Detection of human immunodeficiency virus type 1 clinical isolates with reduced sensitivity to zidovudine and dideoxyinosine by RNA.RNA hybridization

In vitro Selection of Human Immunodeficiency Virus Type 1 Resistant to Ro 31-8959 Proteinase Inhibitor

Human immunodeficiency virus type I (HIV-1) resistant to the proteinase inhibitor Ro 31–8959 has been obtained by growing HIV-IIIB in C8166 cell cultures in the presence of inhibitory concentrations of the drug.

Ro 31-8959-resistant HIV-1 obtained after five passages in the presence of the drug was capable of replicating, as measured by infectious virus yield, in the presence of concentrations of Ro 31-8959 able to inhibit completely the replication of the parental HIV-IIIB.

Cloning of the Ro 31–8959-resistant viral preparations revealed that a number of different strains of HIV-1 with various degrees of sensitivity to Ro 31–8959 emerged during propagation of HIV-IIIB in C8166 cells in the presence of the drug. The most resistant strain shows an ID50 value of 21.8 nM that is 13.2 fold higher than the value observed in HIV-IIIB infected C8166 cells.

Drug Combination of AZT and ddl: Synergism of Action and Prevention of Appearance of AZT-Resistance

Both 3′-azido-3′-deoxythymidine (AZT) and 2′,3′-dideoxynosine (ddl) strongly inhibit the replication of human immunodeficiency virus type 1 (HIV-1). Here, it is shown that combination of AZT and ddl at concentrations that are readily achievable in vivo synergistically inhibit HIV-1 replication in C8166 cells and peripheral blood mononuclear cells. The synergism is significant even when the effect of AZT and ddl alone was negligible.

Our findings show that AZT-resistance is less likely to occur when a combination of AZT and ddl is used. Particularly, generation of AZT-resistant strains by in vitro selection is prevented, or delayed, by the combination of AZT plus ddl.

Taken together these observations provide a rationale for combination of AZT and ddl in the therapy of AIDS patients.

In vitro Resistance to an Inhibitor of HIV Proteinase (Ro 31-8959) Relative to Inhibitors of Reverse Transcriptase (AZT and TIBO)

Serial passage of cell-free human immunodeficiency virus type 1 (HIV-1) strain GB8 on CEM cells was carried out in the presence of increasing concentrations of the HIV proteinase inhibitor Ro 31-8959 in parallel with representative reverse transcriptase (RT) inhibitors (AZT and the TIBO compound, R82150). In all instances, a significant increase in the concentration of compound required to produce a 90% reduction of syncytium formation (IC90) was found after seven to nine passages of virus. Reduced sensitivity to Ro 31 −8959 was confirmed by p24 ELISA. Virus passaged in the presence of RT inhibitors did not show a significant change in either the ability to grow in culture supplemented with step-wise increments of inhibitor concentration or the rate of growth in the presence of compound. In contrast, virus passaged in the presence of Ro 31 −8959 required, on average, more than 2.5-times the normal passage time and often did not replicate when increases in inhibitor concentration were applied during the initial passages.

The results show that reduced sensitivity to an inhibitor of HIV proteinase, Ro 31–8959, can be generated, but it would seem to arise less readily than that found with either of the RT inhibitors studied. While this study indicates the potential for a reduction in sensitivity to proteinase inhibitors, it does not necessarily reflect the ability of mutant virus either to be selected for or to be propagated in the clinic.

Antiviral Synergy between Inhibitors of HIV Proteinase and Reverse Transcriptase

Ro 31–8959, a potent and specific inhibitor of HIV proteinases, is shown to interact synergistically in combination with nucleoside analogue inhibitors of HIV reverse transcriptase (AZT, ddC, 2′-FddC). Ninety per cent inhibition endpoints (IC90), obtained from checkerboard titrations of compound mixtures on CEM-T4 cells infected with HIV-i strain GB8, have been further analysed. Compared with concentrations needed when inhibitors are used individually, reductions of between 2- and 30-fold were observed in combination, depending on the nucleoside analogue and the ratio of concentrations employed. The results suggest that combinations of AZT or ddC with Ro 31–8959 should be considered for development as candidate antiviral treatments of patients with HIV infection.

An Assay for HIV RNA in Infected Cell Lysates, and its use for the Rapid Evaluation of Antiviral Efficacy

A rapid, high-capacity assay for evaluating the potency of anti-HIV compounds was devised. This assay measures cell associated viral RNA levels 3 days after infection of susceptible T-cell lines grown in individual microtitre plate wells. Viral RNA was quantified by a sandwich hybridization assay, the first step of which was performed directly in crude infected cell lysates prepared in guanidinium isothio-cyanate. Levels of cell-associated viral RNA were shown to correlate with the yield of infectious virus and this correlation formed the basis of the test. Antiviral potencies of a large series of compounds tested in this RNA hybridization assay correlated closely with potency values determined by a sensitive but slower and more labour-intensive yield-reduction assay. Both laboratory strains and selected clinical isolates of HIV can be detected in this RNA hybridization assay.

Viral resistance patterns selected by antiretroviral drugs and their potential to guide treatment choice

Massive viral turnover and reverse transcriptase's high error rate create the potential for drug-resistant viral variants to appear rapidly under the selective pressure of antiretroviral therapy. Loss of antiviral effect in treatment-adherent persons is most commonly coincident with the appearance of viral mutants with reduced drug sensitivity. Thus, detection of viral resistance may represent an early marker of therapy failure. Similarly, control of viral replication in the plasma compartment, as defined by plasma viral load below the levels of assay quantification, is associated with a sustained therapeutic response and delayed development of viral resistance. Information on patterns of resistance to and cross-resistance between antiretroviral agents is increasingly well characterised and represents an important consideration when deciding how to combine and/or sequence antiretrovirals to achieve optimal antiviral effects. Given the limited number of antiretrovirals presently available or in advanced development, it is important not to limit future therapeutic options by using therapies early in the treatment sequence which may select for cross-resistant viral variants and hence potentially reduce the magnitude of therapeutic response when treatment is changed to another member of that drug class. However, no studies using resistance to guide clinical decision making have been reported to date and available sequencing studies have focused largely on switching or adding therapies to patients experienced with zidovudine monotherapy. Thus, no resistance driven treatment algorithm is currently available.

New real-time reverse transcriptase-initiated PCR assay with single-copy sensitivity for human immunodeficiency virus type 1 RNA in plasma

More sensitive assays for human immunodeficiency virus type 1 (HIV-1) RNA are needed to detect, quantify, and characterize persistent viremia in patients who are receiving antiretroviral therapy and whose plasma HIV-1 RNA levels are suppressed to less than 50 to 75 copies/ml. We therefore developed an internally controlled real-time reverse transcriptase-initiated PCR assay that quantifies HIV-1 RNA concentrations down to 1 copy per ml of plasma. This assay with single-copy sensitivity (the single-copy assay) generates a reproducible linear regression plot of input copy number versus threshold cycle by using HIV-1 RNA transcripts at copy numbers ranging from 1 to 10(6) per reaction mixture. The single-copy assay was compared to the ultrasensitive AMPLICOR HIV-1 MONITOR assay and a more sensitive modification of the ultrasensitive assay by repeatedly testing a low-copy-number panel containing 200 to 0.781 copies of HIV-1 RNA per ml of plasma. This comparison showed that the single-copy assay had a greater sensitivity than the other assays and was the only assay that detected HIV-1 RNA at levels as low as 0.781 copies/ml. Testing of plasma samples from 15 patients who were receiving antiretroviral therapy and who had <75 HIV-1 RNA copies/ml revealed persistent viremia in all 15 patients, with HIV-1 RNA levels ranging from 1 to 32 copies/ml (median, 13 copies/ml). The greater sensitivity of the single-copy assay should allow better characterization of persistent viremia in patients who are receiving antiretroviral therapy and whose HIV-1 RNA levels are suppressed to below the detection limits of present assays.

Performance of drug resistance assays in testing HIV-1 non-B subtypes

Antiretroviral susceptibility analyses were performed in plasma samples collected from 32 HIV-1 non-B-infected individuals, most of whom had received antiretroviral drugs. Reverse transcriptase (RT) and protease gene sequences were obtained, and 15 anti-HIV drugs were tested in a recombinant virus phenotypic assay. Phenotypic results were obtained in 25 (78.1%) samples, while genotypic data were recorded in 19 (59.4%). In seven samples (21.9%), neither genotypic nor phenotypic results were obtained. Ten of 13 samples with plasma HIV RNA below 2000 copies/mL did not yield genotypic results. Resistance assays work accurately when testing HIV-1 non-B subtypes. However, as for subtype B variants, a low viral load is the most important factor limiting the application of these tests.

Incidence of discontinuation of highly active antiretroviral combination therapy (HAART) and its determinants

OBJECTIVE

To determine the incidence and determinants for discontinuation of initial highly active antiretroviral therapy (HAART).

DESIGN

In this retrospective follow-up study from hospital files and pharmacy dispensing data, a standard dataset was collected including patient characteristics, therapy characteristics, and HIV-monitoring parameters (e.g., CD4+ lymphocyte counts, viral load determinations). Kaplan-Meier estimates of the cumulative probability of discontinuation of initial HAART were calculated. Cox proportional hazard analysis was used to identify determinants for discontinuation of initial HAART.

PATIENTS

All patients starting HAART (n = 99) during June 1996 to February 1997 at our regional AIDS center.

MAIN OUTCOME MEASURES

Incidence and determinants for discontinuation of HAART.

RESULTS

During the mean follow-up of 450+/-10 days, 27 patients switched initial HAART, 3 patients stopped any antiretroviral therapy. Reasons for switching were increasing viral load (18x), insufficient decrease of viral load (3x), and adverse events (6x). Nonnaivete for antiretroviral therapy and a lower CD4+ lymphocyte count at start were identified as determinants for discontinuation of initial HAART.

CONCLUSIONS

The overall incidence density for discontinuation of initial HAART was 25 per 100 patients/year. The main reason for switching was an increasing viral load. CD4+ lymphocyte counts at start and nonnaivete for antiretroviral therapy were identified as determinants for discontinuation.

Dominance of the E89G substitution in HIV-1 reverse transcriptase in regard to increased polymerase processivity and patterns of pausing

The substitution of a glycine for glutamic acid at position 89 in human immunodeficiency virus-1 (HIV-1) reverse transcriptase (RT) (E89G) confers resistance to several nucleoside and non-nucleoside inhibitors of RT. As residue 89 contacts the template strand, it has been suggested that this mutation may modulate the conformation of the RT.template/primer complex. In addition, certain mutations in RT that confer resistance to nucleoside analogs, such as M184V, are located near the polymerase active site. To characterize further these substitutions, we performed processivity assays alongside an analysis of pausing profiles with wild-type (wt) RT and recombinant RTs containing substitutions at E89G, M184V, or both. We now show that E89G RT has higher processivity than wt enzyme as well as a different pattern of pausing sites. Similar findings were obtained with the doubly mutated RT, although enzyme containing only the M184V mutation had lower processivity than wt. Consistent with these observations, and from a mechanistic standpoint, both E89G-containing as well as doubly mutated RT had decreased dissociation constants from a complex consisting of RT and template-primer, in comparison with either wt RT or M184V-containing RT. No significant differences were observed among the various enzymes in regard to Km values for the heteropolymeric RNA template used in these studies. Viruses containing the E89G mutation synthesized longer strand DNA products than either wt viruses or viruses containing only the M184V mutation in endogenous RT assays. Thus, the E89G substitution is a dominant determinant in regard to each of the koff values from an RT.template/primer complex, RT processivity, and specific patterns of pausing during DNA polymerization.

Imidazo[1,5-b]pyridazine-d4T conjugates: synthesis and anti-human immunodeficiency virus evaluation

In an attempt to combine the human immunodeficiency virus type 1 (HIV-1)-inhibitory capacity of 2',3'-dideoxy-2',3'-didehydronucleoside analogues [nucleoside reverse transcriptase (RT) inhibitors; NRTI] and non-nucleoside RT inhibitors (NNRTI), we have designed, synthesized and evaluated for their anti-HIV activity several heterodimers of the general formula [d4T]-NH-(CH2)n-NH-[imidazo[1,5-b]pyridazine]. The synthesis of these heterodimers was conducted in three parts. The first part focused on the synthesis of the NRTI. The second part was devoted to the NNRTI and the NNRTI linked to appropriate spacers: [NNRTI]-NH-(CH2)n-NH2. In the third part, the condensation between the NRTI and the [NNRTI]-NH-(CH2)n-NH2 was performed. The in vitro inhibitory activities against HIV-1 of the [d4T]-NH-(CH2)n-NH-[imidazo[1,5-b]pyridazine] heterodimers were found to be comparable to that of d4T (stavudine) in HIV-infected cells. Moreover, the heterodimers were endowed with anti-HIV-2 activity and with anti-nevirapine-resistant HIV-1 activity. None of the heterodimers proved markedly cytotoxic to CEM-SS or MT-4 cells. There was not a clear trend toward antiviral potency on lengthening the methylene spacer in the [d4T]-NH-(CH2)n-NH-[imidazo[1,5-b]pyridazine] heterodimers.

Endogenous reverse transcriptase assays reveal synergy between combinations of the M184V and other drug resistance-conferring mutations in interactions with nucleoside analog triphosphates

Resistance of HIV-1 reverse transcriptase (RT) to nucleoside analogs (e.g. AZT, ddC and 3TC) is conferred by various amino acid substitutions or combinations thereof on the RT molecule. The M184V mutation, that confers high and low-level resistance to 3TC and ddC, respectively, can restore sensitivity to AZT when introduced into RT against a background of AZT-resistance. The K65R mutation, that confers low level resistance to both 3TC and ddC, can also restore sensitivity to AZT. This information is of potential utility in choosing combinations of anti-viral drugs for clinical use. To explore this subject further, we have used an endogenous RT reaction to study mutated viruses containing M184V alone or M184V combined with each of the K65R, E89G or both the M41L and T215Y substitutions. Endogenous assays possess the advantage of utilizing genomic RNA as template in a reaction mixture that includes each of tRNALys.3 and viral nucleocapsid protein, necessary for specific initiation of reverse transcription, as well as all other viral proteins that might impact on this process. We now show that viruses containing both M184V and K65R displayed synergistic resistance to 3TC triphosphate (3TCTP), while the same combination yielded the same level of resistance to ddC triphosphate (ddCTP) as that manifested by K65R alone. The combination of M184V and E89G displayed synergistic resistance against ddCTP but not 3TCTP, while viruses containing only E89G were highly resistant to 3TCTP and displayed low-level resistance to ddCTP. The results show that endogenous RT assays can reveal variable synergistic, antagonistic, or neutral effects in regard to drug sensitivity, depending on the presence of specific amino acid substitutions in RT itself.

A recombinant retroviral system for rapid in vivo analysis of human immunodeficiency virus type 1 susceptibility to reverse transcriptase inhibitors

We have developed a new recombinant retroviral system in which a library of infectious molecular clones of human immunodeficiency virus type 1 (HIV-1) is constructed with reverse transcriptase (RT) genes derived from viral RNA sequences in plasma. HIV-1 RT is amplified from plasma HIV-1 RNA by nested RT-PCR and cloned into a RT-defective HIV-1 proviral vector (xxLAI-np), generating 10(3) to 10(4) recombinant proviral clones from each reaction. The bulk cloning products or individual molecular clones are transfected into MT-2 cells to generate infectious virus. The resultant viruses are assayed for drug susceptibility in CD4+ cell lines to determine either the dominant phenotype of the recombinant virus mixture or the phenotypes of the individual viral clones. DNA sequencing of the cloned RT genes can identify mutations associated with phenotypic resistance of clonal mixtures or individual clones. This method can be used to rapidly detect the in vivo emergence of HIV-1 quasispecies resistant to RT inhibitors.

Use of viral resistance patterns to antiretroviral drugs in optimising selection of drug combinations and sequences

High rates of viral replication throughout HIV infection, and the frequency of mutation occurring during each replication cycle due to the inaccuracy of reverse transcriptase, drive the potential for drug-resistant viral variants to appear under the selective pressure of antiretroviral therapy. Loss of antiviral effect with a variety of antiretroviral agents has been reported to coincide with the appearance of viral mutants with reduced drug sensitivity. Additionally, the presence of both phenotypic and genotypic zidovudine resistance is associated with an increased risk of clinical disease progression and death, independent of a change of therapy to didanosine. The patterns of resistance to and cross-resistance between antiretroviral agents are increasingly well characterised, and represent an important consideration when deciding how to combine and/or sequence antiretrovirals to achieve optimal antiviral effects. Given the limited number of antiretrovirals currently available or in advanced development, it is important not to potentially limit future therapeutic options by using, early in the treatment sequence, therapies which may select for cross-resistant viral variants and hence potentially reduce the additional therapeutic response when treatment is changed to another member of that drug class.

Evolution of zidovudine resistance-associated genotypes in human immunodeficiency virus type 1-infected patients

Substantial differences have been described in the response of individual patients to zidovudine (ZDV) therapy, both in the clinical impact and in virus load. Genotypic changes associated with the appearance of drug resistance may also be different or occur at different rates. We have obtained the nucleotide sequence of the RT domain of individual HIV-1 genomes extracted from 10 plasma and peripheral blood mononuclear cell (PBMC) samples donated by two haemophiliac patients before, during, and after long-term ZDV therapy. Although the plasma virus load was similar throughout, the order and timing of appearance of resistance-associated substitutions differed in the two patients. In patient p74, K70R appeared after 4 months, T215Y at 5.5 months, and M41L at 13 months. In p87, K70R also appeared at 4 months, but T215Y and K219Q were not observed until 18 months and M41L not at all. Much greater sequence change overall occurred in p74. The evolution of the viral population in that patient was dominated by the unique appearance of T215Y and subsequently M41L, with all sequences from the last time point being descended by a single path from the pretreatment samples. However, in p87, several different lineages of RT sequences were found to persist throughout treatment. We propose that these differences in outcome may be determined by differences in genetic background at sites other than the five generally considered to be associated with ZDV sensitivity.

Synergistic inhibition of human immunodeficiency virus isolates (including 3'-azido-3'-deoxythymidine-resistant isolates) by foscarnet in combination with 2',3'-dideoxyinosine or 2',3'-dideoxycytidine

The combinations of foscarnet plus 2',3'-dideoxyinosine and foscarnet plus 2',3'-dideoxycytidine synergistically inhibit the replication of human immunodeficiency virus type 1 isolates, including two 3'-azido-3'-deoxythymidine-resistant isolates. The combination of 2',3'-dideoxyinosine plus 2',3'-dideoxycytidine showed additive inhibition of the majority of the human immunodeficiency virus isolates tested. All three combinations showed pronounced antagonistic cytotoxicity and thus were less toxic to the growth of peripheral blood mononuclear cells than the separate drugs.

Intravirion reverse transcripts in the peripheral blood plasma on human immunodeficiency virus type 1-infected individuals

Variable levels of viral DNA have been demonstrated within human immunodeficiency virus type 1 (HIV-1) virions purified from cell cultures. In the present studies, it is demonstrated that DNase-resistant viral DNA is associated with HIV-1 virions purified from the peripheral blood plasma of both symptomatic and asymptomatic HIV-1-infected individuals. The differences in viral DNA copy numbers, detected by quantitative PCR in various regions of the HIV-1 genome, indicated that the intravirion HIV-1 DNA is frequently, but perhaps not totally, the result of partial reverse transcription. These in vivo data suggest that it may be valuable to further investigate the impact of virion-associated viral DNA upon the efficiency of intra- and interhost HIV-1 transmission modes.

Susceptibilities of zidovudine-resistant variants of human immunodeficiency virus type 1 to inhibition by acyclic nucleoside phosphonates

The acyclic purine nucleoside phosphonates, a newly described class of broad-spectrum antiviral agents, effectively inhibit human immunodeficiency virus type 1 (HIV-1) replication in vitro and in animal AIDS models. 9-(2-Phosphonylmethoxyethyl)adenine (PMEA) is currently being evaluated in clinical trials in patients with AIDS. In this study, we investigated the efficacy of PMEA and a related analog, 9-(2-phosphonylmethoxypropyl)diaminopurine (PMPDAP), against HIV-1 isolates exhibiting various degrees of resistance to zidovudine (azidothymidine [AZT]). HIV isolates highly (approximately 50 to 200-fold) resistant to AZT were found to be about two- to eightfold less susceptible to PMEA. A comparable degree of cross-resistance to PMPDAP, a structurally related analog of PMEA, was also observed. However, the 50% effective dose values of PMEA or PMPDAP against a panel of HIV isolates showing intermediate levels (approximately 8 to 25-fold) of AZT resistance was indistinguishable from the 50% effective dose values of PMEA (0.7 to 1.7 versus 2 microM) or PMPDAP (0.4 to 1.4 versus 0.8 to 1 microM) against HIV isolates from patients who had not previously used AZT. In addition, we were unable to generate PMEA- (or PMPDAP)-resistant HIV-1 variants by > 30 serial passages of the virus in the presence of increasing concentrations of PMEA. Careful analysis of HIV-1 isolates from patients previously treated with AZT for cross-resistance to PMEA are needed to evaluate the significance of these observations.

Sensitivity of wild-type human immunodeficiency virus type 1 reverse transcriptase to dideoxynucleotides depends on template length; the sensitivity of drug-resistant mutants does not

Analysis of the three-dimensional structure of human immunodeficiency virus type 1 (HIV-1) reverse transcriptase (RT) complexed with double-stranded DNA indicates that while many nucleoside-resistance mutations are not at the putative dNTP binding site, several are in positions to interact with the template-primer. Wild-type HIV-1 RT and two nucleoside-resistant variants, Leu74-->Val and Glu89-->Gly, have been analyzed to determine the basis of resistance. The ability of the wild-type enzyme to incorporate, or reject, a 2',3'-dideoxynucleoside triphosphate (ddNTP) is strongly affected by interactions that take place between the enzyme and the extended template strand 3-6 nt beyond the polymerase active site. Inspection of a model of the enzyme with an extended template suggests that this interaction involves the fingers subdomain of the p66 subunit in the vicinity of Leu74. These data provide direct evidence that the fingers subdomain of the p66 subunit of HIV-1 RT interacts with the template strand. The wild-type enzyme is resistant to ddITP if the template extension is 3 nt or less and becomes sensitive only when the template extends more than 3 or 4 nt beyond the end of the primer strand. However, the mutant enzymes are resistant with both short and long template extensions. Taken together with the three-dimensional structure of HIV-1 RT in complex with double-stranded DNA, these data suggest that resistance to the dideoxynucleotide inhibitors results from a repositioning or change in the conformation of the template-primer that alters the ability of the enzyme to select or reject an incoming dNTP.

Nucleotide composition as a driving force in the evolution of retroviruses

All complete retrovirus sequences in the GenEMBL database were examined with the goal of assessing possible relationships between the nucleotide composition of retroviral genomes, the amino acid composition of retroviral proteins, and evolutionary strategies used by retroviruses. The results demonstrated that the genome of each viral lineage has a characteristic base composition and that the variations between groups are related to retroviral phylogeny. By analogy to microbial species, we suggest that the variations arise from group-specific patterns of directional mutations where the bias can be exerted on any of the four nucleotides. It is most likely that the mutational patterns are introduced during reverse transcription, and a direct participation of reverse transcriptase in the process is suspected. A straightforward strategy was used to analyze the compositional relationship between nucleotides and encoded amino acids. The procedure entailed calculations of amino acid frequencies from nucleotide content and the comparison of the calculated values to the observed amino acid frequencies in retroviruses. The results revealed an excellent correspondence between variation in genomic base composition and variation in amino acid composition of proteins with the compositional differences extending into all major coding regions of the viruses. Because of the magnitude and dispersion of these effects, and because of the nonconservative nature of many of the substitutions between groups with different genomic biases, we suggest that the variations in protein composition driven by biased nucleotide frequencies are an important factor in shaping the characteristic phenotypes of the different viral lineages. A clue to the nature of the evolutionary forces that are responsible for the generation of nucleotide biases was provided by the observation that viruses with radically different base frequencies most often inhabit the same cell type. This observation, along with analysis of amino acid and nucleotide replacement patterns between and within reverse transcriptase sequences from the various groups, permitted us to advance a model for the evolution of retroviruses. According to the model, speciation could initiate when daughter virions from a single progenitor vary in the direction of their mutational bias. These variations would exert a pleiotropic effect on the frequencies of nucleotides in all viral genes and consequently on the frequencies of amino acids in the encoded proteins. The variants with the most extreme compositional differences would have a selective advantage because their different precursor requirements would enable them to occupy different ecological niches within a single cell.(ABSTRACT TRUNCATED AT 400 WORDS)

Dideoxynucleoside resistance emerges with prolonged zidovudine monotherapy. The RV43 Study Group

Human immunodeficiency virus type 1 (HIV-1) isolates resistant to zidovudine (ZDV) have previously been demonstrated to exhibit in vitro cross-resistance to other similar dideoxynucleoside agents which contain a 3'-azido group. However, cross-resistance to didanosine (ddI) or dideoxycytidine (ddC) has been less well documented. ZDV, ddI, and ddC susceptibility data have been collected from clinical HIV-1 isolates obtained by five clinical centers and their respective retrovirology laboratories. All subjects were treated only with ZDV. Clinical HIV-1 isolates were isolated, amplified, and assayed for drug susceptibility in standardized cultures of phytohemagglutinin-stimulated donor peripheral blood mononuclear cells obtained from healthy seronegative donors. All five cohorts showed a correlation between decreased in vitro susceptibility to ZDV and decreased susceptibility to ddI and ddC. For each 10-fold decrease in ZDV susceptibility, an average corresponding decrease of 2.2-fold in ddI susceptibility was observed (129 isolates studied; P < 0.001, Fisher's test of combined significance). Similarly, susceptibility to ddC decreased 2.0-fold for each 10-fold decrease in ZDV susceptibility (82 isolates studied; P < 0.001, Fisher's test of combined significance). These data indicate that a correlation exists between HIV-1 susceptibilities to ZDV and ddI or ddC for clinical HIV-1 isolates.

Rapid phenotypic reversion of zidovudine-resistant feline immunodeficiency virus without loss of drug-resistant reverse transcriptase

We have selected and plaque purified zidovudine (3'-azido-3'-deoxythymidine [AZT])-resistant mutants from an infectious molecular clone of feline immunodeficiency virus (FIV). The patterns of cross-resistance and drug susceptibilities of these mutants were similar to those of the AZT-resistant FIV that we previously selected in vitro from a wild-type FIV population and to those of the most common AZT-resistant clinical isolates of human immunodeficiency virus type 1. Two AZT-resistant mutants of FIV, one selected from a normal population and one selected from the molecular clone, each reverted rapidly to an AZT-sensitive phenotype when passaged in the absence of drug. Sequence analysis of the reverse transcriptase (RT)-encoding region from the plaque-purified AZT-resistant FIV revealed a single base change at position 2939, resulting in a Glu-to-Lys substitution at amino acid 202 of the RT. Similar analyses of plaque-purified revertants showed that the phenotypic reversion was not the result of a genotypic reversion at this position and that no additional mutations existed within the RT-encoding region of the revertants. Moreover, RTs purified from the mutant and revertant were both resistant to the 5'-triphosphate of AZT. These results indicate the complexity of AZT resistance and suggest the presence of additional factors, outside the RT-encoding region, which may contribute to AZT resistance.

Antiviral drug resistance mutations in human immunodeficiency virus type 1 reverse transcriptase occur in specific RNA structural regions

A statistically significant correlation exists between the locations of drug resistance mutations (DRMs) observed for various reverse transcriptase inhibitors and features of the secondary structure predicted for the RNA coding for human immunodeficiency virus type 1 reverse transcriptase. The known DRMs map onto "unstable" bases, which are predominantly nonhelical regions (i.e., loops, bulges, and bends) of the predicted RNA secondary structure, whereas codons for the key conserved residues of polymerase sequence motifs map onto "stable" paired bases involved in helical regions. On the basis of these results, we hypothesize that the secondary structure of the RNA template (in this case, the reverse transcriptase gene itself) may be a previously unrecognized factor contributing to base misincorporation errors during reverse transcription and that, rather than being randomly distributed, mutations are more likely to occur in specific regions of the genome. The results suggest that these "mutation-prone" regions can be predicted by using a standard algorithm for RNA secondary structure.

Recombinant virus assay: a rapid, phenotypic assay for assessment of drug susceptibility of human immunodeficiency virus type 1 isolates

Antiviral drug susceptibility assays for clinical human immunodeficiency virus type 1 (HIV-1) isolates are required to monitor the development of drug resistance during clinical trials and antiretroviral drug therapy. First-generation phenotypic assays possess a number of drawbacks, not least the selection of unrepresentative virus populations during cocultivation. Here we describe a rapid phenotypic assay for the assessment of the susceptibility of clinical isolates to reverse transcriptase (RT) inhibitors. This procedure, called the recombinant virus assay, allows the generation of viable virus by homologous recombination of a PCR-derived pool of RT coding sequences into an RT-deleted, noninfectious proviral clone, pHIV delta BstEII. A nested PCR procedure has been optimized to allow the amplification of an RT pool from both uncultured and cocultured infected patient peripheral blood lymphocyte (PBL) DNA for subsequent use in the creation of recombinant viruses. Analysis of two patients during the course of zidovudine therapy showed that this approach produced viruses which accurately exhibited the same genotype and phenotype as that of the original infected PBL DNA. The recombinant virus assay can be performed in approximately 3 weeks without the use of donor PBLs and therefore represents a rapid, nonselective procedure for the assay of clinical isolates.

Development and significance of zidovudine resistance in children infected with human immunodeficiency virus

Decreasing susceptibility to zidovudine (ZDV) has been described in persons infected with human immunodeficiency virus (HIV) type 1 who are receiving ZDV therapy. However, the clinical significance of decreased ZDV susceptibility remains unclear. In this study, HIV isolates obtained from children with symptomatic HIV infection treated with ZDV were monitored for their susceptibility to the antiretroviral agent and correlated with disease progression. Using a peripheral blood mononuclear cell-based assay to measure ZDV susceptibility, we evaluated HIV isolates from 19 children (mean age, 6.8 years; range, 5 months to 12 years) during ZDV therapy for susceptibility to ZDV. Of the 19 children studied, 10 continued to have susceptible HIV strains during ZDV treatment, and 9 acquired resistant viruses. All eight isolates from children without previous exposure to ZDV were initially susceptible. After a median of 11 months of ZDV therapy, three (38%) of these eight children had acquired resistant HIV strains (defined as ZDV susceptibility > or = 10 mumol/L). Children with resistant strains had worse clinical outcomes than children whose viruses remained susceptible, as determined by a 50% decline in absolute CD4+ cell counts after 1 year of treatment, failure to thrive, or death. Children with resistant viruses who were given alternative antiretroviral therapy frequently responded to the new treatment with improved growth and stabilization of their HIV-related disease. These data suggest that, in HIV-infected children, ZDV-resistant HIV strains are associated with diminished drug efficacy and more rapid disease progression.

Standardized peripheral blood mononuclear cell culture assay for determination of drug susceptibilities of clinical human immunodeficiency virus type 1 isolates. The RV-43 Study Group, the AIDS Clinical Trials Group Virology Committee Resistance Working Group

A standardized antiviral drug susceptibility assay for clinical human immunodeficiency virus type 1 (HIV-1) isolates has been developed for use in clinical trials. The protocol is a two-step procedure that first involves cocultivation of patient infected peripheral blood mononuclear cells (PBMC) with seronegative phytohemagglutinin-stimulated donor PBMC to obtain an HIV-1 stock. The virus stock is titrated for viral infectivity (50% tissue culture infective dose) by use of serial fourfold virus dilutions in donor PBMC. A standardized inoculum of 1,000 50% tissue culture infective doses per 10(6) cells is used in the second step of the procedure to acutely infect seronegative donor PBMC in a 7-day microtiter plate assay with triplicate wells containing zidovudine (ZDV) concentrations ranging from 0 to 5.0 microM. The ZDV 50% inhibitory concentrations (IC50) for reference ZDV-susceptible and ZDV-resistant HIV-1 isolates ranged from 0.002 to 0.113 microM and from 0.15 to > 5.0 microM, respectively. Use of this consensus protocol reduced interlaboratory variability for ZDV IC50 determinations with reference HIV-1 isolates. Among eight laboratories, the coefficient of variation ranged from 0.85 to 1.25 with different PBMC protocols and was reduced to 0.39 to 0.98 with the standardized assay. Among the clinical HIV-1 isolates assayed by the standardized drug susceptibility assay, the median ZDV IC50 increased gradually with more ZDV therapy. This protocol provides an efficient and reproducible means to assess the in vitro susceptibility to antiretroviral agents of virtually all clinical HIV-1 isolates.

Preclinical evaluation of antiviral activity and toxicity of Abbott A77003, an inhibitor of the human immunodeficiency virus type 1 protease

A synthetic, symmetry-based inhibitor of the human immunodeficiency virus type 1 (HIV-1) protease, A77003, was evaluated for antiviral activity and cytotoxicity in vitro in human peripheral blood lymphocytes or cell lines H9, CEM, and U937. Toxicity and antiviral activity of the HIV-1 protease inhibitor were compared with those of the reverse transcriptase inhibitors zidovudine and 2',3'-dideoxy-2',3'-didehydrothymidine and human recombinant alpha and beta interferons. Production of infectious virus particles, cell-free p24 antigen, and cell-associated viral proteins was reduced 50% by the HIV-1 protease inhibitor at concentrations of 0.12 to 0.26 microM (50% effective concentration [EC50]) in acute infection and 0.2 to 1.7 microM (EC50) in persistent infection. Fluorescence-activated cell sorter analysis of U937 cells persistently infected with HIVIIIB using a monoclonal antibody to HIV also showed a reduction of cell-associated viral protein in A77003-treated cells. Furthermore, toxicity of A77003 assessed by 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide assay was not observed at greater than 100 times the EC50. A77003 was more effective in persistent HIV-1 infection than alpha and beta interferons (1,000 U/ml), while zidovudine and 2',3'-dideoxy-2',3'-didehydrothymidine were not active.

Generation of drug-resistant variants of human immunodeficiency virus type 1 by in vitro passage in increasing concentrations of 2',3'-dideoxycytidine and 2',3'-dideoxy-3'-thiacytidine

We selected in vitro human immunodeficiency virus type 1 variants that are resistant to each of 2',3'-dideoxycytidine (ddC) and the racemic mixture of 2',3'-dideoxy-3'-thiacytidine (BCH-189). The median effective concentrations of ddC and BCH-189 obtained for the resistant viruses ranged between 10 and 50 times above those for parental wild-type strains, and extensive cross-resistance was observed against 2',3'-dideoxyinosine (ddI) but not 3'-azido-3'-deoxythymidine (AZT). Two dimer compounds, in which either AZT and ddI or AZT and BCH-189 were linked through phosphodiester linkages, did not permit the emergence of variants resistant to BCH-189, ddI, or AZT but were ineffective at inhibiting the replication of AZT-resistant viruses.

Resistance of HIV-1 to AZT might also involve the cellular expression of multidrug resistance P-glycoprotein

Resistance of tumor cells to the antigrowth activity of several cytotoxic compounds has been associated with the expression of the so-called multidrug resistance protein or P-glycoprotein. This article addresses the question whether the expression of such protein could also affect the sensitivity of HIV to AZT. Our data indicate that this possibility does exist. In fact, multidrug-resistant CEM VBL100 cells, which express high levels of P-glycoprotein, are less sensitive to both the antiproliferative activity and the antiviral action of AZT. Additionally, our data suggest that this phenomenon is specifically mediated by P-glycoprotein since trifluoroperazine, which is known to circumvent multidrug resistance due to the action on P-glycoprotein, increases the intracellular accumulation of AZT and affects the sensitivity of HIV to AZT. Although the biological and clinical significance of these observations has still to be established, this study suggests that cellular factors, other than virus itself, should be taken into account to address the phenomenon of drug resistance of HIV.

Didanosine and zidovudine resistance patterns in clinical isolates of human immunodeficiency virus type 1 as determined by a replication endpoint concentration assay

Reports of in vitro resistance of human immunodeficiency virus type 1 (HIV-1) to zidovudine (AZT) have raised concerns about the development of resistance to other dideoxynucleosides in clinical use. To address this, we have developed a screening assay which supports the growth of clinical isolates and have applied this to a series of paired isolates from patients entered into a phase I trial of didanosine (DDI). Thirteen patients (10 with AIDS, 3 with AIDS-related complex) who had been exposed to AZT for a mean of 6.5 months (range, 1 to 13 months) were treated with DDI at 750 mg/day. Paired isolates were obtained pretherapy and after a mean of 58 weeks (range, 21 to 90) of DDI therapy by coculture of peripheral blood mononuclear leukocytes (PBLs) with phytohemagglutinin-stimulated donor PBLs. Isolates were passaged only one additional time in PBLs and then tested in parallel in a microtiter assay with phytohemagglutinin-stimulated donor PBLs as targets. PBLs were infected with 10(5) 50% tissue culture infectious doses per 10(7) cells and exposed to DDI (1 to 50 microM) or AZT (0.01 to 100 microM), and supernatants were assayed for the HIV p24 antigen at 7 days postinfection. Control AZT-susceptible and resistant isolates were included. The median pre- and posttherapy DDI susceptibilities of the 13 pairs of isolates were 10.0 microM (range, 1 to 25 microM) and 17.5 microM (range, 2.5 to 50 microM), respectively (P = 0.036; Wilcoxon signed-rank test). These studies thus indicated that (i) the susceptibility to DDI tends to mildly decrease with drug exposure; (ii) the susceptibility to AZT improves with time off AZT; (iii) baseline susceptibilities to DDI have a wide range, and the CD4 response may correlate with the initial susceptibility; and (iv) a PBL-based microtiter assay is useful for screening clinical isolated for dideoxynucleoside susceptibility profiles.

Susceptibility testing by polymerase chain reaction DNA quantitation: a method to measure drug resistance of human immunodeficiency virus type 1 isolates

Polymerase chain reaction (PCR) DNA quantitation (PDQ) susceptibility testing rapidly and directly measures nucleoside sensitivity of human immunodeficiency virus type 1 (HIV-1) isolates. PCR is used to quantitate the amount of HIV-1 DNA synthesized after in vitro infection of peripheral blood mononuclear cells. The relative amounts of HIV-1 DNA in cell lysates from cultures maintained at different drug concentrations reflect drug inhibition of virus replication. The results of PDQ susceptibility testing of 2- or 3-day cultures are supported by assays measuring HIV-1 p24 antigen production in supernatants of 7- or 10-day cultures. DNA sequence analyses to identify mutations in the reverse transcriptase gene that cause resistance to 3'-azido-3'-deoxythymidine also support the PDQ results. With the PDQ method, both infectivity titration and susceptibility testing can be performed on supernatants from primary cultures of peripheral blood mononuclear cells. PDQ susceptibility testing should facilitate epidemiologic studies of the clinical significance of drug-resistant HIV-1 isolates.

Fifth mutation in human immunodeficiency virus type 1 reverse transcriptase contributes to the development of high-level resistance to zidovudine

It is recognized that high-level resistance to 3'-azido-3'-deoxythymidine (AZT, zidovudine, or Retrovir) is conferred by the presence of four mutations in the human immunodeficiency virus (HIV) reverse transcriptase [RT; deoxynucleoside-triphosphate:DNA deoxynucleotidyltransferase (RNA-directed), EC 2.7.7.49] coding sequence. However, a number of clinical isolates have been observed that exhibit high-level resistance but contain only three of the four identified mutations (Asn-67, Arg-70, and Tyr-215). Construction of a molecular clone with this genotype gave rise to only a partially resistant virus, raising the possibility that an additional mutation existed in some clinical isolates. Using an HIV marker rescue system, we have mapped and identified a fifth mutation conferring resistance to zidovudine, namely, methionine to leucine at codon 41 of HIV RT. An infectious molecular clone containing this mutation together with three previously identified mutations in the RT coding sequence yielded highly resistant HIV after transfection of T cells. Direct detection of the fifth mutation in DNA samples from cocultured peripheral blood lymphocytes by the PCR revealed that it occurred relatively early in the development of zidovudine resistance. However, this mutation was only detected after the appearance of the codon 215 change in the RT coding sequence. Identification of this mutation in addition to the other known mutations conferring resistance enables rapid and direct correlation between an RT genotype and sensitivity of the virus.

Antiviral therapy: current concepts and practices

Drugs capable of inhibiting viruses in vitro were described in the 1950s, but real progress was not made until the 1970s, when agents capable of inhibiting virus-specific enzymes were first identified. The last decade has seen rapid progress in both our understanding of antiviral therapy and the number of antiviral agents on the market. Amantadine and ribavirin are available for treatment of viral respiratory infections. Vidarabine, acyclovir, ganciclovir, and foscarnet are used for systemic treatment of herpesvirus infections, while ophthalmic preparations of idoxuridine, trifluorothymidine, and vidarabine are available for herpes keratitis. For treatment of human immunodeficiency virus infections, zidovudine and didanosine are used. Immunomodulators, such as interferons and colony-stimulating factors, and immunoglobulins are being used increasingly for viral illnesses. While resistance to antiviral drugs has been seen, especially among AIDS patients, it has not become widespread and is being intensely studied. Increasingly, combinations of agents are being used: to achieve synergistic inhibition of viruses, to delay or prevent resistance, and to decrease dosages of toxic drugs. New approaches, such as liposomes carrying antiviral drugs and computer-aided drug design, are exciting and promising prospects for the future.

In vitro selection of variants of human immunodeficiency virus type 1 resistant to 3'-azido-3'-deoxythymidine and 2',3'-dideoxyinosine

Drug-resistant variants of human immunodeficiency virus type 1 (HIV-1) have been isolated by in vitro selection. MT-4 cells were infected with either a laboratory strain (HIV-IIIB) or a clinical isolate (no. 187) of HIV-1 and maintained in medium containing subeffective concentrations of the drugs 3'-azido-3'-deoxythymidine (AZT) and 2',3'-dideoxyinosine (ddI). By gradually increasing the drug concentration in the culture medium during propagation of the virus on fresh MT-4 cells, we were able to isolate variants of HIV-IIIB and clinical isolate 187 which showed up to 100-fold increases in resistance to the drugs. The drug resistance phenotypes remained stable after propagation of the variants in the absence of drug pressure for over 2 months. However, variants resistant to one drug showed little or no cross-resistance to the other, suggesting that the genetic bases for resistance to the compounds differed. Genotypic analysis of these nucleoside-resistant variants by polymerase chain reaction (PCR) with primer pairs previously shown to correspond to mutations responsible for resistance to AZT was also carried out. A heterogeneity of genotypes was observed, with known mutations at pol codons 70 and 215 occurring in most of the AZT-resistant variants generated from either HIV-IIIB or clinical strain 187. However, mutations in codons 67 and 219 were less frequently detected, and none of these changes were observed in each of four variants resistant to ddI. Cloning and sequencing studies of the reverse transcriptase coding region of two of the isolates were also performed and confirmed the PCR data that had been obtained. In addition to previously described mutation sites responsible for resistance to AZT, an HIV-IIIB-resistant variant was shown to be mutated at positions 108 (Val----Ala) and 135 (Ile----Thr), while a resistant variant of strain 187 was mutated at positions 50 (Ile----Val) and 135 (Ile----Val).