double-mutation K101Q/Y181C resulted in significant increase in NVP resistance (1253.9-fold and 986.4-fold), while only K101Q/Y181C/H221Y brought a 5.00-fold significant increase in EFV resistance. Remarkably, K101Q/H221Y was hypersusceptible to EFV (FC = 0.04), but was significantly resistant to the three NRTIs. Then, the interaction analysis suggested the interaction was not significant to NVP (F = 0.77, P = 0.4061) but significant to EFV and other three NRTIs.

CONCLUSION

Copresence of mutations reported to be associated with NNRTIs confers significant increase to NVP resistance. Interestingly, some may increase the susceptibility to EFV. Certainly, the double mutation (K101Q/H221Y) also changes the susceptibility of viruses to NRTIs. Interaction between two different sites makes resistance more complex.

Molecular basis of the association of H208Y and thymidine analogue resistance mutations M41L, L210W and T215Y in the HIV-1 reverse transcriptase of treated patients

Thymidine analogue resistance mutations (TAMs) in HIV-1 reverse transcriptase (RT) associate in two clusters: (i) TAM1 (M41L, L210W and T215Y) and TAM2 (D67N, K70R, K219E/Q, and sometimes T215F). The amino acid substitution H208Y shows increased prevalence in patients treated with nucleoside analogues and is frequently associated with TAM1 mutations. We studied the molecular mechanism favoring the selection of H208Y in the presence of zidovudine, tenofovir and other nucleoside RT inhibitors (NRTIs). NRTI susceptibility was not affected by the addition of H208Y in phenotypic assays carried out in MT-4 cells using recombinant HIV-1 containing wild-type (subtype B, BH10), H208Y, M41L/L210W/T215Y or M41L/H208Y/L210W/T215Y RTs. However, enzymatic studies carried out with purified RTs revealed that in the presence of M41L/L210W/T215Y, H208Y increases the RT's ability to unblock and extend primers terminated with zidovudine, tenofovir and in a lesser extent, stavudine. These effects were observed with DNA/DNA complexes (but not with RNA/DNA) and resulted from the higher ATP-dependent excision activity of the M41L/H208Y/L210W/T215Y RT compared with the M41L/L210W/T215Y mutant. The increased rescue efficiency of the M41L/H208Y/L210W/T215Y RT was observed in the presence of ATP but not with GTP or ITP. Molecular dynamics studies predict an alteration of the stacking interactions between Tyr(215) and the adenine ring of ATP due to long-distance effects caused by tighter packaging of Tyr(208) and Trp(212). These studies provide a mechanistic explanation for the association of TAM-1 and H208Y mutations in viral isolates from patients treated with NRTIs.

Comparable long-term efficacy of Lopinavir/Ritonavir and similar drug-resistance profiles in different HIV-1 subtypes

BACKGROUND

Analysis of potentially different impact of Lopinavir/Ritonavir (LPV/r) on non-B subtypes is confounded by dissimilarities in the conditions existing in different countries. We retrospectively compared its impact on populations infected with subtypes B and C in Israel, where patients infected with different subtypes receive the same treatment.

METHODS

Clinical and demographic data were reported by physicians. Resistance was tested after treatment failure. Statistical analyses were conducted using SPSS.

RESULTS

607 LPV/r treated patients (365 male) were included. 139 had HIV subtype B, 391 C, and 77 other subtypes. At study end 429 (71%) were receiving LPV/r. No significant differences in PI treatment history and in median viral-load (VL) at treatment initiation and termination existed between subtypes. MSM discontinued LPV/r more often than others even when the virologic outcome was good (p = 0.001). VL was below detection level in 81% of patients for whom LPV/r was first PI and in 67% when it was second (P = 0.001). Median VL decrease from baseline was 1.9±0.1 logs and was not significantly associated with subtype. Median CD4 increase was: 162 and 92cells/µl, respectively, for patients receiving LPV/r as first and second PI (P = 0.001), and 175 and 98, respectively, for subtypes B and C (P<0.001). Only 52 (22%) of 237 patients genotyped while under LPV/r were fully resistant to the drug; 12(5%) were partially resistant. In48%, population sequencing did not reveal resistance to any drug notwithstanding the virologic failure. No difference was found in the rates of resistance development between B and C (p = 0.16).

CONCLUSIONS

Treatment with LPV/r appeared efficient and tolerable in both subtypes, B and C, but CD4 recovery was significantly better in virologically suppressed subtype-B patients. In both subtypes, LPV/r was more beneficial when given as first PI. Mostly, reasons other than resistance development caused discontinuation of treatment.

Sequence and structure based models of HIV-1 protease and reverse transcriptase drug resistance

BACKGROUND

Successful management of chronic human immunodeficiency virus type 1 (HIV-1) infection with a cocktail of antiretroviral medications can be negatively affected by the presence of drug resistant mutations in the viral targets. These targets include the HIV-1 protease (PR) and reverse transcriptase (RT) proteins, for which a number of inhibitors are available on the market and routinely prescribed. Protein mutational patterns are associated with varying degrees of resistance to their respective inhibitors, with extremes that can range from continued susceptibility to cross-resistance across all drugs.

RESULTS

Here we implement statistical learning algorithms to develop structure- and sequence-based models for systematically predicting the effects of mutations in the PR and RT proteins on resistance to each of eight and eleven inhibitors, respectively. Employing a four-body statistical potential, mutant proteins are represented as feature vectors whose components quantify relative environmental perturbations at amino acid residue positions in the respective target structures upon mutation. Two approaches are implemented in developing sequence-based models, based on use of either relative frequencies or counts of n-grams, to generate vectors for representing mutant proteins. To the best of our knowledge, this is the first reported study on structure- and sequence-based predictive models of HIV-1 PR and RT drug resistance developed by implementing a four-body statistical potential and n-grams, respectively, to generate mutant attribute vectors. Performance of the learning methods is evaluated on the basis of tenfold cross-validation, using previously assayed and publicly available in vitro data relating mutational patterns in the targets to quantified inhibitor susceptibility changes.

CONCLUSION

Overall performance results are competitive with those of a previously published study utilizing a sequence-based strategy, while our structure- and sequence-based models provide orthogonal and complementary prediction methodologies, respectively. In a novel application, we describe a technique for identifying every possible pair of RT inhibitors as either potentially effective together as part of a cocktail, or a combination that is to be avoided.

Elucidating a relationship between conformational sampling and drug resistance in HIV-1 protease

Enzyme targets in rapidly replicating systems, such as retroviruses, commonly respond to drug-selective pressure with mutations arising in the active site pocket that limit inhibitor effectiveness by introducing steric hindrance or by eliminating essential molecular interactions. However, these primary mutations are disposed to compromising pathogenic fitness. Emerging secondary mutations, which are often found outside of the binding cavity, may or can restore fitness while maintaining drug resistance. The accumulated drug pressure selected mutations could have an indirect effect in the development of resistance, such as altering protein flexibility or the dynamics of protein-ligand interactions. Here, we show that accumulation of mutations in a drug-resistant HIV-1 protease (HIV-1 PR) variant, D30N/M36I/A71V, changes the fractional occupancy of the equilibrium conformational sampling ensemble. Correlations are made among populations of the conformational states, namely, closed-like, semiopen, and open-like, with inhibition constants, as well as kinetic parameters. Mutations that stabilize a closed-like conformation correlate with enzymes of lowered activity and with higher affinity for inhibitors, which is corroborated by a further increase in the fractional occupancy of the closed state upon addition of inhibitor or substrate-mimic. Cross-resistance is found to correlate with combinations of mutations that increase the population of the open-like conformations at the expense of the closed-like state while retaining native-like occupancy of the semiopen population. These correlations suggest that at least three states are required in the conformational sampling model to establish the emergence of drug resistance in HIV-1 PR. More importantly, these results shed light on a possible mechanism whereby mutations combine to impart drug resistance while maintaining catalytic activity.

Subtype-specific differences in the development of accessory mutations associated with high-level resistance to HIV-1 nucleoside reverse transcriptase inhibitors

OBJECTIVES

To identify accessory mutations associated with high-level resistance to reverse transcriptase (RT) inhibitors in HIV-1 subtypes B and C.

METHODS

Changes relative to the wild-type for codons 1-400 of RT were analysed from treatment-experienced patients infected with subtypes B (5464 patients) and C (1920 patients). Positions associated with the accumulation of mutations conferring resistance to thymidine analogues and to non-nucleoside RT inhibitors (NNRTIs) were identified. A subtype-specific single-replication cycle drug susceptibility assay was used to determine whether some of the mutations affected drug susceptibility or viral infectivity.

RESULTS

In subtype B, mutations at 31 and 26 positions were associated with the accumulation of thymidine analogue mutations (TAMs) and NNRTI mutations, respectively; in subtype C, 18 and 13 positions were identified, respectively. Amino acid changes at the following positions were differentially associated with (i) the accumulation of 0-4+ TAMs in subtypes B and C (away from consensus): 43 (27.0% B versus 2.5% C); 118 (36.4% B versus 16.2% C); 135 (12.5% B versus 28.0% C); and 326 (2.6% towards consensus in B versus 7.6% away in C) and (ii) the accumulation of 0-3+ NNRTI mutations (away from consensus): 43 (10.2% B versus 0.5% C); and 68 (5.2% B versus 10.3% C). Codon changes K43E, E44D and V118I were found to have no effect on susceptibility to three NRTIs with or without TAMs in either subtype; however, some accessory mutations had subtype-specific effects on viral infectivity.

CONCLUSIONS

Differences between subtypes B and C were observed in the development and effect of accessory mutations associated with high-level resistance to RT inhibitors.

Prediction of mutational tolerance in HIV-1 protease and reverse transcriptase using flexible backbone protein design

Predicting which mutations proteins tolerate while maintaining their structure and function has important applications for modeling fundamental properties of proteins and their evolution; it also drives progress in protein design. Here we develop a computational model to predict the tolerated sequence space of HIV-1 protease reachable by single mutations. We assess the model by comparison to the observed variability in more than 50,000 HIV-1 protease sequences, one of the most comprehensive datasets on tolerated sequence space. We then extend the model to a second protein, reverse transcriptase. The model integrates multiple structural and functional constraints acting on a protein and uses ensembles of protein conformations. We find the model correctly captures a considerable fraction of protease and reverse-transcriptase mutational tolerance and shows comparable accuracy using either experimentally determined or computationally generated structural ensembles. Predictions of tolerated sequence space afforded by the model provide insights into stability-function tradeoffs in the emergence of resistance mutations and into strengths and limitations of the computational model.

Many related protein sequences can be consistent with the structure and function of a given protein, suggesting that proteins may be quite robust to mutations. This tolerance to mutations is frequently exploited by pathogens. In particular, pathogens can rapidly evolve mutated proteins that have a new function - resistance against a therapeutic inhibitor - without abandoning other functions essential for the pathogen. This principle may also hold more generally: Proteins tolerant to mutational changes can more easily acquire new functions while maintaining their existing properties. The ability to predict the tolerance of proteins to mutation could thus help both to analyze the emergence of resistance mutations in pathogens and to engineer proteins with new functions. Here we develop a computational model to predict protein mutational tolerance towards point mutations accessible by single nucleotide changes, and validate it using two important pathogenic proteins and therapeutic targets: the protease and reverse transcriptase from HIV-1. The model provides insights into how resistance emerges and makes testable predictions on mutations that have not been seen yet. Similar models of mutational tolerance should be useful for characterizing and reengineering the functions of other proteins for which a three-dimensional structure is available.

Prediction of HIV-1 protease/inhibitor affinity using RosettaLigand

Predicting HIV-1 protease/inhibitor binding affinity as the difference between the free energy of the inhibitor bound and unbound state remains difficult as the unbound state exists as an ensemble of conformations with various degrees of flap opening. We improve computational prediction of protease/inhibitor affinity by invoking the hypothesis that the free energy of the unbound state while difficult to predict is less sensitive to mutation. Thereby the HIV-1 protease/inhibitor binding affinity can be approximated with the free energy of the bound state alone. Bound state free energy can be predicted from comparative models of HIV-1 protease mutant/inhibitor complexes. Absolute binding energies are predicted with R = 0.71 and SE = 5.91 kJ/mol. Changes in binding free energy upon mutation can be predicted with R = 0.85 and SE = 4.49 kJ/mol. Resistance mutations that lower inhibitor binding affinity can thereby be recognized early in HIV-1 protease inhibitor development.

Ultra-deep sequencing of HIV-1 reverse transcriptase before start of an NNRTI-based regimen in treatment-naive patients

There are conflicting data on the impact of low frequency HIV-1 drug-resistant mutants on the response of first-line highly active antiretroviral therapy (HAART), more specifically containing a NNRTI. As population sequencing does not detect resistant viruses representing less than 15-25% of the viral population, more sensitive techniques have been developed but still need clinical validation. We evaluated ultra-deep sequencing (UDPS), recently more available and affordable, as a tool for the detection of HIV-1 minority species carrying drug resistant mutation (DRM) in a clinical setting. A retrospective analysis of the reverse transcriptase (RT) gene of plasma HIV-1 from 70 patients starting a NNRTI based regimen was performed. Minority populations were defined as representing > 1% and < 20% of the total viral population. Using UDPS, we could not confirm an association between the presence of low minority variants harbouring RT mutations at the start of therapy and primary or secondary therapeutic failure.

A multifaceted analysis of HIV-1 protease multidrug resistance phenotypes

Background

Great strides have been made in the effective treatment of HIV-1 with the development of second-generation protease inhibitors (PIs) that are effective against historically multi-PI-resistant HIV-1 variants. Nevertheless, mutation patterns that confer decreasing susceptibility to available PIs continue to arise within the population. Understanding the phenotypic and genotypic patterns responsible for multi-PI resistance is necessary for developing PIs that are active against clinically-relevant PI-resistant HIV-1 variants.

Results

In this work, we use globally optimal integer programming-based clustering techniques to elucidate multi-PI phenotypic resistance patterns using a data set of 398 HIV-1 protease sequences that have each been phenotyped for susceptibility toward the nine clinically-approved HIV-1 PIs. We validate the information content of the clusters by evaluating their ability to predict the level of decreased susceptibility to each of the available PIs using a cross validation procedure. We demonstrate the finding that as a result of phenotypic cross resistance, the considered clinical HIV-1 protease isolates are confined to ~6% or less of the clinically-relevant phenotypic space. Clustering and feature selection methods are used to find representative sequences and mutations for major resistance phenotypes to elucidate their genotypic signatures. We show that phenotypic similarity does not imply genotypic similarity, that different PI-resistance mutation patterns can give rise to HIV-1 isolates with similar phenotypic profiles.

Conclusion

Rather than characterizing HIV-1 susceptibility toward each PI individually, our study offers a unique perspective on the phenomenon of PI class resistance by uncovering major multidrug-resistant phenotypic patterns and their often diverse genotypic determinants, providing a methodology that can be applied to understand clinically-relevant phenotypic patterns to aid in the design of novel inhibitors that target other rapidly evolving molecular targets as well.

Added value of deep sequencing relative to population sequencing in heavily pre-treated HIV-1-infected subjects

OBJECTIVE

To explore the potential of deep HIV-1 sequencing for adding clinically relevant information relative to viral population sequencing in heavily pre-treated HIV-1-infected subjects.

METHODS

In a proof-of-concept study, deep sequencing was compared to population sequencing in HIV-1-infected individuals with previous triple-class virological failure who also developed virologic failure to deep salvage therapy including, at least, darunavir, tipranavir, etravirine or raltegravir. Viral susceptibility was inferred before salvage therapy initiation and at virological failure using deep and population sequencing genotypes interpreted with the HIVdb, Rega and ANRS algorithms. The threshold level for mutant detection with deep sequencing was 1%.

RESULTS

7 subjects with previous exposure to a median of 15 antiretrovirals during a median of 13 years were included. Deep salvage therapy included darunavir, tipranavir, etravirine or raltegravir in 4, 2, 2 and 5 subjects, respectively. Self-reported treatment adherence was adequate in 4 and partial in 2; one individual underwent treatment interruption during follow-up. Deep sequencing detected all mutations found by population sequencing and identified additional resistance mutations in all but one individual, predominantly after virological failure to deep salvage therapy. Additional genotypic information led to consistent decreases in predicted susceptibility to etravirine, efavirenz, nucleoside reverse transcriptase inhibitors and indinavir in 2, 1, 2 and 1 subject, respectively. Deep sequencing data did not consistently modify the susceptibility predictions achieved with population sequencing for darunavir, tipranavir or raltegravir.

CONCLUSIONS

In this subset of heavily pre-treated individuals, deep sequencing improved the assessment of genotypic resistance to etravirine, but did not consistently provide additional information on darunavir, tipranavir or raltegravir susceptibility. These data may inform the design of future studies addressing the clinical value of minority drug-resistant variants in treatment-experienced subjects.

Structures of reverse transcriptase pre- and post-excision complexes shed new light on HIV-1 AZT resistance

HIV-1 resistance to 3′-azido-2′,3′-deoxythymidine (AZT, zidovudine) results from mutations in reverse transcriptase that increase the ability of the enzyme to excise AZT-monophosphate after it has been incorporated. Crystal structures of complexes of wild type and mutant reverse transcriptase with double-stranded DNA with or without the excision product, AZT adenosine dinucleoside tetraphosphate (AZTppppA), have recently been reported [1]. The excision-enhancing mutations dramatically change the way the enzyme interacts with the excision product.

Mining protein dynamics from sets of crystal structures using "consensus structures"

In this work, we describe two novel approaches to utilize the dynamic structure information implicitly contained in large crystal structure data sets. The first approach visualizes both consistent as well as variable ligand-induced changes in ligand-bound compared with apo protein crystal structures. For this purpose, information was mined from B-factors and ligand-induced residue displacements in multiple crystal structures, minimizing experimental error and noise. With this approach, the mechanism of action of non-nucleoside reverse transcriptase inhibitors (NNRTIs) as an inseparable combination of distortion of protein dynamics and conformational changes of HIV-1 reverse transcriptase was corroborated (a combination of the previously proposed "molecular arthritis" and "distorted site" mechanisms). The second approach presented here uses "consensus structures" to map common binding features that are present in a set of structures of NNRTI-bound HIV-1 reverse transcriptase. Consensus structures are based on different levels of structural overlap of multiple crystal structures and are used to analyze protein-ligand interactions. The structures are shown to yield information about conserved hydrogen bonding interactions as well as binding-pocket flexibility, shape, and volume. From the consensus structures, a common wild type NNRTI binding pocket emerges. Furthermore, we were able to identify a conserved backbone hydrogen bond acceptor at P236 and a novel hydrophobic subpocket, which are not yet utilized by current drugs. Our methods introduced here reinterpret the atom information and make use of the data variability by using multiple structures, complementing classical 3D structural information of single structures.

HIV-1 protease mutations and protease inhibitor cross-resistance

The effects of many protease inhibitor (PI)-selected mutations on the susceptibility to individual PIs are unknown. We analyzed in vitro susceptibility test results on 2,725 HIV-1 protease isolates. More than 2,400 isolates had been tested for susceptibility to fosamprenavir, indinavir, nelfinavir, and saquinavir; 2,130 isolates had been tested for susceptibility to lopinavir; 1,644 isolates had been tested for susceptibility to atazanavir; 1,265 isolates had been tested for susceptibility to tipranavir; and 642 isolates had been tested for susceptibility to darunavir. We applied least-angle regression (LARS) to the 200 most common mutations in the data set and identified a set of 46 mutations associated with decreased PI susceptibility of which 40 were not polymorphic in the eight most common HIV-1 group M subtypes. We then used least-squares regression to ascertain the relative contribution of each of these 46 mutations. The median number of mutations associated with decreased susceptibility to each PI was 28 (range, 19 to 32), and the median number of mutations associated with increased susceptibility to each PI was 2.5 (range, 1 to 8). Of the mutations with the greatest effect on PI susceptibility, I84AV was associated with decreased susceptibility to eight PIs; V32I, G48V, I54ALMSTV, V82F, and L90M were associated with decreased susceptibility to six to seven PIs; I47A, G48M, I50V, L76V, V82ST, and N88S were associated with decreased susceptibility to four to five PIs; and D30N, I50L, and V82AL were associated with decreased susceptibility to fewer than four PIs. This study underscores the greater impact of nonpolymorphic mutations compared with polymorphic mutations on decreased PI susceptibility and provides a comprehensive quantitative assessment of the effects of individual mutations on susceptibility to the eight clinically available PIs.

Co-lethality studied as an asset against viral drug escape: the HIV protease case

BACKGROUND

Co-lethality, or synthetic lethality is the documented genetic situation where two, separately non-lethal mutations, become lethal when combined in one genome. Each mutation is called a "synthetic lethal" (SL) or a co-lethal. Like invariant positions, SL sets (SL linked couples) are choice targets for drug design against fast-escaping RNA viruses: mutational viral escape by loss of affinity to the drug may induce (synthetic) lethality.

RESULTS

From an amino acid sequence alignment of the HIV protease, we detected the potential SL couples, potential SL sets, and invariant positions. From the 3D structure of the same protein we focused on the ones that were close to each other and accessible on the protein surface, to possibly bind putative drugs. We aligned 24,155 HIV protease amino acid sequences and identified 290 potential SL couples and 25 invariant positions. After applying the distance and accessibility filter, three candidate drug design targets of respectively 7 (under the flap), 4 (in the cantilever) and 5 (in the fulcrum) amino acid positions were found.

CONCLUSIONS

These three replication-critical targets, located outside of the active site, are key to our anti-escape strategy. Indeed, biological evidence shows that 2/3 of those target positions perform essential biological functions. Their mutational variations to escape antiviral medication could be lethal, thus limiting the apparition of drug-resistant strains.

REVIEWERS

This article was reviewed by Arcady Mushegian, Shamil Sunyaev and Claus Wilke.

Resistance pathways of human immunodeficiency virus type 1 against the combination of zidovudine and lamivudine

A better understanding of human immunodeficiency virus type 1 drug-resistance evolution under the selective pressure of combination treatment is important for the design of long-term effective treatment strategies. We applied Bayesian network learning to sequences from patients treated with the reverse transcriptase inhibitor combination of zidovudine (AZT) and lamivudine (3TC) to identify the role of many treatment-selected mutations in the development of resistance. Based on the Bayesian network structure, an in vivo fitness landscape was built, reflecting the necessary selective pressure under treatment, to evolve naive sequences to sequences obtained from patients treated with the combination. This landscape, combined with an evolutionary model, was used to predict resistance evolution in longitudinal sequence pairs. In our analysis, mutations 41L, 70R, 184V and 215F/Y were identified as major resistance mutations to the combination of AZT and 3TC, as they were associated directly with treatment experience. The network also suggested a possible role in resistance development for a number of novel mutations. Estimated fitness, using the landscape, correlated significantly with in vitro resistance phenotype in genotype-phenotype pairs (R(2)=0.70). Variation in predicted evolution under selective pressure correlated significantly with observed in vivo evolution during AZT plus 3CT treatment. In conclusion, we confirmed current knowledge on resistance development to the combination of AZT and 3CT, but additional novel mutations were identified. Moreover, a model to predict resistance evolution during AZT and 3CT treatment has been built and validated.

New approaches to HIV protease inhibitor drug design II: testing the substrate envelope hypothesis to avoid drug resistance and discover robust inhibitors

PURPOSE OF REVIEW

Drug resistance results when the balance between the binding of inhibitors and the turnover of substrates is perturbed in favor of the substrates. Resistance is quite widespread to the HIV-1 protease inhibitors permitting the protease to process its 10 different substrates. This processing of the substrates permits the virus HIV-1 to mature and become infectious. The design of HIV-1 protease inhibitors that closely fit within the substrate-binding region is proposed to be a strategy to avoid drug resistance.

RECENT FINDINGS

Cocrystal structures of HIV-1 protease with its substrates define an overlapping substrate-binding region or substrate envelope. Novel HIV-1 protease inhibitors that were designed to fit within this substrate envelope were found to retain high binding affinity and have a flat binding profile against a panel of drug-resistant HIV-1 proteases.

SUMMARY

The avoidance of drug resistance needs to be considered in the initial design of inhibitors to quickly evolving targets such as HIV-1 protease. Using a detailed knowledge of substrate binding appears to be a promising strategy for achieving this goal to obtain robust HIV-1 protease inhibitors.

Nonpolymorphic human immunodeficiency virus type 1 protease and reverse transcriptase treatment-selected mutations

The spectrum of human immunodeficiency virus type 1 (HIV-1) protease and reverse transcriptase (RT) mutations selected by antiretroviral (ARV) drugs requires ongoing reassessment as ARV treatment patterns evolve and increasing numbers of protease and RT sequences of different viral subtypes are published. Accordingly, we compared the prevalences of protease and RT mutations in HIV-1 group M sequences from individuals with and without a history of previous treatment with protease inhibitors (PIs) or RT inhibitors (RTIs). Mutations in protease sequences from 26,888 individuals and in RT sequences from 25,695 individuals were classified according to whether they were nonpolymorphic in untreated individuals and whether their prevalence increased fivefold with ARV therapy. This analysis showed that 88 PI-selected and 122 RTI-selected nonpolymorphic mutations had a prevalence that was fivefold higher in individuals receiving ARVs than in ARV-naïve individuals. This was an increase of 47% and 77%, respectively, compared with the 60 PI- and 69 RTI-selected mutations identified in a similar analysis that we published in 2005 using subtype B sequences obtained from one-fourth as many individuals. In conclusion, many nonpolymorphic mutations in protease and RT are under ARV selection pressure. The spectrum of treatment-selected mutations is changing as data for more individuals are collected, treatment exposures change, and the number of available sequences from non-subtype B viruses increases.

Molecular characterization of clinical isolates of human immunodeficiency virus resistant to the protease inhibitor darunavir

Darunavir is the most recently approved human immunodeficiency virus (HIV) protease (PR) inhibitor (PI) and is active against many HIV type 1 PR variants resistant to earlier-generation PIs. Darunavir shows a high genetic barrier to resistance development, and virus strains with lower sensitivity to darunavir have a higher number of PI resistance-associated mutations than viruses resistant to other PIs. In this work, we have enzymologically and structurally characterized a number of highly mutated clinically derived PRs with high levels of phenotypic resistance to darunavir. With 18 to 21 amino acid residue changes, the PR variants studied in this work are the most highly mutated HIV PR species ever studied by means of enzyme kinetics and X-ray crystallography. The recombinant proteins showed major defects in substrate binding, while the substrate turnover was less affected. Remarkably, the overall catalytic efficiency of the recombinant PRs (5% that of the wild-type enzyme) is still sufficient to support polyprotein processing and particle maturation in the corresponding viruses. The X-ray structures of drug-resistant PRs complexed with darunavir suggest that the impaired inhibitor binding could be explained by change in the PR-inhibitor hydrogen bond pattern in the P2' binding pocket due to a substantial shift of the aminophenyl moiety of the inhibitor. Recombinant virus phenotypic characterization, enzyme kinetics, and X-ray structural analysis thus help to explain darunavir resistance development in HIV-positive patients.

Genotypic analysis of the protease and reverse transcriptase of non-B HIV type 1 clinical isolates from naïve and treated subjects

One hundred and ninety-two pol sequences of drug-naïve and drug-experienced subjects infected with non-B HIV-1 subtypes were analyzed to identify treatment-related amino acid changes which might be relevant for drug-resistance and possibly not included in the accepted mutation list for the B subtype. The correspondence analysis identified non-B-specific and subtype-specific polymorphisms which should not be mistaken for mutations. Multiple chi(2) were performed to detect the differences between naïve vs treated subjects and between different subtypes. To verify the contribution of each single mutation to the resistance levels as predicted by the Virtual Phenotype-LM, simple univariate linear regression was used with fold resistance as a dependent variable and individual mutations as predictors. Commonly accepted protease (PR) and reverse transcriptase (RT) positions along with mutants at RT positions 118 and 90 were significantly associated with treatment. Two unusual PR (K14R and I66F) and five RT positions (E28K, S68G, H221Y, L228R/H and P294A) were also associated with treatment (p<0.01). Only minimal variations were observed with respect to commonly accepted amino acid changes. All amino acid changes correlated with treatment influenced the resistance levels to each single drug. Our findings demonstrate that there are no substantial differences regarding known resistance-associated mutations and the newly emergent substitutions between non-B and B subtype strains.

Viral protease inhibitors

This review provides an overview of the development of viral protease inhibitors as antiviral drugs. We concentrate on HIV-1 protease inhibitors, as these have made the most significant advances in the recent past. Thus, we discuss the biochemistry of HIV-1 protease, inhibitor development, clinical use of inhibitors, and evolution of resistance. Since many different viruses encode essential proteases, it is possible to envision the development of a potent protease inhibitor for other viruses if the processing site sequence and the catalytic mechanism are known. At this time, interest in developing inhibitors is limited to viruses that cause chronic disease, viruses that have the potential to cause large-scale epidemics, or viruses that are sufficiently ubiquitous that treating an acute infection would be beneficial even if the infection was ultimately self-limiting. Protease inhibitor development is most advanced for hepatitis C virus (HCV), and we also provide a review of HCV NS3/4A serine protease inhibitor development, including combination therapy and resistance. Finally, we discuss other viral proteases as potential drug targets, including those from Dengue virus, cytomegalovirus, rhinovirus, and coronavirus.

National survey for drug-resistant variants in newly diagnosed antiretroviral drug-naive patients with HIV/AIDS in South Korea: 1999-2005

We investigated the prevalence of drug-resistant variants and assessed their severity against antiretroviral drugs among patients in South Korea. Three hundred antiretroviral drug-naive patients were collected and drug-resistant variants were analyzed using the Stanford database with sequences and mutation data of the HIV-1 genes for protease (codons 1-99) and reverse transcriptase (codons 1-250). Of this group, 199 isolates (66.3%) showed at least 1 or more sites related to drug resistance. However, the average prevalence of drug resistance for patients newly diagnosed with HIV-1 but still treatment-naive between 1999 and 2005 was very low (4.3%, by "SIR" interpretation) compared with other countries. Most of the newly infected patients carried HIV subtype B (96%, n = 288) based on phylogenetic analysis of the conserved pol region. In summary, there has been no significant increase in the prevalence of drug resistance among antiretroviral drug-naive patients infected with HIV-1 for the last 7 years in South Korea. This study is quite significant regarding its larger scale of prevalence study for drug-resistant variants comparing to other drug-resistant studies using small scale of populations in South Korea. It is also important to provide suitable guidelines of genotyping assays for Korean drug-naive patients.

Bayesian network analyses of resistance pathways against efavirenz and nevirapine

OBJECTIVE

To clarify the role of novel mutations selected by treatment with efavirenz or nevirapine, and investigate the influence of HIV-1 subtype on nonnucleoside reverse transcriptase inhibitor (nNRTI) resistance pathways.

DESIGN

By finding direct dependencies between treatment-selected mutations, the involvement of these mutations as minor or major resistance mutations against efavirenz, nevirapine, or coadministrated nucleoside analogue reverse transcriptase inhibitors (NRTIs) is hypothesized. In addition, direct dependencies were investigated between treatment-selected mutations and polymorphisms, some of which are linked with subtype, and between NRTI and nNRTI resistance pathways.

METHODS

Sequences from a large collaborative database of various subtypes were jointly analyzed to detect mutations selected by treatment. Using Bayesian network learning, direct dependencies were investigated between treatment-selected mutations, NRTI and nNRTI treatment history, and known NRTI resistance mutations.

RESULTS

Several novel minor resistance mutations were found: 28K and 196R (for resistance against efavirenz), 101H and 138Q (nevirapine), and 31L (lamivudine). Robust interactions between NRTI mutations (65R, 74V, 75I/M, and 184V) and nNRTI resistance mutations (100I, 181C, 190E and 230L) may affect resistance development to particular treatment combinations. For example, an interaction between 65R and 181C predicts that the nevirapine and tenofovir and lamivudine/emtricitabine combination should be more prone to failure than efavirenz and tenofovir and lamivudine/emtricitabine.

CONCLUSION

Bayesian networks were helpful in untangling the selection of mutations by NRTI versus nNRTI treatment, and in discovering interactions between resistance mutations within and between these two classes of inhibitors.