Tebit, Denis M PhD*; Ganame, Jean MD†; Sathiandee, Kanokporn MSc*; Nagabila, Youssouf MD†; Coulibaly, Boubacar PhD†; Krausslich, Hans-Georg MD*
Sub-Saharan Africa is the region most severely affected by HIV. It is home to 25.4 million (60%) of the 40 million people living with HIV. In 2004, approximately 3.1 million individuals in the region were newly infected, with women and children being most frequently affected.1 Burkina Faso is a West African country from which relatively little information on viral infections is available. This is indicated by the low number of reports on seroprevalence from this country and the fact that existing reports have been limited to the urban regions of Ouagadougou and Bobo Dioulasso, where HIV prevalence reaches up to 7%.2,3 Recently, the government of Burkina Faso set up a national program to prevent the transmission of HIV from mother to child (PMTCT). For logistic reasons, this program was implemented only in the major cities, with the exception of Nouna, a town in western Burkina Faso close to the boarder of Mali, which was included as pilot rural site (approximately 300 km from the capital, Ouagadougou). The Nouna PMTCT site has been operational since July 2003.
Sequence analysis classifies HIV into 2 types: HIV-1 and HIV-2. Type 1 is more widely distributed in the pandemic and has been divided into 3 groups; M (major), N (non-M, non-O), and O (outlier). Although these 3 groups have been well documented in Central Africa, HIV seems to be less diverse in West Africa.4-6 Group M viruses are divided into 9 subtypes (A-D, F-H, J, and K), 5 sub-subtypes (A1, A2, A3, F1, and F2), and 18 circulating recombinant forms (CRFs; CRF01-CRF16, CRF18m, and CRF19).7 CRFs arise through recombination events involving viruses from different clades and have been shown to circulate in epidemic proportions. The most dominant form of HIV-1 in West Africa and western Central Africa is CRF02_AG.4-6,8 HIV-2 (divided into subtypes A-G) is also prevalent in West Africa, but its prevalence has remained constant compared with HIV-1.9 To date, only a single study has been done concerning the genetic diversity of HIV in Burkina Faso.10 This study reported CRF06.cpx as the dominant HIV strain circulating in Ouagadougou, the capital city of Burkina Faso.
The high genetic variability of HIV and its potential for recombination are clearly important for escape from the immune system and also play a major role in the development of resistance against antiretroviral drugs. Various drug resistance mutations to protease, reverse transcriptase, and fusion inhibitors have been described for HIV-1 subtype B isolates, depending on the drug regimen used.11,12 Some of these mutations have been observed in drug-naive individuals as well, but their prevalence and potential for resistance development in non-B-type strains and recombinant forms of HIV-1 are currently largely unknown. Also, new drugs to treat HIV are continuously being introduced, but information on their efficacy in non-subtype B viruses in terms of resistance development is usually lacking.
Accordingly, studies with complete data sets, including the serologic types, genetic types and subtypes, CD4 counts, viral load, and drug resistance polymorphisms, are needed for Burkina Faso. Such studies are especially important before the introduction of antiretroviral therapy (ART). Here, we describe the serologic and molecular analysis of HIV from drug-naive pregnant women attending the Nouna PMTCT site and from drug-naive HIV-infected individuals attending the Dedougou Regional Hospital (DRH) located in Dedougou, a semiurban area some 50 km from Nouna.
MATERIALS AND METHODS
Subjects and Samples
Whole-blood samples were collected from HIV-positive pregnant women and other HIV-positive subjects attending the PMTCT site in Nouna as well as from HIV-positive individuals presenting at the DRH in Dedougou. The University of Heidelberg Ethics Committee, the Nouna Ethics Committee, and the Ministry of Health in Burkina Faso approved the study. Informed consent was obtained from all participants, who routinely underwent individual counseling before HIV testing. Samples were collected between July 2003 and October 2004. At the time of sampling, all subjects were reported to be drug naive.
All samples were centrifuged at a speed of 3000 rpm for 10 minutes, separated into plasma and buffy coats, and frozen until further analysis. An aliquot was used to screen for HIV-seropositive samples using a rapid assay, Determine (Abbott Diagnostics) and confirmed by a second rapid assay, Genie II (Bio-Rad), which distinguishes between HIV-1 and HIV-2. For quality control, we initially screened all samples using fourth-generation enzyme-linked immunoassays (ELISAs): Enzygnost HIV Integral (Dade Behring) or Abbott Murex Ab/Ag (Abbott Diagnostics). All positive samples were confirmed using the AxSYM automated ELISA (Abbott Diagnostics) and the Inno-Line Immuno Assay (LIA) HIV Confirmation test (Innogenetics) at the University of Heidelberg or at the Centre de Recherche en Santé de Nouna (CRSN) in Nouna.
CD4 Counts and Viral Load Determination
Whole blood less than 6 hours old was used to determine the CD4 cell counts of infected subjects using a FACScount fluorescence cytometer (Becton Dickinson) in Nouna. Quantification of viral RNA was carried out at the Virology Unit of the Center Hospitalier Universitaire Yalgado Ouedraogo (CHUYO) in Ouagadougou using the LcX HIV RNA Quantitative Probe system (Abbott Diagnostics) on fresh or frozen plasma. Both assays were done as recommended by the manufacturers.
DNA Isolation, Polymerase Chain Reaction, and Sequence Analysis
DNA was extracted from buffy coats using the Qiagen DNA extraction kit (Qiagen, Hilden, Germany) as recommended by the manufacturer. The following distinct regions of gag, pol, and env were amplified by nested polymerase chain reaction (PCR) assay: a 775-base pair (bp) gag region (complete matrix and 5′ capsid coding regions, positions 836-1611 in HxB2), a 752-bp pol region (complete protease [PR] and 5′ reverse transcriptase [RT] coding regions, positions 2241-2993 in HxB2), and a 675-bp region of env (V3-V5, positions 7003-7678 in HxB2). The primer pairs used for the nested PCR assay have been described earlier.6 New sets of pol and env primers were used to amplify pol fragments that could not be amplified by the first pol primer set or to amplify the gp41 region of env. These primers were as follows: pol first-round primers: BF1 forward, AGACACAGGAGCAGATGATACAGT (positions 2324-2347 in HxB2) and BF2 reverse, GAAAAATATGCATCCCCCACATC (positions 2877-2899 in HxB2); and pol second-round primers: BF3 forward, ATGATAGGGGGAATTGGAGG (positions 2388-2407 in HxB2) and BF4 reverse, AACCCAGCAGGATGTGGTAT (positions 2829-2848 in HxB2).
These primers amplify a 400-bp fragment within the PR and RT coding regions. The env gp41 primers used were Env5+ (TCAGACCTGGAGGAGGAGATATGA (positions 7627-7650 in HxB2) and GP41R1 (positions 8347-8375) in the first-round PCR assay and GP40F1 (positions 7789-7816 in HxB2) and GP47R2 (positions 8265-8294 in HxB2) in the second-round PCR assay, which amplified a 500-bp fragment of gp41 as described.13 Samples positive for HIV-2 by serology were amplified by PCR assay using the following long terminal repeat (LTR) primers: outer forward: nef2For, TGGCAATAGATATGTCACATTTTATAAA (positions 8776-8804 in HIV-2 ROD), outer reverse: RLTR2rev, TGGTGAGAGTCTAGCAGGG (positions 9449-9468 in ROD); inner forward: U3HIV2F, CCAGAAGAGTTTGGGCACAAGTCAGG (positions 9242-9269 in ROD), and inner reverse: U3HIV2R, AACACCCAGGCTCTACCTGCT (positions 9552-9573 in ROD). This nested PCR assay enabled the amplification of a 330-bp region of the HIV-2 LTR. PCR assay was performed using standard conditions. The amplimers were purified using the Qiaquick PCR purification kit (Qiagen). Nucleotide sequences were obtained by direct sequencing of both strands of the PCR products. Sequencing was performed by the fluorescent dye terminator method using a CEQ 2000 automated capillary sequencer (Beckman Coulter, CA).
Phylogenetic and Drug Resistance Analyses
Sequence alignment of gene fragments was performed automatically using CLUSTAL-X, with minor manual adjustments taking into consideration protein-coding sequences.14 New sequences were aligned with the recent compilation of nonrecombinant and CRF HIV-1 sequences from the Los Alamos database. Phylogenetic analyses of these sequences were performed using the neighbor-joining method of the CLUSTAL X program.14 To prevent the possibility of new sequences obscuring genetic relations, unknown sequences were phylogenetically analyzed individually. During the final analysis, gaps that were introduced to create the alignment were stripped. The statistical robustness of the neighbor-joining tree and reliability of the branching pattern were confirmed by bootstrapping (1000 replicates).
Putative PR and RT drug resistance mutations were determined using the Drug Resistance Algorithm from the Stanford HIV database.15 This tool compares the input sequences with sequences of HIV-1 subtype B isolates shown to confer resistance to anti-HIV drugs. Codons 9 through 99 of the PR sequence and 1 through 130 of the RT sequence were analyzed. To analyze the gp41 amino acid sequences, the heptad repeat regions 1 (HR1) and 2 (HR2) were aligned with that of HxB2 as reference. Interpretation of the T-20 resistance mutation was based on results from published reports of phenotypic and genotypic assays.16-21
Serologic Analysis, CD4 Cell Counts, and Viral Load
HIV-positive pregnant women and some HIV-infected persons before initiation of ART were included in this study. Women attending the PMTCT site in Nouna undergo counseling and, if they consent, are screened for HIV. If the result is positive, it is confirmed by a second rapid assay. Subsequently, they are clinically followed up before and after delivery and receive single-dose nevirapine for mother and child, as recommended by the World Health Organization and stated in the PMTCT program of the National Ministry of Health.
The algorithm proposed by the Ministry of Health in Burkina Faso for the PMTCT program recommends Determine for initial HIV screening and a second rapid assay, Genie II, for confirmation of reactive samples. In case of discordant results, an ELISA should be used as a third test. Of the 935 samples tested by Determine, 44 were HIV reactive. Only 34 of these were reactive by Genie II, however. For routine quality control purposes, we screened all 935 samples with the Enzygnost HIV Integral ELISA. Unexpectedly, a higher number of reactive samples was observed, including all 44 Determine-positive samples and 16 samples that were negative by both Determine and Genie II. Based on this result, all Enzygnost HIV Integral reactive samples were further tested using (1) the Murex ELISA (Abbott Diagnostics), (2) the AxSYM ELISA (Abbott Diagnostics), and (3) the LIA (Innogenetics). All 34 Genie II reactive samples were confirmed by the 2 ELISAs and the LIA, whereas all Enzygnost HIV Integral-positive samples but Genie II-negative samples were negative in the other 2 ELISAs and indeterminate or negative in the LIA. To confirm that the negative and indeterminate results were reliable, DNA was obtained from buffy coats of subjects with discordant results and PCR amplification of HIV sequences was attempted using the primer sets described in the section on materials and methods. No positive result was observed, confirming that these women were uninfected. This was also confirmed when a second set of samples was collected from these subjects after 3 to 6 months. Accordingly, the prevalence of HIV infection among pregnant women in the rural area of Nouna was 3.6% (34 of 935 women), which is lower than for the major towns of Ouagadougou and Bobo-Dioulasso, where rates of up to 7% have been reported.2,3 The age range of the HIV-positive women in this study was 21 to 45 years, with a mean age of 30.2 years (Table 1).
In addition to HIV-positive pregnant women, we obtained samples (n = 18) from drug-naive HIV-positive individuals who presented at the DRH before initiation of ART. All 18 persons had previously been found to be HIV-positive at the DRH, and this was confirmed using the Murex ELISA. In addition, we obtained samples from 3 confirmed HIV-positive individuals from Nouna who were not enrolled in the PMTCT program.
Of the total of 55 HIV-positive samples, 43 underwent further serologic, virologic, and molecular analyses (see Table 1). Twelve samples from PMTCT participants from the initial phase of the study were excluded because of the lack of stored buffy coats. The remaining 43 HIV-positive study samples were distributed as follows: pregnant women in the PMTCT program in Nouna (n = 22), HIV-positive individuals not enrolled in the PMTCT program in Nouna (n = 3), and drug-naive individuals in Dedougou (n = 18). The CD4 cell counts were determined for all 43 subjects in this study. CD4 values among pregnant women ranged from 93 to 747 cells/μL, whereas CD4 counts of patients from the DRH ranged from 1 to 718 cells/μL (see Table 1). The viral load values were not available for all subjects but were generally high, ranging between 10,000 and 1,000,000 copies/mL, as expected for HIV drug-naive individuals (see Table 1).
DNA samples from buffy coats of the selected 43 HIV-positive individuals from Nouna and Dedougou were genotypically analyzed. Using a set of conserved gag primers, sequences could be obtained from 38 of 43 samples.6 The distribution of strains and subtypes is shown in Figure 1A and Table 1. Most sequences grouped with CRF02_AG and more closely with each other compared with CRF02_AG reference strains from West Africa. Within the CRF06_cpx cluster, the Burkina Faso sequences clustered almost evenly with reference sequences from Mali and Senegal and with the prototypic CRF06_cpx-BFP90 strain from Burkina Faso. The subtype distribution in gag was as follows: CRF02_AG (n = 23), CRF06_cpx (n = 12), sub-subtype A3 (n = 1), subtype D (n = 1) and unclassified (n = 1) (see Fig. 1A; see Table 1).
In the pol region, 33 sequences were amplified with the Pol 1 through 4 primer set, whereas 8 more could be amplified using the BFpol 1 through 4 primer set. Two sequences could not be amplified by either of the 2 primer sets. A similar distribution of strains and CRFs was observed as for gag, with CRF02_AG as the dominant subtype (n = 20), closely followed by CRF06_cpx (n = 18). Within the CRF06_cpx cluster, there was a subcluster of 5 sequences with a bootstrap value of 72% (see Table 1). CRF09_cpx, subtype D, and unclassified sequences were observed, respectively, in 1 case each (see Figs. 1B-i, -ii). In the env gene, 2 regions (V3-V5 of gp120 and gp41) were amplified. In total, 34 sequences could be amplified with the Env 1 through 4 primers (see Table 1). Sequences A0061, A675, and A101 clustered with sub-subtype A3. Only A0061 clustered with a significant bootstrap value of >70% (data not shown), however. A675 and A101 were thus tentatively classified as A3-like viruses. Using the gp41 primers, 38 sequences, including 7 not amplifiable in the V3 through V5 region, could be amplified (see Fig. 1C; see Table 1). Two sequences could not be amplified using any of the envelope primers. The genotypic distribution in V3 through V5/gp41 regions showed a dominance of CRF06_cpx/CRF06_cpx (n = 19), followed by CRF02_AG/CRF02_AG (n = 16) and sub-subtype A3/CRF02_AG (n = 2). Two CRF09_cpx sequences (D009 and D007) clustered closely with CRF09_cpx reference sequences from Senegal in the gp41 tree (see Fig. 1C). These sequences showed discordant genotypes in gag, clustering, respectively, with CRF02_AG and CRF06_cpx (see Figs. 1A, C; see Table 1). Sequences that grouped as sub-subtype A3 and unclassified were observed in 1 case each (see Fig. 1C; see Table 1). In general, there was no difference in the pattern of gag, pol, and env sequences from Nouna and Dedougou.
Amplification and sequence analysis of more than 1 gene in the HIV genome facilitates the identification of potential recombinant genomes. Based on this, the gag, pol, and env sequences were grouped (Table 2) into concordant and discordant genotypes, whereas genomes with incomplete gene profiles were grouped separately. Twenty samples were genotypically concordant in gag, pol, and env; of these, 11 and 9 were “pure” CRF02_AG and CRF06_cpx genotypes, respectively. Sequences from 18 of the viruses analyzed showed a discordant genotype. The most common putative recombinants between CRFs were CRF02_AGgag/CRF06_cpxpol/env, which were found in 6 cases (see Table 2). CRF02_AG was more frequently observed among viruses with discordant genotypes than was CRF06_cpx. This was also the case for the single subtype Dgag/pol sequence, which was CRF02_AG in env.
Serologic analysis with the Genie II assay allows differentiating HIV-1 from HIV-2. Only 1 HIV-2 singly reactive sample was observed among the 43 samples, and subsequently confirmed by the LIA (see Table 1). Surprisingly, 5 (12%) of 43 samples were serologically dually reactive against HIV-1 and HIV-2 (see Table 1). This result was supported by the LIA, where these samples showed a pattern of strong reactivity with all HIV-1 proteins but reactivity only with gp36 and not with the gp105 envelope proteins of HIV-2. All HIV-2-reactive samples (including the dually reactive samples) were further analyzed using the HIV-2-specific LTR primer set. A 330-bp fragment could be amplified and sequenced in the case of samples Ou169 and A440, which had been serologically determined as dually reactive for HIV-1 and HIV-2 or singly reactive for HIV-2, respectively (see Table 1). The sequences obtained were too short to enable subtype-specific phylogenetic classification. They were, however, confirmed to be HIV-2 after a BLAST search. Accordingly, only Ou169 could be confirmed to be doubly infected by HIV-1 and HI-2, whereas the other 4 serologically dually reactive samples are likely to have been cross-reactive with the gp36 protein of HIV-2.
V3 Loop Sequences and Predicted Biologic Phenotypes
Previous studies have shown that the net charge of the V3 loop and the polarity of the amino acids located at positions 11 and 25 are important determinants of biologic phenotype of HIV isolates.6,22,23 Accordingly, we compared the V3 loop coding sequences from samples in which positive results were obtained (Fig. 2). The V3 loop sequences ranged from 31 to 35 amino acids in length and were quite homogeneous. The apex of the V3 loop was marked by the following amino acid sequences: GPGQAF (n = 11), GPGQTF (n = 7), GPGRVF (n = 3), GPGQAL (n = 2), RPGRAF (n = 1), GPGRAF (n = 1), GPGQAY (n = 1), APGQAF (n = 1), APGRAFY (n = 1), GPGHTW (n = 1), GRWQTF (n = 1), GAGQAI (n = 1), GPGQMF (n = 1), GPGQSF (n = 1), and GPGQIF (n = 1). Net charges of the V3 loop were comparatively low (+2 to +5), except for 1 virus with a net charge of +6. No positively charged amino acid was found on position 11 or 25 (see Fig. 2). Based on these results, all but 1 virus were predicted to use CCR5 as the main coreceptor.
Drug Resistance-Associated Polymorphisms
Different subtypes of HIV carry polymorphisms that may play a significant role in the development of drug resistance. We therefore looked for resistance-associated polymorphisms in the PR, 5′ RT, and gp41 regions by comparing known resistance-associated mutations found in the HIV sequence database and checked whether some of these were present in drug-naive non-subtype B samples from Burkina Faso. The PR coding region was analyzed from codon 9, and the RT coding region was analyzed up to codon 130. The most commonly observed polymorphisms in the PR region were K20I and M36I, which were found in more than 95% of all samples (data not shown). The L10V, L33I, I50M, L63P, V82I, and N88D polymorphisms occurred in 4, 2, 1, 10, 2, and 3 patients, respectively. These polymorphisms do not directly cause drug resistance but only in association with other changes within the PR region. They have thus been termed secondary mutations. Primary mutations, conversely, are directly associated with drug resistance even in the absence of other mutations. We found polymorphisms conforming to primary mutations in the RT region, notably K103N (n = 2) and V118I (n = 1). These are commonly observed in individuals exposed to nonnucleoside reverse transcriptase inhibitors (NNRTIs). V118I is one of the mutations that have been implicated in cross-resistance to nucleoside reverse transcriptase inhibitors (NRTIs). V106L and V116Y were the most common alterations observed in 3 samples, respectively. The prevalence of an RT region-associated polymorphism in the subjects was generally less than 10% (data not shown). We further examined the frequencies of multiple resistance polymorphisms in the study population in the PR and RT regions. The most common polymorphism combinations in the PR region were: K20I + M36I (n = 15), K20I + M36I + L63P (n = 7), and L10V + K20I + M36I (n = 4). In the RT region, 4 individuals were found to carry 2, 2, 3, and 6 polymorphisms, respectively. Searching for multiple polymorphisms in the PR and RT regions showed that each of these 6 subjects carried at least 2 polymorphisms in the PR region. The highest multiple polymorphisms were noted in subject D039, who carried K20I + M36I in the PR region and F77L + F116Y + V118I + K101I + V106L in the RT region.
T-20 (enfuvirtide) is a 36-amino acid peptide derived from the HR2 region of the gp41 transmembrane domain of HIV-1 subtype B. The amino acid region 36 through 43 (GIVQQQNN) of HR1 primarily determines enfuvirtide responsiveness. Changes within the GIV motif are sufficient to cause drug resistance to T-20.16 We determined T-20 drug resistance-associated polymorphisms in the HR1 and HR2 regions of 38 gp41 sequences (data not shown). In the HR1 region, the GIV motif was conserved among all sequences. The most common polymorphism observed within this region was N42S (n = 36), which was dominant among CRF02_AG and CRF06_cpx viruses. The Q54K alteration was observed among sub-subtype A3, CRF02_AG, and CRF09_cpx viruses but not in CRF06_cpx viruses. The combination L54M/Q56K was observed in 4 sequences, with 3 of them being CRF02_AG strains, whereas a Q56K/R polymorphism was found in all CRF02_AG, sub-subtype A3, and CRF09_cpx viruses. Ten amino acid positions in HR2 are known to be involved in the sensitivity to T-20. Amino acids at 6 of these positions were totally conserved in all strains. Meanwhile, S138A (n = 4) was the most common polymorphism found at any of the remaining 4 nonconserved amino acid loci. We also observed the polymorphism N126K (n = 3) and N126S (n = 1), with the former having recently been shown to be highly insensitive to T-20 in the context of the HIVBaL isolate.24 This polymorphism has also been reported rarely in subtype F and CRF01_AE.25 Generally, the amino acid sequence of the HR2 region was more heterogeneous compared with that of the HR1 region.
Nucleotide Sequences Accession Numbers
Genbank accession numbers of the sequences reported in this study are: Gag: DQ682788-DQ68285; Pol: DQ682709-DQ682741; Polbf: DQ682780-DQ682787; Env V2-V3: DQ682780-DQ682787; Gp41: DQ682742-DQ682779.
In this study, we present the first molecular characterization of HIV-1 subtypes circulating in rural Burkina Faso. This study also provides important preliminary information on drug resistance-associated polymorphisms in the PR, RT, and gp41 regions of viruses from drug-naive Burkinabes.
The use of a combination of Determine as screening test and Genie II for confirmation of HIV infection can be highly recommended, because all samples positive in both assays were later confirmed by all other methods and none of the 901 negative samples was found to be a false-negative result. Thus, this combination is sensitive and specific, whereas using only Determine gives a significant number of false-positive results. Screening with fourth-generation ELISAs from different manufacturers showed an unexpected variability of false-positive results. Specifically, using the Enzygnost HIV Integral assay, 1.7% (16 of 935) of all samples represented false-positive results, as shown by a combination of other methods. In contrast, results of the Murex ELISA were concordant with those of the Determine/Genie II, immunoblot, and PCR assay. The choice of ELISA type seems to be important in this setting, with the Murex ELISA being the primary choice, whereas the LIA or other immunoblot assays do not seem to be necessary.
It is important to know what type of HIV is infecting an individual, because this is crucial information for therapy. We observed 5 HIV-1 and HIV-2 serologically dually reactive samples among the study subjects, which was surprisingly high. Four of these dually reactive samples were not confirmed by PCR assay, however, suggesting that they were false-positive results for HIV-2 or that the primer sets used were not sensitive enough for their amplification. One dually positive sample was confirmed, however.
The HIV epidemic in rural Burkina Faso is dominated by 2 main CRFs: CRF02_AG and CRF06_cpx, a mosaic virus composed of 4 subtypes, namely, A, G, J, and K. When the 3 genomic regions gag, pol, and env were considered, it became clear that these 2 sets of CRFs are circulating almost equally, with CRF02_AG having a slight advantage in distribution. Eleven pure (same genotype in all 3 genes sequenced) CRF02_AG samples were found compared with 9 CRF06_cpx samples. It is interesting to note that the sequences in this rural area were different compared with what was reported previously in Ouagadougou, where CRF06_cpx dominates.10 A preliminary study of 24 viruses from a cohort of commercial sex workers in Bobo-Dioulasso showed CRF02_AG viruses comprising almost 50% of the circulating viruses, whereas CRF06_cpx comprised 30%.26 It is important to carry out further studies to confirm this difference in genotypes between different regions in Burkina Faso.
Sequencing the 3 major genes of HIV allowed the identification of a high number of putative recombinants, although one cannot rule out the possibility of dual infection in some of the individuals who carried discordant subtypes. CRF09_cpx viruses have been described predominantly in West African countries, including Ghana, Côte d'Ivoire, and Senegal.27 A recent study found two cases of CRF09_cpx viruses in the northwestern region of Cameroon.28 Although these 2 Cameroonian isolates were pure CRF09_cpx in pro and env, we observed discordance in the 3 genes sequenced in the 2 Burkina Faso samples (see Table 2). One of them was an A3gag/CRF09_cpxpol/env, and the other was CRF02_AGgag/pol/CRF09_cpxenv. Recently, a CRF02_AG/CRF09_cpx virus was described in neighboring Côte d'Ivoire.29 Likewise, sub-subtype A3 viruses, which we describe in this study for the first time in Burkina Faso, were reported earlier to be circulating in other parts of West and Central Africa, where they had also recombined with CRF02_AG.30 Most interesting is the fact that the A3 virus in our study seemed to have recombined, respectively, with CRF02_AG and CRF06_cpx.
CRF02_AG and CRF06_cpx, which were the vastly predominant strains in our study, also circulate in neighboring countries, albeit with different relative prevalences.5,31,32 In Ghana, Côte d'Ivoire, and Mali, for example, CRF02_AG is the most dominant strain. In Niger, however, CRF02_AG and CRF06_cpx cocirculate. In fact, similar recombinants like we report here (ie, discordant CRF02_AG and CRF06_cpx in env and gag, or vice versa) were also seen earlier in Niger.33,34
ART is gradually being introduced in Burkina Faso, thereby prompting monitoring for drug resistance. A good basis in such a setup is to monitor the circulating resistance-associated polymorphisms in drug-naive individuals before ART introduction. Our study clearly shows that among naive individuals, some secondary as well as important primary drug resistance polymorphisms occur. Some of these polymorphisms have been described earlier in B and non-B viruses.35 The V82I mutation, for example, was described in approximately 1% of B subtypes and 5% to 10% of non-B subtypes. This is also true for L10V, K20I, and M36I, with the last 2 being found in more than 95% of the study subjects. The alteration L63P occurs more frequently in B subtypes than in non-B subtypes.35 The primary drug resistance-associated variants in the RT sequence of some patients are likely to have been transmitted from drug-exposed individuals or to be because of unreported prior use of ART. Two of these mutations (K103N in the RT region and N88D in the PR region) were previously reported in neighboring Côte d'Ivoire, confirming that viruses carrying drug resistance-associated polymorphisms are already prevalent in some countries in West Africa, where therapy is just being introduced.31 Introduction of ART requires adequate planning and infrastructure. The National Ministry of Health in Burkina Faso has developed structures to facilitate implementation of ART; however, important structures, such as monitoring for drug resistance and compliance, still have to be improved. We have recently noticed several discrepancies in PR and RT sequences and compliance data from patients with high viral loads who were supposedly under treatment but showing “naive” sequences (unpublished data).
Phenotypic and genotypic studies on drug resistance-associated polymorphisms and mutations in gp41 of non-subtype B viruses are rare.36,37 Several mutations located at positions 36 through 46 of the HR1 region notably, G36D, V38A, N43D, L44V, and L45Q, are prominent in subtype B-infected patients undergoing treatment with T-20; however, only G36D and N43D have been associated with treatment failure.18,38 In general, sequencing all 38 samples did not show any major polymorphisms in this region, confirming previous findings that natural T-20 genotypic resistance is rare in drug-naive individuals.20 There were a few polymorphisms observed among our sequences whose phenotypic implications are largely unknown. The L54M and Q56K polymorphisms that were observed in 38% and 75% of all CRF02_AG samples, respectively, were recently found to cause approximately 5-fold reduced sensitivity to inhibition by T-20 using a subtype B viral backbone.37 Further studies using a more closely related CRF02_AG recombinant backbone are therefore needed. Asparagine occurs naturally at position 42 in subtypes B, D, and F1.20 An N42S polymorphism was found in 37 gp41 sequences studied, however. Polymorphism at this position is subtype specific. It needs to be determined whether this polymorphism leads to phenotypic changes in these CRFs. In HIV-1 group O, for example, an N42D polymorphism occurs but does not compromise T-20 activity.39
This study shows that serologically, the currently used rapid tests in the PMTCT program in Burkina Faso are highly reliable. Molecularly, CRF02_AG and CRF06_cpx dominate the HIV epidemic in rural Burkina Faso, with the former having a slight edge. These 2 viruses seem to have recombined to form complex recombinants that need constant monitoring. The presence of some “primary” drug resistance mutations among supposedly naive individuals calls for an extended screening of drug-naive persons to confirm whether these are natural subtype-specific polymorphisms or transmitted resistant viruses.
The authors thank Richard Sourougou and Kanazoe Rasmane of the Centre Medical avec Antenne Chirugicale (CMA) in Nouna and Nestor Dembele of the CRSN in Nouna for help with sampling. Sandrine Tchatchou helped with PCR analysis. The authors are also grateful to Bocar Kouyaté and the authorities of the Ministry of Health in Burkina Faso for initiating the PMTCT program in Nouna and for constant support of this study.
2. Meda N, Cartoux M, Dabis F, et al. Stabilization of HIV infection rates in Burkina Faso, 1995-1999. Int J STD AIDS. 2001;12:460-462.
3. Simpore J, Granato M, Santarelli R, et al. Prevalence of infection by HHV-8, HIV, HCV and HBV among pregnant women in Burkina Faso. J Clin Virol. 2004;31:78-80.
4. Heyndrickx L, Janssens W, Zekeng L, et al. Simplified strategy for detection of recombinant human immunodeficiency virus type 1 group M isolates by gag/env heteroduplex mobility assay. Study Group on Heterogeneity of HIV Epidemics in African Cities. J Virol. 2000;74:363-370.
5. Peeters M, Esu-Williams E, Vergne L, et al. Predominance of subtype A and G HIV type 1 in Nigeria, with geographical differences in their distribution. AIDS Res Hum Retroviruses. 2000;16:315-325.
6. Tebit DM, Zekeng L, Kaptue L, et al. Genotypic and phenotypic analysis of HIV type 1 primary isolates from western Cameroon. AIDS Res Hum Retroviruses. 2002;18:39-48.
8. Montavon C, Toure-Kane C, Liegeois F, et al. Most env and gag subtype A HIV-1 viruses circulating in West and West Central Africa are similar to the prototype AG recombinant virus IBNG. J Acquir Immune Defic Syndr. 2000;20:363-374.
9. Kanki PJ, Travers KU, MBoup S, et al. Slower heterosexual spread of HIV-2 than HIV-1. Lancet. 1994;16;943-946.
10. Ouedraogo-Traore R, Montavon C, Sanou T, et al. CRF06-cpx is the predominant HIV-1 variant in AIDS patients from Ouagadougou, the capital city of Burkina Faso. AIDS. 2003;17:441-442.
11. Condra JH, Schleif WA, Blahy OM, et al. In vivo emergence of HIV-1 variants resistant to multiple protease inhibitors. Nature. 1995;374:569-571.
12. Xu L, Hue S, Taylor S, et al. Minimal variation in T-20 binding domain of different HIV-1 subtypes from antiretroviral-naive and -experienced patients. AIDS. 2002;16:1684-1686.
13. Yang C, Pieniazek D, Owen SM, et al. Detection of phylogenetically diverse human immunodeficiency virus type 1 groups M and O from plasma by using highly sensitive and specific generic primers. J Clin Microbiol. 1999;37:2581-2586.
14. Thompson JD, Gibson TJ, Plewniak F, et al. The CLUSTAL X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 1997;25:4876-4882.
16. Reeves JD, Lee FH, Miamidian JL, et al. Enfuvirtide resistance mutations: impact on human immunodeficiency virus envelope function, entry inhibitor sensitivity, and virus neutralization. J Virol. 2005;79:4991-4999.
17. Sista PR, Melby T, Davison D, et al. Characterization of determinants of genotypic and phenotypic resistance to enfuvirtide in baseline and on-treatment HIV-1 isolates. AIDS. 2004;18:1787-1794.
18. Poveda E, Rodes B, Labernardiere JL, et al. Evolution of genotypic and phenotypic resistance to enfuvirtide in HIV-infected patients experiencing prolonged virologic failure. J Med Virol. 2004;74:21-28.
19. Heil ML, Decker JM, Sfakianos JN, et al. Determinants of human immunodeficiency virus type 1 baseline susceptibility to the fusion inhibitors enfuvirtide and T-649 reside outside the peptide interaction site. J Virol. 2004;78:7582-7589.
20. Roman F, Gonzalez D, Lambert C, et al. Uncommon mutations at residue positions critical for enfuvirtide (T-20) resistance in enfuvirtide-naive patients infected with subtype B and non-B HIV-1 strains. J Acquir Immune Defic Syndr. 2003;33:134-139.
21. Carmona R, Perez-Alvarez L, Munoz M, et al. Natural resistance-associated mutations to enfuvirtide (T20) and polymorphisms in the gp41 region of different HIV-1 genetic forms from T20 naive patients. J Clin Virol. 2005;32:248-253.
22. Chesebro B, Wehrly K, Nishio J, et al. Mapping of independent V3 envelope determinants of human immunodeficiency virus type 1 macrophage tropism and syncytium formation in lymphocytes. J Virol. 1996;70:9055-9059.
23. Xiao L, Owen SM, Goldman I, et al. CCR5 coreceptor usage of non-syncytium-inducing primary HIV-1 is independent of phylogenetically distinct global HIV-1 isolates: delineation of consensus motif in the V3 domain that predicts CCR-5 usage. Virology. 1998;240:83-92.
24. Lohrengel S, Hermann F, Hagmann I, et al. Determinants of human immunodeficiency virus type 1 resistance to membrane-anchored gp41-derived peptides. J Virol. 2005;79:10237-10246.
25. Aghokeng AF, Ewane L, Awazi B, et al. Enfuvirtide binding domain is highly conserved in non-B HIV type 1 strains from Cameroon, West Central Africa. AIDS Res Hum Retroviruses. 2005;21:430-433.
26. Manigart O, Courgnaud V, Sanou O, et al. HIV-1 super infections in a cohort of commercial sex workers in Burkina Faso as assessed by an autologous heteroduplex mobility procedure. AIDS. 2004;18:1645-1651.
27. McCutchan FE. Understanding the genetic diversity of HIV-1. AIDS. 2000;14:31-44.
28. Burda ST, Konings FA, Williams CA, et al. HIV-1 CRF09_cpx circulates in the North West Province of Cameroon where CRF02_AG infections predominate and recombinant strains are common. AIDS Res Hum Retroviruses. 2004;20:1358-1363.
29. Toni T, Adje-Toure C, Vidal N, et al. Presence of CRF09_cpx and complex CRF02_AG/CRF09_cpx recombinant HIV type 1 strains in Cote d'Ivoire, West Africa. AIDS Res Hum Retroviruses. 2005;21:667-672.
30. Meloni ST, Kim B, Sankale JL, et al. Distinct human immunodeficiency virus type 1 subtype A virus circulating in West Africa: sub-subtype A3. J Virol. 2004;78:12438-12445.
31. Toni TD, Recordon-Pinson P, Minga A, et al. Presence of key drug resistance mutations in isolates from untreated patients of Abidjan, Cote d'Ivoire: ANRS 1257 study. AIDS Res Hum Retroviruses. 2003;19:713-717.
32. Fischetti L, Opare-Sem O, Candotti D, et al. Molecular epidemiology of HIV in Ghana: dominance of CRF02_AG. J Med Virol. 2004;73:158-166.
33. Mamadou S, Vidal N, Montavon C, et al. Emergence of complex and diverse CRF02-AG/CRF06-cpx recombinant HIV type 1 strains in Niger, West Africa. AIDS Res Hum Retroviruses. 2003;19:77-82.
34. Mamadou S, Montavon C, Ben A, et al. Predominance of CRF02-AG and CRF06_cpx in Niger, West Africa. AIDS Res Hum Retroviruses. 2002;18:723-726.
35. Gonzales MJ, Machekano RN, Shafer RW, et al. Human immunodeficiency virus type 1 reverse-transcriptase and protease subtypes: classification, amino acid mutation patterns, and prevalence in a northern California clinic-based population. J Infect Dis. 2001;184:998-1006.
36. Cilliers T, Patience T, Pillay C, et al. Sensitivity of HIV type 1 subtype C isolates to the entry inhibitor T-20. AIDS Res Hum Retroviruses. 2004;20:477-482.
37. Chinnadurai R, Munch J, Kirchhoff F. Effect of naturally-occurring gp41 HR1 variations on susceptibility of HIV-1 to fusion inhibitors. AIDS. 2005;19:1401-1405.
38. Cooper DA, Lange JM. Peptide inhibitors of virus-cell fusion: enfuvirtide as a case study in clinical discovery and development. Lancet Infect Dis. 2004;4:426-436.
39. Poveda E, Barreiro P, Rodes B, et al. Enfuvirtide is active against HIV type 1 group O. AIDS Res Hum Retroviruses. 2005;21:583-585.
© 2006 Lippincott Williams & Wilkins, Inc.