Role of Semen in HIV-1 Transmission: Inhibitor or facilitator?


Gustavo F. Doncel, CONRAD-Department of Obstetrics and Gynecology, Eastern Virginia Medical School, Norfolk, VA 23507, USA.


Citation Doncel GF, Joseph T, Thurman AR. Role of semen in HIV-1 transmission: inhibitor or facilitator? Am J Reprod Immunol 2011; 65: 292–301

Sexual transmission of human immunodeficiency virus type 1 (HIV-1) accounts for 60-90% of new infections, especially in developing countries. During male-to-female transmission, the virus is typically deposited in the vagina as cell-free and cell-associated virions carried by semen. But semen is more than just a carrier for HIV-1. Evidence from in vitro and in vivo studies supports both inhibitory and enhancing effects. Intrinsic antiviral activity mediated by cationic antimicrobial peptides, cytotoxicity, and blockage of HIV–dendritic cell interactions are seminal plasma properties that inhibit HIV-1 infection. On the contrary, neutralization of vaginal acidic pH, enhanced virus–target cell attachment by seminal amyloid fibrils, opsonization by complement fragments, and electrostatic interactions are factors that facilitate HIV-1 infection. The end result, i.e., inhibition or enhancement of HIV mucosal infection, in vivo, likely depends on the summation of all these biological effects. More research is needed, especially in animal models, to dissect the role of these factors and establish their relevance in HIV-1 transmission.

Mechanisms of genital mucosal infection by HIV-1 and role of semen

Sexual transmission of human immunodeficiency virus type 1 (HIV-1) accounts for 60–90% of new infections, especially in developing countries.1 During male-to-female transmission, the virus is typically deposited in the vagina as cell-free (CF) and cell-associated (CA) virions carried by semen. The efficiency of transmission is variable, ranging from 0.1 to 0.001% depending on co-existing risk factors such as stage of disease in the male, seminal viral load, and sexually transmitted infections (STIs) and other cervico-vaginal (CV) infections in the female. The surface of the CV mucosa provides a large portal of entry for HIV-1. The virus has been shown to penetrate several layers from the luminal surface into the thin gaps between squamous epithelial cells.2 This penetration may bring the virus in direct contact with two key cell types presumably involved in the initial stages of mucosal infection: intraepithelial Langerhans cells and CD4+ T lymphocytes. In addition, the virus may reach basal epithelial cells that are susceptible to viral binding, endocytosis, or transcytosis, or may penetrate even further, reaching subepithelial targets, such as T cells and dendritic cells (DCs), through breaches in the epithelium caused by microabrasions.3,4

Utilizing single-genome amplification and mathematical modeling, it has been reported in several patient cohorts and non-human primates that most (60–90%) mucosal infections originate from single-variant transmissions.5,6 The small, focally infected population is initially composed mainly of resting CD4+ T cells lacking conventional markers of activation.7 HIV-1 expands locally in these ‘resting’ and in activated CD4+ T cells, and then disseminates, initially to the draining lymph node and subsequently to secondary lymphoid organs, to generate a systemic infection. Exposure of reproductive tract epithelium to virus increases the expression of chemokines that recruit plasmacytoid dendritic cells (pDCs).8 They in turn recruit, through secretion of additional chemokines, more CD4+ T cells that fuel local expansion. Interferons and chemokines from the pDCs also suppress viral replication, but the balance is tipped in favor of the virus by the cells that fuel the local expansion necessary for dissemination and establishment of systemic infection.

Pre-existing inflammation, caused by lower genital tract infections such as bacterial vaginosis (BV) and trichomoniasis, also facilitates infection by thinning and disrupting the multilayered lining, recruiting a pool of target cells for local HIV expansion, initiating clinical or sub-clinical inflammation, and interfering with innate antimicrobial activity.9 Recruitment and activation of new HIV-1 target cells increase the chances of infection as they provide more permissive cells expressing receptors and co-receptors for HIV.10 Furthermore, cellular products generated during inflammation, e.g., nuclear factor-kB (NF-kB) and IL-6, directly facilitate HIV-1 replication by interacting with genetic elements controlling proviral DNA transcription.11

Semen represents the main vector for HIV-1 transmission worldwide. It contains three major sources of infectious virus: free virions, infected leukocytes, and spermatozoa-associated virions. It is difficult to separate the contribution of CF and CA HIV-1 to sexual transmission, as sexual exposure in humans includes both. The infectiousness of semen is influenced by several factors including stage of the disease and duration of infection in the male, with viral loads peaking in the very early stages of infection or end-stage disease.12,13 Semen viral load typically peaks to about 4.5 ± 0.4 log10 copies/mL after initial infection and stabilizes after approximately 16 weeks of infection.13 Other factors such as coexisting herpes simplex virus type 2 (HSV-2)14 also increase genital shedding and seminal viral load of HIV-1. Highly active antiretroviral therapy (HAART) serves to decrease viral load in the blood and to some extent in semen,15 but a non-detectable viral load in the serum does not guarantee that HIV-1 will be absent from the semen. This is in part because of the anatomical sites, which are the source of seminal HIV-1. Anatomical features of the male reproductive tract and the limited access of the immune system to compartments containing germ cells suggest that HIV-1 in semen may originate from different compartments. Most CF HIV-1 in seminal plasma arises from sites distal to the vas deferens.16 Therefore, vasectomized men are still able to transmit HIV-1. HIV-1-infected leukocytes in semen do not parallel those found in serum and appear to arise from a genetically distinct compartment. Recent studies indicate that HIV-1 in men without urethritis or prostatitis comes from sources in the male genital tract, which are distal to the prostate, further supporting a separate viral reservoir for seminal fluid and plasma HIV-1.

Is semen anything more than a carrier for HIV?

Unprotected sexual intercourse between discordant couples is the most common route of HIV-1 transmission.3 Despite this, it is known that the transmission of HIV-1 without other cofactors is poorly efficient. Several cofactors such as genital ulcer disease, BV,17 HSV-218 trichomoniasis9, and male circumcision19,20 have been shown to alter the efficiency of a productive HIV-1 infection. Other cofactors including race, age, menopausal status, parity, and environmental exposures such as hormones (e.g. contraceptive methods) and tobacco use likely affect the susceptibility of a host to HIV-1 infection, but less evidence exists regarding these variables. The fact that the risk of infection is low and highly variable suggests that several processes are involved in sexual transmission of the virus. At the biological level, enhancing and inhibitory factors are present in semen and female genital tract secretions. The summation of their effects acting in concert with seminal viral load, viral fitness, and the structural and functional state of the CV mucosa will determine the chances for HIV-1 to establish a productive infection. Semen itself is clearly more than a vector for HIV-1. Seminal factors facilitating or inhibiting viral infection include cationic peptides with antiviral activity, cytotoxic molecules, amyloid fibrils derived from seminal phosphatases, complement fragments and prostaglandin E2 (PGE2) and bioactive peptides responsible for inducing mucosal inflammatory reactions (Table I). All of these interacting processes need to be considered to better understand HIV-1 mucosal transmission and devise strategies for prevention. The effect of semen and seminal plasma (SP) warrants further investigation into in vitro and in vivo models of sexual transmission of HIV-1 to elucidate their role, relevance, and mechanisms of action.

Table I.   Evidence for the role of semen and/or seminal plasma (SP) in the efficiency of HIV-1 male-to-female transmission
In vitro. Cationic polypeptides contained in SP contribute to its aggregate anti-HIV-1 activity 32
In vitro. SP/semen inhibited HIV-1 (RF and IIIB) infection of T-lymphocyte cell line C8166. Inhibitory concentrations were 35- to 50-fold lower than cytotoxic concentrations 31
In vitro. SP contains a potent inhibitor of the attachment of HIV-1 to DC-specific intercellular adhesion molecule 3-grabbing non-integrin. Significant inhibition was observed using SP dilutions as high as 1:105. Inhibitor was greater than 100 kDa, heat stable and trypsin resistant 33
In vivo. Rhesus macaque model. Titration study to determine viral inoculum concentration and effect of seminal plasma on the efficiency of intravaginal infection with SIV. There wasa trend toward an inhibitory effect of SP at lower-titer viral inocula 36
In vitro. After ejaculation, the pH of the vaginal fluid increases to neutral or higher pH within 30 s. In vitro incubation of HIV-1 (RF and IIIB) in a range of buffer systems (pH = 3.5–8.0) with C8166 cells found that HIV-1 strains were uniformly stable at pH of 5.0–8.0, with mild reduction in infectivity (25%) at pH 4.5 31, 44, 46, 47
In vitro. Semen-derived enhancer of virus infection (SEVI), amyloid fibrils formed by prostatic acidic phosphatase fragments, capture HIV virions and promote their attachment to HIV-1 target cells, such as CD4+ T lymphocytes and dendritic cells. Appropriate co-receptors (CCR5, CXCR4) are required 48, 54
In vitro. R5- and X4- HIV-1 strains activate complement in seminal fluid and generate C3 cleavage fragments, which enhance infection in colorectal cell lines (HT-29) 61

Evidence supporting an inhibitory effect of semen on HIV transmission

Redox Activity of Seminal Plasma

It is thought that the oxidation of SP polyamines by diamine oxidase,21 augmented by peroxidases present in a healthy vaginal environment, produces radicals that inactivate HIV-1. The virus, in particular the lipids contained in its envelope, is highly sensitive to oxygen radicals.22 Semen produces reactive oxygen species,23 which can alter the infectivity of HIV. A normal healthy vagina also contains lactobacilli-produced hydrogen peroxide (H2O2), which maintains a low level of virucidal activity.24In vitro studies demonstrate that at concentrations where H2O2-producing lactobacilli levels are not virucidal, the addition of peroxidase, such as myeloperoxidase or eosinophil peroxidase and a halide (chloride, iodide, bromide, thiocyanate), can restore anti-HIV-1 activity.25 Data from the 1970s also support that several viruses are inactivated by polyamine oxidation products.26–29

Natural Antiviral Activity of Semen is Attributed to Cationic Polypeptides Contained in SP

Cationic antimicrobial polypeptides, such as secretory leukoprotease inhibitor, defensins and lactoferrin, produced by mucosal surfaces from the oral and CV tracts, have been identified and found to have varying levels of antibacterial and anti-HIV-1 activity.30 O’Connor et al.31 demonstrated in vitro that semen, and specifically SP, had antiviral activity against HIV-1. Semen showed consistent activity against HIV-1, and the inhibitory concentration was between 35- and 50-fold lower than the cytotoxic concentration.31 In further experiments, Martellini et al.32 demonstrated that SP contained 52 individual cationic polypeptides, which contributed to its aggregate anti-HIV-1 activity, and that SP maintained anti-HIV-1 activity, even when diluted 3200-fold. However, this phenomenon was transient, as whole SP incubated for over 24 hr exhibited a reduction in anti-HIV-1 activity.

SP Interferes with the Attachment of HIV-1 to DC-SIGN-Positive Cells

In order for a male-to-female HIV-1 exposure to become a productive infection, the virus must cross an epithelial surface to interact with T lymphocytes, macrophages, and DCs, which are the main targets of infection. DC-specific intercellular adhesion molecule 3-grabbing non-integrin (DC-SIGN) is expressed by DCs at mucosal surfaces and appears to be a universal pathogen receptor, which protects HIV-1 from degradation and efficiently promotes infection of T CD4+ cells.33,34 DC projections may extend to, or near, the luminal surface and present antigens to lamina propria target cells. This is why genital ulcerations35 or any breach of epithelial integrity, including micro-trauma that can exist after consensual intercourse,4 heightens the risk of HIV-1 transmission.

SP contains a potent inhibitor of the attachment of HIV-1 to DC-SIGN, which inhibits the capture and transmission of HIV-1 to T CD4+ cells.33 A significant inhibition of HIV-1 capture was observed for both HIV-1 IIIB (CXCR4) and HIV-1 BaL (CCR5) using SP dilutions as high as 1:104.33 The effect of SP was not related to cell cytotoxicity, as cell viability was higher than 90% in these models.33 This group also incubated HIV-1 with B-THP-DC-SIGN cells and found that SP in dilutions up to 1:103 diminished capture of HIV-1 IIIB and HIV-1 BaL to the levels observed for DC-SIGN negative cells, while significant levels of inhibition were observed even at SP dilutions as great as 1:105.33 Monocytes, activated PBMCs, and the T cell line SupT-1 (all of which do not express DC-SIGN) were used as negative controls. Capture of HIV-1 by these cell populations was not inhibited by SP, supporting that CD4-dependent mechanisms of HIV-1 capture are not inhibited by SP. Using structural analysis, it was determined that the component of SP with inhibitory effects on DC-SIGN had a molecular weight greater than 100 kDa and was heat stable and resistant to the action of trypsin.33 SP, like HIV-1, can gain access to sub-epithelial target cellsand decrease the efficiency of HIV-1 transmission via DC-SIGN.

Effects of SP in Animal Models

Using a rhesus macaque model, Miller et al.36 tested the effects of SP on the efficiency of CF SIV transmission. In general, higher viral inoculums produced persistent viremia in monkeys, with or without the presence of SP. At lower viral load inoculums (e.g., 102 or 10 TCID50), the addition of SP showed a trend toward increasing the efficiency of persistent viremia among animals inoculated with SIV-mac251 grown in huPBMC stock. However, this trend was not clearly demonstrated among animals receiving SIV-mac251 grown in rhPMBcs.36

CA virus is also believed to be an important source of HIV-1 sexual transmission, but may be less efficient at crossing the CV mucosa when compared to CF virus.37,38 Semen of treatment-naïve infected men contains a significant number of infected leukocytes (from 3 × 104 to 5.6 × 107 cells/mL, between 10 000 and 80 000 HIV DNA copies/mL).39 Recently, Salle et al.37investigated intravaginal administration of CA SIV prepared from spleen cells obtained directly from two cynomolgus macaques infected with SIVmac251. This experimental design was thought to more accurately reflect the CA HIV-1 present in semen of infected men. Inoculated macaques (n = 9) were pre-treated with depot medroxyprogesterone acetate to thin the vaginal epithelium. The three macaques inoculated with 107 cells became persistently infected, but persistent infection was not detected in animals inoculated with lower concentrations of CA SIV (4.2 × 105–3.5 × 104), which is in agreement with previous findings.38 CA HIV-1, such as infected leukocytes in semen, needs to migrate and penetrate between epithelial cells to infect underlying HIV-1 target cells. This has been demonstrated in vitro and in vivo in a mouse model.40 The macaque data parallel epidemiologic evidence which shows that the efficiency of HIV-1 transmission is increased 10-fold during acute infection, when the semen viral load provided by CF and CA virus is at its highest.41

Evidence supporting an HIV enhancing effect of semen on HIV transmission

Neutralization of Normal Acidic Vaginal pH

The healthy vagina is colonized with lactobacilli, which produce lactic acid and H2O2. H2O2-producing lactobacilli have been shown to play a crucial role in maintaining normal vaginal flora and inhibiting the growth of pathogens.24,42,43 Lactobacillus-produced lactic acid creates an acidic pH in the normal vagina, which helps maintain the resident microbiome and combat pathogens.42 CF and CA HIV-1 are rapidly inactivated in vitro at acidic pH levels.44 O’Connor et al.31 demonstrated that laboratory strains of HIV-1 were uniformly stable at pH of 5.0–8.0, with mild reduction in infectivity (25%) at pH 4.5. The pH of semen is 7.0–8.4.45 After ejaculation, semen increases the pH of the vaginal fluid to neutral or higher levels within 30 s, maintaining an increased pH level for up to 2 hr.46,47 Thus, semen can facilitate HIV-1 infection by raising vaginal pH, allowing CF and CA HIV-1 to survive in a less acidic vagina.

Semen-derived Enhancer of Virus Infection ‘SEVI’

Screening a complex peptide/protein library derived from human seminal fluid to determine possible inhibitors and enhancers of HIV-1 infection, Munch et al.48 found semen-derived enhancer of virus infection (SEVI), or semen-derived enhancer of virus infection, a term used for amyloid fibrils formed by the abundant semen marker prostatic acidic phosphatase (PAP) fragments. These amyloid fibrils are similar to amyloid fibrils associated with Alzheimer’s disease, which have also been previously shown to enhance HIV-1 infection.49 PAP is a protein produced by the prostatic gland and secreted in large amounts (1–2 mg/mL) in seminal fluid.48 Elevated levels of PAP can be detected in the vagina for up to 24 hr after sexual intercourse.50 The predominant form of the PAP fragment in the amyloid fibrils was a 4551-Dalton peptide, which corresponded to amino acids 248–286 of PAP. This fragment has eight basic residues, which make it highly cationic (isoelectric point = 10.21), an important property for its attachment effects.51,52 These amyloid fibrils appear to capture HIV virions and promote their attachment to HIV-1 target cells, thereby enhancing the infectiousness of the virus by orders of magnitude.48 The positively charged SEVI fibrils bind to both target cells and HIV-1 virions and augment infection by increasing physical contact between these entities, similar to the manner in which synthetic cationic polymers promote retrovirus attachment to target cells.51,53

SEVI significantly enhances binding of wild-type HIV-1 particles and virions lacking Env, although the absolute levels of CA p24 are about 30-fold lower in the absence of Env.48 SEVI enhances in vitro HIV infection in a dose- and time-dependent manner, and its effects are seen across different envelopes.54 Infection enhancement, however, appears to be donor dependent.54 Further experiments showed that SEVI enhanced infection with R5-, X4- and dual-tropic HIV-1 clones. Importantly, the enhancing effect of SEVI was most pronounced at low concentrations of virus, resembling conditions of sexual HIV-1 transmission.48 In general, these authors stated that SEVI may promote virus attachment to genital surfaces, penetration of the mucosal barrier, and subsequent dissemination to lymphoid organs by increasing HIV-1 virion binding to epithelial cells and to migrating DCs.48 This is in accordance with confocal microscopy data that shows the presence of seminal fluid enhances binding of virions to epithelial cells in ex vivo CV tissue.55 Using dose/response assays, it was determined that 1–3 virions, in the presence of SEVI, are sufficient for productive HIV-1 infection of PBMCs.48

The effect of SEVI enhancement was tested in hCD4/hCCR5-transgenic rats inoculated with either HIV-1 YU2 or SEVI-treated HIV-1.48 Tail vein inoculation with SEVI-treated HIV-1 increased the cDNA copy numbers in splenectomy extracts by fivefold.48 Further testing of SEVI in animal models is warranted, as reproducibility of the enhancing effect in vitro varies according to the laboratory and assay conditions employed, casting doubts about the relevance of this phenomenon.

Role of Electrostatic Interactions in HIV-1 Enhancing Effects of Semen

Another possible enhancing effect of semen is mediated by electrostatic interaction of spermatozoa with HIV-1 virions, involving negatively charged heparin sulfate. This complex can transmit virus directly to DC-SIGN on DCs.56 Once the spermatozoa are internalized by DCs, the DCs undergo phenotypic maturation and produce IL-10.56 Other receptors on spermatozoa may also be involved. Roan et al.51 hypothesized that SEVI, because of its highly cationic nature, may bind to target cells by interacting with cell-surface heparan sulfate proteoglycans (HSPG), naturally occurring anionic carbohydrate polymers that are closely related in structure to heparin sulfate. They hypothesized that HSPG antagonists would inhibit the viral enhancing effects of SEVI.51 Surfen, a HSPG antagonist, induced a dose-dependent inhibition of SEVI at concentrations of 6.25 μm with the maximal inhibitory plateau occurring at 50–100 μm.57 Surfen appeared to directly inhibit SEVI and not compromise the infectivity of the virions.57

Electrostatic interactions between SP and microbicides may also hamper the efficacy of HIV-1 prevention products. The antiviral activity of several anionic polymer microbicide candidates (e.g. cellulose sulfate, carrageenan) was reduced 4- to 73-fold in the presence of SP.58 The reduction in antiviral capacity in the presence of SP may in part be explained by electrostatic interactions between cationic SP polyamines and the polyanions of the microbicide candidates. This reduction in the inhibitory activity of polyanionic microbicides has also been observed in clinical trials.59,60

Opsonization by Complement Fragments Enhances HIV Infection

Semen from HIV-1-positive individuals contains CF HIV-1 particles and soluble complement components.61 Opsonization with complement was previously shown to enhance HIV-1 infection of T and B cells, monocytes and macrophages.61 Complement receptors are expressed on the apical surface of epithelial cells, DCs, and macrophages.61 Bouhlal et al.61 showed that both R5- and X4-tropic HIV-1 strains can activate complement in seminal fluid in vitro. They found that enhancement of HIV-1 infection in colorectal cell lines (HT-29) was complement dependent. Infection of HT-29 cells with HIV-1 that was pre-opsonized with complement (C3 and C9) in seminal fluid resulted in an enhanced (1.5–2-fold) rate of HIV-1 infection compared to infection of these cells in the presence of virus alone.61 R5- and X4- strains activate complement in seminal fluid and generate C3 cleavage fragments (C3a/C3adesArg).61

Proinflammatory Effects of Semen

The immediate reaction of semen deposition into the mammalian reproductive tract is a dramatic influx of inflammatory cells.62–64 Changes in the leukocyte population of the female reproductive tract (FRT) after introduction of the male ejaculate have been well documented in mice, pigs, rabbits, and women.63,65–67 Most of these pro-inflammatory effects in animals are attributed to the presence of transforming growth factor (TGF)-β in SP.68,69 The majority of TGF-β present in male SP is synthesized in latent form and appears to be activated by plasmin and other enzymes in the FRT.69

Women respond to semen deposition with a similar influx of leukocytes, especially to the cervix, called leukocytic reaction. These leukocytes predominantly include neutrophils and to a lesser extent macrophages and T lymphocytes.63,64 SP is also considered a cause of recurrent vaginitis in certain sexually active women, a condition possibly related to SP protein allergy and associated with localized irritation and inflammation.70 The etiology of this inflammatory response, however, is not well understood.

The semen-induced leukocyte influx to the FRT is believed to be mediated by chemoattracting factors released by the epithelial lining of the FRT in response to sperm and SP.62 Although a transient, semen-induced inflammation of the FRT is probably necessary for a successful establishment of pregnancy, it also recruits and activates HIV target cells to the portals of virus entry, thus facilitating mucosal infection and HIV transmission.

SP induces differential expression of inflammatory genes in human cervical and vaginal epithelial cells.71 In ectocervical cells, these genes include IL-8, IL-6, CSF2, CCL2, GM-CSF, and MCP-1. Our laboratory has demonstrated that human SP also enhances the secretion of proinflammatory factors such as GM-CSF, IFN-γ, IL-12p70, IL-1β, IL-6, IL-2 and IL-8 by human vaginal cells (Joseph et al., manuscript in preparation). We and Berlier et al.72 have demonstrated that SP also induces the expression of CCL20, a key chemotactic factor involved in recruitment and maturation of Langerhans cells and dendritic cells, which, together with intraepithelial T lymphocytes, are considered to be the first target cells for HIV genital mucosal infection.73–75

A common gene overexpressed in pathological conditions involving mucosal inflammation is cyclooxygenase (COX)-2. Semen exposure leads to overexpression of COX-2 in pig and mare endometrium.76,77 COX-2 catalyzes the rate-limiting step in the synthesis of prostaglandins from arachidonic acid.78 Prostaglandins are considered to be important biological modulators of inflammation. They attract immune cells to the area of inflammation. They also act in an autocrine/paracrine manner to elevate COX-2 expression.79,80 Seminal plasma contains 1000-fold higher concentration of prostaglandins, mainly PGE2, compared to normal endometrium.81 Seminal plasma PGE2 has been reported to induce COX-2 in immortalized human endocervical cells.82 This induction is because of the intracellular activation of cAMP pathway via PGE2 receptor subtypes, EP2 and EP4.

Our laboratory has demonstrated that SP also induces COX-2 in human vaginal cells (Joseph et al., manuscript in preparation). Furthermore, it potentiates COX-2 induction by microbial products such as bacterial lipopeptides (Fig. 1). This enhanced expression of COX-2 could be one of the main causes of inflammation associated with STIs and CV infections.

Figure 1.

 Seminal plasma induces expression of cyclooxygenase (COX)-2 in human vaginal cells. Representative immunoblot showing COX-2 expression in vaginal cells induced by seminal plasma (1%) alone and in combination with bacterial lipopeptide. Seminal plasma potentiates COX-2 induction by the bacterial lipopeptide.

In addition, SP has been shown to activate multiple signal transduction pathways, which are involved in inflammatory responses. In cervical cells, SP induces the phosphorylation of extracellular signal-regulated kinase (ERK1/2) via EP4 receptor.83 In endometrial cells, SP induces the phosphorylation of c-Src, ERK, and activation of cAMP pathway via EP2 receptor.84 SP has also been shown to activate NF-kB signaling pathway in vaginal cells. This pathway is considered central to inflammation and is involved in the control of numerous proinflammatory genes including COX-2 and multiple chemokines and cytokines. NF-kB activation has also been linked to the enhancement of HIV replication.11

Summary and conclusions

The role of semen in HIV-1 transmission is defined by a complex array of factors and processes involved in semen, virus, and female genital tract interactions. Semen carries CF and CA virus and is believed to be the main vector for HIV-1 in male-to-female sexual transmission. Seminal viral load varies with multiple factors such as stage of infection and disease in the male, presence of reproductive tract inflammation, and whether or not the man is on antiretroviral therapy. However, semen is more than a carrier for HIV. Evidence from in vitro and in vivo studies supports both inhibitory and enhancing effects (Table I and Fig. 2). Intrinsic antiviral activity mediated by cationic antimicrobial peptides, cytotoxicity, and interference of HIV-DC interaction are seminal properties that inhibit HIV infection. On the opposite side, neutralization of vaginal acidic pH increased viral attachment by amyloid fibrils (SEVI), opsonization by complement fragments, and recruitment and activation of HIV target cells to mucosal portals of virus entry are factors that facilitate HIV infection. The end result, i.e., inhibition or enhancement of HIV-1 mucosal infection, in vivo, depends on the summation of all these biological effects. More research is needed, especially in animal models, to elucidate the role of these factors and establish their relevance for sexual transmission of HIV-1.

Figure 2.

 Semen-associated factors enhancing or inhibiting male-to-female transmission of HIV-1. This cartoon depicts a series of semen-associated factors inhibiting (red) or enhancing (green) cervicovaginal mucosal infection by cell-free and cell-associated HIV-1 carried by semen. Among inhibiting factors are semen’s intrinsic antiviral activity, leukocyte toxicity, and interference with binding of virions to DC-specific intercellular adhesion molecule 3-grabbing nonintegrin. On the contrary, vaginal pH neutralization, enhanced HIV attachment by semen-derived enhancer of virus infection (SEVI) fibrils, HIV opsonization by complement, and induction of proinflammatory mediators would favor HIV transmission and mucosal infection. The bottom right and left insets show a cartoon of potential portals of entry for the virus. Local amplification of a small population of founder viruses gives rise to disseminating virus that reaches the draining lymph nodes (LN) and, through blood vessels (B), systemic circulation, determining the initial stage of an established HIV-1 infection.


This work was supported by CONRAD intramural funds (GD) from the US Agency for International Development (grant GPO-8-00-08-00005-00) and the Bill and Melinda Gates Foundation (grant 41266). The views of the authors do not necessarily represent those of their funding agencies. The authors are also grateful to Nancy Gonyea for her assistance in the preparation of this manuscript.