However relevant the virologic events occurring in blood and lymph nodes may be to HIV-1 pathogenesis and disease progression, genital and other mucosal fluids are the primary vehicles of viral transmission. Since viral populations are not evenly distributed throughout the body (reviewed in ), understanding the processes involved in viral colonization and persistence in mucosal fluids remains a central concern for devising rational HIV-1 prevention and therapeutic strategies.
As has been confirmed in genital and anorectal secretions, HIV-1 RNA, proviral DNA and viral particles are readily detected in oral fluids during both acute and long-term HIV infection, often at titers that meet or exceed corresponding titers in blood and/or genital fluids [2,3]. The observation that HIV is inefficiently transmitted through oral fluids (reviewed in ) is probably related to the high concentration of endogenous inhibitors that are secreted into these mixed fluids by salivary glands . Yet, the potential cellular sources for viral shedding into mixed oral fluids are common to the fluids that lubricate other mucosal tissues, such as exocrine glands and gut-associated lymphoid tissue that are abundant in the oropharynx. Oral fluids may therefore serve as a relevant and particularly convenient model for studying the dynamics of mucosal HIV-1 shedding.
The biology of viral colonization at mucosal sites is not well understood. Virus is detected initially only in localized tissue following mucosal exposure and rapidly disseminates throughout the lymphatic system within days of infection . In some individuals, the genital tract may comprise unique tissue compartments of HIV-1 replication, based on disparities in viral load, envelope sequence diversity, chemokine coreceptor usage and/or syncytium-inducing (SI) capacity compared to blood (reviewed in ). Tissue compartmentalization is thought to arise from differences in target cell types, migration of infected cells, local selective immune pressures and/or reduced penetration of antiretroviral drugs in the protected site compared to blood. The degree to which viral compartmentalization extends to non-genital mucosal sites, such as the oral cavity, has not been examined in primary HIV-1 infection (PHI). Knowledge of sequence heterogeneity, chemokine coreceptor usage and SI patterns of HIV-1 in oral fluids during PHI will further our understanding of viral dissemination to non-genital mucosal surfaces during this critical stage of infection.
We have previously shown that viral titers in oral fluids are highly correlated with corresponding titers in plasma during the early acute phase of primary HIV-1 infection . Given that viremia is in rapid flux during this time, these observations suggest parallel viral dynamics and possibly rapid trafficking of the virus between blood and oral fluids. We therefore hypothesized that genetically related HIV-1 populations are present in oral fluids and blood during PHI. We further hypothesized that viral populations of R5 and non-SI (NSI) phenotypes predominate in oral fluids early in infection and reflect the biological phenotypes of virus in blood. To test these hypotheses, heteroduplex tracking assays (HTAs) and sequence analyses targeting the V1/V2 and V3 regions of gp120 and the MT-2 assay were used to characterize the genetic diversity, coreceptor usage, X4/R5 genotype, and SI phenotype of HIV-1 variants in paired oral fluids and plasma during PHI.
Materials and methods
Study participants were adults who presented at one of three south-eastern US medical centers [University of North Carolina (UNC) Health Care, Chapel Hill, North Carolina; Duke University Medical Center, Durham, North Carolina; Emory University, Atlanta, Georgia] with clinical symptoms of PHI  and agreed to provide both blood and saliva samples. PHI was defined as a positive p24 antigen or viral RNA and either enzyme-linked immunosorbent assay negativity or an evolving Western blot (two or fewer bands) within 30 days of study enrollment. Plasma HIV-1 RNA was determined using the Monitor® (Roche, Branchburg, New Jersey, USA) HIV-1 RNA assay. Informed consent was obtained from each participant according to the Institutional Review Boards of the respective centers. Data describing socio-demographics, transmission risk factors and onset of PHI-related symptoms were gathered by interview.
Collection and processing of body fluids
Peripheral blood was collected in EDTA anti- coagulated tubes. Cell-free plasma was obtained by centrifugation (1000 × g for 20 min) and stored at −70°C until tested. Whole unstimulated saliva (5 ml) was collected by expectoration into a sterile container kept chilled on ice, remixed and stored at −70°C until use. Subjects abstained from eating or brushing their teeth for at least 30 min prior to saliva collection.
Viral load testing and CD4+ cell enumeration
HIV-1 RNA levels were measured in blood plasma using either the Monitor® system with a lower detection limit of 2.60 log10 copies/ml or the Monitor Ultradirect® system with a lower detection limit of 1.40 log10 copies/ml. Salivary viral loads were measured by Nuclisens® assay (BioMerieux, Durham, NC) with a lower detection limit of 2.60 log10 copies/ml. CD4+ cell count was determined by four-color flow cytometry using monoclonal antibodies from Becton Dickinson (San Jose, California, USA).
Characterization of HIV-1 variants by HTAs
V1/V2- and V3-specific HTAs were performed on saliva and plasma samples as we have described . Briefly, total RNA was extracted using the QIAmp® Viral RNA Extraction Kit (Qiagen Inc., Valencia, California, USA). The regions of interest were amplified by reverse transcriptase (RT)-polymerase chain reaction (PCR) using AMV-RT (Roche) and Titan® RT-PCR Kit (Roche). HTAs specific for the gp120 V1/V2  and V3 [8,10] encoding regions were performed as detailed using single-stranded probes. The regions correspond to amino acids 123–222 and 292–327 of the HIV-1 JR-FL sequence, respectively . Published criteria  based on the calculated mobility ratios of heteroduplex bands were used to assign R5/X4 phenotype. To confirm results and improve RNA template sampling, each sample was subjected to two or more independent RT-PCR amplifications and the products were analyzed on at least two independent gels. RT-PCR products from multiple independent RNA isolations were also examined for selected subjects. The relative abundances of viral variant populations were determined by phosphorimagery analysis using FragmeNT Analysis Software (Molecular Dynamics, Sunnyvale, California, USA) and confirmed in two or more separate experiments. Only variants representing greater than 10% of the total population and detected in at least two independent RT-PCR reactions were analyzed. HTA analyses were limited to paired fluids from which heteroduplex bands specific for V1/V2 (n = 17 subjects) and V3 (n = 11 subjects) were reproducibly detected.
Genotype and sequence analyses
Nucleotide sequences of the V1/V2 and V3 regions were determined by either direct sequencing of RT-PCR products resolving as single HTA bands or sequencing of cloned products representing the predominant HTA variants. Products were sequenced using the ABI Prism dye terminator cycle sequencing kit on an ABI 3100 instrument (Applied Biosystems, Foster City, California, USA). Sequences were aligned and compared to the probe sequence using the PileUp program of the University of Wisconsin Genetics Computer Group sequence analysis software package . Alignments were refined manually. R5/X4 classifications were determined following V3 amino acid sequence criteria . Briefly, the accumulation of basic amino acids, particularly at positions 11 and/or 25 of the V3-encoding region is indicative of X4 variants .
The SI/NSI phenotypes of HIV-1 in plasma samples were determined in the MT-2 cell fusion assay as described .
Viral load data were log10-transformed prior to analysis. The difference between the detection of multiple V1/V2 HTA variants and the number of days from PHI-related symptoms onset to sample collection was tested by χ2 test (significance at P < 0.05).
Genbank accession numbers
Nucleotide sequences have been assigned GenBank accession numbers AF536871-AF536930.
Characteristics of study participants
The study population included 17 antiretroviral therapy-naive men with acute subtype B infection who resided in the south-eastern US (Table 1). All subjects described unprotected sex as the most likely transmission risk factor of infection. Eleven subjects reported homosexual sex as their sole risk factor, five subjects acknowledged heterosexual sex as their sole risk factor, and one subject identified both homosexual and heterosexual sex as risk factors. Median age of subjects was 28 years and median CD4+ cell count was 408 × 106 cells/l. All subjects had high viral loads in both blood (median = 5.49 log10 copies/ml) and saliva (median = 3.78 log10 copies/ml). The median duration from the onset of PHI-related symptoms to sample collection was 24 days.
Blood and saliva contain genetically related viral variants during PHI
To determine whether blood and oral fluids contain similar viral populations, HIV-1 RNA was extracted from paired samples and analyzed by V1/V2-specific HTA. As shown in Figure 1, both single and multiple V1/V2 populations were detected in paired fluids. A single viral population was identified by V1/V2-HTA in both fluids from 10 (58.8%) subjects. Two or more discrete populations were detected in the blood of seven subjects (41.2%) and in the saliva of five subjects (29.4%). Multiple variants were detected in plasma but not saliva of two subjects (Z10 and Z16), likely due to the presence of salivary PCR inhibitors that limit the amplification of low copy number templates in oral secretions . Overall, the number of discrete viral populations identified in saliva mirrored plasma in the study cohort.
The relationship between the timing of PHI-related onset of symptoms and the detection of multiple variants was assessed by grouping subjects as either early (25 days or less following symptoms onset) or late (more than 25 days following symptoms onset) onset and then comparing the V1/V2 HTA banding patterns of the two groups. This cut-off point corresponds to a natural breakpoint in the frequency distribution of the duration between symptoms onset and sample collection for the cohort (data not shown), and approximates the median number of days between these two events. The cut-off point also reflects the typical duration between the appearance of clinical symptoms of acute infection (estimated at 14 days following initial infection)  and the development of HIV-1 antibodies (estimated at 6 weeks following initial infection) among the majority of newly infected individuals [7,16]. In this analysis, seven subjects in the early group had single V1/V2 bands compared to three subjects in the later group, while two subjects in the early group had multiple bands compared to five subjects in the later group (compare Table 1 and Fig. 1). The non- significant trend (P = 0.088, χ2 test) of subjects in the early group having single rather than multiple variants supports the evolution of a founder strain rather than transmission of multiple variants in this small cohort.
To determine the genetic relatedness of the variants in paired samples, the V1/V2-encoding region of the predominant variant was sequenced for each sample. Amino acid sequences deduced from the nucleotide sequences were aligned to the HIV-1 JR-FL V1/V2 probe sequence and compared within and between subjects. As summarized in Table 2, the amino acid sequences of major bands were genetically identical or similar in paired fluids of all subjects. Sequences were unique for each subject and did not correspond to laboratory-adapted HIV-1 isolates (data not shown), verifying the lack of PCR contamination. These data suggest that highly related viral populations are established in blood and oropharyngeal tissues early in infection.
Relative abundance of major variants in saliva and blood
To determine whether the overall viral quasispecies varied in oral and systemic compartments, the relative abundance of major V1/V2-HTA bands was compared in saliva and blood for subjects Z11 and Z18. These individuals were selected because they harbored at least two distinct V1/V2 populations present at apparent different relative abundances in the fluids (Fig. 1). As shown in Figure 1, Z11 had two major variants that migrated as higher and lower heteroduplex bands. The higher band averaged 78.3% of the variants in blood and 57.2% of the variants in saliva, while the lower band was more abundant in saliva (40.5%) than blood (17.0%). The deduced V1/V2 amino acid sequences of the lower bands were identical in saliva and plasma, and the sequences of the higher bands differed by only one amino acid in the two fluids (Table 2). The sequences of the high and low bands in saliva varied at only two positions, confirming the restricted sequence heterogeneity of the major variants in the oral cavity of this subject.
Samples from Z18 contained four discrete V1/V2-HTA bands whose relative abundances (from top-to-bottom of the gel) averaged 24.7, 5.9, 8.5, and 60.9% in blood and 30.4, 20.7, 17.6, and 31.2% in saliva (Fig. 1). Sequences of the higher and lower migrating bands were each identical in saliva and blood. Sequences of the higher and lower bands differed by a four amino acid insertion, clustered substitutions in V1 and single substitutions in V2 (Table 2), for a total divergence of 1.2%.
Genetically distinct variants are present in a sexual transmission pair
The study cohort included a transmission pair involving subjects Z10 and Z11 . Z11 acquired his infection following a sexual encounter with Z10 who had documented PHI. The transmission pair was confirmed by phylogenetic analysis of HIV-1 reverse transcriptase sequences . Z10 and Z11 samples analyzed in the current study were obtained 53 and 57 days, respectively, following the reported onset of acute retroviral syndrome. As shown in Figure 1, the V1/V2 HTA banding patterns of Z10 and Z11 were distinct and did not contain a detectable variant shared between the partners. Although the V1/V2-encoding sequences from the couple were more closely related to each other than to other study participants, the major variants in Z10 and Z11 exhibited considerable sequence heterogeneity with one another (Table 2). HIV-1 sequences from plasma and saliva differed at 11 and 12 positions of the 70 amino acid V1/V2 loop, creating an overall sequence diversity of 1.6 to 1.7%. In addition, the predominant Z10 variant included a three-residue insertion in V1 that was not detected in Z11 (Table 2). The discordant banding patterns may be due to either rapid evolution of the Z11 variant following acquisition, or the acquisition of a minor variant representing less than 3% (lower level of HTA sensitivity in blood) of the total viral population of Z10.
R5 NSI variants predominate in both fluids early in infection
To examine coreceptor usage of saliva-derived virus, we analyzed HIV-1 from samples using a V3-HTA [8,10]. As shown in Figure 2 and summarized in Table 3, eight subjects appeared to have R5-like variants and three subjects had X4-like variants. To verify the X4 banding pattern, the V3 sequences of RT-PCR products from subjects Z13, Z16 and Z27 were compared with the probe sequence (Table 3). The V3 sequences of Z13 and Z16 differed by a single amino acid deletion at position 24 relative to the probe, whereas the V3 sequence of Z27 contained multiple amino acid substitutions at non-key positions. These differences are sufficient to cause the R5-like sequences to migrate similar to ‘true’ X4/SI variants [8,10]. Further analysis of plasma samples by MT-2 cell fusion assay and V3 sequence analyses of plasma and saliva samples (Table 3) confirmed the presence of R5/NSI variants in both fluids of all PHI subjects, supporting reports from blood-based studies [18–20].
In this first report of gp120 diversity, predicted coreceptor usage and SI phenotype of HIV-1 in oral fluids during PHI, we found that these fluids contain V1/V2 and V3 populations that were identical or highly related to those in blood plasma in the 17 subjects studied. The high degree of sequence concordance strengthens our earlier proposition that viral quasi-species in plasma and oral fluids share a common viral source . These findings are in agreement with previous analyses of V3 env sequences from chronically infected individuals, which have shown with rare exceptions  that the same variants usually exist in both compartments [8,21]. Together with previous data showing that viral loads in oral fluids generally mirrored those in plasma during acute  and chronic [2,22,23] HIV-1 infection, these findings support a view that there is rapid trafficking of virus from systemic sources into oral fluids during acute infection.
The majority (58.8%) of subjects harbored single, unique R5 NSI variants in both fluids. This finding adds further support to previous descriptions of restricted viral heterogeneity in blood near the time of seroconversion among subtype B-infected men [18–20,24–30] and, to a lesser extent, African women infected with subtypes A and C [31–33].
Although the V1/V2 and V3 sequences were highly related or identical within all subjects, we did detect slight differences in the relative abundance of related variants in saliva and blood in two subjects. This finding does not appear to be due to sampling, as the differences were consistently observed in multiple HTA gels using two or more independent RT-PCR reactions for each sample (data not shown). It is possible that site-specific pressures (e.g., local immune responses, endogenous antiviral factors) influence the overall viral composition in oral and systemic compartments. These local pressures, however, do not exclude blood-derived variants from the oral cavity and seem to have little effect on the overall sequence heterogeneity of virus in the two compartments very early in HIV-1 infection. Non-specific and cytotoxic T-cell responses may contribute to the apparent differences in the abundance of predominant viral populations in the fluids of these subjects, and this possibility will be pursued in further studies.
In all of the subjects, complete agreement was found between the coreceptor usage pattern predicted by the V3 amino acid sequence and the MT-2 assay (Table 3). In contrast, the V3-HTA analysis of samples from three subjects (Z13, Z16, Z27) demonstrated discordance between the coreceptor usage pattern predicted by the HTA and these methods of determining X4/R5 genotype and SI phenotype. Sequence analyses revealed that the discrepancies were due to coding changes corresponding to either single (Z13, Z16) or multiple (Z27) amino acid substitutions at non-predicted sites within V3. The base mismatches formed kinks in the heteroduplex secondary structure that slowed its migration through the non-denaturing gel and resulted in the misclassification of the R5 sequences as X4 sequences (compare Fig. 2 and Table 3). In our earlier comparative study of a chronically infected cohort , three of 11 (27.3%) HIV-1- positive subjects demonstrated discordance between X4/R5 genotype predicted by V3-HTA and the V3 sequences. Other investigators [10,34] have reported that X4/R5 genotype can be predicted by V3-HTA analysis in most, but not all, subjects studied to date. These collective findings stress the need of confirming the findings from V3-HTA analysis with more standard methods of determining X4/R5 genotype and SI phenotype.
The rapid shedding of concentrated and generally homogeneous populations of R5 HIV-1 variants into oral fluids may have important implications for sexual transmission. Although oral fluids are themselves associated with low transmission potential due to the presence of endogenous inhibitors in saliva , the mechanisms of HIV-1 shedding into oral fluids may be similar to the mechanisms active at other mucosal surfaces, where virus can be more efficiently transmitted via sexual intercourse. Since homogeneous populations closely related to a recently transmitted strain are presumably infectious, high level HIV-1 shedding on mucosal surfaces may in part explain high sexual transmission potential associated with acute HIV-1 infection postulated in epidemiologic studies [35,36] and documented in transmission case series [17,37]. Further studies are needed to understand the viral kinetics, evolution and responses to antiretroviral therapy in oral tissues compared to mucosal tissues at anorectal and genital sites.
We thank the Duke-UNC-Emory Acute HIV Consortium for patient recruitment and CD4 cell determinations, the UNC Core Retrovirology lab for specimen processing and viral load testing, and L. Gray for technical assistance.
Sponsorship: Funding support was provided by National Institutes of Health (AI07001, DE13603, K24AI01608, K23AI01781); the University of North Carolina, Duke, and Emory Centers for AIDS Research (NICHD/NIAID 9P30-AI50410, 5P30-AI28662, P30-AI12121); and unrestricted grants by Bristol Myers-Squibb and Boehringer Ingelheim.
1. Blankson JN, Persaud D, Siliciano RF. The challenge of viral reservoirs in HIV-1 infection. Ann Rev Med
2. Shugars DC, Slade GD, Patton LL, Fiscus SA. Oral and systemic factors associated with increased levels of human immunodeficiency virus type 1 RNA in saliva. Oral Surg Oral Med Oral Pathol Oral Radiol Endod
3. Pilcher CD, Shugars DC, Fiscus SA, Miller WC, Menezes P, Giner J, et al
. HIV in body fluids during primary HIV infection: implications for pathogenesis, treatment and public health. AIDS
4. Cohen MS, Shugars DC, Fiscus SA. Limits on oral transmission of HIV-1. Lancet
5. Shugars DC, Wahl SM. The role of the oral environment in HIV-1 transmission. J Am Dent Assoc
6. Zhang Z, Schuler T, Zupancic M, Wietgrefe S, Staskus KA, Reimann KA, et al
. Sexual transmission and propagation of SIV and HIV in resting and activated CD4+ T cells
7. Kinloch-de Loes S, de Saussure P, Saurat JH, Stalder H, Hirschel B, Perrin LH. Symptomatic primary infection due to human immunodeficiency virus type 1: review of 31 cases. Clin Infect Dis
8. Freel SA, Williams JM, Nelson JAE, Patton LL, Fiscus SA, Swanstrom R, et al
. Characterization of human immunodeficiency virus type 1 in saliva and blood plasma by V3-specific heteroduplex tracking assay and genotype analyses. J Virol
9. Gorry PR, Bristol G, Zack JA, Ritola K, Swanstrom R, Birch CJ, et al
. Macrophage tropism of human immunodeficiency virus type 1 isolates from brain and lymphoid tissues predicts neurotropism independent of coreceptor specificity. J Virol
10. Nelson JAE, Fiscus SA, Swanstrom R. Evolutionary variants of the human immunodeficiency virus type 1 V3 region characterized by using a heteroduplex tracking assay. J Virol
11. Korber B, Kuiken C, Foley B, Hahn B, McCutchan F, Mellors J, et al
. Human Retroviruses and AIDS: 1998
. Theoretical Biology and Biophysics
1998, Los Alamos National Laboratory: Los Alamos, NM.
12. Devereux J, Haeberli P, Smithies O. A comprehensive set of sequence analysis programs for the VAX
. Nucl Acids Res
13. Milich L, Margolin B, Swanstrom R. V3 loop of the human immunodeficiency virus type 1 Env protein: interpreting sequence variability. J Virol
14. Division of AIDS, National Institute of Allergy and Infectious Diseases. AC7G virology manual for HIV laboratories.
Publication NIH-94-3828. Bethesda, Maryland: National Institutes of Health; 1994.
15. Ochert AS, Boulter AW, Birnbaum W, Johnson NW, Teo CG. Inhibitory effect of salivary fluids on PCR: potency and removal. PCR Methods Appl
16. Kahn JO, Walker BD. Acute human immunodeficiency virus type 1 infection. N Engl J Med
17. Pilcher CD, Eron JJ Jr, Vernazza PL, Battegay M, Harr T, Yerly S, et al
. Sexual transmission during the incubation period of primary HIV infection. JAMA
18. Roos MT, Lange JM, de Goede RE, Coutinho RA, Schellekens PT, Miedema F, et al
. Viral phenotype and immune response in primary human immunodeficiency virus type 1 infection. J Infect Dis
19. Zhang LQ, MacKenzie P, Cleland A, Holmes EC, Leigh Brown AJ, Simmons P. Selection for specific sequences in the external envelope protein of human immunodeficiency virus type 1 upon primary infection. J Virol
20. Zhu T, Mo H, Wang N, Nam DS, Cao Y, Koup RA, et al
. Genotypic and phenotypic characterization of HIV-1 in patients with primary infection. Science
21. Kakizawa J, Ushijima H, Morishita Y, Oka S, Ikeda Y, Muller WE. Diversity of HIV type 1 envelope V3 loop region in saliva. AIDS Res Hum Retroviruses
22. Phillips J, Qureshi N, Barr C, Henrard DR. Low level of cell-free virus detected at high frequency in saliva from HIV-1-infected individuals. AIDS
23. Liuzzi G, Chirianni A, Clementi M, Bagnarelli P, Valenza A, Cataldo PT, et al
. Analysis of HIV-1 load in blood, semen and saliva: evidence for different viral compartments in a cross-sectional and longitudinal study. AIDS
24. McNearney T, Hornickova Z, Markham R, Birdwell A, Arens M, Saah A, et al
. Relationship of human immunodeficiency virus type 1 sequence heterogeneity to stage of disease. Proc Natl Acad Sci USA
25. Wolfs TG, Zwart G, Bakker M, Goudsmit J. HIV-1 genomic RNA diversification following sexual and parenteral virus transmission. Virology
26. Delwart EL, Sheppard HW, Walker BD, Goudsmit J, Mullins JI. Human immunodeficiency virus type 1 evolution in vivo tracked by DNA heteroduplex mobility assays. J Virol
27. Wolinsky SM, Korber BT, Neumann AU, Daniels M, Kunstman KJ, Whetsell AJ, et al
. Adaptive evolution of human immunodeficiency virus-type 1 during the natural course of infection. Science
28. Karlsson AC, Lindback S, Gaines H, Sonnerborg A. Characterization of the viral population during primary HIV-1 infection. AIDS
29. Delwart E, Magierowska M, Royz M, Foley B, Peddada L, Smith R, et al
. Homogeneous quasispecies in 16 out of 17 individuals during very early HIV-1 primary infection. AIDS
30. Learn GH, Muthui D, Brodie SJ, Zhu T, Diem K, Mullins JI, et al
. Viral population homogenization following acute human immunodeficiency virus type 1 infection. J Virol
31. Poss M, Martin HL, Kreiss JK, Granville L, Chohan B, Nyange P, et al
. Diversity in virus populations from genital secretions and peripheral blood in women recently infected with human immunodeficiency virus type 1. J Virol
32. Poss M, Rodrigo AG, Gosink JJ, Learn GH, de Vange Panteleeff D, Martin HL Jr, et al
. Evolution of envelope sequences from the genital tract and peripheral blood of women infected with clade A human immunodeficiency virus type 1. J Virol
33. Long EM, Martin HL Jr, Kreiss JK, Rainwater SM, Lavreys L, Jackson DJ, et al
. Gender differences in HIV-1 diversity at time of infection. Nature Med
34. Ping L-H, Nelson JAE, Hoffman IF, Schock J, Lamers SL, Goodman M, et al
. Characterization of V3 sequence heterogeneity in subtype C human immunodeficiency virus type 1 isolates from Malawi: underrepresentation of X4 variants. J Virol
35. Jacquez JA, Koopman JS, Simon CP, Longini IM Jr. Role of the primary infection in epidemics of HIV infection in gay cohorts
. J Acquir Immune Defic Synd
r 1994; 7
36. Leynaert B, Downs AM, de Vincenzi I. Heterosexual transmission of human immunodeficiency virus: variability of infectivity throughout the course of infection
. European Study Group on Heterosexual Transmission of HIV. Am J Epidemiol
37. Yerly S, Vora S, Rizzardi P, Chave JP, Vernazza PL, Flepp M, et al
. Acute HIV infection: impact on the spread of HIV and transmission of drug resistance