Infection of an individual with two genetically distinct HIV-1 variants is not an exceptional phenomenon. It can be classified as either a co-infection when a host is infected by two distinct HIV strains at or around the time of infection, or as a super-infection, in the case of the sequential acquisition of two [1,2] or more  HIV variants. Several cases of HIV-1 co-infections or super-infections have been reported , but only limited data are available on the prevalence of this phenomenon .
We investigated the occurrence of HIV-1 co-infections from frozen cells collected from HIV-1-infected seroconverters enrolled in the French ANRS SEROCO and PRIMO cohorts between 1988 and 2004. These are two well documented cohort studies of recently infected individuals with a known date of infection [6,7]. Ninety-five per cent of patients from the SEROCO cohort and 77% of patients (82% in non-African patients) from the PRIMO cohort are infected with a B subtype of HIV-1. As in the United States  and other European countries, considering the epidemiology and the distribution of HIV-1 subtypes in France, the occurrence of dual infections in that population is more likely to involve HIV-1 subtype B than to involve two distinct subtypes or circulating recombinant forms. The detection of HIV-1 co-infections or super-infections is currently based on subcloning and sequencing of the viral genome, a time-consuming and labour-intensive approach. In the present study, the large-scale identification of HIV-1 intrasubtype B dual infections was assessed by means of an improved heteroduplex mobility assay (HMA) . HMA is a reliable technique that allows the sensitive detection of divergent sequences in a polymerase chain reaction product, and is also well adapted for the screening of large numbers of samples . A 630 base pair fragment in the C2-V5 region of the HIV-1 env gene was amplified by nested polymerase chain reaction from DNA extracted from the peripheral blood mononuclear cells of patients . HMA conditions were optimized as previously described . Migration patterns that allow the detection of intrasubtype B mixed infections were set up through DNA analysis of 20 known HIV-1 B subtypes obtained from clinical samples and mixed in different amounts in order to distinguish patterns of heteroduplex generated by intrapatient quasispecies populations from those observed from different strains. The analytical sensitivity of the HMA was estimated at 10%, because minority strains representing a proportion as small as 10% of the whole viral population could be detected.
Three hundred and seventy and 290 seroconverters were tested from the SEROCO and PRIMO cohorts, respectively. All samples were collected at the inclusion visit between 1988 and 1995 for the SEROCO cohort study, and between 1996 and 2004 for the PRIMO cohort study. Time from HIV-1 infection to this first blood collection ranged from 0.7 to 23.8 months (median 8.3 months) in the SEROCO cohort and from 0.6 to 7.9 months (median 1.7 months) in the PRIMO cohort. Therefore, the calendar period of infection ranged from 1986 to 2004. Among the 660 samples tested by HMA, 75 heteroduplex migration patterns compatible with a co-infection or super-infection were identified and subjected to further analyses. Heteroduplexes were gel excised, subcloned and sequenced. In all cases, sequence analyses showed that a slow migration pattern of heteroduplexes may be explained by deletions or insertions (ranging from 10 to 70 bp) in the env gene.
No evidence of co-infection was thus detected in the baseline samples from two French seroconverter cohorts. The first HIV-1 dual infection case reported in 1995 was in a patient with repeated exposure through multiple sexual partners . The prevalence of HIV-1 co-infections has been estimated at approximately 5% or less in previous studies [5,9,13]. These estimates were, however, based on highly selected individuals with a high probability of multiple HIV exposure or features indicative of co-infections or super-infections. In our study, co-infections were for the first time systematically looked for at enrolment, without focusing on particular high-risk exposures or features predictive of co-infections. Not surprisingly in these settings, the prevalence was expected to be low, and indeed our calculated 95% confidence interval around the prevalence estimate is [0, 0.5%].
We cannot, however, rule out the possibility that co-infections involving minority viral populations representing less than 10% of the whole HIV-1 population were missed by our technique. This 10% sensitivity is consistent with previous reports [9,10]; it could be significantly improved through subcloning and sequencing of a large number of clones, but this strategy is not suitable for large-scale studies. Too early or too late sampling might also have prevented us from capturing a co-infection, but this is very unlikely because the range of delays between HIV-1 infection and blood sample collection was large (from 0.6 months to 23.8 months). Finally, we cannot exclude the possibility that this HMA approach, targeting the sole C2V5 region, could have missed a recombinant strain sharing the same env region as the initial virus.
In our analysis, HMA migration patterns suggesting co-infections or super-infections were indeed associated with insertions/deletions in the env gene of HIV-1 strains (‘false-positive HMA’). These insertions/deletions observed in the env gene are unlikely to result from recombination events because the flanking regions with aligned sequences displayed almost 100% homology. In the early phase of the infection, viral populations are more homogeneous than in chronically infected patients. The interpretation of these deletions remains unclear and deserves further investigation. In our study, no significant association could be observed between the presence of the deletion in the env gene of the HIV-1 strain infecting the patient and the biological or clinical characteristics recorded at enrolment, including viral subtype and delay since infection.
Our study suggests that, at least in the context of the SEROCO and PRIMO cohort studies, and until 2004, HIV-1 co-infections or early super-infections were an infrequent phenomenon. Further studies focusing on subjects with likely re-exposure to HIV-1 selected from the same cohorts are in process to estimate the incidence of super-infections occurring later in the course of HIV-1 infection.
Sponsorship: This work was supported by a grant from the French National Agency for Research on AIDS (ANRS contract 2005-070).
1. Jost S, Bernard MC, Kaiser L, Yerly S, Hirschel B, Samri A, et al
. A patient with HIV-1 superinfection. N Engl J Med 2002; 347:731–736.
2. Ramos A, Hu DJ, Nguyen L, Phan KO, Vanichseni S, Promadej N, et al
. Intersubtype human immunodeficiency virus type 1 super-infection following seroconversion to primary infection in two injection drug users. J Virol 2002; 76:7444–7452.
3. Van Der Kuyl AC, Kozacynska K, Van Den Burg R, Zorgdrager F, Back N, Jurrians S, et al
. Triple HIV-1 infection. N Engl J Med 2005; 352:2557–2559.
4. Smith DM, Richman DD, Little SJ. HIV superinfection. J Infect Dis 2005; 192:438–443.
5. Goettlieb GS, Nickle DC, Jensen MA, Wong KG, Grobler J, Li F, et al
. Dual HIV-1 infection associated with rapid disease progression. Lancet 2004; 303:619–622.
6. Carré N, Meyer L, Boufassa F, Deveau C, Persoz A, Rouzioux C, Vildé JL, and the SEROCO study group. High risk of HIV disease progression after infection through a sexual partner with AIDS. AIDS 1996; 10:77–80.
7. Desquilbet L, Goujard C, Rouzioux C, Sinet M, Deveau C, Pellegrin I, et al
, for the PRIMO and SEROCO study groups. Does transient HAART during primary infection lower the virological set-point? AIDS 2004; 18:2361–2369.
8. Koelsch KK, Smith DM, Little SJ, Ignacio CC, Macaranas TR, Leigh Brown AJ, et al
. Clade B HIV-1 superinfection with wild-type virus after primary infection with drug-resistant clade B virus. AIDS 2003; 17:F11–F16.
9. Manigart O, Courgnaud V, Sanou O, Valéa D, Nagot N, Meda N, et al
. HIV-1 superinfections in a cohort of commercial sex workers in Burkina Faso as assessed by an autologous heteroduplex mobility procedure. AIDS 2004; 18:1–7.
10. White PA, Li Z, Zhai X, Marinos G, Rawlinson WD. Mixed viral infection identified using heteroduplex mobility analysis (HMA). Virology 2000; 271:382–389.
11. Delwart E, Shaper EG, Louwagi J, McCutchan FE, Grez M, Rubsamen-Waignann H, et al
. Genetic relationships determined by a heteroduplex mobility assay: analysis of HIV env
genes. Science 1993; 262:1257–1261.
12. Zhu T, Wang N, Carr R, Wolinsky SM, Ho DD. Evidence for coinfection by multiple strains of human immunodeficiency virus type 1 subtype B in acute seroconvertor. J Virol 1995; 69:1324–1327.
13. Grobler J, Gray CM, Rademeyer C, Seoighe C, Ramjee G, Karim S, et al
. Incidence of HIV-1 dual infection and its association with increased viral load set point in a cohort of HIV-1 subtype-C-infected female sex workers. J Infect Dis 2004; 190:1355–1359.