Medical Device Link.

Originally Published IVD Technology April 2004

Molecular Diagnostics

The Viroseq HIV-1 genotyping system, as used with the ABI Prism 3100 genetic analyzer, is capable of sequencing up to 25 patient samples simultaneously.

Human immunodeficiency virus type I (HIV-1), a member of the Retroviridae family, has a single-stranded RNA genome. The replication cycle of HIV-1 requires integration of the viral genome into the host genomic DNA. This integration is preceded by reverse transcription, the process of converting viral RNA into single-stranded DNA.¹ 

Due to the lack of DNA proofreading, HIV-1 reverse transcriptase is a highly error-prone enzyme introducing one single-nucleotide substitution about every 10,000 nucleotides. Given the high rate of HIV-1 replication, estimated at 1–10 billion infective virus particles, or virions, each day, all possible single-point mutations occur 104–105 times per day in an untreated HIV-infected individual. Subpopulations of viral variants (quasispecies) are created that continue to evolve, resulting in a high level of HIV genetic diversity.

HIV-1 strains are divided into three phylogenetically distinct groups designated M, N, and O. Group M strains are further subdivided into nine phylogenetically pure subtypes, which are identified by the letters A–D, F–H, J, and K.²  The intermixture of HIV-1 variants that circulate together within a geographical region provides an opportunity for recombination of virus strains within dually infected individuals. Fifteen circulating recombinant forms (CRFs) and numerous unique mosaic strains of HIV-1 have been identified.³  Although subtype B viruses predominate in North America and Western Europe, they represent only 12% of HIV-1 infections worldwide.⁴ As the HIV-1 epidemic continues to evolve, increasing numbers of non–subtype B infections are being identified in Western Europe and the United States.

Selective drug pressure can induce further sequence diversification in patients undergoing antiretroviral therapy. In the presence of a drug, viral strains carrying mutations conferring drug resistance have a replicative advantage over wild-type viruses. Quasispecies harboring resistance-associated mutations increase in absolute number and in proportion relative to wild-type strains. The effectiveness of drug treatment is monitored by measuring plasma HIV-1 RNA levels—this is viral load testing—and by detecting drug-resistant mutations in the viral genome, which is called resistance testing.

A sample-preparation procedure that could serve for both viral load determination and drug resistance monitoring would be highly desirable to increase laboratory efficiency. This article suggests that such a procedure exists. The following discussion presents technology available for performing analysis of patient HIV status and genotypic determination of drug resistance. This provides a context for the subsequent presentation of a study showing that assays for viral load determination and resistance to antiretroviral drug therapy can be performed using a single sample preparation.

HIV-1 RNA Quantification

Several different technologies have been developed for quantifying HIV-1 RNA in plasma. They are reverse transcription–polymerase chain reaction (RT-PCR), isothermal nucleic acid sequence–based amplification (NASBA), and signal amplification using nucleic acid probes.

Examples of RT-PCR–based tests are the Amplicor HIV-1 Monitor version 1.5 (Monitor v1.5) manufactured by Roche Molecular Systems (Branchburg, NJ) and the LCx HIV RNA quantitative assay (not available in the United States) from Abbott Laboratories (Abbott Park, IL), a focus of this article. These assays have several processes in common: sample preparation, that is, extraction of HIV RNA from plasma; competitive RT-PCR amplification using HIV-1-specific oligonucleotide primers; hybridization of oligonucleotide probes; and detection of probe/ target hybrids. Monitor v1.5 targets a highly conserved region of gag p24.^5,6 The LCx assay targets a highly conserved region of pol integrase and is read out on an automated LCx analyzer.⁷

The NucliSens HIV-1 QT test from bioMérieux (Durham, NC) is based on an alternative target amplification technology. This assay targets the gag p24 region of HIV-1 RNA extracted from plasma and utilizes NASBA, an isothermal amplification procedure.⁸

The final technology category is represented by the Versant HIV-1 RNA 3.0 assay developed by Bayer Diagnostics (Tarrytown, NY), which directly quantitates HIV-1 RNA from signal amplification rather than using target amplification technology. This assay is based on a set of oligonucleotide probes complementary to conserved regions of the HIV-1 pol gene.⁹

All of these technologies rely on primer and/or probe hybridization to HIV-1-specific target sequences. Thus, the genetic diversity of HIV-1—that is, the natural polymorphisms within the primer or probe sites—has the potential to reduce the efficiency of hybridization and to compromise the accuracy of quantification.⁶ Implementation of alternative primers and probes and modified amplification and hybridization conditions have led to significant improvements in currently available tests.^6,10 Nevertheless, the genetic heterogeneity of HIV-1 continues to challenge viral load assay performance. Of all commercial assays, only the LCx assay has been shown to reliably quantify both group M strains and genetically diverse group O strains.¹⁰

HIV Drug Resistance Testing

HIV drug resistance testing is a valuable tool for implementing and reevaluating antiretroviral treatment regimens.¹¹ Resistance assays are recommended for patients with acute HIV infection prior to the initiation of drug therapy. They are also recommended for those failing antiretroviral therapy, for those with suboptimal viral suppression for whom new treatment regimens are indicated, and for drug-naive patients with suspected drug-resistant infections.¹¹ There are two main approaches to determining whether drug-resistant mutations are present in patient samples: genotyping and phenotyping.

Genotyping assays are more commonly employed in the clinical setting than phenotyping tests because of their lower cost, faster turnaround, and ability to more readily detect minor quasispecies carrying mutations associated with drug resistance. They typically use DNA sequencing or hybridization techniques to detect resistance-associated mutations in the reverse transcriptase or protease genes.^12-15

Most clinical testing for HIV drug resistance is conducted using automated dideoxynucleotide-cycle sequencing. Sequence-based assays require 2–3 days to perform and provide the most accurate and complete information in analyzed regions of the HIV-1 genome. Interpretation requires knowledge of the mutations commonly induced by each of the various antiretroviral drugs.

Two sequence-based genotyping assays have been cleared by FDA for HIV-1 drug resistance analysis through the 510(k) notification process. They are the Viroseq HIV-1 genotyping system that Celera Diagnostics (Alameda, CA) developed and the Trugene HIV-1 genotyping system and OpenGene DNA sequencing system of Bayer Diagnostics.

The Viroseq assay, a second focus of this article, provides an integrated method for identifying HIV-1 mutation profiles through sequencing-based genotyping in patients infected with HIV-1 subtype B virus strains.¹² The assay includes optimized reagents for sample preparation; RT-PCR amplification of a 1.8-kb pol gene product spanning the entire protease gene and 335 codons of the reverse transcriptase gene; sequence analysis; and dedicated software from which a final HIV genotyping report is generated. 

Sample preparation involves high-speed centrifugation of 0.5 ml of plasma to pellet the virus particles, lysis of viral particles with a chaotropic agent to release the viral RNA, and isopropanol/ ethanol precipitation for RNA recovery. This assay includes a contamination control system and has been FDA-cleared for use with the ABI Prism 377 DNA sequencer and high-throughput ABI Prism 3100 genetic analyzer and 3700 DNA analyzer, instruments for capillary electrophoresis from Applied Biosystems (Foster City, CA).

The Trugene assay is designed to amplify a 1.3-kb product using RT-PCR. Four fragments are generated in a coupled amplification and sequencing step.13 A gel-based system is used for sequence resolution. Sequence assembly and analysis is performed with dedicated software, after which an interpretive resistance report is generated. This assay provides mutation analysis of protease codons 1–99 and reverse transcriptase codons 41–247.

DNA sequencing assays have a clear advantage over hybridization technologies in overall coverage and in their high specificity. Because the complete sequence is available, inclusion of new or secondary resistance-associated mutations via updating of the analysis algorithm is relatively easy. An additional advantage of sequence-based techniques relative to hybridization-based assays is that sequence information obtained for drug resistance analysis can also be used for group and subtype determination.

In the study reported in the remainder of this article, the performance of the Viroseq sequence-based genotyping assay was evaluated using RNA extracted by an alternative sample-preparation method, that of the LCx HIV viral load assay. The objective of the research was to determine whether a single RNA sample preparation could be used to monitor both HIV-1 viral load and drug resistance. Geographically dispersed specimens were included in the analysis to provide additional information on the impact of HIV-1 genetic diversity on the performance of the genotyping assay.

Experimental Methods

Figure 1. Phylogenetic relationships of discordant samples for the protease gene panel (a) and the 5' reverse transcriptase region (b). SIVcpz is used as an out-group and is designated by a closed box. Subtypes are listed in capital letters at the ends of major clusters. Samples whose subtype as derived by gag/integrase/env did not match those of the Stanford subtyping tool are boxed (click to enlarge).

HIV-1-infected plasma specimens were obtained from 91 individual asymptomatic blood donors from geographically diverse populations. The panel was composed of samples from Brazil (18 samples), Cameroon (44), Uganda (14), Thailand (5), and South Africa (10). All specimens tested positive for antibodies to HIV by at least one commercially available HIV-1 antibody enzyme immunoassay. 

In order to assess the level of genetic diversity and assign group and subtype designations, each specimen was characterized molecularly by RT-PCR amplification and by sequence analysis of three independent regions of the viral genome: 399 nucleotides of gag p24 (the gag sequence), 864 nucleotides of pol integrase (the integrase sequence), and 369 nucleotides of the env gp41 immunodominant region (the env sequence), as described in an earlier publication.16 Specimens for which the subtype designation differed between the genetic regions analyzed were categorized as mosaics.

RNA was extracted for evaluation in both the viral load and genotyping assays using the sample preparation procedure of the LCx HIV RNA quantitative assay. That involved the QIAamp viral nucleic acid extraction kit from Qiagen GmbH (Hilden, Germany). Total nucleic acid was extracted from 1 ml of plasma. 

Samples were processed by guanidine hydrochloride/protease treatment to disrupt the viral particles. Internal-standard transcript was added to each sample and processed simultaneously with the clinical specimen; this is an HIV-1 RNA transcript similar to native HIV-1 except for a short unique internal sequence. Viral nucleic acids were purified by binding to silica-based QIAamp spin columns, then subjected to buffer washes to remove contaminants that inhibit PCR, and fully eluted with water.

The investigators quantified the viral load of each RNA sample using the LCx assay in accordance with the manufacturer’s instructions. Employing a calibration curve, the LCx analyzer calculates RNA concentrations from a logarithm of the ratio of measured rate counts for HIV divided by rate counts for the internal standard. The 1-ml-sample preparation protocol that was used provided upper and lower limits of quantitation of 1 million (6 log10) copies per milliliter and 50 (1.7 log10) copies per milliliter, respectively.

Then, the Viroseq HIV-1 genotyping assay version 2.0 was performed following the manufacturer’s instructions. PCR product was purified and size-verified by agarose gel analysis. Both strands of the PCR product were sequenced directly using seven overlapping sequencing primers. Sequencing data were collected on the ABI Prism 3100 genetic analyzer and the presence of drug-resistant mutations determined with version 2.5 of the Viroseq HIV-1 genotyping system software. The Stanford HIV RT and Protease Sequence Database were used to make subtype determinations on 1.3 kb of the pol sequence.17 The results were compared with the HIV-1 group M subtype assigned on the basis of phylogenetic analysis of the protease and reverse transcriptase gene regions.

Table I. HIV-1 subtype distribution of the 91 samples in the experimental panel, as determined by gag/integrase/env genotyping (click to enlarge).

The protease and reverse transcriptase sequences were aligned with representative reference strains using the Clustal W method performed with the MegAlign Lasergene program version 5.01 from DNAStar Inc. (Madison, WI). Phylogenetic analysis was conducted using default parameters of version 3.5c of the PHYLIP phylogeny inference software package developed by Joe Felsenstein at the University of Washington (Seattle). The simian immunodeficiency virus of chimpanzee genome (SIVcpz) was used as the out-group. Genetic distance was estimated using Dnadist software (the Kimura two-parameter model), phylogeny was determined with the Neighbor program (the neighbor-joining method), trees were drawn by means of Drawtree, and determination of branch reproducibility was performed on 100 replicates using Seqboot and Consense.

Results

From gag/integrase/env genotyping, the panel of 91 HIV-1 group M samples was found to consist of representatives of 6 phylogenetically pure subtypes (A-D, F, and G), 2 CRFs, and 8 mosaics (of the forms A/D, A/F, A/G, and B/F), as specified in Table I. Viral loads for the RNA samples ranged from 2.92 log10 to 6 log10 copies per milliliter. Eighty-seven specimens, or 95.6% of the total, were successfully amplified and sequenced using Viroseq. The four samples that could not be amplified were three subtype D specimens and one CRF02_AG, all from Cameroon. Two of these samples had viral loads close to the assay’s lower level of detection of 3.3 log10 copies per milliliter: one subtype D carried 3.38 log10 RNA copies per milliliter and the CRF02_AG measured 3.49 log10 copies per milliliter.

Genotypes on the protease and reverse transcriptase sequences obtained using the Stanford database subtyping tool were concordant with the subtype assignment derived from gag/ integrase/env sequencing except for six samples. One subtype A and one CRF02_AG specimen were classified by the Stanford tool as K/G mosaics. In addition, one subtype F was typed as a D/F1 mosaic, two CRF02_AG specimens were typed as K/F2, and one subtype A was typed as D. Since most of the samples used in the analysis were obtained from countries in which multiple subtypes cocirculate, it is not surprising that a more complex genomic composition was identified for some isolates. The sequence assigned on the basis of gag/integrase/env is determined from approximately 1.6 kb of the genome. Therefore, the Viroseq-derived 1.3-kb pol fragment nearly doubles the phylogenetic information available for each sample.

To substantiate the Stanford database subtyping tool, the assay-derived pol sequence information from the six discordant samples was subjected to PHYLIP analysis. The phylogenetic relationships are shown in Figure 1. Bootstrap values are shown at the major branch nodes.

The additional sequence information gained from the 1.3-kb pol fragment supports a more complex mosaic composition of these samples, as demonstrated in the figure by basal branching and bootstrap support of less than 70%. Subtype classification is complicated by the high degree of conservation and relatively short fragment size of the protease gene (297 nt). As can be seen, the genotype of two samples, ABB002/96 and ABB105/98, remained unresolved by PHYLIP. In the protease region, no phylogenetic evidence of subtype K was indicated. Although differences in genotype determination between the methods used for subtype analysis were observed, the high level of concordance of subtype determination based on gag/ integrase/env with the assay-derived protease/reverse transcriptase sequence (81 of 87 samples, or 93%) demonstrates that the Viroseq technology could be used to determine subtype across the region of the genome that is monitored for mutations exhibiting drug resistance.

Drug resistance analysis detected a high frequency of the M36I substitution, ranging from 86 to 100%, in non-B subtypes versus 30% in subtype B viruses (see Table II). Interestingly, presence of this particular polymorphism is observed in less than 50% of subtype B specimens monitored in the United States. The K20R mutation was present in 50% of subtype A specimens and 20% of subtype C and CRF01_AE samples. These results suggest that polymorphic mutations in non-B HIV-1 isolates potentially could affect a response to protease inhibitors. In contrast, no polymorphisms in the positions associated with drug resistance were detected in the reverse transcriptase gene.

Discussion

Table II. The frequency (in %) of natural polymorphisms occurring in positions associated with protease inhibitors. L10I/V/R, M361, L63P, and V77I are associated with resistance to protease inhibitors when present with other mutations. K20R is associated with resistance to IDV, RTV, and LPV when present with other mutations. (click to enlarge).

As more diagnostic assays become available for the management of HIV infection, it would be desirable that a single sample preparation could be used to perform all testing. The research study reported here demonstrated that the RNA-extraction procedure used for the LCx HIV assay for viral load determination could also be applied to Viroseq drug resistance monitoring. Drug-resistance mutation analysis was suc- cessful for 95.6% of the evaluation panel using RNA extracted by this alternate method. There was no evidence of interference by the internal standard used in the LCx assay.

The manual sample-preparation method used for Viroseq at this time is time-consuming and labor-intensive. Thus, employment of a single sample-preparation procedure for both viral load measurement and genotyping by Viroseq could potentially reduce laboratory costs related to technical time significantly, and would serve to streamline laboratory work flow. An additional advantage to using the same specimen for both viral load testing and genotyping is that it provides the opportunity to characterize rebounding virus at the earliest possible moment. Moreover, it would establish a direct linkage between these two very important patient-monitoring technologies.

In the present study, the genotyping assay showed excellent performance on a panel of genetically diverse strains of HIV-1 group M subtypes A through G and recombinant viral genomes. This is consistent with results of a previous study using a panel of non–subtype B viral isolates.¹⁸

Conclusion

Natalia Marlowe, PhD, is manager and HIV/HLA group leader at Celera Diagnostics (Alameda, CA).

Priscilla Swanson is principal scientist, AIDS Research and Retrovirus Laboratory at Abbott Laboratories (Abbott Park, IL).

Birgit Drews is senior associate scientist, product development at Celera Diagnostics.

Catherine Brennan, PhD, is section manager,

Sushil G. Devare, PhD, is manager,

John Hackett Jr., PhD, is section manager in the AIDS research and retrovirus discovery unit of Abbott Laboratories. The authors can be reached at natalia.marlowe@ celeradiagnostics.com, priscilla. swanson@abbott.com, birgit.drews@ celeradiagnostics.com, catherine. brennan@abbott.com, sushil.devare@abbott. com, and john.hackett@abbott .com, respectively.

As antiretroviral drugs become increasingly available in regions where non–subtype B strains predominate, genotyping assays that perform in a largely subtype-independent manner will be invaluable. Significantly, the additional sequence data derived from the 1.3-kb pol fragment can be used for phylogenetic analysis. Such analysis will improve understanding of the molecular composition of these genetically diverse viral strains and may aid in discovery of the influence of subtype on the evolution of drug-resistant viruses.

In the future, the application of automated sample-preparation technology to viral load and genotyping assays would make possible greater efficiency in the laboratory evaluation of patient samples, and could reduce cost. Studies are in progress to evaluate whether automated RNA-extraction instrumentation being developed for use with the LCx HIV quantification assay can also be used in the 96-well Viroseq format.

References

1. JM Coffin, “Retroviridae: The Viruses and Their Replication,” in Fields Virology, 3rd, ed. B Fields et al. eds.(Philadelphia: Lippincott-Raven, 1996), 1767–1847.

2. CL Kuiken et al., eds., Human Retroviruses and AIDS 2000: A Compilation and Analysis of Nucleic Acid and Amino Acid Sequences (Los Alamos, NM: Theoretical Biology and Biophysics Group, Los Alamos National Laboratory, 2000).

3. M Peeters, “Recombinant HIV Sequences: Their Role in the Global Epidemic,” in: Human Retroviruses and AIDS 2000: A Compilation and Analysis of Nucleic Acid and Amino Acid Sequences, CL Kuiken et al., eds. (Los Alamos, NM: Theoretical Biology and Biophysics Group, Los Alamos National Laboratory, 2000), 54–72.

4. S Osmanov et al., “Estimated Global Distribution and Regional Spread of HIV-1 Genetic Subtypes in the Year 2000,” Journal of Acquired Immune Deficiency Syndromes 29 (2002): 184–190.

5. R Sun et al., “Ultrasensitive Reverse Transcription-PCR Assay for Quantitation of Human Immunodeficiency Virus Type 1 RNA in Plasma,” Journal of Clinical Microbiology 36 (1998): 2964–2969.

6. K Triques et al., “Efficiencies of Four Versions of the Amplicor HIV-1 Monitor Test for Quantification of Different Subtypes of Human Immunodeficiency Virus Type 1,” Journal of Clinical Microbiology 37 (1999): 110–116.

7. J Johanson et al., “A New Ultrasensitive Assay for Quantitation of HIV-1 RNA in Plasma,” Journal of Virological Methods 95 (2001): 81–92.

8. B van Gemen et al., “A One-Tube Quantitative HIV-1 RNA NASBA Nucleic Acid Amplification Assay Using Electrochemiluminescent (ECL) Labelled Probes,” Journal of Virological Methods 49 (1994): 157–167.

9. ML Collins et al., “A Branched DNA Signal Amplification Assay for Quantification of Nucleic Acid Targets below 100 Molecules/ ml,” Nucleic Acids Research 25 (1997): 2979–2984.

10. P Swanson et al., “Comparative Performance of Three Viral Load Assays on Human Immunodeficiency Virus Type 1 (HIV-1) Isolates Representing Group M (Subtypes A to G) and Group O: LCx HIV RNA Quantitative, Amplicor HIV-1 Monitor Version 1.5, and Quantiplex HIV-1 RNA Version 3.0,” Journal of Clinical Microbiology 39 (2001): 862–870.

11. MS Hirsch et al., “Antiretroviral Drug Resistance Testing in Adults Infected with Human Immunodeficiency Virus Type 1: 2003 Recommendations of an International AIDS Society–USA Panel,” Clinical Infectious Disease 37 (2003): 113–128.

12. SB Cunningham et al., “Performance of the Applied Biosystems Viroseq Human Immunodeficiency Virus Type 1 (HIV-1) Genotyping System for Sequence-Based Analysis of HIV-1 in Pediatric Plasma Samples,” Journal of Clinical Microbiology 39 (2001): 1254–1257.

13. RM Grant et al., “Accuracy of the Trugene HIV-1 Genotyping Kit,” Journal of Clinical Microbiology 41 (2003): 1586–1593.

14. L Stuyver et al., “Line Probe Assay for Rapid Detection of Drug-Selected Mutations in the Human Immunodeficiency Virus Type 1 Reverse Transcriptase Gene,” Antimicrobial Agents and Chemotherapy 41 (1997): 284–291.

15. M Vahey et al., “Performance of the Affymetrix GeneChip HIV PRT 440 Platform for Antiretroviral Drug Resistance Genotyping of Human Immunodeficiency Virus Type 1 Clades and Viral Isolates with Length Polymorphisms,” Journal of Clinical Microbiology 37 (1999): 2533–2537.

16. P Swanson, SG Devare, and J Hackett Jr, “Molecular Characterization of 39 HIV-1 Isolates Representing Group M (Subtypes A–G) and Group O: Sequence Analysis of gag p24, pol Integrase, and env gp41,” AIDS Research and Human Retroviruses 19 (2003): 625–629.

17. Stanford HIV RT and Protease Sequence Database (accessed 4 March 2004); available from Internet: http://hivdb.stanford.edu.

18. LL Jagodzinski et al., “Performance Characteristics of Human Immunodeficiency Virus Type 1 (HIV-1) Genotyping Systems in Sequence-Based Analysis of Subtypes Other Than HIV-1 Subtype B,” Journal of Clinical Microbiology 41 (2003): 998–1003. 

Copyright ©2004 IVD Technology