env Gene Typing of Human Immunodeficiency Virus Type 1 Strains on Electronic Microarrays
http://www.100md.com
微生物临床杂志 2005年第4期
Genomics Proteomics Bioinformatics Unit
Sexually Transmitted Bacteria Laboratory
Sexually Transmitted and Blood-borne Virus Laboratory, Centre for Infections, London, United Kingdom
ABSTRACT
The NanoChip system was used for subtyping human immunodeficiency virus type 1 (HIV-1) strains using probes complementary to the V1 region of the env gene. Probes for six subtypes (A to D, F, and G) and two circulating recombinant forms (AG and AE) of HIV-1 group M were included. The specificity of these oligonucleotides had been evaluated previously in a DNA enzyme immunoassay. Samples from 112 patient sera were used as templates in a nested reverse transcription-PCR to produce amplicons that were applied to the array. The array was then hybridized successively to pairs of oligonucleotide probes. The strains were assigned a subtype on the basis of their probe hybridization patterns. One strain gave a contradictory pattern and was designated as untypeable by the NanoChip assay. Eighty-eight strains gave hybridization patterns that allowed a correct subtype designation to be made by the NanoChip assay compared to either the sequence or the heteroduplex mobility assay (HMA)-determined subtypes. Thirteen strains that reacted with the subtype A probe (SA2) were incorrectly assigned to subtype A, or to one of the related circulating recombinant types (AE or AG), on the basis of reactions with probe SAE1 or SAG1. The results indicate that these oligonucleotides have relatively low specificities. The probe subtypes of three strains matched the subtypes determined for the gag and pol genes but not the env gene, suggesting that a recombination event may have occurred within the env gene. Overall, the NanoChip assay gave results comparable to those for HMA and sequencing and provides a convenient and cost-effective means by which to subtype HIV-1.
INTRODUCTION
Human immunodeficiency virus type 1 (HIV-1) sequences are highly diverse and exhibit considerable genetic variation both between and within viral strains. This is an important complicating factor in the accuracy of diagnostic assays (4) and may alter the susceptibility to antiretroviral drugs (2, 10). Such variability also provides a means of monitoring the epidemiology of the pandemic, can indicate the origin of an infection, and provides a means of tracking the spread of infection between exposure groups (15, 26, 29). Given the potential public health impact of HIV-1 genetic diversity, the monitoring of subtype trends remains an essential component of HIV surveillance (24). Consequently, systems for subdividing HIV-1 strains on the basis of sequences are of great interest and importance for diagnosis and in epidemiological studies of the AIDS pandemic. Phylogenetic analysis has allowed the classification of HIV-1 into three groups: the major (M) group, the outlier (O) group, and the non-M, non-O (N) group. Group M is responsible for the majority of infections in the pandemic; the other two groups are highly diverse and less prevalent (18). Group M isolates can be subdivided into nine subtypes (A to D, F to H, J, and K) and a number of circulating recombinant forms (CRFs), which have identical mosaic genomes and are assumed to have arisen by recombination between different subtypes. The diversity and widespread nature of the pandemic have contributed to an increase in the subtyping of HIV-1 isolates; hence, more rapid and reliable methods are required.
Several methods for HIV subtyping have been described. DNA sequencing provides the most information but is time-consuming and laborious. The heteroduplex mobility assay (HMA) can be used to compare a test sample against many references, enabling visualization of quasispecies, but is also laborious and technically complex. Recently, Plantier and colleagues (17) have reported a DNA enzyme immunoassay (DEIA) for typing HIV-1 using subtype-specific oligonucleotide probes. The assay uses 20 to 30 base probes which hybridize to a 250-bp PCR amplicon derived from the first codons of the env gene, i.e., between codon 1 and the middle of the conserved region C1 of gp120. The probes were designed to have similar melting temperatures (Tms) and at least 20% sequence differences with other types. It was not possible to fulfill the latter criterion for strains of types B and D or for types A, CRF_AE, and CRF_AG, and additional probes were therefore selected. Probe SBD3 identifies strains of both types B and D. These strains can then be identified with type-specific probes. Probe SA2 hybridizes to strains of types A, CRF_AE, and CRF_AG. Subsequently, CRF_AE and CRF_AG can be identified with subtype-specific probes. Biotinylated probes were immobilized in microtiter tray wells and hybridized to the denatured env amplicon from test samples. Hybridization was detected using an antibody specific for double-stranded DNA. Three different sets of probes were used, corresponding to the subtypes commonly found in the geographic region that was the source of the HIV-1 strains examined. As many as six probes, each in a separate well of the microtiter plate, were hybridized to amplicons concurrently. The DEIA method was found to be both accurate and reliable compared to HMA or sequencing, allowing correct identification of 84% of the test samples.
The NanoChip system (Nanogen Inc., San Diego, Calif.) provides a convenient platform for oligonucleotide hybridization assays. The system is comprised of the loader, which applies the samples to specific test sites on the microarray, and a two-channel fluorescent reader, which can resolve measurements of fluorescence at the emission peaks of two different fluors. Both components can perform repetitive liquid handling operations automatically so that the hands-on technical effort is minimized. In addition, both instruments can apply a voltage selectively at one or more of the pads. The NanoChip cartridges consist of 10-by-10 arrays of microelectrode test sites coated with a gel permeation layer that contains streptavidin. The array format allows up to 100 separate analyses to be done in parallel, one at each pad. Each test site can be independently electronically activated using the NanoChip Molecular Biology Workstation.
Here we describe the use of the DEIA probes to type HIV-1 env gene amplicons using the NanoChip system (11-14, 19, 21, 25, 27). The use of previously described DEIA probes for genotyping on the NanoChip system allowed the identification of 79% of the test samples. Electronic microarrays thus offer an alternative methodology for HIV subtyping.
MATERIALS AND METHODS
Strains. Residual serum specimens, obtained from heterosexual individuals attending sexually transmitted infection clinics participating in a national Unlinked Anonymous HIV Prevalence Monitoring Program, were submitted to the Sexually Transmitted and Blood-borne Virus Laboratory for subtyping. Subtypes were determined by HMA in the gag and env gene regions, with sequencing where necessary, as previously described (3, 5, 9, 23). All pol subtypes were assigned by sequencing, as previously described (3). One hundred twelve samples were included in the study.
Nucleic acid extraction and PCR. HIV-1 viral RNA was extracted from residual serum by the Boom method (6), and a mixed specific-primer reverse transcription reaction was performed, as described previously (3). First- and second-round PCR conditions were as previously described (3), with the exception that the first-round forward env primer ED3 was used in place of ED5, and the env primers ES4 (5'-AAG AGC AGA AGA YAG TGG CAA TG-3') and ES5 (5'-TAC ACA GGC ATG TGT RGC CC-3') were used in the second-round PCR, as described by Plantier and colleagues (17). Primer ES5 was 5' end modified by addition of a biotin residue. All native and modified oligonucleotides used in this study were supplied by MWG Ltd., Milton Keynes, United Kingdom. The biotinylated amplicon was purified and desalted on Millipore MultiScreen 96 PCR cleanup plates [Millipore (UK) Ltd., Watford, United Kingdom] and reconstituted in 50 mM histidine (200 μl).
Loading of NanoChip cartridges for HIV-1 subtyping. NanoChip cartridges were prepared for loading by filling with 0.3 M NaOH for 5 min and washing with deionized water. The desalted biotinylated amplicons were resuspended in 50 mM histidine buffer (60 μl) and placed into wells of a Nunc 384 square-well polypropylene microtiter plate (Scientific Laboratory Supplies Ltd., Nottingham, United Kingdom) which was placed into the NanoChip loader. Each amplicon was loaded sequentially onto duplicate nonadjacent NanoChip test sites for 3 min at a nominal 2 V. Negative-control pads were addressed with the histidine buffer. Acceptable levels of current at each electrode were in the range of 300 to 600 nA. Following loading the nonbiotinylated strand was removed by filling the NanoChip with 0.3 M NaOH for 5 min. The NanoChip was then washed through with hybridization buffer (50 mM NaPO4 [pH 7.4], 0.5 M NaCl). For storage, loaded NanoChip cartridges were filled with deionized water and placed in a sealed bag at 4°C.
NanoChip HIV-1 subtyping. Subtype-specific oligonucleotide probes (17) were hybridized to the HIV-1 env gene amplicons immobilized on duplicate test sites on the NanoChip cartridge. Each probe was either Cy3 or Cy5 end labeled, as listed in Table 1. Stock solutions (100 μM) of one Cy3- and one Cy5-labeled probe were diluted to final concentrations of 2 μM in hybridization buffer, and five probe pairs were hybridized sequentially to the cartridge (SB2-SBD3, SA2-SAG1, SC1-SAE1, SF6-SG1, and SB2-SD1.1). Fifty microliters of the probe solution was introduced into the NanoChip cartridge. The NanoChip cartridge was then inserted into the reader. Hybridization was performed by heating the cartridge to 45°C for 2 min, cooling it to 40°C for 2 min, and then cooling it to 35°C for a further 2 min. Unbound probe was removed by three washing cycles with hybridization buffer at 25°C (each wash cycle comprised 500 μl at 75 μl/s). Fluorescent signal remaining at the electrodes was detected in the red (Cy5) and green (Cy3) channels of the reader at 25°C. The temperature was then raised to 40 and then to 50°C, with three wash cycles at each temperature. The remaining fluorescent signal was measured after washing at each temperature. Pairs of probes were hybridized sequentially to the HIV-1 NanoChip microarray. Between hybridizations, the array was regenerated by being washed three times for 5 min with 0.3 M NaOH at ambient temperature and rescanned to ensure that the signal had returned to background. Subtypes were assigned by consideration of the pattern of hybridization of the env amplicon to all nine probes.
RESULTS
Each biotinylated ES4/5 amplicon was electronically addressed to two microarray test sites, where it was anchored via the interaction with the streptavidin in the gel permeation layer. The array was then hybridized to the nine fluorescent oligonucleotide probes described by Plantier and colleagues (17), and the patterns of reactivity were observed. The Tms of the probes were in the range of 55 to 74°C as determined by the nearest neighbor method assuming a local probe concentration of 50 nM (Table 1). The probe concentration is 2 μM during initial hybridization but is then reduced to very low levels during the wash steps. Under the high-salt conditions used for hybridization and washing steps, all of the probes are likely to form stable homoduplexes at up to 50°C. The NanoChip system was used to measure fluorescence retained on the chip after washing at 25, 40, and 50°C. The scan at 25°C was performed to indicate the level of hybridization at low stringency, and the higher-temperature scans were intended to measure changes in probe binding at closer to their Tms. For a subtyping call to be made, the fluorescent signal of the sample (40°C) had to be at least fivefold higher than the negative controls (pads mock addressed with histidine) for at least one probe. Figure 1 shows the hybridization values for five strains. A sample was considered positive for a particular subtype if the fluorescent signal decreased less than 50% between 25 and 50°C, indicating stable hybridization of the probe. When the fluorescent signal was less than fivefold higher than the negative control at 40°C or when the loss of fluorescent signal between 25 and 50°C was greater than 50%, the reaction was considered negative.
There was no apparent decrease in signal levels or in signal-to-noise ratios between successive rounds of hybridization. There was some variation in fluorescent signal obtained on replicate pads, but this was not greater than a twofold difference and was usually within 20% for negatives. Very occasionally, fluorescent particles adhering to the pads caused higher levels of variation, but these could be identified using the NanoChip instrument, and these measurements were repeated. There was sometimes considerable variation (up to two- to threefold) between the levels of fluorescent signal observed for replicate hybridizations, but this made only a small difference to the positive/negative ratio and did not affect the overall result.
Probe SD1.1 gave nonspecific results under the conditions used for the array, and its reactions were therefore disregarded. The 112 strains included in the study were assigned to type according to probe reactions as shown in Table 2. Table 2 was compiled by following a set of rules formulated by examination of the results of typing on the array for the first 40 strains as follows. (i) All HIV-1 types give positive hybridization signals with their corresponding probes. (ii) Strains of CRF_AE or CRF_AG always hybridize additionally to probe SA2. (iii) Type AG strains may cross-react with probe SAE1. (iv) Probe SF6 reactions are assumed to be cross-reactions when another probe gives a positive reaction (except for SAE1 or SAG1 when SA2 is negative). (v) Probe SAE1 or SAG1 reactions are assumed to be cross-reactions when another probe gives a positive reaction (except when SA2 is positive). (vi) Type G amplicons may cross-react with probe SA2.
The results were compared with the env gene subtypes determined by HMA or sequencing and are summarized in Table 3. The subtypes determined by HMA and/or sequencing and probe hybridization on the NanoChip system were identical for 88 strains (79%). One strain gave an inconsistent pattern of probe hybridization by the NanoChip system and was designated as nontypeable. The remaining 23 strains were assigned different subtypes by different methods.
Thirteen of the 23 discrepant strains were identified as type A or CRF_AE or CRF_AG by HMA and/or sequencing. When typed on NanoChip microarrays, all 13 hybridized to the subtype A probe (SA2); however, hybridization to the probes SAE1 and SAG1 did not match the HMA and/or sequencing subtype designation. This represents 32% of all SA2-reactive strains. In eight instances the mismatch involved an A-AE pair, and in the remaining five, it involved an A-AG pair. In one case the probe subtype (AG) was in agreement with the HMA or sequence subtype for gag and pol genes, respectively, but discordant with the env subtype as determined by HMA. The results are summarized in Table 3.
Three of the remaining 10 subtyping discrepancies can be best explained by recombination. In these cases, the probe subtype corresponded with the sequence subtype for both the gag and pol genes but not for the env gene. The region of the env gene used in the HMA and sequencing assays did not include the probed amplicon. It is possible, therefore, that recombination events affecting the first codons of env may explain these observed discrepancies. The HMA and/or sequencing subtypes of the remaining seven samples were C (n = 3), B (n = 2), and D and G (one sample each). There was no consistent pattern of mismatching between subtypes. The results for the samples with mismatching subtypes are shown in Table 4.
The overall accuracy of the method was 79%, but this increased to 91% if no attempt was made to distinguish between type A and the CRFs. The corresponding false-positive rates were therefore 21 and 9%. Coverage of the method was over 99%. False-negative and false-positive rates for each probe are given in Table 5.
Overall, the data suggest that the use of these probes on the NanoChip system may provide a promising alternative method for reliable, high-throughput subtyping of HIV-1 samples obtained from heterosexuals in England and Wales.
DISCUSSION
Two different assay designs were considered for the NanoChip HIV-1 typing system. The first was to immobilize the probes at the test sites, with each probe occupying a different subset of sites. Since 100 sites are available, each of the nine probes could be directed to 11 test sites, leaving one control pad. The chip could then be hybridized with up to 22 different amplicons on the loader. This design requires the sequential introduction of differentially labeled amplicon pairs to the cartridge, with a different set of nine sites electronically activated at each step. The electrophoretic concentration at activated test sites would enable rapid hybridization of the amplicon to the immobilized probe (i.e., active hybridization). Once all amplicons are hybridized, the array would be transferred to the reader for washing and reading at the desired temperatures. This method is analogous to the DEIA method previously used with this probe set (17). The major disadvantage of this approach is that the two strands of the amplicon must be separated to allow hybridization to the probe. Since there are no practical means of maintaining the amplicons in denatured form on the NanoChip loader, it is best to apply only the strand that is complementary to the probes. Single-strand amplicons can be produced by asymmetric PCR, but this may compromise the sensitivity of the assay. The second method, and the one that was actually implemented for the study, involves electronic addressing of biotinylated amplicons to test sites, using the NanoChip loader. Each probe pair, one Cy3 and one Cy5 labeled, was then passively hybridized to the amplicons on the microarray. Since it is not yet possible to automate the addition of different hybridization solutions to the cartridge on the NanoChip reader, the pairing of probes in this way reduces hands-on time for each experiment. Subsequently, thermal discrimination and fluorescent signal detection were performed on the reader. In this format, a single cartridge could be used to type a total of 49 specimens at one time, so that the assay is relatively inexpensive compared with HMA or sequencing, with an approximate materials cost per strain of $4.
Another consideration in assay design was selection of the probes. Several previous studies using the NanoChip system have described probes that have estimated Tms in the range of 30 to 50°C (12, 19, 20, 27). The probes for typing HIV-1 by DEIA (17) all have calculated Tms greater than 55°C. Trimming terminal bases might have reduced the Tms of the probes, possibly resulting in higher specificities and in elimination of many cross-reactions observed at the low wash temperatures. Although shortened probes may have theoretical advantages for HIV-1 typing on NanoChip microarrays, the full-length sequences were used for this work. The main reason for this choice was to enable comparison with the previous work (17); however, it may also be argued that probes with high Tms are required to maintain overall strain coverage in this application, due to the highly variable nature of the target sequence. If it had been desirable to use more stringent conditions, it would have been possible to reduce the salt concentration of the wash buffer. Optimization of the hybridization stringency for probe SD1.1 on the NanoChip platform might have resulted in it showing better specificity. Probe SD1.1 has the highest calculated Tm of the oligonucleotides in this probe set, and this may explain its poor performance under the selected assay conditions. It should be noted that in the DEIA (17) probe SD1.1 was used only for analysis of strains that reacted with probe SBD3. However, probe SBD3 was adequate in identifying type D strains. Several significant cross-reactions were noted for other probes included in the set. The SF6 probe was found to cross-react with one CR_AE strain and 40% of the type B strains. In the same way, probe SA2 cross-reacted with a single type G strain and the probes for the circulating recombinant types cross-reacted with a proportion of the type F and G strains (Table 2). These cross-reactions are caused by sequence similarities between the types at the probe sites.
As noted by Plantier and colleagues (17), typing based on probes is inherently less reliable and less informative than that based on sequencing. This is inevitable, because the sequences sampled by the probes are relatively short, and recombinant forms of HIV-1 arise frequently. Nevertheless, development of rapid and simple methods for subtyping will be valuable for screening analyses. The results presented here demonstrate that DEIA and microarray-based subtyping show comparable accuracy compared to HMA or sequencing, with 79% of samples correctly subtyped on microarrays, versus 84% by DEIA. Experience in this laboratory has previously shown that approximately 90% of samples can be assigned a subtype by using HMA alone. Given the possibility of recombination between HIV-1 strains, the result of any assay that assigns a subtype based on a single gene region may not be applicable to other gene regions. The difficulty of identifying recombinant viruses is a problem common to all methods other than the comparison of full genome sequences. Increased accuracy may thus be achieved by incorporation of additional probes targeting other env gene regions and/or an alternative target gene(s). Such an approach would increase the likelihood of detecting recombinant viruses that are becoming more prevalent in Europe (3). However, it may not be feasible to design specific probes for each subtype based on the relatively conserved gag and pol genes.
The figure of 79% strains correctly subtyped on the array should be considered in the context that not all strains were tested as unknowns. The raw data for 40 strains were first used to develop a consistent set of rules for the interpretation of hybridization patterns, and then types were designated using this basis. There was a small bias in favor of correct identification for the unblinded samples, but this was not significant (2, P = 0.22). On the other hand, three of the strains were recombinant viruses that matched the probe type in the gag and pol genes but were incorrectly identified because their env genes were of different types.
Thirteen of the discrepant calls involved discrimination among subtypes A, CRF_AG, and CRF_AE. Sequencing and phylogenetic analysis might resolve some of these discrepancies in favor of the probe methods, since some subtype A samples may have been misclassified by HMA. No amplicon to detect either CRF_AG or CRF_AE was included in the gag or env HMA, so the discrepancies observed with the chip assay may be due to misclassification of these samples by HMA. Alternatively, it may be that the probes designed to detect the CRFs (SAE1 and SAG1) are unreliable when used in areas where these viral strains cocirculate with other HIV-1 subtypes, such as the heterosexual epidemic in England and Wales (3, 22).
While the likely accuracy of the probe system was similar for the NanoChip and DEIA platforms, the NanoChip assay has several advantages over the DEIA method (17). It is possible to distinguish between probe-amplicon duplexes that have different Tms on the chip system because the fluorescent endpoint is compatible with the collection of data at multiple temperature values. The number of manipulations and the time required to assess fluorescent signals at different temperatures are lower for the chip assay. This is due to the simple assay format, combined with the automated liquid handling provided by the NanoChip instrument. The design of the chip assay also enables the use of a single format for all specimens. Plantier and colleagues (17) used different probe sets for specimens from France, Cameroon, and Senegal in order reduce the number of test wells required for each sample; in contrast, multiple probe sets can be sequentially hybridized to immobilized amplicons on a single electronic microarray. In addition, some probes in each set were used hierarchically in the DEIA analyses, so that two steps were required to identify certain subtypes. These differences in the testing protocol were, at least in part, responsible for the relatively poor performance of the AE and AG probes in the microarray format.
Recent studies have shown the heterosexual HIV-1 epidemic in England and Wales to be highly diverse, with all group M subtypes and several CRFs having been detected (3, 22). This situation is mirrored in other Western European countries, where the diversity is thought to have arisen due to the importation of infections from areas with diverse HIV-1 epidemics (1, 3, 7, 8, 16, 22, 28). Such diversity may require additional probe optimization for increased accuracy in hybridization-based assay formats. Given the continued global expansion of HIV-1 infection, there is an increasing need to monitor the molecular diversity of the epidemic. The NanoChip assay has been shown to produce results comparable to the widely used HMA and has the potential to provide a relatively high-throughput, reliable, and cost-effective alternative means of subtyping HIV-1.
Present address: WW Epidemiology (Neurosciences), GlaxoSmithKline Research and Development, Harlow, United Kingdom.
REFERENCES
Alaeus, A., T. Leitner, K. Lidman, and J. Albert. 1997. Most HIV-1 genetic subtypes have entered Sweden. AIDS 11:199-202.
Apetrei, C., D. Descamps, G. Collin, I. Loussert-Ajaka, F. Damond, M. Duca, F. Simon, and F. Brun-Vezinet. 1998. Human immunodeficiency virus type 1 subtype F reverse transcriptase sequence and drug susceptibility. J. Virol. 72:3534-3538.
Barlow, K. L., I. D. Tatt, P. A. Cane, D. Pillay, and J. P. Clewley. 2001. Recombinant strains of HIV type 1 in the United Kingdom. AIDS Res. Hum. Retrovir. 17:467-474.
Barlow, K. L., J. H. Tosswill, J. V. Parry, and J. P. Clewley. 1997. Performance of the Amplicor human immunodeficiency virus type 1 PCR and analysis of specimens with false-negative results. J. Clin. Microbiol. 35:2846-2853.
Belda, F. J., K. L. Barlow, and J. P. Clewley. 1997. Subtyping HIV-1 by improved resolution of heteroduplexes on agarose gels. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 16:218-219.
Boom, R., C. J. Sol, M. M. Salimans, C. L. Jansen, P. M. Wertheim-van Dillen, and J. van der Noordaa. 1990. Rapid and simple method for purification of nucleic acids. J. Clin. Microbiol. 28:495-503.
Clewley, J. P., C. Arnold, K. L. Barlow, P. R. Grant, and J. V. Parry. 1996. Diverse HIV-1 genetic subtypes in UK. Lancet 347:1487.
Couturier, E., F. Damond, P. Roques, H. Fleury, F. Barin, J. B. Brunet, F. Brun-Vezinet, F. Simon, et al. 2000. HIV-1 diversity in France, 1996-1998. AIDS 14:289-296.
Delwart, E. L., B. Herring, A. G. Rodrigo, and J. I. Mullins. 1995. Genetic subtyping of human immunodeficiency virus using a heteroduplex mobility assay. PCR Methods Appl. 4:S202-S216.
Descamps, D., C. Apetrei, G. Collin, F. Damond, F. Simon, and F. Brun-Vezinet. 1998. Naturally occurring decreased susceptibility of HIV-1 subtype G to protease inhibitors. AIDS 12:1109-1111.
Edman, C. F., P. Mehta, R. Press, C. A. Spargo, G. T. Walker, and M. Nerenberg. 2000. Pathogen analysis and genetic predisposition testing using microelectronic arrays and isothermal amplification. J. Investig. Med. 48:93-101.
Erali, M., B. Schmidt, E. Lyon, and C. Wittwer. 2003. Evaluation of electronic microarrays for genotyping factor V, factor II, and MTHFR. Clin Chem. 49:732-739.
Ewalt, K. L., R. W. Haigis, R. Rooney, D. Ackley, and M. Krihak. 2001. Detection of biological toxins on an active electronic microchip. Anal. Biochem. 289:162-172.
Heller, M. J., A. H. Forster, and E. Tu. 2000. Active microelectronic chip devices which utilize controlled electrophoretic fields for multiplex DNA hybridization and other genomic applications. Electrophoresis 21:157-164.
Kunanusont, C., H. M. Foy, J. K. Kreiss, S. Rerks-Ngarm, P. Phanuphak, S. Raktham, C. P. Pau, and N. L. Young. 1995. HIV-1 subtypes and male-to-female transmission in Thailand. Lancet 345:1078-1083.
Perez-Alvarez, L., M. T. Cuevas, M. L. Villahermosa, J. D. Pedreira, N. Manjon, I. Herrero, S. Lopez-Calvo, E. Delgado, E. V. de Parga, L. Medrano, M. M. Thomson, J. A. Taboada, and R. Najera. 2001. Prevalence of drug resistance mutations in B, non-B subtypes, and recombinant forms of human immunodeficiency virus type 1 in infected individuals in Spain (Galicia). J. Hum. Virol. 4:35-38.
Plantier, J. C., L. Vergne, F. Damond, S. MBoup, E. MPoudi-NGole, L. Buzelay, I. Farfara, D. Brand, M. Peeters, F. Brun-Vezinet, E. Delaporte, and F. Barin. 2002. Development and evaluation of a DNA enzyme immunoassay method for env genotyping of subtypes A through G of human immunodeficiency virus type 1 group M, with discrimination of the circulating recombinant forms CRF01_AE and CRF02_AG. J. Clin. Microbiol. 40:1010-1022.
Robertson, D. L., J. P. Anderson, J. A. Bradac, J. K. Carr, B. Foley, R. K. Funkhouser, F. Gao, B. H. Hahn, M. L. Kalish, C. Kuiken, G. H. Learn, T. Leitner, F. McCutchan, S. Osmanov, M. Peeters, D. Pieniazek, M. Salminen, P. M. Sharp, S. Wolinsky, and B. Korber. 2000. HIV-1 nomenclature proposal. Science 288:55-56.
Santacroce, R., A. Ratti, F. Caroli, B. Foglieni, A. Ferraris, L. Cremonesi, M. Margaglione, M. Seri, R. Ravazzolo, G. Restagno, B. Dallapiccola, E. Rappaport, E. S. Pollak, S. Surrey, M. Ferrari, and P. Fortina. 2002. Analysis of clinically relevant single-nucleotide polymorphisms by use of microelectronic array technology. Clin. Chem. 48:2124-2130.
Schrijver, I., M. J. Lay, and J. L. Zehnder. 2003. Diagnostic single nucleotide polymorphism analysis of factor V Leiden and prothrombin 20210G > A. A comparison of the Nanogen Electronic Microarray with restriction enzyme digestion and the Roche LightCycler. Am. J. Clin. Pathol. 119:490-496.
Sosnowski, R., M. J. Heller, E. Tu, A. H. Forster, and R. Radtkey. 2002. Active microelectronic array system for DNA hybridization, genotyping and pharmacogenomic applications. Psychiatr. Genet. 12:181-192.
Tatt, I. D., K. L. Barlow, J. P. Clewley, O. N. Gill, and J. V. Parry. 2004. Surveillance of HIV-1 subtypes among heterosexuals in England and Wales, 1997-2000. J. Acquir. Immune Defic. Syndr. 36:1092-1099.
Tatt, I. D., K. L. Barlow, and J. P. Clewley. 2000. A gag gene heteroduplex mobility assay for subtyping HIV-1. J. Virol. Methods 87:41-51.
Tatt, I. D., K. L. Barlow, A. Nicoll, and J. P. Clewley. 2001. The public health significance of HIV-1 subtypes. AIDS 15:S59-S71.
Thistlethwaite, W. A., L. M. Moses, K. C. Hoffbuhr, J. M. Devaney, and E. P. Hoffman. 2003. Rapid genotyping of common MeCP2 mutations with an electronic DNA microchip using serial differential hybridization. J. Mol. Diagn. 5:121-126.
Trask, S. A., C. A. Derdeyn, U. Fideli, Y. Chen, S. Meleth, F. Kasolo, R. Musonda, E. Hunter, F. Gao, S. Allen, and B. H. Hahn. 2002. Molecular epidemiology of human immunodeficiency virus type 1 transmission in a heterosexual cohort of discordant couples in Zambia. J. Virol. 76:397-405.
Westin, L., C. Miller, D. Vollmer, D. Canter, R. Radtkey, M. Nerenberg, and J. P. O'Connell. 2001. Antimicrobial resistance and bacterial identification utilizing a microelectronic chip array. J. Clin. Microbiol. 39:1097-1104.
Yirrell, D. L., D. J. Goldberg, J. Whitelaw, C. McSharry, F. Raeside, and G. Codere. 1999. Viral subtype and heterosexual acquisition of HIV infections diagnosed in Scotland. Sex. Transm. Infect. 75:392-395.
Yirrell, D. L., P. Robertson, D. J. Goldberg, J. McMenamin, S. Cameron, and A. J. Leigh Brown. 1997. Molecular investigation into outbreak of HIV in a Scottish prison. BMJ 314:1446-1450.(N. A. Saunders, S. Alexan)
Sexually Transmitted Bacteria Laboratory
Sexually Transmitted and Blood-borne Virus Laboratory, Centre for Infections, London, United Kingdom
ABSTRACT
The NanoChip system was used for subtyping human immunodeficiency virus type 1 (HIV-1) strains using probes complementary to the V1 region of the env gene. Probes for six subtypes (A to D, F, and G) and two circulating recombinant forms (AG and AE) of HIV-1 group M were included. The specificity of these oligonucleotides had been evaluated previously in a DNA enzyme immunoassay. Samples from 112 patient sera were used as templates in a nested reverse transcription-PCR to produce amplicons that were applied to the array. The array was then hybridized successively to pairs of oligonucleotide probes. The strains were assigned a subtype on the basis of their probe hybridization patterns. One strain gave a contradictory pattern and was designated as untypeable by the NanoChip assay. Eighty-eight strains gave hybridization patterns that allowed a correct subtype designation to be made by the NanoChip assay compared to either the sequence or the heteroduplex mobility assay (HMA)-determined subtypes. Thirteen strains that reacted with the subtype A probe (SA2) were incorrectly assigned to subtype A, or to one of the related circulating recombinant types (AE or AG), on the basis of reactions with probe SAE1 or SAG1. The results indicate that these oligonucleotides have relatively low specificities. The probe subtypes of three strains matched the subtypes determined for the gag and pol genes but not the env gene, suggesting that a recombination event may have occurred within the env gene. Overall, the NanoChip assay gave results comparable to those for HMA and sequencing and provides a convenient and cost-effective means by which to subtype HIV-1.
INTRODUCTION
Human immunodeficiency virus type 1 (HIV-1) sequences are highly diverse and exhibit considerable genetic variation both between and within viral strains. This is an important complicating factor in the accuracy of diagnostic assays (4) and may alter the susceptibility to antiretroviral drugs (2, 10). Such variability also provides a means of monitoring the epidemiology of the pandemic, can indicate the origin of an infection, and provides a means of tracking the spread of infection between exposure groups (15, 26, 29). Given the potential public health impact of HIV-1 genetic diversity, the monitoring of subtype trends remains an essential component of HIV surveillance (24). Consequently, systems for subdividing HIV-1 strains on the basis of sequences are of great interest and importance for diagnosis and in epidemiological studies of the AIDS pandemic. Phylogenetic analysis has allowed the classification of HIV-1 into three groups: the major (M) group, the outlier (O) group, and the non-M, non-O (N) group. Group M is responsible for the majority of infections in the pandemic; the other two groups are highly diverse and less prevalent (18). Group M isolates can be subdivided into nine subtypes (A to D, F to H, J, and K) and a number of circulating recombinant forms (CRFs), which have identical mosaic genomes and are assumed to have arisen by recombination between different subtypes. The diversity and widespread nature of the pandemic have contributed to an increase in the subtyping of HIV-1 isolates; hence, more rapid and reliable methods are required.
Several methods for HIV subtyping have been described. DNA sequencing provides the most information but is time-consuming and laborious. The heteroduplex mobility assay (HMA) can be used to compare a test sample against many references, enabling visualization of quasispecies, but is also laborious and technically complex. Recently, Plantier and colleagues (17) have reported a DNA enzyme immunoassay (DEIA) for typing HIV-1 using subtype-specific oligonucleotide probes. The assay uses 20 to 30 base probes which hybridize to a 250-bp PCR amplicon derived from the first codons of the env gene, i.e., between codon 1 and the middle of the conserved region C1 of gp120. The probes were designed to have similar melting temperatures (Tms) and at least 20% sequence differences with other types. It was not possible to fulfill the latter criterion for strains of types B and D or for types A, CRF_AE, and CRF_AG, and additional probes were therefore selected. Probe SBD3 identifies strains of both types B and D. These strains can then be identified with type-specific probes. Probe SA2 hybridizes to strains of types A, CRF_AE, and CRF_AG. Subsequently, CRF_AE and CRF_AG can be identified with subtype-specific probes. Biotinylated probes were immobilized in microtiter tray wells and hybridized to the denatured env amplicon from test samples. Hybridization was detected using an antibody specific for double-stranded DNA. Three different sets of probes were used, corresponding to the subtypes commonly found in the geographic region that was the source of the HIV-1 strains examined. As many as six probes, each in a separate well of the microtiter plate, were hybridized to amplicons concurrently. The DEIA method was found to be both accurate and reliable compared to HMA or sequencing, allowing correct identification of 84% of the test samples.
The NanoChip system (Nanogen Inc., San Diego, Calif.) provides a convenient platform for oligonucleotide hybridization assays. The system is comprised of the loader, which applies the samples to specific test sites on the microarray, and a two-channel fluorescent reader, which can resolve measurements of fluorescence at the emission peaks of two different fluors. Both components can perform repetitive liquid handling operations automatically so that the hands-on technical effort is minimized. In addition, both instruments can apply a voltage selectively at one or more of the pads. The NanoChip cartridges consist of 10-by-10 arrays of microelectrode test sites coated with a gel permeation layer that contains streptavidin. The array format allows up to 100 separate analyses to be done in parallel, one at each pad. Each test site can be independently electronically activated using the NanoChip Molecular Biology Workstation.
Here we describe the use of the DEIA probes to type HIV-1 env gene amplicons using the NanoChip system (11-14, 19, 21, 25, 27). The use of previously described DEIA probes for genotyping on the NanoChip system allowed the identification of 79% of the test samples. Electronic microarrays thus offer an alternative methodology for HIV subtyping.
MATERIALS AND METHODS
Strains. Residual serum specimens, obtained from heterosexual individuals attending sexually transmitted infection clinics participating in a national Unlinked Anonymous HIV Prevalence Monitoring Program, were submitted to the Sexually Transmitted and Blood-borne Virus Laboratory for subtyping. Subtypes were determined by HMA in the gag and env gene regions, with sequencing where necessary, as previously described (3, 5, 9, 23). All pol subtypes were assigned by sequencing, as previously described (3). One hundred twelve samples were included in the study.
Nucleic acid extraction and PCR. HIV-1 viral RNA was extracted from residual serum by the Boom method (6), and a mixed specific-primer reverse transcription reaction was performed, as described previously (3). First- and second-round PCR conditions were as previously described (3), with the exception that the first-round forward env primer ED3 was used in place of ED5, and the env primers ES4 (5'-AAG AGC AGA AGA YAG TGG CAA TG-3') and ES5 (5'-TAC ACA GGC ATG TGT RGC CC-3') were used in the second-round PCR, as described by Plantier and colleagues (17). Primer ES5 was 5' end modified by addition of a biotin residue. All native and modified oligonucleotides used in this study were supplied by MWG Ltd., Milton Keynes, United Kingdom. The biotinylated amplicon was purified and desalted on Millipore MultiScreen 96 PCR cleanup plates [Millipore (UK) Ltd., Watford, United Kingdom] and reconstituted in 50 mM histidine (200 μl).
Loading of NanoChip cartridges for HIV-1 subtyping. NanoChip cartridges were prepared for loading by filling with 0.3 M NaOH for 5 min and washing with deionized water. The desalted biotinylated amplicons were resuspended in 50 mM histidine buffer (60 μl) and placed into wells of a Nunc 384 square-well polypropylene microtiter plate (Scientific Laboratory Supplies Ltd., Nottingham, United Kingdom) which was placed into the NanoChip loader. Each amplicon was loaded sequentially onto duplicate nonadjacent NanoChip test sites for 3 min at a nominal 2 V. Negative-control pads were addressed with the histidine buffer. Acceptable levels of current at each electrode were in the range of 300 to 600 nA. Following loading the nonbiotinylated strand was removed by filling the NanoChip with 0.3 M NaOH for 5 min. The NanoChip was then washed through with hybridization buffer (50 mM NaPO4 [pH 7.4], 0.5 M NaCl). For storage, loaded NanoChip cartridges were filled with deionized water and placed in a sealed bag at 4°C.
NanoChip HIV-1 subtyping. Subtype-specific oligonucleotide probes (17) were hybridized to the HIV-1 env gene amplicons immobilized on duplicate test sites on the NanoChip cartridge. Each probe was either Cy3 or Cy5 end labeled, as listed in Table 1. Stock solutions (100 μM) of one Cy3- and one Cy5-labeled probe were diluted to final concentrations of 2 μM in hybridization buffer, and five probe pairs were hybridized sequentially to the cartridge (SB2-SBD3, SA2-SAG1, SC1-SAE1, SF6-SG1, and SB2-SD1.1). Fifty microliters of the probe solution was introduced into the NanoChip cartridge. The NanoChip cartridge was then inserted into the reader. Hybridization was performed by heating the cartridge to 45°C for 2 min, cooling it to 40°C for 2 min, and then cooling it to 35°C for a further 2 min. Unbound probe was removed by three washing cycles with hybridization buffer at 25°C (each wash cycle comprised 500 μl at 75 μl/s). Fluorescent signal remaining at the electrodes was detected in the red (Cy5) and green (Cy3) channels of the reader at 25°C. The temperature was then raised to 40 and then to 50°C, with three wash cycles at each temperature. The remaining fluorescent signal was measured after washing at each temperature. Pairs of probes were hybridized sequentially to the HIV-1 NanoChip microarray. Between hybridizations, the array was regenerated by being washed three times for 5 min with 0.3 M NaOH at ambient temperature and rescanned to ensure that the signal had returned to background. Subtypes were assigned by consideration of the pattern of hybridization of the env amplicon to all nine probes.
RESULTS
Each biotinylated ES4/5 amplicon was electronically addressed to two microarray test sites, where it was anchored via the interaction with the streptavidin in the gel permeation layer. The array was then hybridized to the nine fluorescent oligonucleotide probes described by Plantier and colleagues (17), and the patterns of reactivity were observed. The Tms of the probes were in the range of 55 to 74°C as determined by the nearest neighbor method assuming a local probe concentration of 50 nM (Table 1). The probe concentration is 2 μM during initial hybridization but is then reduced to very low levels during the wash steps. Under the high-salt conditions used for hybridization and washing steps, all of the probes are likely to form stable homoduplexes at up to 50°C. The NanoChip system was used to measure fluorescence retained on the chip after washing at 25, 40, and 50°C. The scan at 25°C was performed to indicate the level of hybridization at low stringency, and the higher-temperature scans were intended to measure changes in probe binding at closer to their Tms. For a subtyping call to be made, the fluorescent signal of the sample (40°C) had to be at least fivefold higher than the negative controls (pads mock addressed with histidine) for at least one probe. Figure 1 shows the hybridization values for five strains. A sample was considered positive for a particular subtype if the fluorescent signal decreased less than 50% between 25 and 50°C, indicating stable hybridization of the probe. When the fluorescent signal was less than fivefold higher than the negative control at 40°C or when the loss of fluorescent signal between 25 and 50°C was greater than 50%, the reaction was considered negative.
There was no apparent decrease in signal levels or in signal-to-noise ratios between successive rounds of hybridization. There was some variation in fluorescent signal obtained on replicate pads, but this was not greater than a twofold difference and was usually within 20% for negatives. Very occasionally, fluorescent particles adhering to the pads caused higher levels of variation, but these could be identified using the NanoChip instrument, and these measurements were repeated. There was sometimes considerable variation (up to two- to threefold) between the levels of fluorescent signal observed for replicate hybridizations, but this made only a small difference to the positive/negative ratio and did not affect the overall result.
Probe SD1.1 gave nonspecific results under the conditions used for the array, and its reactions were therefore disregarded. The 112 strains included in the study were assigned to type according to probe reactions as shown in Table 2. Table 2 was compiled by following a set of rules formulated by examination of the results of typing on the array for the first 40 strains as follows. (i) All HIV-1 types give positive hybridization signals with their corresponding probes. (ii) Strains of CRF_AE or CRF_AG always hybridize additionally to probe SA2. (iii) Type AG strains may cross-react with probe SAE1. (iv) Probe SF6 reactions are assumed to be cross-reactions when another probe gives a positive reaction (except for SAE1 or SAG1 when SA2 is negative). (v) Probe SAE1 or SAG1 reactions are assumed to be cross-reactions when another probe gives a positive reaction (except when SA2 is positive). (vi) Type G amplicons may cross-react with probe SA2.
The results were compared with the env gene subtypes determined by HMA or sequencing and are summarized in Table 3. The subtypes determined by HMA and/or sequencing and probe hybridization on the NanoChip system were identical for 88 strains (79%). One strain gave an inconsistent pattern of probe hybridization by the NanoChip system and was designated as nontypeable. The remaining 23 strains were assigned different subtypes by different methods.
Thirteen of the 23 discrepant strains were identified as type A or CRF_AE or CRF_AG by HMA and/or sequencing. When typed on NanoChip microarrays, all 13 hybridized to the subtype A probe (SA2); however, hybridization to the probes SAE1 and SAG1 did not match the HMA and/or sequencing subtype designation. This represents 32% of all SA2-reactive strains. In eight instances the mismatch involved an A-AE pair, and in the remaining five, it involved an A-AG pair. In one case the probe subtype (AG) was in agreement with the HMA or sequence subtype for gag and pol genes, respectively, but discordant with the env subtype as determined by HMA. The results are summarized in Table 3.
Three of the remaining 10 subtyping discrepancies can be best explained by recombination. In these cases, the probe subtype corresponded with the sequence subtype for both the gag and pol genes but not for the env gene. The region of the env gene used in the HMA and sequencing assays did not include the probed amplicon. It is possible, therefore, that recombination events affecting the first codons of env may explain these observed discrepancies. The HMA and/or sequencing subtypes of the remaining seven samples were C (n = 3), B (n = 2), and D and G (one sample each). There was no consistent pattern of mismatching between subtypes. The results for the samples with mismatching subtypes are shown in Table 4.
The overall accuracy of the method was 79%, but this increased to 91% if no attempt was made to distinguish between type A and the CRFs. The corresponding false-positive rates were therefore 21 and 9%. Coverage of the method was over 99%. False-negative and false-positive rates for each probe are given in Table 5.
Overall, the data suggest that the use of these probes on the NanoChip system may provide a promising alternative method for reliable, high-throughput subtyping of HIV-1 samples obtained from heterosexuals in England and Wales.
DISCUSSION
Two different assay designs were considered for the NanoChip HIV-1 typing system. The first was to immobilize the probes at the test sites, with each probe occupying a different subset of sites. Since 100 sites are available, each of the nine probes could be directed to 11 test sites, leaving one control pad. The chip could then be hybridized with up to 22 different amplicons on the loader. This design requires the sequential introduction of differentially labeled amplicon pairs to the cartridge, with a different set of nine sites electronically activated at each step. The electrophoretic concentration at activated test sites would enable rapid hybridization of the amplicon to the immobilized probe (i.e., active hybridization). Once all amplicons are hybridized, the array would be transferred to the reader for washing and reading at the desired temperatures. This method is analogous to the DEIA method previously used with this probe set (17). The major disadvantage of this approach is that the two strands of the amplicon must be separated to allow hybridization to the probe. Since there are no practical means of maintaining the amplicons in denatured form on the NanoChip loader, it is best to apply only the strand that is complementary to the probes. Single-strand amplicons can be produced by asymmetric PCR, but this may compromise the sensitivity of the assay. The second method, and the one that was actually implemented for the study, involves electronic addressing of biotinylated amplicons to test sites, using the NanoChip loader. Each probe pair, one Cy3 and one Cy5 labeled, was then passively hybridized to the amplicons on the microarray. Since it is not yet possible to automate the addition of different hybridization solutions to the cartridge on the NanoChip reader, the pairing of probes in this way reduces hands-on time for each experiment. Subsequently, thermal discrimination and fluorescent signal detection were performed on the reader. In this format, a single cartridge could be used to type a total of 49 specimens at one time, so that the assay is relatively inexpensive compared with HMA or sequencing, with an approximate materials cost per strain of $4.
Another consideration in assay design was selection of the probes. Several previous studies using the NanoChip system have described probes that have estimated Tms in the range of 30 to 50°C (12, 19, 20, 27). The probes for typing HIV-1 by DEIA (17) all have calculated Tms greater than 55°C. Trimming terminal bases might have reduced the Tms of the probes, possibly resulting in higher specificities and in elimination of many cross-reactions observed at the low wash temperatures. Although shortened probes may have theoretical advantages for HIV-1 typing on NanoChip microarrays, the full-length sequences were used for this work. The main reason for this choice was to enable comparison with the previous work (17); however, it may also be argued that probes with high Tms are required to maintain overall strain coverage in this application, due to the highly variable nature of the target sequence. If it had been desirable to use more stringent conditions, it would have been possible to reduce the salt concentration of the wash buffer. Optimization of the hybridization stringency for probe SD1.1 on the NanoChip platform might have resulted in it showing better specificity. Probe SD1.1 has the highest calculated Tm of the oligonucleotides in this probe set, and this may explain its poor performance under the selected assay conditions. It should be noted that in the DEIA (17) probe SD1.1 was used only for analysis of strains that reacted with probe SBD3. However, probe SBD3 was adequate in identifying type D strains. Several significant cross-reactions were noted for other probes included in the set. The SF6 probe was found to cross-react with one CR_AE strain and 40% of the type B strains. In the same way, probe SA2 cross-reacted with a single type G strain and the probes for the circulating recombinant types cross-reacted with a proportion of the type F and G strains (Table 2). These cross-reactions are caused by sequence similarities between the types at the probe sites.
As noted by Plantier and colleagues (17), typing based on probes is inherently less reliable and less informative than that based on sequencing. This is inevitable, because the sequences sampled by the probes are relatively short, and recombinant forms of HIV-1 arise frequently. Nevertheless, development of rapid and simple methods for subtyping will be valuable for screening analyses. The results presented here demonstrate that DEIA and microarray-based subtyping show comparable accuracy compared to HMA or sequencing, with 79% of samples correctly subtyped on microarrays, versus 84% by DEIA. Experience in this laboratory has previously shown that approximately 90% of samples can be assigned a subtype by using HMA alone. Given the possibility of recombination between HIV-1 strains, the result of any assay that assigns a subtype based on a single gene region may not be applicable to other gene regions. The difficulty of identifying recombinant viruses is a problem common to all methods other than the comparison of full genome sequences. Increased accuracy may thus be achieved by incorporation of additional probes targeting other env gene regions and/or an alternative target gene(s). Such an approach would increase the likelihood of detecting recombinant viruses that are becoming more prevalent in Europe (3). However, it may not be feasible to design specific probes for each subtype based on the relatively conserved gag and pol genes.
The figure of 79% strains correctly subtyped on the array should be considered in the context that not all strains were tested as unknowns. The raw data for 40 strains were first used to develop a consistent set of rules for the interpretation of hybridization patterns, and then types were designated using this basis. There was a small bias in favor of correct identification for the unblinded samples, but this was not significant (2, P = 0.22). On the other hand, three of the strains were recombinant viruses that matched the probe type in the gag and pol genes but were incorrectly identified because their env genes were of different types.
Thirteen of the discrepant calls involved discrimination among subtypes A, CRF_AG, and CRF_AE. Sequencing and phylogenetic analysis might resolve some of these discrepancies in favor of the probe methods, since some subtype A samples may have been misclassified by HMA. No amplicon to detect either CRF_AG or CRF_AE was included in the gag or env HMA, so the discrepancies observed with the chip assay may be due to misclassification of these samples by HMA. Alternatively, it may be that the probes designed to detect the CRFs (SAE1 and SAG1) are unreliable when used in areas where these viral strains cocirculate with other HIV-1 subtypes, such as the heterosexual epidemic in England and Wales (3, 22).
While the likely accuracy of the probe system was similar for the NanoChip and DEIA platforms, the NanoChip assay has several advantages over the DEIA method (17). It is possible to distinguish between probe-amplicon duplexes that have different Tms on the chip system because the fluorescent endpoint is compatible with the collection of data at multiple temperature values. The number of manipulations and the time required to assess fluorescent signals at different temperatures are lower for the chip assay. This is due to the simple assay format, combined with the automated liquid handling provided by the NanoChip instrument. The design of the chip assay also enables the use of a single format for all specimens. Plantier and colleagues (17) used different probe sets for specimens from France, Cameroon, and Senegal in order reduce the number of test wells required for each sample; in contrast, multiple probe sets can be sequentially hybridized to immobilized amplicons on a single electronic microarray. In addition, some probes in each set were used hierarchically in the DEIA analyses, so that two steps were required to identify certain subtypes. These differences in the testing protocol were, at least in part, responsible for the relatively poor performance of the AE and AG probes in the microarray format.
Recent studies have shown the heterosexual HIV-1 epidemic in England and Wales to be highly diverse, with all group M subtypes and several CRFs having been detected (3, 22). This situation is mirrored in other Western European countries, where the diversity is thought to have arisen due to the importation of infections from areas with diverse HIV-1 epidemics (1, 3, 7, 8, 16, 22, 28). Such diversity may require additional probe optimization for increased accuracy in hybridization-based assay formats. Given the continued global expansion of HIV-1 infection, there is an increasing need to monitor the molecular diversity of the epidemic. The NanoChip assay has been shown to produce results comparable to the widely used HMA and has the potential to provide a relatively high-throughput, reliable, and cost-effective alternative means of subtyping HIV-1.
Present address: WW Epidemiology (Neurosciences), GlaxoSmithKline Research and Development, Harlow, United Kingdom.
REFERENCES
Alaeus, A., T. Leitner, K. Lidman, and J. Albert. 1997. Most HIV-1 genetic subtypes have entered Sweden. AIDS 11:199-202.
Apetrei, C., D. Descamps, G. Collin, I. Loussert-Ajaka, F. Damond, M. Duca, F. Simon, and F. Brun-Vezinet. 1998. Human immunodeficiency virus type 1 subtype F reverse transcriptase sequence and drug susceptibility. J. Virol. 72:3534-3538.
Barlow, K. L., I. D. Tatt, P. A. Cane, D. Pillay, and J. P. Clewley. 2001. Recombinant strains of HIV type 1 in the United Kingdom. AIDS Res. Hum. Retrovir. 17:467-474.
Barlow, K. L., J. H. Tosswill, J. V. Parry, and J. P. Clewley. 1997. Performance of the Amplicor human immunodeficiency virus type 1 PCR and analysis of specimens with false-negative results. J. Clin. Microbiol. 35:2846-2853.
Belda, F. J., K. L. Barlow, and J. P. Clewley. 1997. Subtyping HIV-1 by improved resolution of heteroduplexes on agarose gels. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 16:218-219.
Boom, R., C. J. Sol, M. M. Salimans, C. L. Jansen, P. M. Wertheim-van Dillen, and J. van der Noordaa. 1990. Rapid and simple method for purification of nucleic acids. J. Clin. Microbiol. 28:495-503.
Clewley, J. P., C. Arnold, K. L. Barlow, P. R. Grant, and J. V. Parry. 1996. Diverse HIV-1 genetic subtypes in UK. Lancet 347:1487.
Couturier, E., F. Damond, P. Roques, H. Fleury, F. Barin, J. B. Brunet, F. Brun-Vezinet, F. Simon, et al. 2000. HIV-1 diversity in France, 1996-1998. AIDS 14:289-296.
Delwart, E. L., B. Herring, A. G. Rodrigo, and J. I. Mullins. 1995. Genetic subtyping of human immunodeficiency virus using a heteroduplex mobility assay. PCR Methods Appl. 4:S202-S216.
Descamps, D., C. Apetrei, G. Collin, F. Damond, F. Simon, and F. Brun-Vezinet. 1998. Naturally occurring decreased susceptibility of HIV-1 subtype G to protease inhibitors. AIDS 12:1109-1111.
Edman, C. F., P. Mehta, R. Press, C. A. Spargo, G. T. Walker, and M. Nerenberg. 2000. Pathogen analysis and genetic predisposition testing using microelectronic arrays and isothermal amplification. J. Investig. Med. 48:93-101.
Erali, M., B. Schmidt, E. Lyon, and C. Wittwer. 2003. Evaluation of electronic microarrays for genotyping factor V, factor II, and MTHFR. Clin Chem. 49:732-739.
Ewalt, K. L., R. W. Haigis, R. Rooney, D. Ackley, and M. Krihak. 2001. Detection of biological toxins on an active electronic microchip. Anal. Biochem. 289:162-172.
Heller, M. J., A. H. Forster, and E. Tu. 2000. Active microelectronic chip devices which utilize controlled electrophoretic fields for multiplex DNA hybridization and other genomic applications. Electrophoresis 21:157-164.
Kunanusont, C., H. M. Foy, J. K. Kreiss, S. Rerks-Ngarm, P. Phanuphak, S. Raktham, C. P. Pau, and N. L. Young. 1995. HIV-1 subtypes and male-to-female transmission in Thailand. Lancet 345:1078-1083.
Perez-Alvarez, L., M. T. Cuevas, M. L. Villahermosa, J. D. Pedreira, N. Manjon, I. Herrero, S. Lopez-Calvo, E. Delgado, E. V. de Parga, L. Medrano, M. M. Thomson, J. A. Taboada, and R. Najera. 2001. Prevalence of drug resistance mutations in B, non-B subtypes, and recombinant forms of human immunodeficiency virus type 1 in infected individuals in Spain (Galicia). J. Hum. Virol. 4:35-38.
Plantier, J. C., L. Vergne, F. Damond, S. MBoup, E. MPoudi-NGole, L. Buzelay, I. Farfara, D. Brand, M. Peeters, F. Brun-Vezinet, E. Delaporte, and F. Barin. 2002. Development and evaluation of a DNA enzyme immunoassay method for env genotyping of subtypes A through G of human immunodeficiency virus type 1 group M, with discrimination of the circulating recombinant forms CRF01_AE and CRF02_AG. J. Clin. Microbiol. 40:1010-1022.
Robertson, D. L., J. P. Anderson, J. A. Bradac, J. K. Carr, B. Foley, R. K. Funkhouser, F. Gao, B. H. Hahn, M. L. Kalish, C. Kuiken, G. H. Learn, T. Leitner, F. McCutchan, S. Osmanov, M. Peeters, D. Pieniazek, M. Salminen, P. M. Sharp, S. Wolinsky, and B. Korber. 2000. HIV-1 nomenclature proposal. Science 288:55-56.
Santacroce, R., A. Ratti, F. Caroli, B. Foglieni, A. Ferraris, L. Cremonesi, M. Margaglione, M. Seri, R. Ravazzolo, G. Restagno, B. Dallapiccola, E. Rappaport, E. S. Pollak, S. Surrey, M. Ferrari, and P. Fortina. 2002. Analysis of clinically relevant single-nucleotide polymorphisms by use of microelectronic array technology. Clin. Chem. 48:2124-2130.
Schrijver, I., M. J. Lay, and J. L. Zehnder. 2003. Diagnostic single nucleotide polymorphism analysis of factor V Leiden and prothrombin 20210G > A. A comparison of the Nanogen Electronic Microarray with restriction enzyme digestion and the Roche LightCycler. Am. J. Clin. Pathol. 119:490-496.
Sosnowski, R., M. J. Heller, E. Tu, A. H. Forster, and R. Radtkey. 2002. Active microelectronic array system for DNA hybridization, genotyping and pharmacogenomic applications. Psychiatr. Genet. 12:181-192.
Tatt, I. D., K. L. Barlow, J. P. Clewley, O. N. Gill, and J. V. Parry. 2004. Surveillance of HIV-1 subtypes among heterosexuals in England and Wales, 1997-2000. J. Acquir. Immune Defic. Syndr. 36:1092-1099.
Tatt, I. D., K. L. Barlow, and J. P. Clewley. 2000. A gag gene heteroduplex mobility assay for subtyping HIV-1. J. Virol. Methods 87:41-51.
Tatt, I. D., K. L. Barlow, A. Nicoll, and J. P. Clewley. 2001. The public health significance of HIV-1 subtypes. AIDS 15:S59-S71.
Thistlethwaite, W. A., L. M. Moses, K. C. Hoffbuhr, J. M. Devaney, and E. P. Hoffman. 2003. Rapid genotyping of common MeCP2 mutations with an electronic DNA microchip using serial differential hybridization. J. Mol. Diagn. 5:121-126.
Trask, S. A., C. A. Derdeyn, U. Fideli, Y. Chen, S. Meleth, F. Kasolo, R. Musonda, E. Hunter, F. Gao, S. Allen, and B. H. Hahn. 2002. Molecular epidemiology of human immunodeficiency virus type 1 transmission in a heterosexual cohort of discordant couples in Zambia. J. Virol. 76:397-405.
Westin, L., C. Miller, D. Vollmer, D. Canter, R. Radtkey, M. Nerenberg, and J. P. O'Connell. 2001. Antimicrobial resistance and bacterial identification utilizing a microelectronic chip array. J. Clin. Microbiol. 39:1097-1104.
Yirrell, D. L., D. J. Goldberg, J. Whitelaw, C. McSharry, F. Raeside, and G. Codere. 1999. Viral subtype and heterosexual acquisition of HIV infections diagnosed in Scotland. Sex. Transm. Infect. 75:392-395.
Yirrell, D. L., P. Robertson, D. J. Goldberg, J. McMenamin, S. Cameron, and A. J. Leigh Brown. 1997. Molecular investigation into outbreak of HIV in a Scottish prison. BMJ 314:1446-1450.(N. A. Saunders, S. Alexan)