|Home | About | Journals | Submit | Contact Us | Français|
As antiretroviral therapy (ART) is scaled up in resource-limited countries, surveillance for HIV drug resistance (DR) is vital to ensure sustained effectiveness of first-line ART. We have developed and applied a broadly sensitive dried-blood-spot (DBS)-based genotyping assay for surveillance of HIV-1 DR in international settings. In 2005 and 2006, 171 DBS samples were collected under field conditions from newly diagnosed HIV-1-infected individuals from Malawi (n = 58), Tanzania (n = 60), and China (n =53). In addition, 30 DBS and 40 plasma specimens collected from ART patients in China and Cameroon, respectively, were also tested. Of the 171 DBS analyzed at the protease and RT regions, 149 (87.1%) could be genotyped, including 49 (81.7%) from Tanzania, 47 (88.7%) from China, and 53 (91.4%) from Malawi. Among the 70 ART patient samples analyzed, 100% (30/30) of the Chinese DBS and 90% (36/40) of the Cameroonian plasma specimens were genotyped, including 8 samples with a viral load of <400 copies/ml. The results of phylogenetic analyses indicated that the subtype, circulating recombinant form (CRF), and unique recombinant form (URF) distribution was as follows: 73 strains were subtype C (34%), 37 were subtype B (17.2%), 24 each were CRF01_AE or CRF02_AG (11.2% each), 22 were subtype A1 (10.2%), and 9 were unclassifiable (UC) (4.2%). The remaining samples were minor strains comprised of 6 that were CRF07_BC (2.8%), 5 that were CRF10_CD (2.3%), 3 each that were URF_A1C and CRF08_BC (1.4%), 2 each that were G, URF_BC, and URF_D/UC (0.9%), and 1 each that were subtype F1, subtype F2, and URF_A1D (0.5%). Our results indicate that this broadly sensitive genotyping assay can be used to genotype DBS collected from areas with diverse HIV-1 group M subtypes and CRFs. Thus, the assay is likely to become a useful screening tool in the global resistance surveillance and monitoring of HIV-1 where multiple subtypes and CRFs are found.
Filter papers have been useful for the collection, storage, and testing of blood specimens in the diagnostic screening of metabolic and inherited disorders in newborn babies in the United States for many years (13, 14). More recently, many resource-limited countries have been using them for HIV-related molecular assays, including early infant diagnosis using Roche Amplicor HIV-1 DNA PCR testing (supported by the U.S. President's Emergency Plan for AIDS Relief [PEPFAR]) (31), real-time PCR-based HIV-1 diagnosis (22, 27) and viral load measurement (6, 17, 18, 20, 26), and HIV-1 drug resistance (DR) genotyping (5, 7, 8, 12, 16, 19, 23-25, 28, 30, 35-37).
Dried blood spots (DBS) offer several advantages over conventional plasma or serum for sample collection and storage. First, DBS circumvent the need for phlebotomy and reduce the risk of needle-stick-related HIV exposure when they are collected using finger or heel stick. Second, they do not require cold-chain transportation of specimens from collection sites to the testing laboratories; thus, they can be transported using standard postal services, resulting in the reduction of cost for storage and transportation when dried and processed appropriately. Overall, DBS provide resource-limited countries with opportunities for sustainable laboratory services and surveillance programs for HIV diagnosis and DR genotyping. Very few studies have been conducted using DBS collected under real field conditions. Furthermore, the DR genotyping assays used were not evaluated for multiple HIV-1 subtypes and circulating recombinant forms (CRFs) (5, 16, 19, 23, 25, 37).
The aim of the current study was to determine the performance of a broadly sensitive genotyping assay for detecting multiple HIV-1 group M subtypes and CRFs that cocirculate in different countries, using DBS collected under field conditions.
In 2005 and 2006, 171 DBS were collected under field conditions from individuals from Malawi (n = 58), Tanzania (n = 60), and China (n = 53) who were newly diagnosed with HIV-1 infections. In addition, specimens from antiretroviral therapy (ART) patients, including 30 DBS collected from Chinese patients in 2008 and 40 plasma specimens collected from HIV-1-infected Cameroon women in 2007, were also analyzed. To determine the specificity of the assay, 42 HIV-1-negative DBS specimens collected from women attending an antenatal clinic (ANC) in Tanzania were tested. All study protocols were approved by local and CDC institutional review boards. Anonymous genotyping was performed at the International Laboratory Branch, Division of Global AIDS, Centers for Disease Control and Prevention (CDC), Atlanta, GA. The detailed demographic and clinical information on the participants from Malawi, Tanzania, and China who were newly diagnosed with HIV-1 infection has been described elsewhere (19, 30, 36).
Dried-blood-spot specimens were made by spotting 100 μl of whole blood collected into a Vacutainer containing K3-EDTA onto Whatman 903 paper (Whatman, Inc., Piscataway, NJ) and drying overnight at ambient temperature. After drying, each DBS card was wrapped with a folded sheet of glassine paper, and 5 to 10 glassine paper-wrapped cards were packaged into a low-gas-permeable, zip closure plastic bag containing 8 to 10 desiccant bags and one humidity indicator (Fisher Scientific, Inc., Pittsburgh, PA). The DBS were stored at −70°C or −20°C before being shipped to Atlanta and were stored at −70°C in Atlanta until testing. The quality of the plasma samples received from Cameroon might have been compromised, since the samples were received under ambient-temperature conditions.
HIV serology tests for DBS specimens collected from newly diagnosed HIV-1-infected patients from Tanzania, Malawi, and China were repeated using the Genetic System HIV-1/HIV-2 peptide EIA kit (Bio-Rad Laboratories, Redmond, WA) and the Cambridge Biotech HIV-1 Western blot kit (Calypte Biomedical Corp., Rockville, MD). The HIV-1 viral loads in Cameroon plasma specimens were determined with an Amplicor HIV-1 monitor, version 1.5, which has a detection limit of 400 copies/ml (Roche Diagnostics, Indianapolis, IN). For samples collected from Chinese ART patients, the plasma viral loads were measured with the NASBA HIV-1 quantitation system, which has a detection limit of 50 copies/ml (Organon Teknika Boxtel, Netherlands), at Shandong Provincial CDC, Shandong, China. Due to the lack of available plasma specimens to pair with DBS collected from the newly HIV-1 diagnosed patients from Malawi, Tanzania, and China, the plasma viral loads in these patients were not determined.
HIV-1 viral RNA was extracted from 140-μl amounts of the Cameroon plasma samples using a QIAamp viral RNA kit, following the manufacturer's protocol (Qiagen, Valencia, CA). Total nucleic acids (TNA) were extracted from the DBS using a modified NucliSens silicon-based extraction procedure (bioMerieux, Inc., Durham, NC). In brief, one or two whole spots containing 100 μl of whole blood were cut with scissors and added to 9 ml of NucliSens lysis buffer. Following a 2-hour incubation at room temperature under gentle rotation, the tubes were centrifuged at 250 × g for 5 min and supernatant was transferred to a new 15-ml conical tube. TNA were extracted using NucliSens isolation reagents according to the manufacturer's procedure, resuspended in 50 μl of elution buffer, and stored at −70°C before use.
Primers were designed and modified (25) with the intention to amplify all HIV-1 group M subtypes and CRFs of the pol gene region associated with DR in the protease and reverse transcriptase (RT) regions of the HIV-1 genome. The outer primers for RT-PCR were Prt-F1 (forward, 5′-CCT CAA ATC ACT CTT TGG CAR CG, nucleotides [nt] 2253 to 2275 based on HXBII) and RT-R1 (reverse, 5′-ATC CCT GCA TAA ATC TGA CTT GC, nt 3370 to 3348), and the nested-PCR primers were Prt-F2 (forward, 5′-CTT TGG CAA CGA CCC CTY GTC WCA, nt 2265 to 2288) and RT-R2 (reverse, 5′-CTT CTG TAT GTC ATT GAC AGT CC, nt 3326 to 3304). For RT-PCR, 10 μl of RNA or TNA extracts was used with primers Prt-F1 and RT-R1 (4 μM each/50 μl of RT-PCR mixture) and the SuperScript III one-step RT-PCR system with Platinum Taq DNA polymerase high fidelity, following the manufacturer's protocol (Invitrogen, Carlsbad, CA). For nested PCR, 4 to 8 μl of each of the RT-PCR products was added to 100 μl of the PCR mixture, containing 0.12 μM (each) Prt-F2 and RT-R2 primers, 1× GeneAmp gold buffer II, 2 mM MgCl2, and 2.5 U of AmpliTaq gold LD DNA polymerase (Applied Biosystems, Foster City, CA). After initial denaturation at 94°C for 4 min, 40 cycles of PCR were performed in a GeneAmp 9700 thermocycler with PCR conditions of 94°C for 15 s, 55°C for 20 s, and 72°C for 2 min following an extension at 72°C for 10 min. Nested PCR products of 1,062 bp were confirmed by agarose gel electrophoresis, and purified products were used for sequencing.
The sensitivity of the genotyping primer set was determined using DBS and plasma specimens collected from ART patients with known plasma viral loads. To determine the specificity of the primer set, 42 HIV-negative DBS collected from pregnant women at a Tanzanian ANC were tested.
Sequencing analysis of HIV-1 pol was performed using either an ABI 3100 or a 3730 DNA sequencer (Applied Biosystems, Foster City, CA). ChromasPro, version 1.42 (Technelysium Pty. Ltd., Tewantin, Australia), was used for editing sequences and producing consensus sequences. The protease and RT genotypes were interpreted using the Stanford Genotypic Resistance Interpretation Algorithm for specimens collected from ART patients and the Calibrated Population Resistance (CPR) tool for specimens collected from ART-naïve patients (http://hivdb.stanford.edu/index.html), as well as the WHO list for determination of transmitted HIV-1 DR (3).
Phylogenetic and molecular evolutionary analyses were conducted using MEGA software, version 4 (32), and BioEdit (15) along with reference sequences from the HIV sequence database. After all gaps were trimmed, phylogenetic trees were constructed by the neighbor-joining method included in MEGA 4. The reliability of the tree topology was tested with 1,000 bootstrap analyses, and bootstrap values of ≥70% were considered significant.
The newly obtained sequences were submitted to GenBank with accession numbers HQ009559 to HQ009690.
The HIV status of the DBS specimens collected from patients newly diagnosed with HIV-1 was retested and confirmed. The median viral loads were 11,605 copies/ml (range, <400 to 367,875 copies/ml) for the 40 plasma samples from Cameroon and 4,620 copies/ml (range, 290 to 72,000 copies/ml) for DBS samples from China (Tables (Tables11 and and22).
First, we evaluated the assay's sensitivity with 40 plasma specimens from ART patients from Cameroon, of which 36 (90%) were successfully genotyped. Of the 4 specimens that were not genotyped, two had viral loads of 1,341 and 1,391 copies/ml and the remaining two had viral load of 15,287 and 119,260 copies/ml. The inability to genotype these four specimens could likely be attributed to the compromised quality of the specimens (see Materials and Methods for details). However, it could also be a result of sequencing variations at the primer locations, since we were able to genotype two of the four specimens (2007693335 and 2007693340) (Table (Table1)1) with the Trugene HIV-1 genotyping system (Siemans Medical Solutions Diagnostics, Tarrytown, NY) after the broadly sensitive genotyping assay had failed. Sequence comparison analysis revealed that there are two nucleotide variations at positions 4 and 9 for the nested-PCR forward primer (Prt-F2) at the 3′ side. However, it is worth noting that the assay was able to genotype all 6 specimens with a viral load of <1,000 copies/ml (Table (Table11).
Next, we measured the assay's sensitivity with DBS specimens collected from ART patients from Shandong Province, China. All 30 DBS specimens tested were genotyped, although some of the DBS specimens had viral loads below 1,000 copies/ml of plasma (Table (Table22).
Genotyping analyses using the Stanford Genotypic Resistance Interpretation Algorithm (http://hivdb6.stanford.edu/asi/deployed/HIVdb.html) indicated that 6 of the 38 Cameroon patients and 22 of the 30 Chinese ART patients had mutations associated with resistance to protease inhibitors (PI), nucleoside RT inhibitors (NRTI), and nonnucleoside RT inhibitors (NNRTI) (Table (Table33).
The phylogenetic analyses revealed that the predominant strains from Cameroon samples were CRF02_AG (24/36, 66.7%), followed by A1 and unclassifiable (UC) (4/36 of each, 11.1%), G (2/36, 5.6%), and F1 and F2 (1/36 of each, 2.8%). The subtype distributions from DBS specimens collected from China were 90% (27/30) B and 10% (3/30) CRF01_AE.
To evaluate the specificity of the genotyping assay, we tested 42 HIV-1-negative DBS collected from women attending an ANC in Tanzania. These specimens were confirmed to be HIV-1 negative with serologic tests at CDC, Atlanta, GA. Genotyping analysis revealed that they were all negative by RT-PCR using the broadly sensitive genotyping assay.
Surveillance of transmitted DR in resource-limited settings has become an urgent task with the scaling up of ART in these countries. The initial characterization of the new assay has revealed that this assay might be suitable for populations infected with non-B HIV-1 subtypes and CRFs. We received 171 DBS specimens from 3 countries where transmitted-DR surveys (19, 30, 36) were conducted using the WHO HIV DR Threshold Survey (HIVDR-TS) protocol (4). Of the 149 (149/171, 87.1%) successfully genotyped DBS samples, 49 (49/60, 81.7%) were from Tanzania, 47 (47/53, 88.7%) from China, and 53 (53/58, 91.4%) from Malawi. Genotyping analysis revealed that all the specimens had no transmitted-DR mutations, except that one specimen from China carried the K101E mutation to NNRTI based on the WHO list (3) and CPR criteria. Phylogenetic analyses revealed that the predominant subtype was C (49.0%), followed by CRF01_AE (14.1%), A1 (12.1%), B (6.7%), CRF07_BC (4.0%), CRF10_CD (3.4%), and UC (3.4%). The remaining strains were unique recombinant form A1C (URF_A1C), CRF08_BC, URF_D/UC, URF_BC, and URF_A1D, at ≤2% prevalence (Table (Table4).4). It is noticeable that all the HIV-1 strains from Malawi belonged to subtype C, while the strains from China and Tanzania had very diverse subtypes, CRFs, and URFs.
In this study, we report the development of a broadly sensitive RT-PCR-based genotyping assay for surveillance and monitoring of HIV-1 DR in the pol gene region. By applying this assay, we genotyped DBS from 3 countries for transmitted HIV-1 DR. This study confirms and extends previous reports that DBS are a useful sample collection device for HIV-1 DR genotyping analysis (5, 7, 8, 12, 16, 19, 23-25, 28, 30, 35-37). However, the current study provides additional support for the use of DBS for HIV DR genotyping in several ways. (i) Samples were collected under routine clinical conditions and transported to CDC Atlanta for genotyping. (ii) We genotyped multiple HIV-1 group M subtypes and CRFs in samples from China, Cameroon, Malawi, and Tanzania, reported in this study, and also from many other countries, including Botswana, Honduras, Ethiopia, Kenya, Nigeria, Thailand, and Zambia (data not shown). Together, these results indicated that this genotyping assay is broadly sensitive for all the major HIV-1 subtypes and CRFs. (iii) The assay is one-step RT-PCR-based and covers both the protease (amino acids [aa] 13 to 99) and RT (aa 1 to 251) regions, which results in the reduction of genotyping cost and the possibility of laboratory-related cross-contaminations because fewer hands-on manipulations are needed. Compared to commercially available genotyping assays, our assay reduces the cost for genotyping reagents by almost 50%. (iv) Lastly, the assay can be successful in genotyping at least 90% of samples collected under resource-limited field conditions and with a viral load of ≥400 copies/ml of plasma. However, the current 5′ nested-PCR primer (Prt-F2) for the protease region starting at codon 13 is a few amino acids shorter than those in commonly used genotyping assays, such as the Trugene HIV-1 genotyping system (Siemans Medical Solutions Diagnostics, Tarrytown, NY) and the Viroseq HIV-1 genotyping system (Celera Diagnostics, Alameda, CA), and does not cover all the amino acid positions associated with DR against protease inhibitors. Currently, it will not affect the genotyping results, as many countries are only using the first-line regimens recommended by WHO (34), which do not include protease inhibitors. We are currently evaluating a new primer set to overcome this issue.
The unusually high viral diversity of HIV-1 poses problems historically for HIV diagnosis using nucleic acid-based tests (1, 9, 21). The two commercially available genotyping assays, the Viroseq HIV-1 genotyping system (Celera, Alameda CA) and the TruGene HIV-1 genotyping system (Siemans Medical Solutions Diagnostics, Tarrytown, NY), were designed and approved by the FDA for HIV-1 subtype B only. Although studies have indicated that these assays can be used for other HIV-1 group M subtypes, the reported sensitivities vary (2, 10, 11, 29, 33). The assay we report here and the one reported recently by Buckton et al. (7) are the only assays that have directly addressed subtype-genotyping-efficiency issues. In the analysis of Buckton et al., using plasmid clones representing major HIV-1 group M subtypes, they found that the genotyping assay was less efficacious for subtypes A2, CRF01_AE, CRF02_AG, F, and H in the RT region and for CRF02_AG, F, and H in the protease region, respectively. We have genotyped DBS and plasma specimens from more countries than reported here using our broadly sensitive genotyping assay. We found that the genotyping efficacies ranged from 90% to 100% for plasma samples (data not shown) and from 82% to 100% for DBS samples representing HIV-1 group M subtypes A1, A2, B, C, D, F1, F2, and G, circulating recombinant forms CRF01_AE, CRF02_AG, CRF07_BC, CRF08_BC, and CRF10_CD, and many unique recombinant forms (Tables (Tables1,1, ,2,2, and and44).
In summary, we have developed a broadly sensitive genotyping assay that is restricted to the pol gene region. Using this assay, we were able to genotype diverse HIV-1 group M viral strains from DBS samples collected under field conditions in several countries where multiple subtypes and CRFs are cocirculating. The assay is likely to become a useful tool for HIV DR surveillance in many developing countries.
Jing Zhang was a recipient of an 2008-2009 International Emerging Infectious Diseases (IEID) fellowship sponsored by the Association of Public Health Laboratories (APHL) and the CDC.
We express our sincere appreciation and thanks to all the field staff and participants for their support for the study and to the staff of the HIV Incidence and Serology Team of the International Laboratory Branch for performing HIV-1 confirmation tests at the CDC.
The use of trade names is for identification only and does not constitute endorsement by the U.S. Department of Health and Human Services, the Public Health Service, or the Centers for Disease Control and Prevention.
Published ahead of print on 21 July 2010.