Detection of minor populations of drug-resistant HIV-1 in acute seroconverters

Objective: The transmission of drug-resistant HIV-1 is a major health concern. To date, most clinical studies have relied on sequencing techniques for genotypic analyses which do not allow quantification of minority viral populations below 25%. As minor populations of drug-resistant HIV-1 could impact the efficiency of antiretroviral therapy, this study was performed to determine the prevalence of minor populations of drug-resistant HIV-1 in acute seroconverters.

Design and methods: Forty-nine acute seroconverters from two clinical centers in Germany were included in the study. Individuals were identified between June 1999 and March 2003, and none had received antiretroviral therapy prior to sampling. Minor populations of drug-resistant variants were detected by quantitative real-time polymerase chain reaction using allele-discriminating oligonucleotides for three key resistance mutations: L90M (protease), K103N and M184V (reverse transcriptase). The approximate discriminative power was between 0.01 and 0.2%.

Results: Drug-resistant variants were detected in 10 of 49 patients (20.4%). The L90M mutation was found in one of 49 (2%), the K103N mutation in five of 49 (10.2%) and the M184V mutation in six of 49 (12.2%) patients, respectively. In five of the 10 individuals with detectable drug-resistant virus (50%), the detected population represented a minor viral quasi-species (< 25% of viruses) and was not detected by direct sequencing.

Conclusions: The prevalence of minor populations of drug-resistant HIV-1 in acute seroconverters can be frequently detected and may impact the success of antiretroviral therapy.

Introduction

Antiretroviral therapy (ART) has significantly reduced morbidity and mortality in HIV-infected individuals ^[1,2]. However, in a substantial fraction of treated patients, control of virus replication is incomplete or temporary, and resistance to antiretroviral drugs develops. The presence of drug-resistant HIV-1 is associated with impaired response to subsequent ART. After the first report of primary infection with drug-resistant HIV-1 ^[3], transmission of drug-resistant viral variants has been frequently observed (reviewed in ^[4]). In addition to the development of resistance during ART, transmission of drug-resistant HIV-1 has become a major health concern. Although a large number of studies aimed at quantifying HIV-1 drug-resistant variants have been performed, these studies are usually small in size and vary considerably in methodology and interpretation. Large, long-term studies are currently ongoing to determine whether the prevalence of drug-resistant viral variants in acute seroconverters is actually increasing (reviewed in ^[4]).

In general, the presence of HIV-1 drug resistance is evaluated by sequencing techniques that demonstrate the presence of typical mutations in the viral genome known to be associated with reduced sensitivity to an individual antiretroviral compound. These genotypic analyses are unable to detect and quantify minorities of drug-resistant viral quasi-species below 25% ^[5,6]. The clinical importance of minor populations of drug-resistant variants is still not clear. Certain clinical observations suggest that minor viral variants can emerge as the major viral population ^[7,8]. By using a methodology that allows the quantification of minor viral populations, we evaluated the prevalence of drug-resistant HIV-1, with emphasis on identification of drug-resistant variants as minor viral species in acutely HIV-infected patients.

Methods

Study design and patients

Patients referred to two clinical centers in Germany (Hamburg and Berlin) were included in this study. Acute seroconversion was confirmed as follows: At the first visit, all patients showed clinical signs and symptoms of an acute retroviral syndrome associated with positive plasma HIV-1 RNA levels. Most patients showed an incomplete pattern on western blot which changed to a complete pattern on subsequent visits. Patients with a complete pattern on western blot had a negative HIV-1 enzyme-linked immunosorbent assay test result within 3 months prior to presentation. Patients were identified between June 1999 and March 2003 and none had received ART before sampling. Study patients from Hamburg were participants in an observational cohort (HIV-AKUT study) that had been approved by the ethics committee of the Ärztekammer Hamburg. Written informed consent was obtained from each patient prior to inclusion.

HIV-1 quantification

Blood samples were obtained on the first visit. Plasma HIV-1 RNA was quantified using the Roche Amplicor PCR assay (Roche, Grenzach, Germany) according to the manufacturer's instructions. If HIV RNA was above the upper limit of quantification, the samples were re-analyzed following 1: 10 dilution with HIV-negative human plasma.

Construction of DNA standards for quantitative real-time polymerase chain reaction for differential amplification of L90M, K103N and M184V mutations

Primers for polymerase chain reaction (PCR) and reverse transcriptase (RT)-PCR were designed based on published sequences within the pol region of HIV-1_HXB2. Oligonucleotides were synthesized by biomers.net (Ulm, Germany). The construction of DNA standards for quantitative real-time PCR for differential amplification of L90M and M184V mutations have been described previously ^[8]. Part of the pol gene of HIV-1_HXB2 was cloned into pGEM-T (Promega, Madison, Wisconsin, USA) according to the manufacturer's instructions. The K103N mutation (AAA to AAC transversion in codon 103 of the HIV-1 RT gene) was introduced into the pGEM-T HIV-1_HXB2-L90M+M184V plasmid by site-directed mutagenesis using a QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, California, USA) and the oligonucleotides tgK103N 5′-CCCGCAGGGTTAAAAAAGAACAAATCAGTAACAGTACTGGATGTGG-3′ (nt 2839-2883) and tgK103Nrc 5′-CCACATCCAGTACTGTTACTGATTTGTTCTTTTTTAACCCTGCGG-3′ (nt 2839-2883) according to the manufacturer's instructions. The presence of each mutation in the pGEM-T HIV-1_HXB2-L90M+K103N+M184V template DNA was confirmed by sequencing. DNA standards for quantification were prepared by PCR from plasmid DNA constructed by in-vitro mutagenesis as previously described ^[8].

Selective quantification of wild-type, and L90M, K103N and M184V drug-resistant HIV-1 variants

Plasma samples collected from whole blood in ethylenediaminetetraacetic acidA anticoagulant were centrifuged at 2000 × g for 5 min to pellet cell debris. Particle-associated HIV-1 RNA was purified from 1 ml cell-free plasma using a QIAamp UltraSens Virus Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Viral RNA was used for cDNA synthesis immediately after extraction.

Reverse transcription with viral RNA from plasma samples was performed as previously described ^[8]. PCR reactions consisted of 20 μl of cDNA, 1 × PCR buffer II (Applied Biosystems, Foster City, California, USA), 1.5 mmol/l MgCl₂, 0.2 mmol/l dNTPs (Eppendorf, Hamburg, Germany), 0.4 μmol/l of each upstream primer L90 EP1 5′-GAAGCTCTATTAGATACAGG-3′ (nt 2313-2332), K103 EP1 5′-CCCGCAGGGTTAAAAAAG-3′ (nt 2838-2855), and M184 EP1 5′-AATCCAGACATAGTTATCTATC-3′ (nt 3072-3093), 0.4 μmol/l of each downstream primer L90 EP2 5′-TTTAAAGTGCAACCAATCTGAG-3′ (nt 2524-2545), K103 EP2 5′-CCTGTGGAAGCACATTG-3′ (nt 2987-3003), and M184 EP2 5′-TTTTTTGTCTGGTGTGGTAAATC-3′ (nt 3187-3209) and Taq DNA polymerase in a final volume of 50 μl. Amplification was carried out for 20 cycles (94°C for 15 s, 43°C for 1 min, 72°C for 30 s) in a Mastercycler gradient (Eppendorf). The PCR products were purified using a NucleoSpin-Extract kit (Macherey-Nagel, Düren, Germany) and eluted with 200 μl 10 mmol/l Tris-HCl, pH 8.5 following the manufacturer's protocol.

Nested real-time amplification of HIV-1 DNA was performed separately for each region of interest (either wild-type or mutant) using 10 μl of purified first round PCR product, 1 × ROX-PCR buffer, 3.5 mmol/l MgCl₂, 0.5 mmol/l dNTPs, 0.1 × SYBR green (Molecular Probes; Invitrogen, Karlsruhe, Germany), 0.4 μmol/l of each primer pol 3002 5′-CTGTGGAAGCACATTGTACTG-3′ (nt 2982-3002) and IN K103 5′-CCGCAGGGTTAAAAAAGAIA-3′ (nt 2839-2858) or IN K103N 5′-CGCAGGGTTAAAAAAGAIC-3′ (nt 2840-2858), and Taq DNA polymerase in a final volume of 50 μl. The oligonucleotides used for the L90/M and M184/V assays are described elsewhere ^[8]. Fifty cycles of amplification (94°C for 15 s, 60°C for 30 s, and 72°C for 30 s) were performed in an Applied Biosystems 7700 Prism spectrofluorometric thermal cycler (Applied Biosystems).

DNA standards were tested in duplicate for each experiment. The standards were prepared by serial dilution from 10⁷ to 10² copies per reaction. Viral first round PCR products were also tested in duplicate. Copy numbers were calculated by interpolation of the experimentally determined threshold cycle for the test specimen onto a control standard regression curve ^[9]. The ratio of wild-type and mutant sequences was calculated based on copy numbers for each population. Nested real-time PCR assays have a detection limit of 10 HIV-1 DNA copies per reaction with a linear dynamic range of > 6 logs. For each sample an individual cut-off was estimated based on the viral load, because low viral loads diminish the discriminatory ability of each assay ^[8].

Genotypic analysis by sequencing

For genotype resistance testing the relevant part of the pol gene was amplified from each patient's plasma. Genotypic analysis was performed by direct sequencing of the amplicon using the ViroSeqTM HIV-1 Sequencing System (Abbott Diagnostics, Wiesbaden, Germany).

Statistical analysis

To investigate the significance of changes in the incidence of transmission of drug-resistant HIV-1 within the time period of this study, statistical analysis was performed using the one-tailed Fisher's exact probability test. P-values < 0.05 were regarded as significant.

Results

Discriminatory ability of quantitative real-time PCR for differential amplification of HIV-1 variants

The quantitative real-time PCR assays for differential amplification used in this study were based on the amplification refractory mutation system (ARMS) ^[10]. Evaluation of the discriminatory abilities and validation of these assays for detection of L90M and M184V mutations have been previously described in detail ^[8]. Quantitative real-time PCR for differential amplification of the K103N mutation was designed and tested following the same procedures (data not shown). Taken together, the estimated discriminatory abilities for the different wild-type and mutant sequences were as follows: 0.01% for detection of L90 and K103 wild-type as minority populations, 0.01% for L90M and K103N, 0.1% for M184 and 0.2% for M184V, respectively.

Quantification of minor populations of drug-resistant HIV-1 in acute seroconverters

Plasma samples from 49 acute seroconverters were analyzed by quantitative real-time PCR for differential amplification of L90M, K103N and M184V, respectively. All patients were acutely infected with HIV-1 as confirmed either by HIV-1 RNA and/or western blot test results and clinical criteria. Plasma samples were obtained at first presentation, prior to administration of any antiretroviral therapy. All except one patient were subsequently treated with ART. Viral load ranged from 4.9 × 10² to 2.2 × 10⁷ HIV-1 RNA copies/ml plasma (median 4.4 × 10⁵ HIV-1 RNA copies/ml plasma) (Table 1).

Table 1

All 49 patients were tested for the presence of L90M, K103N and M184V mutations by quantitative real-time PCR for differential amplification. One or two drug-resistant mutations were found in nine of 49 (18.4%) patients. In one additional patient, as described below, one mutation was only detected by direct sequencing. The M184V mutation was detected in six of 49 (12.2%) at frequencies ranging from 1.6 to 38.1% in five patients who carried this mutation as a minor viral species. In only one patient the majority of viruses contained the M184V mutation (98.5 ± 0.1%). Genotypic analysis by conventional sequencing methods confirmed this result and also identified additional mutations associated with drug resistance to zidovudine in this patient (M41L and T215F, patient no. 3, data not shown). In four patients, viral populations carrying the K103N mutation were detected with prevalence from 9.78 to 99.5%. This mutation was present as a minor viral population in two patients (9.78 and 12.85%, respectively). In the remaining two patients, the K103N mutation was detected as the major viral quasi-species. The L90M mutation was found as a minor viral quasi-species in one patient (0.48 ± 0.05%, Table 1, patient no. 35).

To confirm our results, sequencing was performed on viruses from all nine patients with evidence of drug-resistant mutations, and in five additional patients in which the L90M, K103N and M184V mutations were not detected. With the exception of patient no. 42, the presence of the mutations K103N and M184V mutations was confirmed by sequencing when these mutations were carried by the majority of viral species (Table 1, patients no. 3, 35 and 47). Conversely, genotypic analysis by sequencing was unable to detect mutations present in a minority of viruses with the exception of patient no. 26 (Table 1). In this patient, sequencing revealed the presence of the K103N mutation, which represents a minor viral population (9.78 ± 0.1%) using quantitative real-time PCR for differential amplification. Sequence analysis in this patient revealed a transversion in codon 103 from AAA to AAT (amino acid change from K to N). The allele-discriminating oligonucleotides used in the quantitative real-time PCR assay were designed to distinguish between the wild-type codon AAA and the mutant codon AAC and were therefore unable to detect the AAA to AAT mutation. However, this means that a minor population in this patient is carrying the K103N mutation conferred by the codon AAC. In five patients for whom differential amplification by quantitative real-time PCR did not reveal any evidence of L90M, K103N or M184V mutations, these mutations were also not detected by direct sequencing, with the exception of patient no. 42. Furthermore in this patient, sequence analysis revealed a transversion in codon 103 from AAA to AAT (amino acid change from K to N) which was undetectable by the quantitative real-time PCR assay used for differential amplification.

Taken together, in 10 of 49 patients (20.4%) one or two tested mutations (L90M, K103N and M184V) were clearly detected as either minor or major viral populations (Table 1). In all other patients, no evidence was found for these mutations within the range of detectable minor HIV-1 variants.

Discussion

Transmission of drug-resistant HIV-1 is a major health concern and currently large clinical studies are ongoing to determine the prevalence of drug-resistant viral variants in acute seroconverters (reviewed in ^[4]). In these studies, drug-resistant HIV-1 is detected by direct sequencing. A prevalence of 25% is regarded as the limit of detection of genotypic analysis by sequencing ^[5,6]. This limitation could lead to an underestimation of transmission of drug-resistant viral variants. Here, we investigated the prevalence of minor populations of drug-resistant HIV-1 variants in 49 acutely infected patients identified between June 1999 and March 2003. In this study cohort, minor populations of drug-resistant HIV-1 variants were detected by quantitative real-time PCR for differential amplification of three key resistance mutations: L90M (protease), K103N and M184V (reverse transcriptase).

Overall, drug-resistant HIV-1 variants were detected in 10 of 49 patients (20.4%). We did not observe a significant increase in the prevalence of drug-resistant HIV-1 over the 4-year period of observation for this study. Of the three mutations tested, M184V and K103N were found in six of 49 (12.2%) and five of 49 (10.2%) patients, respectively. In contrast, the L90M mutation was found in only one of 49 patients (2%). The M184V mutation confers high-level resistance to the commonly used antiretroviral drug lamivudine. Thus, it is not unexpected that the rapidly emerging single point mutation M184V is more frequently detected compared with the L90M mutation. Our finding of a high prevalence of the K103N mutation was surprising. It is known that resistance to non-nucleoside reverse transcriptase inhibitors (NNRTIs) can be conferred by the K103N mutation alone, however, NNRTI-resistance is often accompanied by additional mutations, and K103N is not necessarily the first mutation to arise ^[11]. The high number of patients carrying the K103N mutation may therefore be a random effect due to the limited size of the present study.

In five of six patients carrying the M184V mutation this population represented a minority of viruses whereas in only one of five patients harboring the K103N mutation these viruses represented the minority. This may be due to differences in replicative capacity among drug-resistant viruses. It has been shown that the M184V mutation leads to a reduction of replicative capacity ^[12] whereas viral strains carrying the K103N mutation do not show an impaired replicative capacity compared with wild-type viruses ^[13]. Thus, wild-type variants quickly expand to represent the majority of the viral population when the drug-resistant variants are less replication competent. Of note, in those patients with a major population of drug-resistant virus the corresponding wild-type variant was also always detected as a minor population. Assuming wild-type viruses have an advantage in untreated patients, a change in the viral population must be expected. It has often been reported that drug-resistant HIV-1 persists after primary infection ^[14-17]. Thus far, only one study has shown a reversion to wild-type virus after transmission of a multi drug-resistant virus ^[18]. Therefore, it is conceivable that certain drug-resistant variants may have a replicative advantage over wild-type viruses.

We cannot exclude the possibility that minor viral populations of drug-resistant viruses are also present due to natural occurring diversity; namely, that these variants are spontaneously generated by random mutagenesis after primary infection. However, taking into account the reduced viral fitness described for viruses harbouring the M184V mutation ^[12], this seems unlikely, especially in patients with a prevalence of > 13% of viruses carrying the M184V mutation. It is more likely that patients infected by transmission of predominantly drug-resistant HIV-1 may experience a decrease in the prevalence of resistant viral variants in the absence of any selection pressure from antiretroviral drugs. This may be due to a presumed compromise in replicative capacity among drug-resistant viruses, such that wild-type variants quickly expand to represent the majority of the viral population.

Drug-resistant variants representing at least 25% of the viral population were found in four of the nine patients with detectable drug-resistant mutations (except patient no. 42). These patients would be expected to screen positive for drug-resistant variants using conventional sequencing ^[5,6] which was indeed the case for three of the four patients. In the fourth patient, approximately 38% of the viral population carried the M184V mutation, which was not detected by direct sequencing (Table 1, patient no. 41). Assuming transmission of wild-type viruses occurred in all patients in which drug-resistant minorities were not detected using quantitative real-time PCR for differential amplification (and with no additional sequencing information), this equals a transmission rate of drug-resistant viruses of 12.2% (six of 49 patients; the four patients mentioned above and the two additional patients carrying the K103N mutation as majority as revealed by sequencing). This is in close approximation of the currently estimated rate of transmission of drug-resistant viruses in studies using conventional sequencing for genotyping ^[4,19-21]. Of the nine cases in which resistant variants were detected by quantitative real-time PCR for differential amplification, only four were detected using conventional sequencing. No minor variants were detected by direct sequencing at a frequency of less than 38%. Thus, quantitative real-time PCR for differential amplification was slightly superior in screening for transmitted resistance and revealed a higher prevalence of drug-resistant viruses (20.4%), although only three key resistance mutations have been tested. Therefore, based on our findings, the transmission of drug-resistant HIV-1 variants is likely to be underestimated when determined solely by conventional sequencing methods.

Thus far, the impact of minor populations of drug-resistant HIV-1 on the response to antiretroviral therapy remains to be determined. However, the findings of several clinical studies suggest that minorities of drug-resistant HIV-1 could influence the clinical outcome of antiretroviral therapy. Minor drug-resistant variants have been detected especially in the early phase of therapy failure whereas these mutations were not seen by direct sequencing ^[7,22-24]. One study shows very nicely the kinetics of the emergence of drug-resistant viruses from minor to major viral populations in patients failing antiretroviral therapy ^[7]. It could be clinically beneficial to evaluate the prevalence of minor drug-resistant HIV-1 also in drug-naive patients such as acute seroconverters.

In summary, we report here the prevalence of drug-resistant viruses measured in acute seroconverters using a methodology to discriminate and quantify wild-type and drug-resistant variants, and to detect minor viral populations to levels below 1%. In a substantial proportion of patients with detectable drug-resistant viruses, this population represented a minority that was undetected by conventional sequencing techniques. Thus, the prevalence of drug-resistant HIV-1 as detected by sequencing seems likely to yield an underestimation of the transmission rate of these variant viruses. In the future, it may be advantageous to measure minor populations of drug-resistant viruses in acute seroconverters, even though the impact of these findings on the response to antiretroviral therapy remains to be determined.

Acknowledgements

We are grateful to all the patients who participated in this study and for their contribution towards a better understanding of transmission of drug-resistant HIV-1. We thank Bernhard Fleckenstein for his constant support. Moreover, we thank Ingrid Stahmer, Janine Mohn and Kristina Lenz for their intellectual input and their technical support, and Ruth I. Connor for critical reading of the manuscript and helpful suggestions.

Sponsorship: This study was financed by Deutsche Forschungsgemeinschaft (Sonderforschungsbereich 466), by the Wilhem Sander-Stiftung (grant 2002.079.1), by the German HIV Competence Network HIV/AIDS (funded by the Federal Ministry of Research and Technology, grant #01 KI 0211, subproject C6), and supported in part by unrestricted grants of 'big spender' and GlaxoSmithKline (GSK), Germany.