Introduction

Topically applied vaginal microbicide gels have been tested as a strategy for preventing sexual human immunodeficiency virus (HIV)-1 transmission to women for more than 2 decades. Adolescent women in Africa are at particularly high risk of being infected by HIV, and microbicide gels offer an important female-controlled prevention tool for this group. In Southern Africa, HIV incidence rates peak among women aged 15–24 years compared with those aged 25–39 years in their male counterparts. Rates of HIV infection are reported to be three times higher in this group of young women than they are in similarly aged men, with women on average becoming infected 5–7 years earlier than men.

First-generation vaginal microbicide gels, nonoxynol-9 (N-9) and cellulose sulfate, were not effective at preventing HIV infection. In fact, these gels were found to irritate the vaginal epithelium and, upon repeated application, to induce genital inflammatory cytokine responses. Therefore, rather than being protective, these gels may have actually increased the risk of HIV-1 infection among the woman that used them. Contrasting these early microbicial gel failures has been the TVF (1%) gel. This was the first antiretroviral drug containing microbicide to be tested, and to show 39% protection against male-to-female sexual transmission of HIV in a large-scale phase IIb clinical trial.

In this chapter, we focus on the various anatomical and biological factors associated with HIV infection risks in women, and how the onset of adolescence may alter these risks. We explore the potential effect of the female genital tract milieu on the efficacy of candidate microbicides and other interventions that could potentially improve the efficacy of these microbicides.

Anatomy and immunology of the female genital tract relevant to human immunodeficiency virus infection

The female genital tract is a unique environment that has the ability to respond rapidly to infections, while at the same time being tolerant to allogeneic spermatozoa. Several innate immune features of the female genital tract confer protection against pathogens. The epithelium that lines the lower reproductive tract provides a mechanical barrier to infections. The vaginal mucosa and ectocervix are lined with stratified squamous epithelium and the endocervix with a single layer of columnar epithelial cells, which meets the ectocervix at the squamocolumnar junction or transformation zone.

Although tight junctions between the endocervical columnar epithelial cells act as a barrier to infection, the multiple layers of squamous cells in the lower reproductive tract and continuous sloughing off of the superficial layers are thought to provide more effective protection. The greater surface area of the vaginal mucosa and ectocervix, however, may allow greater access for pathogens.

Epithelial cells in the genital tract produce a protective, hydrophilic layer of glycoprotein called glycocalyx and a hydrophobic glycoprotein mucus. Additionally, they release numerous antimicrobial proteins into the mucosal fluid, express toll like receptors and secrete inflammatory cytokines.

The vaginal mucosa of healthy women contain relatively few leukocytes: CD8+ T cells are the most abundant, whereas only small numbers of cluster of differentiation (CD) 4+ T cells, natural killer cells, macrophages and dendritic cells are present. The ectocervical mucosa and transformation zone are potentially more vulnerable to HIV infection, as these regions contain greater numbers of CD4+ T cells, macrophages and dendritic cells than the vaginal mucosa. However, CD8+ T cells and antigen-presenting cells are also abundant in the transformation zone, suggesting the potential for the initiation of cellular immune responses within this zone. The endocervical epithelium contains the lowest numbers of T cells and macrophages, and no dendritic cells.

Although the uterus is a sterile environment, the vagina and ectocervix are not, and contain roughly 10 ⁹ micro-organisms/ml of genital fluid. Lactobacillus species dominate healthy genital-tract microbial populations and metabolise glycogen released by vaginal epithelial cells to lactic acid. Some lactobacilli also produce hydrogen peroxide, a virucidal agent. Lactic-acid production creates an acidic environment that is less conducive to colonisation by gram-negative bacterial pathogens, and also less susceptible to HIV infection.

Anatomy and immunology of the female genital tract relevant to human immunodeficiency virus infection

The female genital tract is a unique environment that has the ability to respond rapidly to infections, while at the same time being tolerant to allogeneic spermatozoa. Several innate immune features of the female genital tract confer protection against pathogens. The epithelium that lines the lower reproductive tract provides a mechanical barrier to infections. The vaginal mucosa and ectocervix are lined with stratified squamous epithelium and the endocervix with a single layer of columnar epithelial cells, which meets the ectocervix at the squamocolumnar junction or transformation zone.

Although tight junctions between the endocervical columnar epithelial cells act as a barrier to infection, the multiple layers of squamous cells in the lower reproductive tract and continuous sloughing off of the superficial layers are thought to provide more effective protection. The greater surface area of the vaginal mucosa and ectocervix, however, may allow greater access for pathogens.

Epithelial cells in the genital tract produce a protective, hydrophilic layer of glycoprotein called glycocalyx and a hydrophobic glycoprotein mucus. Additionally, they release numerous antimicrobial proteins into the mucosal fluid, express toll like receptors and secrete inflammatory cytokines.

The vaginal mucosa of healthy women contain relatively few leukocytes: CD8+ T cells are the most abundant, whereas only small numbers of cluster of differentiation (CD) 4+ T cells, natural killer cells, macrophages and dendritic cells are present. The ectocervical mucosa and transformation zone are potentially more vulnerable to HIV infection, as these regions contain greater numbers of CD4+ T cells, macrophages and dendritic cells than the vaginal mucosa. However, CD8+ T cells and antigen-presenting cells are also abundant in the transformation zone, suggesting the potential for the initiation of cellular immune responses within this zone. The endocervical epithelium contains the lowest numbers of T cells and macrophages, and no dendritic cells.

Although the uterus is a sterile environment, the vagina and ectocervix are not, and contain roughly 10 ⁹ micro-organisms/ml of genital fluid. Lactobacillus species dominate healthy genital-tract microbial populations and metabolise glycogen released by vaginal epithelial cells to lactic acid. Some lactobacilli also produce hydrogen peroxide, a virucidal agent. Lactic-acid production creates an acidic environment that is less conducive to colonisation by gram-negative bacterial pathogens, and also less susceptible to HIV infection.

Human immunodeficiency virus transmission across the female genital mucosa

Several mechanisms for HIV transmission across the female genital epithelium have been suggested. Although the CD4 receptor is absolutely required by HIV for host cell entry, the virus must also use the co-receptors CC chemokine receptor 5 (CCR5) or CXC chemokine receptor 4 (CXCR4). Genital secretions may contain viruses that use one or other of these co-receptors for cell entry (respectively called either R5 or X4-tropic viruses), and all studies on sexual transmission have found that new infections are established only by R5 tropic viruses.

Cell-free virus may either move between the upper layers of the ‘leaky’ squamous epithelium of the lower genital tract, or translocate through the epithelium through physical abrasions.

Transcytosis of HIV may also occur through polarised columnar epithelial cells. Although transcytosis of HIV through squamous epithelial cells has been suggested, it has been found that HIV does not penetrate intact cervical explants. Also, HIV-infected CD4+ T cells and dendritic cells may migrate into the vaginal and ectocervical epithelium.

In the genital tract, resting and activated CD4+ T cells are the predominant targets of both simian immunodeficiency virus (SIV) and HIV. The earliest targets of HIV in cervical tissue culture are memory CD4+ T cells, whereas macrophages are apparently the predominant target cells in the vaginal lamina propria.

Langerhans cells, which reside in the genital epithelium, express CD4 and CCR5, and may also internalise R5-tropic HIV and transfer the virus to CD4+ T cells. Langerhans cells, however, bind HIV through their langerin receptors and may therefore endocytose and degrade viruses rather than transferring them to CD4+ T cells.

Dendritic cells may capture HIV through their expression of dendritic-specific intercellular adhesion molecule 3-grabbing non-integrin, the mannose receptor or heparin sulphate proteoglycan syndecan-3. Once captured by dendritic cells, HIV may be either endocytosed or transferred to CD4 and CCR5 receptors, resulting in viral-cell membrane fusion and infection. Endocytosed HIV may be degraded or transferred to CD4+ T cells or macrophages.

During sexual transmission in about 75% of cases, new infections are established by a single virion. After sexual transmission, viral RNA is not detectable in blood for about 10 days, a period referred to as the eclipse phase. In SIV-macaque studies, it has been shown that, during this time, viral expansion occurs in a small population of cells within the cervical and vaginal tissues. Although SIV RNA is detectable in systemic lymphoid tissues 1 day after infection, little viral replication occurs here until 5–6 days after infection. It has been suggested that continued viral expansion in the genital tissues and perpetual seeding of the systemic compartment is necessary to achieve a threshold of sustainable infection at distal sites.

Inflammation in the female genital tract enhances susceptibility to HIV infection

A large degree of heterogeneity exists in the susceptibility of individuals to sexually transmitted HIV, with some individuals remaining uninfected despite high levels of exposure. Many studies have investigated the biological factors that may influence susceptibility to HIV infection. Among these, inflammation in the female genital tract is a major factor that is thought to increase the risk of sexual HIV transmission.

Pro-inflammatory cytokines that are involved in the genital recruitment, activation and differentiation of immune cells may increase the probability of HIV transmission, as HIV replication is both absolutely dependent on the presence of susceptible immune cells, and is strongly influenced by degrees of immune-cell activation and monocyte differentiation into macrophages or dendritic cells. In macaques, it has been shown that the pro-inflammatory cytokine-associated recruitment of the CD4+ T cells that are targeted by SIV, is an important prerequisite for the establishment of a productive SIV infection after vaginal inoculation.

Unlike dendritic cells and Langerhans cells, whose maturation is accompanied by down-regulation of CCR5, the pro-inflammatory cytokine-induced differentiation of monocytes into macrophages is accompanied by CCR5 up-regulation. Specifically, pro-inflammatory cytokines induce expression of the transcription factor, nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB). In addition to enhancing the expression of various host-cell proteins that are involved in inflammation, NF-κB cells also binds to the HIV-long terminal repeat sequences and directly up-regulates HIV replication.

Finally, pro-inflammatory cytokines may also facilitate the penetration of free virus through the epithelial barrier by disrupting tight junctions between epithelial cells.

Causes of inflammation in the female genital tract

Inflammation in the female genital tract has many potential causes. Perhaps the most significant and well studied are alterations in vaginal microflora (bacterial vaginosis) and STI. Other potential inducers of inflammation include (1) microabrasions in the genital tract epithelia that are caused by sexual activity ; (2) hygiene practices, such as antiseptic douching; and (3) proteins in seminal plasma and the use of lubricants. Changes in oestrogen and progestin concentrations that are associated with adolescence, the use of hormone contraceptives, and even normal hormone cycling, may also influence the genital inflammatory environment.

The adolescent female genital milieu

In adolescent women, puberty leads to the start of adrenal and gonadal maturation, resulting in changes in female genitalia, the acquisition of other secondary sexual characteristics, menstruation and changes in the vaginal microbiome. Younger women seem to have a greater prevalence of anaerobic bacteria, with aerobic bacteria becoming more abundant with age, onset of sexual activity and parity. Oestrogen causes the thickening and stratifying of the vaginal epithelium and an increase in vaginal secretions that protects against pathogens. During adolescence, the squamocolumnar junction may lie on the ectocervix (cervical ectropion) and is larger in size. These changes may result in increased susceptibility to STI, as columnar epithelial cells that are usually associated with the endocervical canal are more easily colonised by pathogens compared with the stratified squamous epithelium of the ectocervix and vagina. Puberty, fluctuations in adolescent sex hormones and cytological changes at the cervix may also influence innate and adaptive responses in the genital tract. Shrier et al. reported lower concentrations of immunoglobulin G in genital secretions during the follicular phase of the menstrual cycle in adolescent females compared with adults that may influence susceptibility to infection. Younger women have been found to have higher concentrations of white blood cells and polymorphonuclear leukocytes in cervicovaginal lavage compared with older women. Studies have also found that physiological concentrations of oestrogen stimulate inflammatory cytokine production by endometrial cells and dendritic cells, whereas progesterone withdrawal during the menstrual cycle results in chemokine up-regulation and recruitment and activation of monocytes and neutrophils. Therefore, cytological alterations and changes in inflammatory and adaptive immune responses that may occur in the female genital tract during adolescence could influence susceptibility to HIV infection and may offer a biological explanation, in addition to behavioural factors, for the high rates of HIV infection that are found in this group of women in Southern Africa.

Microbicide-induced genital inflammation

Early candidate anti-HIV microbicides included agents that (1) non-specifically disrupted cellular and microbial membranes (surfactants); (2) restored the natural acidic protective pH of the vagina (acid buffers); and (3) interfered with interactions between HIV envelope proteins and cellular receptors (anionic polymers ). Over the past 20 years, however, none of these candidate microbicides (N-9, C31G [Savvy ^® ], sodium laurel sulfate, cellulose sulfate, Carraguard ^® containing carrageenan and PRO-2000 ^® ) demonstrated significant protection against HIV in clinical trials. In addition, N-9 and cellulose sulfate were even found to increase risks of HIV infection, probably by either disrupting the vaginal epithelial barrier or by inducing inflammatory cytokine responses.

Nonoxynol-9, a non-ionic surfactant that was commonly included in spermicides and some vaginal lubricants, was shown to disrupt the HIV lipid membrane and inactivate the virus in vitro. Multiple applications of N-9, however, were shown to cause inflammatory cytokine (interleukin [IL]-1α, IL-1β, IL-8) and macrophage inflammatory protein-1β up-regulation and NF-κB activation in the genital tract, resulting in the influx of CD68+ macrophages and increased levels of HIV replication in infected cells. Nonoxynol-9 also disrupted the phospholipid membrane of cells and caused non-specific damage to the vaginal, uterine and cervical epithelia. Nonoxynol-9 was further associated with reduced concentrations of secretory leukocyte peptidase inhibitor.

Similar to N-9, cellulose sulfate, a polyanionic microbicide candidate, was found to prevent infection by HIV and a number of other STI in vitro , but proved ineffective in vivo. Primary among its shortcomings, cellulose sulfate induced NF-κB activity in peripheral blood mononuclear cells. Although it also up-regulated IL-1α and IL-6 in cervical epithelial cell culture supernatants after treatment, it did not detectably induce elevated inflammatory cytokine levels within cervicovaginal lavage from women applying cellulose sulfate.

C31G (Savvy ^® ), another surfactant microbicide, which showed the ability to block HIV infection in pre-clinical studies and entered phase III clinical trials, may have been more effective at preventing HIV infections in vivo than cellulose sulfate and N-9, but its efficacy has so far remained unproven: the trial attempting to test its efficacy was discontinued because the rate of HIV infection among the trial participants was too low.

The acidic pH of the lower female genital tract has been found to inhibit HIV replication, and use of gel-based acid buffers such as Acidform ^® have been investigated as microbicidal acidifying agents. Both semen and bacterial vaginosis can neutralise the pH of the genital tract and can therefore influence the susceptibility of women to HIV transmission. Although a phase I clinical trial found no adverse effects of this gel in women using it, other studies have reported that use of Acidform ^® was associated with mild to moderate vaginal irritation and increased genitourinary symptoms.

The effect of genital tract inflammation on microbicide efficacy

Although the results of these earlier microbicide trials have been frustrating, critical lessons have been learned and, as a result, the evaluation of microbicide safety is now better informed. Following the N-9 and cellulose sulfate clinical trials, thorough studies were conducted to achieve a better understanding of the mechanisms of increased HIV transmission in women using these microbicides. The findings of these studies have indicated that a delicate balance exists between protective immunity in the female genital tract and inappropriate immune responses that can actually facilitate HIV transmission. In this regard, genital inflammation has been found to play a much more significant role in HIV transmission than was previously appreciated.

More recent efforts in microbicide development have focused on gels, including specific antiretroviral compounds. These agents are applied in mild, non-irritating base gels, and seem to be better tolerated by the female genital tract than surfactants, acid buffers and anionic polymers. A prerequisite for future microbicides entering clinical evaluation is that (1) they do not disrupt the genital epithelial barrier; (2) they do not have affect the vaginal microflora greatly; and (3) they do not induce inflammation. It is important to point out, however, that the potentially confounding effects of pre-existing genital inflammation (i.e. inflammation not caused by the application of micobicides) have not yet been adequately addressed.

Combination microbicide strategies to prevent genital inflammation

Genital tract inflammation is a significant concern in the context of the HIV epidemic. The recruitment of activated immune targets for HIV infection to the predominant site of infection in women may likely contribute to the susceptibility of women to infection as well as to the potential for HIV-infected women to transmit the virus to their partners. Until a better understanding of the causes of cervicovaginal inflammation and its association with HIV susceptibility has been achieved, strategies to reduce inflammation in the female genital tract should be intensively investigated. In macaques, the application of a topical anti-inflammatory agent (glycerol-monolaurate) has been found to down-regulate pro-inflammatory chemokine concentrations in the genital tract, and, possibly as a consequence of this, to prevent SIV infection.

Intravaginal application of the steroidal anti-inflammatory hydrocortisone as a suppository or cream is routinely used to treat vaginitis and cervicitis. Many existing broad-spectrum, anti-inflammatory agents that are primarily used systemically, however, are associated with mild-to-severe side-effects that may result in increased, rather than reduced, susceptibility to HIV infection if applied topically directly to the female genital tract. For example, non-steroidal anti-inflammatory agents cause gastrointestinal ulceration by disrupting the mucus layer and causing vasoconstriction that results in local tissue hypoxia and epithelial necrosis. Also, anti-inflammatory agents inhibit the innate immune response and are therefore associated with increased susceptibility to infections.

An alternative to suppressing the genital inflammatory response in high-risk women would be to identify and directly treat the causes of inflammation. In Mwanza, Tanzania, large-scale implementation of syndromic management of STI has successfully reduced HIV incidence. Other intervention strategies aimed at reducing HIV infection by treating STI, however, including mass treatment of bacterial STI, herpes simplex virus-2 suppressive therapy, and two other syndromic management interventions, have not been found to affect HIV incidence significantly. These findings highlight the difficulties inherent in implementing large-scale STI management, and further suggest that laboratory testing to identify asymptomatic STI and the causative agents of symptomatic infections, followed by targeted treatment, may have a more substantial effect on rates of HIV infection.

As lactobacilli are the predominant bacterial commensal in the female genital tract, and colonisation with lactobacilli is associated with anti-viral properties (as a result of low vaginal pH as a result of lactic acid production), it has been suggested that the use of exogenous lactobacilli may improve vaginal health and increase resistance to bacterial vaginosis and STI. Efforts are also under way to further augment the natural anti-viral properties of lactobacilli by bioengineering recombinant organisms that produce antiviral proteins. Such strategies might provide protection against HIV infection on multiple levels: (1) by improving vaginal health; (2) by increasing resistance to colonisation by STI and organisms associated with bacterial vaginosis; and (3) by supplying large amounts of antiviral proteins that are constitutively secreted by the recombinant lactobacilli.

Although treatment of STI and overall improvements in vaginal health may substantially reduce genital inflammation, these efforts may not address other potential causes of genital inflammation. Among these other causes, the genetic, demographic and behavioral factors that influence genital inflammation may prove particularly problematic.