Contraceptive Research and Development: Looking to the Future (1996)

Lourens J.D. Zaneveld, Ph.D., D.V.M.

Department of Obstetrics and Gynecology, Rush University,

Rush-Presbyterian-St. Luke's Medical Center

Deborah J. Anderson, Ph.D.

Fearing Research Laboratory, Department of Obstetrics,

Gynecology, and Reproductive Biology,

Brigham and Women's Hospital and Harvard Medical School

Kevin J. Whaley, Ph.D.

Department of Biophysics, The Johns Hopkins University

This chapter has two parts. The first describes approaches to development of novel vaginal agents and formulations with potential for preventing both conception and STD transmission. To facilitate understanding of these approaches, we present background information on the nature of chemical formulations, the functional activity of spermatozoa, and the infective mechanisms of sexually transmitted pathogens. The second part of the chapter presents some of the theoretical underpinnings of mucosal immunity, an area that offers hope of bridging the large and present gap between contraception and prevention of sexually transmitted disease.

The need to focus on sexually transmitted diseases in connection with contraception reflects a growing, if reluctant, recognition of several large sociomedical facts: that the prevalence of these diseases is mounting in much of the world; that their transmission is not limited to small or distant populations engaging in aberrant sexual behavior; that their immediate and more remote sequelae can be dire; and that, biologically and behaviorally, contraception and disease prevention will be necessary partners, at least sometimes, for significant numbers of people and for the foreseeable future.

While effective contraceptive technology for females has been available for several decades, very few contraceptive methods—the condom and some vaginal

formulations and devices—also protect against sexually transmitted diseases (STDs) (Claypool 1994). Of these, only the condom offers protection against the whole spectrum of these diseases, including HIV, and only the female condom and vaginal formulations and devices can be controlled directly by women. Attempts to limit the spread of HIV and other STDs through behavioral modification, encouraging condom use and fewer sexual partners, for example, have proved insufficient to the large task of stemming transmission of these diseases (Stein 1993). Furthermore, understanding about alternative, nonpenetrative sexual practices (''outercourse") that are both protective and satisfying would appear to be limited, although recent empirical data in large population samples are lacking (Greenwood and Margolis 1981; Norman and Cornett 1995; Norsigian 1994). Women experience particular difficulties owing to underlying gender power inequalities that constrain their ability—sometimes severely—to protect themselves from HIV infection, given the absence of a protective technology they could use, if necessary, without a partner's consent (Stein 1990).

New emphasis on developing contraceptive agents that have the additional, critical characteristic of preventing STDs is also being driven by the realization that sexually transmitted diseases other than HIV elevate the risk of HIV transmission (Berkley 1991; Wasserheit 1992); by the increasing rates of heterosexual transmission of HIV infection; by realization of women's greater vulnerability to infection (Mann et al. 1992; Stein 1993); and by women's need to control their own protection against infection and/or conception.

Globally, most women are at greatest risk of acquiring HIV through heterosexual vaginal intercourse with an infected man (Mayer and Anderson 1995). To avoid infection via this route, a preventive method must establish an effective barrier between the infectious elements in genital secretions and those cells of the female reproductive tract that are susceptible to infection. Such a barrier may be physical (such as that provided by condoms), chemical (such as that provided by an intravaginal microbicide), some combination of a physical and a chemical barrier (such as a condom and an intravaginal microbicide), or immunologic (such as topically applied monoclonal antibodies or mucosal immunogens).

The ideal vaginal microbicide should be colorless, odorless, tasteless, stable at room temperature with a long shelf life, easy to use, fast-acting for an appropriate duration after insertion, effective pre- and postcoition, affordable, available without prescription, and safe for use at least once or twice daily. While it would be desirable to develop some microbicides that do not kill sperm, because women who only sometimes want to prevent pregnancy will always want to prevent STD infection, this paper focuses on chemical barriers that may be able to block both conception and STD transmission, that is, "prophylactic contraception" (Cone and Whaley 1994).

In so doing, however, there is no intention to disparage the present and future value of physical barrier methods. The diaphragm and cervical cap can provide some protection against disease transmission (Rosenberg et al. 1992) and are

appropriate for use by some women, although much is unknown about their efficacy, utilization, and protective power (Stein 1993; Stratton and Alexander 1993).

Vaginal topical formulations had received very little research attention until recently. Advances have relied on serendipitous discovery or relatively minor changes in existing methods and, overall, have failed to solve the most critical problems associated with these formulations. Nevertheless, scientists have accumulated much new knowledge about sperm and genital tract physiology over the past three decades. Combined with recent progress in biochemistry, pharmaceutics, and engineering, such knowledge should make it possible now to compensate for previous lack of progress in this increasingly critical area.

During ejaculation, spermatozoa are often placed near the cervix, trapped within the seminal coagulum. As the coagulum liquefies, the spermatozoa are released and enter cervical mucus, a process of penetration that can occur as rapidly as 1.5 to 3 minutes after ejaculation. The primary mechanism whereby spermatozoa pass through cervical mucus appears to be their motility, although other factors such as enzymatic digestion of the mucus may also play a role. The amount of time viable spermatozoa remain in the vagina is not well studied but appears to vary from two to six hours.

Successful sperm penetration into and through cervical mucus occurs primarily during the mid-cycle of the menstrual period, at which time the cervical mucus may contain micelles that guide spermatozoa toward the uterus. At other times in the cycle, particularly the luteal phase, cervical mucus is "hostile" to spermatozoa and does not allow penetration. Even at the optimal period, only about 0.1 to 1 percent of ejaculated spermatozoa pass through the cervix. Still, many spermatozoa become trapped in the cervical crypts and can be released at a later date, potentially providing a constant source of spermatozoa for about four days. Passage of spermatozoa through the uterus and fallopian tubes (oviducts) relies on sperm motility, contractions of the tract, and the motion of cilia on the endothelial surface. Because the uterotubal junction and the fallopian tube isthmus also present a barrier to sperm transport, only about 5,000 spermatozoa actually reach the site of fertilization.

At the time of fertilization, an oocyte is surrounded by three layers. These are, from the outside inwards, the cumulus oophorus, the corona radiata, and the zona pellucida. The fertilizing spermatozoon has to pass through all these layers to contact the oocyte itself. In addition, it must penetrate the vitelline (egg plasma) membrane and decondense inside the ooplasm. Penetration through the oocyte's protective layers requires the use of lytic enzymes, associated with the acrosome of the spermatozoon. Good evidence suggests that hyaluronidase helps sperm pass through the cumulus; acrosin, a serine proteinase, helps it penetrate

the zona pellucida. Sperm motility, ligand-receptor interactions, and other factors are also required for sperm penetration.

Just before or just after it contacts the zona pellucida, the spermatozoon undergoes a morphologic change called the acrosome reaction. This extracellular or exocytotic event results in the disappearance of the sperm's outer acrosomal membrane and surrounding plasma membrane; dispersal of the acrosome proper; activation of proacrosin; and release of acrosomal contents, with acrosin being of particular importance. While sperm penetration through the cumulus oophorus can occur in the absence of the acrosome reaction—presumably because hyaluronidase is associated with the external membranes—sperm passage through the zona pellucida is not possible without this reaction. An early acrosome reaction, such as one that occurs in the vagina, will result in premature release of acrosomal contents and, most likely, in an inability of the spermatozoon to fertilize.

Proper timing of the acrosome reaction is controlled by capacitation in the female genital tract, an activation process resulting in removal of inhibitory substances from the sperm surface; by possible modification of the sperm's plasma membrane; and by changes in motility patterns. Ejaculated spermatozoa cannot undergo the acrosome reaction unless they are capacitated. Capacitation can occur in all parts of the female genital tract, with the possible exception of the vagina, and requires at least four hours. However, it probably occurs as a continuous process while the spermatozoa are being transported through the cervix, uterus, and fallopian tubes, so that a spermatozoon is fully or mostly capacitated by the time it reaches the oocyte. Final steps in the capacitation process may occur during sperm passage through the cumulus oophorus.

Recently, biochemical aspects of capacitation and the acrosome reaction have received significant attention as possible targets for intervention (Dunbar and O'Rand 1991; see also Appendixes B and C in this volume). The former appears to involve impeding maturation and the acrosome reaction by preventing changes in the sperm's plasma membrane or, in contrast, attempting to induce a premature acrosome reaction that would render sperm unable to meld with an egg. The acrosome reaction, like other exocytotic processes, requires certain physiological conditions such as ligand interaction with surface receptors, activation of second messenger systems, protein phosphorylation, ionic and osmotic changes, and, ultimately, membrane fusion, vesiculation, and disappearance.

Sexually transmitted diseases are caused by a variety of organisms, including bacteria (aerobic and anaerobic), chlamydia, mycoplasmas, ureaplasmas, spirochetes, fungi, flagellates, amoebae, worms, and viruses. Each of these organisms has different biologic properties, conferring widely diverse mechanisms of infection and pathogenesis (see Chapter 2 and Chapter 5).

The most common human STD pathogens and their primary sites of infection are listed in Table D-1. Effective vaginal microbicides must block infection by directly and efficiently killing these organisms or by blocking or inactivating molecular mechanisms underlying infection. Current detergent vaginal formulations have broad-spectrum cytolytic properties that may be effective against several STD pathogens but may also damage epithelial and other genital tract cells and disrupt normal vaginal flora. Ligand/receptor molecules located on the surface of STD pathogens and their host cells—which are responsible for pathogen attachment and entry—are obvious targets for microbicide action. While these molecular structures have not been fully characterized for most STD pathogens, researchers are intensively seeking them today, importantly including HIV-AIDS.

The infectiousness of HIV is highly variable and there are still surprising gaps in understanding of its transmission, including the factors affecting amount and timing of virus shedding, the roles of vaginal and seminal antibodies, and the effects of exogenous or endogenous hormones (Mayer and Anderson 1995; Stratton and Alexander 1994). Partner studies have provided much of the information we do have. They indicate that the following factors affect rates of HIV

transmission: (1) mean numbers of sexual contacts and/or types of sexual practices; (2) differences in infectivity of partners depending on disease stage, symptomatology, and therapeutic drug status; (3) potential differences in infectivity of various clades of HIV-1; and (4) intrinsic biological differences between infecting or susceptible partners. Specific risk factors specific to heterosexual transmission of HIV-1 include: (1) sex during menstruation; (2) anal intercourse; (3) traumatic sex; (4) advanced HIV disease stage of infected partner; (5) zidovudine therapy (negative association); (6) age of the female partner; (7) choice of contraceptive method; and (8) concomitant STD infections (bacterial vaginosis, candidiasis, chancroid, chlamydia, gonorrhea, herpes simplex, human papilloma virus, syphilis, and trichomoniasis have all been associated with HIV transmission) (Mayer and Anderson 1995; Stratton and Alexander 1993).

The cell biology and molecular mechanisms of HIV-1 sexual transmission are areas where much remains to be learned. It is known that HIV-1 is present in semen and cervicovaginal secretions, both in cell-free and cell-associated forms, and data from both in vitro and in vivo primate studies indicate that each of these forms may be capable of transmitting infection. Cell-free HIV-1 in genital-tract secretions may infect Langerhans cells, lymphocytes, and/or macrophages residing within the epithelial layer of mucosal surfaces in the vagina, foreskin, and penile urethra. Infection could occur via CD4, Fc, complement, or other receptors as yet unknown. If genital lesions or microabrasions are present, HIV-1-infected white blood cells (WBCs) in genital-tract secretions could infect by direct access to target cells in connective tissue (resident and extravasated WBCs). Furthermore, in vitro studies indicate that HIV-infected WBCs may have the capacity to adhere to mucosal epithelium and transmit HIV directly to these cells (Mayer and Anderson 1995).

Numerous biologic variables can affect HIV levels in semen and cervicovaginal secretions and may also influence HIV transmission rates. Leukocytospermia (inflammation of the genital tract) in men has been associated with higher HIV-1 titres in semen. This phenomenon may be due to recruitment of HIV-1-infected cells to the genital tract; induction of edema and capillary dilation, both of which increase the potential for erosion and escape of blood lymphocytes and monocytes into the intravascular space and lumen; and/or activation of lymphocytes that could make them more capable of producing HIV- 1. In women, several variables can affect infectious HIV-1 titres in genital-tract secretions, including: (1) inflammation of the genital tract; (2) menstruation, which introduces infected peripheral blood cells into vaginal secretions; (3) factors elevating vaginal pH, which is normally acidic and confers protection against STDs, including HIV-1; (4) cervical ectopy; and (5) hormonal factors that influence the thickness of the epithelial layer and the production of protective mucus (Mayer and Anderson 1995).

An open—and highly significant—question is whether HIV is carried solely by somatic cells in semen or whether it is carried by the sperm themselves. If the

former is the case, contraception is not necessary for prevention of sexually transmitted infection, so that a primary technologic need is for a virucide/microbicide; individuals can then decide whether they wish to contracept as well, but it is a separate decision. However, if the latter is the case, contraception becomes an essential concomitant of disease prevention (Stein 1993).

The healthy human vagina has several natural defense mechanisms against STD pathogens. The stratified squamous epithelium lining of the vagina and ectocervix is generally several cell layers thick, conferring physical protection against most invading organisms. Copious amounts of mucins produced during certain stages of the menstrual cycle also provide a physical barrier to pathogens. In addition, the vagina and cervix are both capable of mounting antibody and cell-mediated immune responses to genital tract infections, and cervicovaginal secretions contain a number of nonspecific antibacterial and antiviral defense agents, including lysozymes, polyamines, zinc, H202, lactoferrin, and B-defensins (Cohen et al. 1990). Finally but quite importantly, the predominant normal vaginal microflora—lactobacillus organisms—help to maintain low pH conditions that are hostile to most viral and bacterial pathogens (Voeller et al. 1992b).

Unfortunately, a number of situations perturb these natural protective mechanisms. Menopause and progestin-dominated hormonal therapies, for example, cause thinning of the squamous epithelial layer and reduction in cervical mucus production. Intercourse and vaginal infections, whose coincidence is not uncommon, can cause an elevation in vaginal pH, disarming this effective protective mechanism. And, although currently marketed vaginal topical contraceptives may, in fact, be buffered at acidic pH, the buffering capacity of semen itself elevates the vaginal pH substantially (Masters and Johnson 1980).

The functional activity of both spermatozoa and pathogenic microbes can be prevented by killing or inactivating them. The former strategy employs spermicides and microbicides. The latter relies on agents that inhibit the functional activity of spermatozoa and microbes so that they are unable to enter, respectively, the oocyte or the vaginal/cervical epithelium and target cells. Although these inactivating agents are more appropriately called spermistats and microbistats, the terms spermicide or microbicide are still frequently used for all antifertility and antimicrobial agents. Candidate classes of these compounds include detergents and surfactants, iodophores, carbohydrates, antibodies, antiviral drugs, defensins, and pyocins.

Many of the chemical formulations that are currently available for use as vaginal contraceptives, in addition to their contraceptive properties, can confer some protection against such STD infections as Neisseria gonorrhea , Trichomonas vaginalis, and Chlamydia trachomatis (Rosenberg et al. 1987, 1992). However, like microbicides, spermicides act as cytotoxins, often destroying vaginal and cervical cell membranes (Patton et al. 1992). As a consequence, the protective capacity of these formulations erodes under frequent and/or prolonged use, largely because their potential effects on normal vaginal flora and on the vaginal/cervical epithelium create perturbations that can actually promote the possibility of STD transmission (Kreiss et al. 1992; Niruthisard et al. 1991, 1992). This means that, if vaginal formulations are to be used more often for STD prevention, they will have to be nonirritating, nontoxic, and not upset the normal vaginal environment. In other words, such formulations must produce no lasting impact on the normal vaginal flora and their work in maintaining the vaginal milieu and its naturally acidic pH, nor should they compromise the vaginal or cervical epithelium. Each formulation will therefore need to be rigorously tested for any proinflammatory effects that could promote transmission of STD pathogens, notably HIV-1.

A topical formulation used for purposes of vaginal contraception consists of one or more active ingredients and a base (carrier) to deliver them. Both are important and each presents particular challenges to researchers and users. Typically, an active ingredient makes up only about 5-10 percent of an entire formulation; at least this is the case for those products that are on today's market, primarily surfactants or detergents such as nonoxynol-9, octoxynol-9, menfegol, and benzalkonium chloride (Haslett 1990; Hatcher and Warner 1992; Mauck et al. 1994). The activity of these products resides in their ability to dissolve lipid components in the cell membrane or viral envelope. Since these agents have approximately the same properties and potency, marketed formulations differ primarily in their base composition—either jellies, creams, foams, suppositories, tablets, films, or sponges. Because the active ingredients in these formulations all act by immobilizing or killing spermatozoa, they are usually referred to as "spermicides."

The contraceptive use-effectiveness of existing formulations varies greatly from study to study (Trussell 1994). Reported failure rates range from 2 percent to over 40 percent, with typical rates falling between 15 and 20 percent (Sobrero 1989), much higher than desirable. The limited efficacy of these formulations, their brief longevity after vaginal placement and effects on coital spontaneity, together with leakage and consequent "messiness," surely constrain consumer interest in available vaginal contraceptives.

Although the relatively poor efficacy of available vaginal contraceptive for-

mulations can be partly blamed on improper use, some animal experiments have shown that the efficacy of many current products may be low even with more reliable use (Homm et al. 1976; Zaneveld et al. 1977). This implies that either their active ingredients and/or their distribution/delivery systems are in some way ineffective; because good comparative studies are rare, it has been difficult to discern where, exactly, the primary limitations reside.

Furthermore, because many pathogens—for instance, HIV, the herpes viruses, and chlamydial elementary bodies—are cell-associated, the margin between toxicity and efficacy is narrow. There is nothing unique about the membrane of an infected cell that permits a surfactant to distinguish it from the membrane of a healthy cell, though the cells of intact genital mucosa may be protected to some extent by a coating of mucus.

Finally, in vitro spermicidal efficacy does not necessarily translate into in vivo contraceptive efficacy (Quigg et al. 1988). Thus, without in vivo studies, it is not possible to state with certainty whether an agent with spermicidal properties is also contraceptive (Zaneveld 1994a, 1994b), nor are in vitro spermicidal comparisons adequate to the task of assessing the relative contraceptive activity of different formulations.

Because an active ingredient is only as effective as its delivery system allows, the formulation's base requires careful attention. Not only can the nature of a given base "make or break" the active ingredients, but can, in itself contribute to prevention of STD infections and conception, an attribute that should enhance the overall acceptability of vaginal topicals to women. The following characteristics of a good base are particularly important:

• A good base should spread rapidly and evenly over the vaginal and cervical surface, forming a slightly adhesive film.
• Optimally, this film should be impenetrable to microbes and, if possible, spermatozoa.
• The film should remain in place for prolonged periods of time, including during intercourse, providing long-acting protection.
• Leakage and consequent "messiness" should be minimal.
• Finally, the formulation should be somehow buffered at an acidic pH so as to retain a pH below 4.5 even in the presence of semen.

Many complicated biochemical processes are required for successful fertilization. Inhibition of any of these processes can lead to infertility, providing a

large number of potential attack sites for new contraceptive agents, both chemical and immunologic. Yet, to date, only a few researchers have taken advantage of advances in knowledge to develop novel vaginal contraceptives.

New information can also be utilized to identify agents that can prevent both conception and STD infection. Tissue and/or cell invasion is required for both spermatozoa and pathogenic microbes to reach their target sites. Such invasion requires the activity of lytic enzymes, binding proteins, receptor-ligand interactions, fusion, endo- and exocytotic events, and a myriad of other processes. Each of these events is susceptible to inhibition, which would in turn prevent invasion. It is possible, if not likely, that spermatozoa and certain pathogenic microbes use some identical mechanisms of invasion. If so, the same agent could be used to prevent both occurrences (see Table D-2).

Several detergents are already licensed in one country or another for intravaginal use as contraceptive agents. These include nonoxynol-9, the most commonly used, as well as menfegol, octoxynol, and benzalkonium chloride. Like the phospholipids that constitute cell membranes, detergents have hydrophobic and hydrophilic domains and exhibit activity that derives from their ability to dissolve lipid components in the cell membrane or viral envelope. Improvements on nonoxynol-9 may come from removing low molecular weight toxic components from the polydispersed N9 mixture (Klebanoff 1992; Walter et al. 1991a, 1991b), from improving dispersion/distribution, by providing for triggered release (Quigg 1991), or by lowering the detergent dose. In order to extend the spectrum of activity or longevity of action, low concentrations of nonoxynol-9 could be combined with other agents, for example, other detergents (Psychoyos 1994), beta-lactoglobulin (Neurath et al. 1996), antivirals (De Clercq 1993; Tsai et al. 1995), or zinc (Krieger and Rein 1992; Mardh et al. 1980; Williams 1980).

Lipids found in human milk and epidermis display antibacterial and antiviral activity (Isaacs et al. 1990). Microbial killing by milk lipids is primarily due to the free fatty acids and monoglycerides that are released from milk triglycerides by lipases. At high concentrations, lipophilic molecules may be expected to have activity on sperm membranes that is similar to their activity on the membranes of other cells (Thormar et al. 1987).

Chlorhexidine is a broad-spectrum antiseptic used in a prescription oral rinse. It is a positively charged molecule at physiological pH and binds to negative sites

on cell surfaces; this leads, in turn, to altered permeability, cellular leakage, and precipitation or coagulation of cytoplasmic proteins. Chlorhexidine has been shown to be spermicidal (Sharman et al. 1986), to inactivate HIV, and to cause changes in cervical mucus that prevent penetration by spermatozoa (Chantler et al. 1989).

C31G is an equimolar mixture of two synthetic, amphoteric surface-active compounds shown to have broad-spectrum antimicrobial properties (Corner et al. 1988). Its active agents are alkyl dimethyl glycine and alkyl dimethyl amine oxide, and the alkyl chain length (C-12 to C-16) is representative of the natural distribution of the fatty acids of coconut oil. The surfactant properties of C31G would suggest potential for activity against sperm and enveloped viruses; it also appears to promote wound healing (Michaels et al. 1983).

1-Docosanol, a 22-carbon-long saturated alcohol, has been shown to inhibit lipid-enveloped viruses (Katz et al. 1994), inhibiting viral replication by interfering with fusion or early intracellular events surrounding viral entry into target cells (Katz et al. 1991). Since such a large percentage of the molecule becomes associated with cell membranes, docosanol can be expected to have some activity on sperm function.

Neem oil extracted from the seeds of Azadirachta indica has been shown to have antiviral, antibacterial, and antifungal activity (National Research Council 1992). The antifertility activity of vaginally applied neem oil has been established in rabbits and monkeys (Talwar et al. 1994). Neem oil can also inhibit spermatogenesis, sperm-egg interaction, implantation, and can act as an abortifacient. Its precise mechanism of action is not clearly understood, but may be due to immunomodulatory activity such as enhanced phagocytic activity and induced production of gamma interferon.

Squalamine is an aminosterol antibiotic first isolated from the shark (Moore et al. 1993). A condensation of an anionic bile salt, it has just recently been chemically synthesized (Sadownik et al. 1995). Squalamine exhibits potent bactericidal activity against both gram-negative and gram-positive bacteria, is fungicidal, and also induces osmotic lysis of protozoa. One mechanism of action may be an inhibition of sodium-hydrogen exchangers on cell surfaces that produces changes in intracellular pH. Such a mechanism may prove to be spermicidal and might influence HIV's ability to infect target cells.

Different types of serine proteinase inhibitors are now known to be able to inhibit HIV replication without cytotoxicity (Bourinbaiar and Nagorny 1994; Hallenberger et al. 1992; Hattori et al. 1989). It is also known that spermatozoa utilize a trypsin-like serine proteinase (acrosin) for penetration through the zona pellucida and that certain acrosin inhibitors are potent contraceptives when placed vaginally in animal models. Aryl 4-guanidinobenzoates such as 4'-acetamidophenyl 4-guanidinobenzoate (AGB) have been found to be more contraceptive than nonoxynol-9 (Kaminski et al. 1985) and it was recently reported that AGB is also a potent anti-HIV agent (Bourinbaiar and Lee-Huang 1994b). N-a-tosyl-Llysyl-chloromethylketone (TLCK) has also been found to inhibit both HIV infec-

tivity (Bourinbaiar and Nagorny 1994) and fertilization when placed vaginally (Zaneveld et al. 1970).

Since sperm are inactivated by acidity (Shedlovsky et al. 1942), it appears that in neutralizing the vagina, the alkalinity of semen enables more sperm to reach and enter the cervix (Masters and Johnson 1980). When this ''window of neutrality" is, in effect, "opened" by semen, it stays open for at least two hours after intercourse. This unfortunately can facilitate the sexual exchange of STD pathogens that otherwise would be killed by the acid and gives those pathogens additional time to reach target cells in the new host. Exposing cell-free HIV to the acid pH ranges normally found in healthy vaginal secretions reduces both its infectivity and the survival of HIV-infected lymphocytes (Voeller and Anderson 1992a, 199b). Even mild acidity (pH 5) has been shown to inhibit other STD pathogens, e.g., Neisseria gonorrhoeae (Zheng et al. 1994).

Magainin, a peptide isolated from frogs, is a cationic peptide that is antibacterial, antifungal, antiprotozoan, antiviral, and antispermatozoal (de Waal et al. 1991; Edelstein et al. 1991). These peptides are able to adopt an amphiphilic alpha-helix structure in a hydrophobic environment and can punch holes in the cell walls of infecting organisms.

Defensins are complexly folded and amphipathic peptides which, while rich in antiparallel beta-sheet, are devoid of alphahelical domains (Ganz and Lehrer 1994). Their unusually broad antibacterial spectrum encompasses gram-positive and gram-negative bacteria, many fungi, mycobacteria, spirochetes, and several enveloped viruses (Lehrer et al. 1993; Schonwetter et al. 1995). The antimicrobial properties of defensins result from their insertion into target cell membranes and the formation of voltage-sensitive channels.

There are many antibacterial peptides that are inhibited by seminal plasma, but magainin II and a synthetic peptide of a 37-Kd cationic antimicrobial protein have been shown to inhibit sperm motility (D'Cruz et al. 1995). Gramicidin, a peptide antibiotic used in the former Soviet Union as an active component of contraceptive gels and foams, also exhibits anti-HIV activity (Bourinbaiar and Lee-Huang 1994a).

We have known for over two decades that polyanionic substances, particularly sulfated polysaccharides, are able to interfere with the virus adsorption process, including HIV (DeClercq 1993). A sulphonated polystyrene polymer has

been shown to be contraceptive in rabbits (Homm et al. 1985); the polymer is known to agglutinate spermatozoa, alter sperm-cervical mucus interaction and inhibit sperm acrosin (Foldesy et al. 1986).

When combined with H₂O₂ and a halide, peroxidases form a powerful antimicrobial system that is effective against a variety of microorganisms, including viruses (Klebanoff and Coombs 1991). The proposal has been made to add hydrogen peroxide and/or peroxidases (derived from animals or produced by genetic engineering) to vaginal formulations. These products may work in conjunction with H₂O₂-generating lactobacilli in the vagina.

Antibodies in mucus play a major role in preventing mucosal transmission of disease. Monoclonal antibodies of human origin, expected to be potent, flexible, specific, sturdy, and inexpensive, can now be developed to serve as natural protective agents for mucosal surfaces (Cone and Whaley 1994; Ma et al. 1995). Both contraception and prevention of viral STDs with MAbs have been demonstrated in animals, and controlled release of antibodies for long-term topical passive immunoprotection permits this technology to be noncoitally related as well (Sherwood et al. 1996). Monoclonal antibodies are a technology almost two decades old yet, while used in many fields, including human health, they have been significantly underutilized in contraception. Nonetheless, MAbs offer a pathway to "prophylactic contraception," a concept within reproductive health that bridges the mechanisms of mucosal transmission of sperm as well as pathogens (see Table D-3).

Deborah J. Anderson, Ph.D., and Alison J. Quayle, Ph.D.

Fearing Research Laboratory, Brigham and

Women's Hospital and Harvard Medical School,

Department of Obstetrics, Gynecology, and Reproductive Biology

Development of effective immunocontraceptives awaits the identification and genetic engineering of appropriate molecular sequences and their expression in immunogenic form. It also depends on the ability to deliver immunogens to achieve the desired effect at the appropriate site within the male or female reproductive tract. Since both the lower female and the male genital tract mucosa are considered part of the common mucosal system, there is an implication that mucosa-targeted immunization would induce effective genital tract immune responses. In this paper, we review the basic principles of mucosal immunology of the male and female genital tracts and explore options that might become available for mucosal immunization, both for purposes of contraception and protection against infectious disease.

Antibody-mediated (Secretory) Mucosal Immunity

The mucosal membranes of the body include the respiratory tract, conjunctiva, intestinal tract, mammary glands, and the urogenital tracts of the male and female. These mucosal surfaces cover a vast surface area, estimated at around 400 square meters, and are constantly exposed to environmental antigens. In contrast to the lymphoid organs, the majority of plasmacytes in mucosal-associated organs are of the immunoglobulin A (IgA) isotype (80 percent, compared to 30 percent in the lymphoid organs) and the predominant antibody in external secretions is usually secretory immunoglobulin A (SIgA). More SIgA is synthesized by the immune system than any other isotype (Ma et al. 1995; Mestecky and McGhee 1993). Approximately 70-80 percent of all Ig-producing plasma

cells are located in the intestinal mucosa, making this organ the largest reservoir of immune cells in the entire body (Brandtzaeg et al. 1993).

Given the importance of the mucosal system and the fact that the human body actually synthesizes and secretes more antibodies into mucus than it does into blood, it is surprising that so little is known about how the secreted antibodies function in mucus to protect the human mucosal surfaces. Some known effector actions are: (1) agglutinating pathogens; (2) blocking pathogen interactions with target cells; (3) trapping pathogens in mucus; (4) helping to regulate protective commensals (e.g., vaginal flora); and (5) secreting or neutralizing pathogens that have penetrated the mucosal surface (Cone and Whaley 1994; Franek et al. 1984; Mazanec et al. 1993). A significant advantage of using polymeric secretory type antibodies for topical application would be the extraordinary potency created by their polyvalency, which increases their avidity and, because of their structure, their particular potency in processes of agglutination (Cone and Whaley 1994; Hornick and Karush 1972; Raff et al. 1991). In addition, sIgAs persist longer in mucosal secretions than IgG because of resistance to proteolysis. Furthermore, they cause less inflammation because of a low capacity to bind complement. These unique attributes of sIgA would seem to be appealing areas for expanded research.

Despite their apparent importance for mucosal protection, few investigators or companies seem to have attempted to develop IgA monoclonal antibodies. This has been particularly the case for sIgA, despite the fact that it is the first line of defense against infectious agents (Ma et al. 1995) and is the isotype least likely to cause the inflammatory responses that IgG can initiate (Cone and Whaley 1994; McGhee et al. 1992). The emphasis has been on IgG, at least partly because of the prominence of blood in concepts of vital functions (Cone and Whaley 1994), and those IgA MAbs that have been developed have been used by most companies for therapeutic purposes rather than for promoting sexual health. This picture may be modified by discoveries of the potential applicability of transgenic plants for large-scale production of recombinant sIgA for passive mucosal immunotherapy (Ma et al. 1995).

Historically, the study of mucosal immunity can be traced back over 200 years (Mestecky and McGhee 1989). Classical scientific studies began in the late eighteenth century, when Pasteur and his pupils demonstrated that oral immunization with the appropriate killed bacterium could protect animals from a number of enteric infections (Gay 1924; Mestecky and McGhee 1989). The concept of an independent local immune defense system was initially proposed by Besredka in 1919 after finding that rabbits orally immunized with killed Shigella were protected against dysentery, irrespective of serum antibody titre (Besredka 1919). The findings from this study were later tested in a clinical setting when thousands of humans were orally immunized against dysentery and typhoid fever (Besredka 1927).

Scientific endeavor in the 1960s and 1970s laid the foundations for contem-

porary understandings of the mucosal immune system. In 1963, Tomasi and Zigelbaum reported a predominance of IgA in human colostrum, saliva, and urine samples in comparison to serum (Tomasi and Ziegelbaum 1963) and immunohistochemical studies suggested that this must be locally produced since IgA-secreting plasma cells were 20 times more abundant than IgG-secreting plasma cells in the human intestine (Crabbe et al. 1965). Later studies showed that the plasma cells in mucosa predominantly produced IgA dimers or polymers (Brandtzaeg 1973) associated with a 15kD glycopeptide "joining" (J) chain that is also a component of pentameric IgM (Halpern and Koshland 1970; Mestecky et al. 1971). This contrasts with the IgA-producing plasma cell populations in the bone marrow, spleen, and lymph nodes, which synthesize the predominantly monomeric IgA found in serum. Immunoglobulin A (Hanson 1961; Tomasi et al. 1965) and IgM (Brandtzaeg 1975) in secretions were found to be associated with secretory component, and both isotypes were shown to be translocated to external secretions by a transepithelial transport mechanism (Brandtzaeg et al. 1968 and 1970). In brief, polymeric IgA binds to secretory component that is expressed on the basolateral surface of epithelial cells. This complex is endocytosed and the resulting vesicle transcytosed, fusing with the apical membrane and finally releasing the secretory component-IgA complex into the lumen (see Brandtzaeg et al. 1993 for review).

Although immune cells are found throughout the mucosae, antigen uptake and induction of immune responses occur predominantly in specialized inductive sites. These are the gut-associated lymphoreticular tissues (GALT) which include the Peyer's patches (PP), the appendix in some animals, solitary lymph nodes and small follicles, and the bronchus-associated lymphoreticular tissue (BALT) (Mestecky and McGhee 1993). The Peyer's patches are distinct nodules, visible to the naked eye and predominantly located in the small intestine; in young adult humans, they number approximately 200 (Cornes 1965). Each PP has a B-cell zone consisting of a number of follicles with germinal centers, a parafollicular or T-cell-dependent area, and a dome region with a lymphocyteand macrophage-rich corona overlain by a specialized epithelium. This epithelium is cuboidal (in contrast to the columnar epithelium lining the remainder of the gut [Bockman and Cooper 19731), with no goblet cells and little mucus, and has special antigen-sampling cells called microfold (M) cells (Owen and Jones 1974). The M cells pinocytose and phagocytose antigen in the lumen of the gut but do not degrade it, delivering it intact to the underlying lymphoid cells. The germinal centers of the follicles of the PP are rich in B lymphocytes committed to producing IgA (Jones and Cebra 1974); they do not, however, mature into IgA-producing plasma cells in the Peyer's patches, but serve importantly as a pool of cells that are capable of migrating to distant mucosal sites.

This ability of B cells from the Peyer's patches to home to, and repopulate, distant mucosal sites in irradiated animals was first noted in the early 1960s (Jacobsen et al. 1961) and the studies were subsequently extended in a number of

laboratories (e.g., Craig and Cebra 1971). In a mouse study, labeled, adoptively transferred mesenteric lymph node (MLN) cells homed to MLN, gut, cervix, vagina, uterus, and mammary glands and were predominantly IgA in their phenotype, while peripheral lymph node (PLN) cells preferentially homed back to PLN and produced IgG (McDermott and Bienenstock 1979). Studies in various animal species have also demonstrated the appearance of specific antibodies in sites remote from oral immunization (Challacombe and Lehner 1979; Forrest et al. 1992; Goldblum et al. 1975; Montgomery et al. 1976). In humans, this work has included the classic studies by Ogra and colleagues using Sabin polio vaccine (Ogra and Karzon 1969; Ogra and Ogra 1973) and by Mestecky and coworkers with bacterial antigen (Mestecky et al. 1978). The term "mucosa-associated lymphoid tissues" (MALT) was introduced by Bienenstock and Befus (1980) to describe the immune cell traffic between mucosal tissues.

In brief, immature surface membrane IgM-positive B cells in the Peyer's patch "switch" to become surface membrane IgA-positive under the influence of cognate antigen, special "switch" T cells, dendritic cells, and T-cell-derived cytokines (McGhee et al. 1993). These committed cells exit the Peyer's patches and travel to the mesenteric lymph nodes; here they mature further prior to entering the bloodstream via the lymph and thoracic duct, thence homing to, and repopulating, distant mucosal sites. An elegant study by Czerkinsky and colleagues provides support for the notion of a common mucosal system in man: Streptococcus mutans-specific, IgA-producing cells were isolated from blood seven days after oral immunization and seven days prior to the appearance of antigen-specific IgA in saliva and tears (Czerkinsky et al. 1987). It should also be noted that, in the mouse, an alternative source of precursor IgA plasmacytes is the peritoneal cavity (Husband and Gowans 1978); these peritoneal cavity-derived B cells have recently been shown to be Ly-1 CD5 positive (Kroese et al. 1988).

The mucosal organs also provide a substantial reservoir of T cells, which are located in three major areas: the organized lymphoid tissue (GALT, BALT), the lamina propria, and the epithelium. The T cells of the Peyer's patches are predominantly of the CD4 phenotype and one of the major roles for these T cells is thought to be regulation of IgA responses (McGhee et al. 1993). In vitro experiments in which Peyer's patch B cells were activated with lipopolysaccharide resulted in a four- to six-fold increase in IgA production in the presence of activated T cells from PP; if the T cells were splenic in origin, however, IgG and IgM production were increased (Elson et al. 1979).

Studies at the University of Alabama are currently dissecting the dynamics of the B and T cell interactions in the Peyer's patch (see Mestecky and McGhee 1993 for review). Cytokines that seem to have a key role in this environment are interleukin 5 (IL-5) and interleukin 6 (IL-6). In the murine model, IL-5 selec-

tively enhanced IgA production (Beagley et al. 1989; Bond et al. 1987; Murray et al. 1987) and IL-6-induced terminal differentiation of activated B cells and increased IgA synthesis in in vitro systems (Beagley et al. 1989). These cytokines are synthesized by T lymphocytes of the Th2 subset (Mosmann and Coffman 1989; Romagnani 1992), a subset shown through ELISPOT assay to be overrepresented in PP T cells and lamina propria compared to blood (McGhee et al. 1993). However, TH1-type cytokines (y-interferon, TNF ß) are also produced by T lymphocytes at mucosal surfaces, especially at sites of infection.

The T cells in the lamina propria are also predominantly CD4-positive memory (CD45RO+) cells (Brandtzaeg et al. 1989); in contrast, approximately 85 percent of the intraepithelial cells (IELs) are of the CD8 phenotype (Janossy et al. 1980; Jarry et al. 1990). In humans, the majority of intestinal T cells express the aß T cell receptor (TCR) (Brandtzaeg et al. 1989; Halstensen et al. 1989); in mice, a substantial proportion of the IELs are TCR positive (Itohara et al. 1990). Mouse studies strongly suggest that the aßTCR cells in the lamina propria and epithelium of the mucosa are derived from T cells in organized lymphoid tissue, indicating that there is committed trafficking of T cells from inductive sites to the various mucosa. Adoptive transfer experiments with labeled T cells from Peyer's patch and mesenteric lymph node have shown a preferential homing back to gut mucosa (Guy-Grand et al. 1978); infection of Peyer's patch with reovirus results in the appearance of reovirus-specific CD8+ cytotoxic IEL (London et al. 1989); and T cells isolated from breast milk respond better to enteric antigens than do peripheral blood T cells from the same donor (Ada 1993).

Recent studies have identified receptors on lymphocytes that are involved in homing to, and possibly maintaining localization within, mucosal-associated organs. The integrin a4ß7 (LPAM- 1), expressed on most naive murine (CD45RA+) T cells and a subset of memory (CD45RO+) T cells, mediates binding to Peyer's patch high endothelial venules (HEV), but not peripheral lymph node HEV (Holzmann et al. 1989). The ligand for LPAM-1 is the immunoglobulin superfamily adhesion molecule MADCAM-1 that is expressed on mucosal venules, and the interaction between the two is thought to play an important role in lymphocyte homing to mucosal sites (Berlin et al. 1993). In both man and mouse, an integrin has been identified that is almost exclusively expressed by T cells residing in the mucosa. This is also a ß7 integrin, but it is co-expressed with a unique a chain. In the mouse it is known as the aM290ß7 integrin (Kilshaw and Murant 199 1) and in humans as the HML- 1 antigen, æß7 integrin, or CD 103 (Cerf-Benussan et al. 1992; Parker et al. 1992). Murine studies suggest that this molecule is induced after mucosal localization, since aM290ß7-positive cells are abundant in the various mucosae but rare in spleen or lymph node (Kilshaw and Murant 1990; Quayle et al. 1994). The in vitro addition of transforming growth factor beta (TGF-ß) to stimulated lymph node cells increases expression of the ß7 subunit and changes the associated a4 subunit to aM290; this has been hypothesized as the possible mechanism of induction in the mucosa (Kilshaw and Murant

1991). Another hypothesis is that αM290v7 plays a role in adhesion, modulating or focusing effector cell activity (Kilshaw and Murant 1991).

Plasma cells are present in the greatest concentration in the subepithelial layers of the human endocervix, but substantial numbers are also seen in the ectocervix, vagina, and fallopian tubes (Kutteh et al. 1993; Kutteh and Mestecky 1994). Approximately 70 percent of the plasmacytes are IgA positive and, of these, 50-75 percent of cells co-label for J chain, indicating that the majority of local IgA synthesis is polymeric. Of the IgA positive cells, 40 percent are of the IgA2 subclass, indicating an A1/A2 ratio similar to that of the lower intestine. Plasmacytes are rarely detected in the ovary, endometrium, or myometrium (Kutteh et al. 1993; Kutteh and Mestecky 1994), but secretory component is expressed on the epithelial cells of the villi of the fallopian tubes and the cervical glandular epithelium, suggesting that all the elements for local production and transport of sIgA are present at those sites (Kutteh et al. 1993; Kutteh and Mestecky 1994).

Infection with Neisseria gonorrhoeae , Trichomonas vaginalis, or Candida albicans also results in an increase in plasma cell numbers on the endocervical region, predominantly of the IgA class (Chipperfield and Evans 1972). Similarly, acute and chronic salpingitis is associated with elevated plasmacyte numbers in the fallopian tubes that are also predominantly IgA-positive (Kutteh et al. 1990); interestingly, plasma cell numbers in tissues with evidence of healed salpingitis were similar to those of normal healthy tissue.

Although there have been a number of studies of immunoglobulin concentrations and isotypes in female genital tract secretions, the data remain conflicted. This may be due to differences among species and other confounding variables such as the secretion collected, technique of collection, hormonal influences, reproductive status, and concomitant infections present in the tract. In addition, cervicovaginal secretions presumably contain "spillover" from uterine secretions. Quantitative analysis of cervical mucus from over 100 healthy women revealed the presence of IgG, IgM, and IgA in this secretion, with IgG the major immunoglobulin; the highly significant correlation between albumin and IgG suggested that this was a serum transudate (Tjokronegoro and Sirisinha 1975). Furthermore, while the mean level of IgA was approximately half that of IgG, the majority of this was associated with secretory component and therefore indicative of local production.

Locally produced antigen-specific IgA has also been found in the cervical secretions of women with gonococccal cervicitis, but only IgG was detected in the serum (O'Reilly et al. 1976). In a series of classic mucosal experiments with polio vaccine performed by Ogra and colleagues, women were immunized vagi-

nally, intrauterinally, or intramuscularly with killed (Salk) poliovirus, or orally with live attenuated (Sabin) polio vaccine; the poliovirus-specific antibody response was then measured (Ogra and Ogra 1973). Vaginal immunization resulted in the appearance of IgA and IgG in vaginal secretions, an occasional low IgG response in uterus, and no detectable specific IgA or IgG in serum. In contrast, intrauterine immunization resulted in the appearance of IgG in uterine washings, an occasional low positive in vaginal secretions, undetectable IgA local secretions, and no antibody response in serum. In the women immunized orally or intramuscularly, specific serum IgG and IgA were detected, but the only locally detectable antibody was low-level IgG; the appearance of this was correlated with increasing serum IgG. This correlates well with a primate study in which animals were vaginally or systemically immunized with lipopolysaccharide (LPS) of Salmonella typhosa. Vaginal immunization resulted in immunoglobulin responses that were greater in cervical secretions than in serum in the majority of animals; in the systemically immunized animals, however, only 10 percent of the serum response was seen locally (Yang and Schumacher 1979).

In summary, studies indicate that, in humans and primates, there is an efficient local production of IgA in the cervix/vagina in response to antigenic challenge; however, the uterine response is predominantly IgG. Whether the IgG is locally produced or is a serum transudate still remains to be established. Since women who are hysterectomized have IgA levels that are 10 percent of normal, but IgG levels are only decreased by 50 percent, it does appear that there may be a substantial serum contribution. This is supported by a study of hysterectomized mice (Parr and Parr 1989a).

Animal studies have also clearly shown that the mucosal immune system in the female urogenital tract is regulated by sex hormones; in an extensive series of experiments by Wira and associates, local immunoglobulin levels in the reproductive tract were shown to be under the control of estradiol and progesterone (Sullivan and Wira 1984; Wira and Sullivan 1985). Immunoglobulin levels have been reported to decrease at mid-cycle (Yang and Schumacher 1979) and an estrogen-dependent increase of IgA has also been documented (Jalanti and Isliker 1977). Nonetheless, the literature on the effects of the menstrual cycle on local immunoglobulin secretion in both primates and humans is conflicted and needs further study.

The T-cell population of the human and primate female reproductive tract varies considerably from tissue to tissue. In the human endometrium, T cells are rarely seen in the proliferative phase of the menstrual cycle but increase in number in the secretory phase, often forming small aggregates (Bulmer et al. 1988). The majority of these endometrial T cells express the CD8 antigen, particularly those found in the lymphoid aggregates and within the epithelium (Bulmer et al. 1988; Kamat and Isaacson 1986). A unique population of phenotypically unusual CD16-, CD56+ natural killer cell-like large granula leukocytes, and CD8positive intra-epithelial T lymphocytes are found in the ectocervix, the vagina,

and the transformation zone (Edwards and Morris 1985; Miller et al. 1992). Substantial infiltrates of CD8- and CD4-positive cells are seen in the stroma of the transformation zone, often in lymphoid aggregates, but T cells are relatively sparse in the stroma of the ectocervix and vagina (Edwards and Morris 1985). The particularly high numbers of T lymphocytes in the transformation zone of the cervix has led some researchers to believe that this is an area of particularly high immunologic activity (Edwards and Morris 1985).

In the endometrium, the predominant antigen presenting cells are macrophages (Bulmer et al. 1988). The vagina and ectocervix are also characterized by the presence of Langerhans cells, bone marrow-derived antigen-presenting cells with dendritic processes that can extend to the epithelial surface. These cells are found only in the epidermis and non-keratinized epithelia (Bjercke et al. 1983; Edwards and Morris 1985; Miller et al. 1992). After activation, Langerhans cells migrate from the epithelium to draining lymph nodes, where they present antigens to T lymphocytes, thereby initiating a specific immune response. Langerhans cells in the epithelium express a phenotype that is characteristically CD 1a, MHC class II, Fcy, CD3 and, to a lesser extent, CD4-positive; following activation, these cells also express a variety of adhesion molecules including ICAM-1 (CD54) and LFA-3 (CD58) (deGraaf et al. 1995). Langerhans cells show marked variation in their distribution, with the highest numbers found in the vulva and ectocervix and lower numbers in the vagina (Edwards and Morris 1985). There is also considerable individual variation in the concentration of Langerhans cells in the human vaginal and ectocervical epithelium, although the variation does not appear to be related to the menstrual cycle.

Until recently, little had been documented on the immunobiology of either the human or primate male reproductive tract. While immunocytochemical evidence for local synthesis of IgA or secretory component has not been seen in normal epididymis, prostate, or seminal vesicles (Brandtzaeg et al. 1993), in autopsy tissues from HIV-positive men with genital-tract inflammation, all three sites stained strongly for secretory component, suggesting that infection/inflammation may up-regulate its expression (Anderson and Pudney 1992).

Further studies are clearly needed to identify the exact origins of immunoglobulin synthesis, as IgA, IgG and IgM can also be detected in semen and sterile prostatic fluid from healthy men (Fowler et al. 1982). Absorption with antisecretory component results in a 64-100 percent decrease in IgA concentration, indicating that this antibody is in the secretory form and thus is locally derived (Fowler et al. 1982). Levels of all three immunoglobulin types are approximately 10 times higher in prostatic fluid than semen, suggesting that a large proportion of immunoglobulin may be derived from the prostate (Fowler et al. 1982; Fowler

and Mariano 1983). However, recent studies suggest that the penile urethra may be the primary source of sIgA in semen from normal men: In penile urethral tissue taken at autopsy from a series of 15 immunologically normal men, epithelial cells were observed to stain strongly for secretory component and numerous IgA positive plasma cells were seen in the submucosa, indicating that the urethra contains all the elements for production of sIgA (Pudney and Anderson 1995). In addition, studies of semen from HIV-positive and normal seronegative men detected high concentrations of sIgA in pre-ejaculatory fluid, suggesting that there is local urethral production and transport of IgA (Haimovici and Anderson 1992). Pathogen-specific secretory IgA has been detected in the semen of men infected with Chlamydia trachomatis (Kojima et al. 1988), Escherichia coli (Fowler and Mariano 1982), and Neisseria gonorrhoeae (McMillan et al. 1979).

On the basis of these findings, it is possible to hypothesize that, while all the components for secretory immunity are present in various areas of the male reproductive tract, only in the presence of overt infection are they brought into play. Such a hypothesis receives indirect support from studies in the mouse: While IgA-positive plasma cells cannot be detected in the urethra of germ-free mice and rats, scattered cells can be seen in the deep urethral glands (Parr and Parr 1989b; Parr et al. 1992). Secretory component staining is also only seen in the urethral glands, suggesting that a low-level constituent production of sIgA may exist. Yet, after oral immunization followed by a local antigen challenge, IgA-secreting plasma cells were readily detectable in the urethra, suggesting that, in fact, local uptake of antigen and/or inflammation can up-regulate production when required (Husband and Clifton 1989). Very importantly, in the same study, chronic drainage of the thoracic lymphatic duct during the postchallenge period abrogated plasma cell appearance in the urethra after local challenge; reinfusion of these cells resulted in the appearance of urinary tract IgA plasmacytes, documenting for the first time that the male genital tract mucosa is part of the common mucosal system.

T lymphocytes are not usually detected in the healthy human testis (ElDemiry et al. 1985 and 1987), but are seen in the rete testis, with CD8+ cells observable within the epithelium and CD4+ cells primarily in the stroma (ElDemiry and James 1988). T cells are also seen in the human urethra, with a predominance of CD8+ cells in the intraepithelial population and a mixture of CD8 and CD4+ cells in the stroma; T cells in the urethra are positive for the integrin a,^Eß7, an antigen exclusively found on mucosa-associated lymphocytes (Pudney and Anderson 1995). A recent study in mice reported similar findings, with scattered T cells in the urethra positive for the murine mucosa-associated integrin αM290ß7 (Quayle et al. 1994).

Macrophages are abundant in the human and primate male urogenital tract: in the testicular interstitium, epididymis, the epithelium and connective tissue of the excurrent ducts and accessory glands, and in the penile urethra (El-Demiry et al. 1987; Fowler et al. 1982; Miller et al. 1994; Wang and Holstein 1983). Unlike

the female lower urogenital tract, Langerhans cells are rarely detected in the penile urethra, but are abundant in the epithelium of the foreskin (Miller et al. 1994).

Leukocytes are commonly found in the semen of healthy, fertile men (Wolff and Anderson 1988a). Although these cells cross the reproductive tract epithelium, neither the mechanism of migration nor where it takes place along the tract is known. The median number of leukocytes in normal semen is 170,000 but the range is wide—from 9,000 to over 100 million (Wolff and Anderson 1988a). Numbers are elevated in samples from men with evidence of genital-tract inflammation, and a leukocyte count of greater than one million is considered pathologic according to criteria established by the World Health Organization and is associated with decreased sperm parameters and subfertility (Wolff and Anderson 1988b). Various cytokines (TNF-a, IL-6, IL-8) have been detected in seminal fluid from infertile men, indicating that mediators of cellular immunity can also enter male genital tract secretions (Anderson and Hill 1995).

A distinguishing and highly relevant feature of the pathways of immune cell traffic that constitute the common mucosal system is the localization of immune cells at mucosal sites that are distant from the site of mucosal immunization. Studies as early as the 1970s demonstrated the appearance of specific humoral and cellular immunity in respiratory and genital tracts after oral immunization (Ogra and Karzon 1969; Ogra and Ogra 1973). However, many studies have also indicated that optimal mucosal immune responses appear to be obtained by immunization at that mucosal site or an adjacent mucosal locale. The reason for this is unknown, but there are at least two appealing hypotheses. The first is that there are tissue-specific homing receptors additional to those that are specific for generic mucosa. The second is that a depot of antigen is retained locally by macrophages or dendritic cells, resulting in the production of soluble factors that can up-regulate the expression of vascular addressins on the endothelial cells and thence increase the number of lymphocytes homing to this area. If this hypothesis is correct, the administration of antigen in a long-lasting adjuvant would be particularly effective when applied locally.

Examples of compartmentalization are numerous in the genital tract. For example, adoptive transfer of T cells isolated from genital-tract-draining (but not other) lymph nodes (including mesenteric) from mice vaginally immunized with heat-killed herpes simplex virus protected naive mice from vaginal challenge with live virus, verifying that there is also a local homing of T cells (McDermott et al. 1989). Rectal immunization has proved more efficient at stimulating secretory immunity in the genital tract than other mucosal routes, including in one study a vaginal route (Haneberg et al. 1994; Lehner et al. 1992, 1993).

A desirable achievement for an immunogen based upon reproductive or STD pathogen antigens would be the induction of both a local and a systemic immune response, the latter appearing to supplement mucosal immunity at least in some tissues. Systemic immunization alone would not fulfill both criteria, except in specific cases such as when an individual has previously been immunized with the antigen ''naturally" via the mucosa. Thus, milk samples from lactating Swedish women who were parenterally immunized with cholera vaccine had undetectable anticholera sIgA, but Pakistani women who had naturally been exposed to the bacterium had strong anticholera secretory IgA responses in their milk (Svennerholm et al. 1980). Similar results have also been seen in mice orally immunized with cholera and boosted parenterally (Bloom and Rowley 1979), although a study in rats recorded a parenteral immunization with an oral boosting, using cholera toxoid as the antigen, as the most successful regime in rats (Pierce and Gowans 1975). Thus, depending on the antigen, if natural exposure has recently occurred, or continually occurs, a subcutaneous or intramuscular immunization regime may be suitable. However, in the majority of situations, mucosal immunization with parenteral or mucosal boosting appears to be a most suitable regime.

A possible prediction is that any orally administered vaccine would have fewer adverse side effects than one that is administered systemically. It would also be easier to administer en masse, relatively inexpensive since costly injection equipment and highly trained personnel would be unnecessary, and therefore eminently suitable for use in many developing countries. There are, however, a number of difficulties inherent in stimulating mucosal immunity.

First, the mucosal immune system has evolved primarily to avoid sensitization of the individual to antigens in food that escape degradation. As a consequence, the majority of antigenic material is not absorbed so that oral immunization requires substantial amounts of antigen to achieve an immune response. Furthermore, administration of some antigens by the oral route may result in the immunologic unresponsiveness of the systemic immune system on parenteral rechallenge with the same antigen, a phenomenon termed "oral tolerance" and demonstrated in rodent models with a variety of antigens (Challacombe and Tomasi 1980; Mowat 1987). Antigens that are "tolerogenic" tend to be soluble proteins, unlike particulate antigens, particularly viable organisms, that are able to effectively elicit systemic immunity when administered mucosally (de Aizpurua and Russell-Jones 1988). Developing tolerance in T cells is far easier than developing tolerance in B cells (Husby et al. 1994; Titus and Chiller 1981); recent studies in the mouse (Burstein et al. 1992) and in humans (Husby et al. 1994) also showed Thl type T cells to be more easily "tolerized" than the Th2 subset.

There are three keys to any successful oral immunogen: (1) The formulation must achieve survival in the mucosal environment; (2) There must be adequate delivery of antigen to inductive sites; and (3) There must be stimulation of an appropriate protective immune response.

Development of new technology may soon offer a number of possible delivery systems for immunogens, and research into vaccines for enteric, respiratory and, particularly, sexually transmitted diseases will inevitably offer insights into options for mucosal immunization for different purposes.

Mucosally active adjuvants (adjuvare = to help) that are administered with antigen increase the immune response to that antigen. This means that development of mucosal vaccines based upon recombinant microbial proteins or synthetically synthesized peptides that otherwise would be only weakly immunogenic becomes a real possibility. Use of an adjuvant decreases the quantity of antigen needed and, therefore, the cost of the vaccine. Finally, different adjuvant formulations can produce qualitatively different immunologic responses. For example, a study in mice showed that while an antigen administered parenterally in Freund's complete antigen stimulated a Th1-type response, in alum the same antigen elicited a Th2 response, indicating both a potential for flexibility among options and the importance of choosing the correct adjuvant for an appropriate immune response (Grun and Maurer 1989).

Live attenuated or genetically engineered organisms that are capable of colonizing mucosal surfaces yet do not cause pathology can provide a continual effective stimulation to the mucosal immune system. For example, live attenuated polio virus (Sabin vaccine) was first reported to effectively induce sIgA in the gastrointestinal tract (Ogra and Karzon 1969) and has been used in successful immunization programs worldwide. Organisms that can be genetically manipulated to express specific antigens from pathogenic species include bacillus Calmette-Guérin (BCG), avirulent Escherichia coli, Salmonella and Shigella strains, and Adenovirus. A number of studies have demonstrated induction of local immunity after administration of these organisms (Mestecky 1987).

Of particular interest for induction of genital tract immunity is a genetically modified Lactobacillus—the dominant comensal bacteria in the human vagina, with no pathogenic potential whatever—expressing Chlamydia trachomatis antigens. Researchers at Queensland University of Technology and the University of Siena have used the guinea pig as an animal model to assess the possibility of exploiting lactobacilli for a human vaccine against pathogens of the female reproductive tract. Preliminary results showed that the recombinant strain had excellent segregational stability in the absence of antibiotic selection and that it also

persisted, though only for five days, when administered to the guinea pig vaginal tract (Hafner et al. 1992). The main significance of the work so far is in demonstrating that naturally occurring vaginal lactobacilli can be genetically manipulated.

Though eminently suitable for mucosal vaccination, the use of a live organism is not without problems. The attraction of a live organism as a mucosal vaccine is its ability to colonize and replicate at a mucosal surface, thus providing a constant source of antigen for immune stimulation with consequent long-term effects. However, in the case of immunocontraception, reversibility may be highly desirable and a long-term effect might well be disadvantageous; that would not be the case for protection from sexually transmitted infection. A second, potentially catastrophic disadvantage in connection with an immunocontraceptive would be induction of herd immunity owing to transmission of the organism in secretions; in this instance, inactivated organisms, while less efficient as inducers of mucosal immunity, would be the clear preference.

This toxin, produced by Vibrio cholera and the cause of the copious secretion of fluid and electrolytes in cholera disease, consists of five binding (B) subunits in a ring with a single toxin-active (A) subunit. The B subunit binds with high affinity to the ganglioside GM 1 expressed on M cells and the A subunit enters the cell where it ADP ribosylates the Gs subunit of adenylate cyclase and increases cAMP formation. Whole cholera toxin is a powerful mucosal adjuvant in mice, promoting isotype switching to IgA and enhancement of IgA production. Unfortunately, in humans, doses as low as 5 µg can promote severe diarrhea. However, recent studies have indicated that, unlike the mouse system, the B subunit can be administered alone, covalently linked to an antigen of choice, and still retain its stimulatory properties (Holmgren et al. 1993). This is probably due to the ability of the B subunit to bind to M cells, thus enhancing uptake of antigen.

Immune-stimulating complexes (ISCOMs) are stable, 30- to 40-nm-diameter particulate complexes of protein antigens incorporated into dodecahedral-shaped structures composed of the adjuvant Quil A (saponin) and lipid. Parenterally administered ISCOMs are highly immunogenic (Mowat et al. 1991; Takahashi et al. 1990) and have been used successfully in animal models to protect against lethal challenge with a number of viruses and parasites (Araujo and Morein 1991; Cook et al. 1990; de Vries et al. 1988). ISCOMs have been shown to elicit antibody, delayed-type hypersensitivity, T-cell proliferative responses, and cytotoxic T lymphocytes (CTLs). The induction of CD8-positive CTLs implies that,

unlike other exogenously administered antigen, ISCOMs are able to enter the endogenous pathway of antigen processing and therefore be presented in the context of MHC class I (Mowat et al. 1991). Furthermore, ISCOMs have been found capable of fusing with endosomes within the cytoplasm (Claasen and Osterhaus 1993). Recent studies also show that ISCOMs administered by the oral (Mowat et al. 1991), respiratory (Jones et al. 1988), and vaginal (Thapar et al. 1991) routes can induce both systemic and local immunity.

An alternative strategy for delivery of antigen to the mucosa is to incorporate antigen into an appropriate "carrier" or delivery system. Two of the most suitable carriers are biodegradable microspheres and liposomes. It is worth noting that there are not always definitive boundaries between adjuvants and delivery systems as categories; many compounds can also probably modulate immune response in addition to providing a depot of antigen.

Biodegradable microspheres are being evaluated clinically as possible vehicles in a mucosal delivery system. These particles may be constructed of poly(DL-lactide-coglycolide), a copolymer made of the same class of material as resorbable sutures. Antigen is dispersed homogeneously within the copolymeric matrix and the resulting product is a stable powder that is easily reconstitutable.

Microspheres can also be made in different sizes for different effects. In an elegant study by Eldridge and colleagues, mice were orally immunized with different diameter particles incorporating staphylococcal enterotoxin B (SEB). The results were that particles less than 5 µm in diameter very quickly entered the systemic circulation, those between 5 and 10 µm in diameter were retained by the macrophages in the Peyer's patch dome region for a substantial time period, and a combination of sphere sizes (1-10 µm plus 20-50 µm) produced a substantially greater plasma IgG anti-toxin titre at 60 days than when either of the two size ranges was administered individually (Eldridge et al. 1991). This ability to effectively prime and boost immune response in a single administration can also be achieved by changing the ratio of lactide to glycolide in the polymer.

Mucosal immunization with biodegradable microspheres has achieved specific systemic and secretory immune responses, as well as the generation of cytotoxic T lymphocytes (CTLs) (Eldridge et al. 1991; O'Hagan et al. 1992, 1993). The microspheres abrogate the need for use of a live organism and, because they lack surface antigen, potential problems with preexisting sIgA are circumvented.

Liposomes are synthesized globules composed of concentric phospholipid bilayers encapsulating the substance to be delivered. They can be made from

biodegradable components common to all mammalian membranes (Edelman 1980), were originally developed for the delivery of biologically active substances, and have been used to deliver a variety of antigens by several parenteral routes (Buiting et al. 1992). Liposomes are avidly phagocytosed by macrophages and in vitro studies have shown that antigen is effectively delivered to CD4+ T cells (Alving and Richards 1990). A small number of studies have indicated that liposomes are effective when administered via the mucosal route; for example, antigens of Streptococcus mutans delivered in liposomes have been found to enhance specific salivary IgA (Michalek et al. 1985; Wachsmann et al. 1983).

Combinations of mucosal delivery systems and adjuvants may prove to be the most effective way of inducing strong mucosal immunity. An example of this is a vaccine consisting of recombinant simian immunodeficient virus (SIV) gag p27 antigen expressed as hybrid Ty-virus-like particles (Ty-VLP) that are chemically coupled to cholera toxin B subunit. The vaccine has been used successfully for oral/vaginal and oral/rectal immunization regimes in macaques, inducing serum and local antibody and T cell proliferative responses in draining lymph nodes after oral and rectal immunization regimes (Lehner et al. 1992, 1993). Presumably the B subunit allows specific binding to the M cell, while the particle formulation of the vaccine allows a continual depot of antigen. Cytokines such as IL-1, IL-2, and IL-6 have also been incorporated into various formulations as "co-adjuvants" for boosting immune responses (Abraham and Shah 1992; Duits et al. 1993).

Two murine studies of intrauterine plus systemic immunization regimes with sperm or sperm antigens have reported reduced fertility parameters (Haimovici and Anderson 1992; Shelton and Goldberg 1986). Numerous reports have indicated that vaginal immunization in both humans and animals results in the production of sIgA (Ogra and Ogra 1973; O'Hagan et al. 1992; McAnulty and Morton 1978; Thapar et al. 1990; Yang and Schumacher 1979). A number of studies have reported that titres of antibody were rather low after vaginal immunization and, in the murine model, it has been found that injection of antigen into the pelvic presacral space resulted in a much higher local antibody titre (Parr and Parr 1989a; Thapar et al. 1990). In another study in which epididymal sperm were injected via the intraperitoneal route, fertility was shown to be decreased, although local antibodies were not measured (Tung et al. 1979).

The success of these regimes may be due to activation of the resident peritoneal Ly- 1 positive cells, which then populate the genital tract as they do in the gut (Kroese et al. 1988). Alternatively, antigen may be gaining access to the draining

lymph nodes of the genital tract. If this is so and an alternative induction site were to be immunized, this would circumvent the need for antigen to cross the vaginal epithelium, which appears to be a significant barrier to uptake (Parr and Parr 1989a). Although lymphoid nodules do seem to be present in the human vagina, lymphoid tissue there is, in total, much sparser than it is in gut, so that induction of immune response through that route would be predicted to be much less efficient. In fact, in mice, rectal immunization with cholera was far more successful at inducing specific IgA in the vagina than was vaginal immunization (Haneberg et al. 1994). A study in rhesus monkeys with an SIV p27 vaccine expressed as Ty-VLP and conjugated to cholera toxin utilized an oral plus vaginal boosting regime and successfully induced vaginal sIgA and IgG, serum IgG, and IgA and T cells located in genital-tract-draining lymph nodes that proliferated to p27 (Lehner et al. 1992). The effectiveness of this approach was probably due to the fact that oral immunization introduced antigen to, and mobilized, the very substantial gut lymphocyte population, while the local application induced a "local" homing. Rectal immunization alone can also induce good sIgA titres in vaginal secretions, for two possible reasons: First, the rectal mucosa is endowed with numerous lymphoid follicles that allow very effective induction of the immune response; second, as discussed earlier, the genital tract and rectum are adjacent mucosae and share draining lymph nodes.

A limited number of studies have demonstrated that infection or inflammation in the male reproductive tract leads to the appearance of specific IgA. Furthermore, numerous T cells are seen in the urethral mucosa and semen, particularly during times of inflammation. It would appear impractical at this time to target the urethra as a route of mucosal vaccination. However, a study of nine male rhesus monkeys, sequentially immunized ororectally with SIV p27 expressed as Ty-VLP and conjugated to cholera toxin, successfully induced serum IgA and IgG, rectal sIgA p27 antibodies, and T cells isolated from draining lymph nodes, blood, and spleen that proliferated specifically to the antigen (Lehner et al. 1993). This was contrasted with intramuscular immunization, which could only induce serum antibodies and antigen-specific proliferation in blood and spleen T cells. Urethral washings were not examined, but data from rectal immunization in the female does suggest that rectal or sequential oralrectal immunization should induce genital tract immunity. Furthermore, in the mouse, the male lower genital tract and the rectum have also been shown to share draining lymph nodes (Quayle et al. 1994).

IgA, as well as IgG-type antisperm and antiovarian antibodies, have been

detected in infertility patients and are implicated as a cause of reproductive failure (see Best and Hill 1995 for review). However, it has become apparent recently that antisperm cellular immunity, mediated by cytokines, may also play a major role in immunologic fertility. Cellular immune responses to sperm have been detected in male and female infertility patients (Anderson and Hill 1995; McShane et al. 1985; Metter and Schirwani 1975) and are known to cause infertility in animal models (Haimovici et al. 1992; Naz and Metha 1989; Shelton and Goldberg 1986). Furthermore, TH1-type cytokines associated with cellular immunity (i.e., y-interferon, tumor necrosis factors a and ß) are toxic to preimplantation embryos, trophoblast cells, and sperm when present in high concentrations and some of these cytokines also inhibit hormone synthesis by granulosa and Leydig cells (see Anderson and Hill 1995 for review). Finally, there has long been a clinical association between genital tract infections and infertility, an effect that may be due to: 1) adjuvant effects of microbial antigens; and/or 2) local production of TH1-type cytokines, which damage "innocent bystander" reproductive cells (Anderson and Hill 1995).

The conclusions that follow relate to both Parts 1 and 2 of this set of paired papers on the subject of barriers to conception and infection and the mucosal immune system as a key entity in the "construction" of such barriers. It is clear that more research is required to determine whether or not a mucosal immunization strategy will produce the effects that are desired, whether contraceptive, anti-infective, or both. However, the following points can be made that suggest avenues for the next wave of research:

1. The uterus does not have a significant secretory IgA defense system, but is permeable to systemic immunoglobulins. Therefore, reproductive function occurring at this site may be better blocked by systemically-administered vaccines. However, other regions of the tract (cervix, vagina, fallopian tube) appear to have a strong local secretory immune component, so that an immunization regime targeting events that occur at these sites (especially sperm migration and fertilization) should have a mucosally targeted component to take advantage of this local defense mechanism.
2. The penile urethra appears to be the principal site of mucosal immunity in the male genital tract. Since it is not yet clear where or how local immunity can be induced at this mucosal site, more studies of the male genital tract will be essential.
3. In both males and females, a primary oral immunization followed by repeated local boosting (rectal, vaginal, or urethral) would appear to have the best chance of being successful. Local boosters could possibly be provided by self-

3. administered suppositories and a boosting effect might also be produced by reproductive processes themselves.
4. Relatively little is known about cell-mediated immune (CMI) responses within the male or female genital tracts and their potential effects on fertility. Although chronic CMI responses may be associated with undesirable side effects such as delayed-type hypersensitivity and irreversibility, CMI responses to infrequent or low antigen exposures could provide a powerful adjunct mechanism for mucosally targeted immunization.
5. More needs to be known about how secreted antibodies function in mucus to protect the human mucosal surfaces. Despite evidence that antibodies can be highly effective at blocking mucosal transmission of both infectious agents and sperm, both sperm and STD pathogens have evolved mechanisms for successful exchange between sexual partners—mechanisms that evade immune defenses in mucosal secretions—of which one of the most challenging is the common evasion mechanism of antigenic variability, which enables pathogens to stay a step ahead of host immune responses. A potential, though surely challenging, research direction would be to seek combinations of pathogen-specific monoclonal antibodies and pan-semen or other anti-human surface antigenic monoclonal antibodies that could defeat these evasive actions and ultimately provide effective broad-spectrum protection against STDs and pregnancy (Cone and Whaley 1994).

Abraham E, S Shah. Intranasal immunization with liposomes containing IL-2 enhances polysaccharide antigen-specific pulmonary secretory antibody response. Journal of Immunology 149:3719-3726, 1992.

Ada GL. The induction of immunity at mucosal surfaces. IN Local Immunity in Reproductive Tract Tissues. PD Griffin, PM Johnson, eds. Oxford, UK: Oxford University Press. 1993.

Alving CR, RL Richards. Liposomes containing lipid A: A potent nontoxic adjuvant for a human malaria sporozoite vaccine. Immunology Letters 25:275-279, 1990.

Anderson DJ. Mechanisms of HIV-1 transmission via semen. Journal of NIH Research 4:104-108, 1992.

Anderson DJ. Cell mediated immunity and inflammatory processes in male infertility. Archives of Immunology and Therapeutic Experiments 38:79-86, 1990.

Anderson DJ, JA Hill. Cytokines in the reproductive tract that affect fertility. Mucosal Immunology Update 2:6-15, 1995.

Anderson DJ, J Pudney. Mucosal immune defense against HIV-1 in the male urogenital tract. Vaccine Research 1:143, 1992.

Anderson DJ, EJ Yunis. "Trojan Horse" leukocytes in AIDS. New England Journal of Medicine 309:984-985, 1983.

Araujo FG, B Morein. Immunization with Trypanosoma cruzi epimastigote antigens incorporated into ISCOMS protects against lethal challenge in mice. Infectious Immunology 59:2909-2914, 1991.

Beagley KW, JH Eldridge, F Lee, et al. Interleukins and IgA synthesis: Human and mouse IL-6 induce high rate of IgA secretion in IgA committed B cells . Journal of Experimental Medicine 169:2133-2148, 1989.

Berkley S. The public health significance of sexually transmitted diseases in HIV infection in Africa. IN AIDS and Women's Health: Science for Policy and Action. L Chen, et al., eds. New York: Plenum Press. 1991.

Berlin C, EL Berg, MJ Briskin, et al. a4ß7 integrin mediates lymphocyte binding to the mucosal vascular addressin MAdCAM- . Cell 74:185-195, 1993.

Besredka A. De la vaccination contre les états typhoides par la voie buccale. Annales de l'Institute Pasteur 33:882-903, 1919.

Besredka A. Local Immunization. Baltimore: Williams and Wilkins. 1927.

Best CL, JA Hill. Natural infertility due to immunological factors. IN Birth Control Vaccines. R Raghupathy, GP Talwar, eds. Austin, TX: RG Landes. 1995.

Bienenstock J, AD Befus. Mucosal immunology. Immunology 41:249-270, 1980.

Bjercke S, H Scott, LR Braathen, et al. HLA-DR expressing Langerhans-like cells in vaginal and cervical epithelium. Acta Obstetrica Gynecologica Scandinavica 585-589, 1983.

Bloom LD, D Rowley. Local immune response in mice to Vibrio cholerae. Australian Journal of Experimental Biology and Medical Science 57:313-323, 1979.

Bockman DE, MD Cooper. Pinocytosis by epithelium associated with lymphoid follicles in the bursa of Fabricus, appendix, and Peyer's patches: An electron microscopic study. American Journal of Anatomy 136: 455-478, 1973.

Bond MW, B Strader, TR Mosmann, et al. A mouse T cell product that preferentially enhances IgA production. 2. Physicochemical characterization. Journal of Immunology 139:3691-3696, 1987.

Bourinbaiar AS, S Lee-Huang. Comparative in vitro study of contraceptive agents with anti-HIV activity: Gramicidin, nonoxynol-9 and gossypol. Contraception 49:131-137, 1994a.

Bourinbaiar AS, S Lee-Huang. Acrosin inhibitor, 4'-acetamidophenyl 4-guanidinobenzoate, an experimental vaginal contraceptive with anti-HIV activity. Contraception 51:319-322, 1994b.

Bourinbaiar AS, R Nagorny. Effect of serine proteinase inhibitor, N-a-tosyl-L-lysyl-chloromethylketone (TLCK) on cell-mediated and cell-free HIV-1 spread. Cellular Immunology 155:230-236, 1994.

Brandtzaeg P. Human secretory immunoglobulin M: An immunochemical and immunohistochemical study. Immunology 29: 559-570, 1975.

Brandtzaeg P. Two types of IgA immunocytes in man. Nature/New Biology 243:142-143, 1973.

Brandtzaeg P, V Bosnes, TS Halstensen, et al. T lymphocytes in human gut epithelium express preferentially the alpha/beta antigen receptor and are CD45/UCHL-I positive. Scandinavian Journal of Immunology 30:23-128, 1989.

Brandtzaeg P, E Christiansen, F Muller, et al. Humoral immune response patterns of human mucosae, including the reproductive tracts: The induction of immunity at mucosal surfaces. IN Local Immunity in Reproductive Tract Tissues. PD Griffin, PM Johnson, eds. Oxford, UK: Oxford University Press. 1993.

Brandtzaeg P, I Fjellanger, ST Gjeruldsen. Human secretory immunoglobulins. I. Salivary secretions from individuals with normal or low levels of serum immunoglobulins. Scandinavian Journal of Haematology, Supplement 12: 1-83, 1970.

Brandtzaeg P, I Fjellanger, ST Gjeruldsen. Immunoglobulin M: Local synthesis and selective secretion in patients with Immunoglobulin A deficiency. Science 160:789-791, 1968.

Buiting AMJ, N van Rooijen, E Claasen. Liposomes as antigen carriers and adjuvants in vivo. Research in Immunology 143:541-548, 1992.

Bulmer JN, DP Lunny, SV Hagin. Immunohistochemical characterization of stromal leukocytes in nonpregnant human endometrium. American Journal of Reproductive Immunology 17: 83-90, 1988.

Burstein HJ, CM Shea, AK Abbas. Aqueous antigens induce in vivo tolerance selectively in IL-2 and IFN-gamma-producing (Thl) cells. Journal of Immunology 148:3687, 1992.

Cerf-Benussan N, B Begue, J Gagnon, et al. The human intraepithelial lymphocyte marker HML-1 is an integrin consisting of a ß7 subunit associated with a distinctive a chain. European Journal of Immunology 22:73-277, 1992.

Challacombe SJ, T Lehner. Salivary antibody responses in rhesus monkeys immunized with Streptococcus mutans by the oral, submucosal or subcutaneous routes . Archives of Oral Biology 24:917-925, 1979.

Challacombe SJ, TB Tomasi. Systemic tolerance and secretory immunity after oral immunization. Journal of Experimental Medicine 152:1459-1472, 1980.

Chantler E, R Sharma, D Sharman. Changes in cervical mucus that prevent penetration by spermatozoa. Society for Experimental Biology 43:325-336, 1989.

Chipperfield EJ, BA Evans. The influence of local infection on immunoglobulin formation in the human endocervix. Clinical Experiments in Immunology 11:219-223, 1972.

Claasen I, A Osterhaus. The ISCOM structure as an immune-enhancing moiety: Experience with viral systems. Research in Immunology 143:531-541, 1993.

Claypool LE. The challenges ahead: Implications of STDs/AIDS for contraceptive research. IN Contraceptive Research and Development, 1984 to 1994: The Road from Mexico City to Cairo and Beyond. PFA Van Look, G Pérez-Palacios, eds. Delhi: Oxford University Press. 1994.

Cohen MS, RD Weber, PA Mardh. Genitourinary mucosal defenses. IN Sexually Transmitted Diseases (2^nd ed.). KK Holmes, P-A Mardh, PF Sparling et al. eds. New York: McGraw-Hill. 1990.

Cone RA, KJ Whaley. Monoclonal antibodies for reproductive health: Part I, Preventing sexual transmission of disease and pregnancy with topically applied antibodies. American Journal of Reproductive Immunology 32:114-131, 1994.

Cook RF, T O'Neill, E Strachan, et al. Protection against lethal equine herpes virus type 1 (subtype 1) infection in hamsters by immune stimulating complexes (ISCOM) containing the major viral glycoproteins. Vaccine 8:491-496, 1990.

Corner A-M, MM Dolan, SL Yankell, et al. C31G, a new agent for oral use with potent antimicrobial and antiadherence properties. Antimicrobiological Agents and Chemotherapy 32:350-353, 1988.

Cornes JS. Number, size and distribution of Peyer's patches in human small intestine. Gut 6:225-233, 1965.

Crabbe PA, AO Carbonara, JF Heremans. The normal human intestinal mucosa a major source of plasma cells containing A-immunoglobulin. Laboratory Investigation 14:235-248, 1965.

Craig SW, JJ Cebra. Peyer's patches: An enriched source of precursors for IgA producing immunocytes in the rabbit. Journal of Experimental Medicine 134:188-200, 1971.

Czerkinsky C, SJ Prince, SM Michalek, et al. IgA antibody-producing cells in peripheral blood after antigen ingestion: Evidence for a common mucosal system. Proceedings of the National Academy of Sciences, USA 84: 2449-2453, 1987.

D'Cruz OJ, HA Pereira, GG Haas. Sperm immobilizing activity of a synthetic bioactive peptide 2044 of 37-kDa cationic antimicrobial protein (CAP37) of human neutrophils. Journal of Andrology 16:432-440, 1995.

de Aizpurua HJ, GJ Russell-Jones. Oral vaccination: Identification of classes of proteins that provoke an immune response upon oral feeding. Journal of Experimental Medicine 167:440-451, 1988.

DeClercq E. Antiviral agents: Characteristic activity spectrum depending on the molecular target with which they interact. Advances in Virus Research 42:1-55, 1993.

deGraaf JH, RYJ Tamminga, WA Kamps, et al. Expression of cellular adhesion molecules in Langerhans cell histiocytosis and normal Langerhans cells. American Journal of Pathology 147:1161-1171, 1995.

de Vries P, RS van Binnendijk, et al. Measles virus fusion protein presented in an immune-stimulating complex (ISCOM) induces haemolysis-inhibiting and fusion-inhibiting antibodies, virus specific T cells and protection in mice. Journal of General Virology 69:549-559, 1988.

de Waal A, AV Goes, A Mensink, et al. Magainins affect respiratory control, membrane potential and motility of hamster spermatozoa. Federation of European Biological Societies Letters 293:219-223, 1991.

Devlin MC, BN Barwin. Barrier contraception. Advances in Contraception 5:197-204, 1989.

Duits AJ, A van Puijenbroek, H Vermuelen, et al. Immunoadjuvant activity of a liposomal IL-6 formulation. Vaccine 11:777-781, 1993.

Dunbar BS, MG O'Rand, eds. A Comparative Overview of Mammalian Fertilization. New York: Plenum Press. 1991.

Edelman R. Vaccine adjuvants. Review of Infectious Disease 2:370, 1980.

Edelstein MC, JE Gretz, TJ Bauer, et al. Studies on the in vitro spermicidal activity of synthetic magainins. Fertility and Sterility 55:647-649, 1991.

Edwards JNT, HB Morris. Langerhans cells and lymphocyte subsets in the female genital tract. British Journal of Obstetrics and Gynaecology 92:974-982, 1985.

El-Demiry MIM, TB Hargreave, A Busuttil, et al. Immunocompetent cells in human testis in health and disease. Fertility and Sterility 48:470-479, 1987.

El-Demiry MIM, TB Hargreave, A Busuttil, et al. Lymphocyte sub-populations in the male genital tract. British Journal of Urology 57:769-774, 1985.

El-Demiry MIM, K James. Lymphocyte subsets and macrophages in the male genital tract in health and disease. European Journal of Urology 14:226-235, 1988.

Eldridge JH, JK Staas, JA Meulbroek, et al. Biodegradable microspheres as a vaccine delivery system. Molecular Immunology 3:287-294, 1991.

Elias CJ, L Heise. The Development of Microbicides: A New Method of HIV Prevention for Women (Working Papers No. 6). New York: The Population Council . 1993.

Elson CO, JA Heck, W Strober. T-cell regulation of murine IgA synthesis. Journal of Experimental Medicine 149:632-643, 1979.

Foldesy RG, RE Homm, SL Levinson, et al. Multiple actions of a novel vaginal contraceptive compound, OR 13904. Fertility and Sterility 45:550-555, 1986.

Forrest BD, DJA Shearman, JT LaBrooy. Specific immune reponses in humans following rectal delivery of live typhoid vaccines. Vaccine 8:209-212, 1992.

Fowler JE, DL Kaiser, M Mariano. Immunologic response of the prostate to bacteriuria and bacterial prostatis: Part 1. Immunoglobulin concentrations in prostatic fluid. Journal of Urology 128:158-164, 1982.

Fowler JE, M Mariano. Immunoglobulin in seminal fluid of fertile, infertile, vasectomy and vasectomy reversal patients. Journal of Urology 129:869-872, 1983.

Fowler JE, M Mariano. Immunologic response of the prostate to bacteriuria and bacterial prostatitis. II. Antigen specific immunoglobulin in prostatic fluid. Journal of Urology 128:165-170, 1982.

Franek J, J Libich, V. Kubin. Mechanisms of antibacterial immunity of mucous membranes. Folia Microbiologia 29:375-384. 1984.

Ganz T, RI Lehrer. Defensins. Current Opportunities in Immunology 6:584-589, 1994.

Gay FP. Local resistance and local immunity to bacteria. Physiology Reviews 4:191-214, 1924.

Goldblum RM, S Ahlstedt, B Carlsson, et al. Antibody forming cells in human colostrum after oral administration. Nature 257:797-799, 1975.

Greenwood S, AJ Margolis. ''Outercourse." Advances in Planning and Parenting 15:126-128, 1981.

Grun JL, PH Maurer. Different T helper cell subsets elicited in mice utilizing two different adjuvant vehicles: The role of endogenous interleukin 1 in proliferative responses. Cell Immunology 121:134-145, 1989.

Guy-Grand D, C Griscelli, P Vassalli. The mouse gut T lymphocyte, a novel type of T cell: Nature, origin and traffic in mice in normal and graft-versus-host conditions. Journal of Experimental Medicine 169:1277-1294, 1978.

Hafner L, C Rush, P Timms. Lactobacilli: Vehicles for antigen delivery to the female urogenital tract. IN Abstract book of the 7^th International Congress of Mucosal Immunology, Prague, Czechoslovakia, August 1992.

Haimovici F, DJ Anderson. Antifertility effects of antisperm cell-mediated immunity in mice. Journal of Reproductive Immunology 22:281-298, 1992.

Hallenberger S, V Bosch, HJ Angliker, et al. Inhibition of cleavage activation of HIV- 1 glycoprotein gp160. Nature 360:358-361, 1992.

Halpern MS, ME Koshland. Novel subunit in secretory IgA. Nature 228:1276-1278, 1970.

Halstensen TS, H Scott, P Brandtzaeg. Intraepithelial T cells of the TCR gamma/delta+ CD8- and V deltal/J delta 1+ phenotypes are increased in coeliac disease. Scandinavian Journal of Gastroenterology 30:665-672, 1989.

Haneberg B, D Kendall, HM Amerongen, et al. Induction of specific IgA in small intestine, colonrectum, and vagina measured with a new method for collection of secretions from local mucosal surfaces. Infectious Immunology 62(1):15-23, 1994.

Hanson LA. Comparative immunological studies of the immune globulins of human milk and of blood serum. International Archives of Allergy and Applied Immunology 18:241-267, 1961.

Haslett S. Barrier methods of contraception. Nursing Standard 4:24-27, 1990.

Hatcher RA, DL Warner. New condoms for men and women, diaphragms, cervical caps, and spermicides: Overcoming barriers to barriers and spermicides. Current Opinions in Obstetrics and Gynecology 4:513-521, 1992.

Hattori T, A Koito, KJ Takatsuki, et al. Involvement of tryptase-related cellular protease(s) in human immunodeficiency virus type 1 infection. Federation of European Biochemical Societies Letters 248:48-52, 1989.

Holmgren J, N Lycke, C Czerkinsky. Cholera toxin and cholera toxin B subunit as oral-mucosal adjuvant and antigen vector systems. Vaccine 11:1179-1184, 1993.

Holzmann B, BW McIntyre, IL Weissman. Identification of a murine Peyer's patch-specific lymphocyte homing receptor as an integrin molecule with an a chain homologous to human VLA-4a . Cell 56:37-46, 1989.

Homm RE, GE Doscher, EG Hummel, et al. Relative antifertility activity of three vaginal contraceptive products in the rabbit: Relationship to in vitro data. Contraception 13:479-488, 1976.

Homm RE, RG Foldesy, DW Hahn. ORF 13904, a new long-acting vaginal contraceptive. Contraception 32:267-274, 1985.

Hornick CL, F. Karush. Antibody affinity-III: The role of multivalency. Immunochemistry 9:325-340, 1972.

Husband AJ, VL Clifton. Role of intestinal immunization in urinary tract defence. Immunology and Cell Biology 67:371, 1989.

Husband AJ, JL Gowans. The origin and antigen-dependent distribution of IgA containing cells in the small intestine. Journal of Experimental Medicine 148:1146, 1978.

Husby S, J Mestecky, Z Moldoveanu, et al. Oral tolerance in humans: T cell but not B cell tolerance after antigen feeding. Journal of Immunology 152:4663-4670, 1994.

Isaacs CE, S Kashyap, WC Heird, et al. Antiviral and antibacterial lipids in human milk and infant formula feeds. Archives of Diseases in Children 65:861-864, 1990.

Itohara S, AG Farr, JJ Lafaille, et al. Homing of a yS thymocyte subset with homogeneous T-cell receptors to mucosal epithelia. Nature 343:754-757, 1990.

Jacobsen LO, EK Marks, EL Simmons, et al. Immune response in irradiated mice with Peyer's patch shielding. Proceedings of the National Academy of Sciences, USA 108:487-493, 1961.

Jalanti R, H Isliker. Immunoglobulins in human cervicovaginal secretions. International Archives of Allergy and Applied Immunology 53:402-409, 1977.

Janossy G, N Tidman, ES Selby, et al. Human T lymphocytes of the inducer and suppressor type occupy different microenvironments. Nature 288:81-84, 1980.

Jarry A, N Cerf-Benussan, N Brousse, et al. Subsets of CD3+ (T cell receptor alpha/beta or gamma/ delta) and CD3-lymphocytes isolated from normal human gut epithelium display phenotypical features different from their counterparts in peripheral blood. European Journal of Immunology 20:1097-1103, 1990.

Jefferson WL. Non-ionic surfactant spermicidal agents. British Journal of Family Planning 11:131-135, 1986.

Jones PD, JJ Cebra. Restriction of gene expression in B lymphocytes and their progeny: 111. Endogenous IgA and IgM on the membranes of different plasma cell precursors. Journal of Experimental Medicine 140:966-976, 1974.

Jones PP, R Tha Hla, B Morein, et al. Cellular immune responses in the murine lung to local immunisation with influenza A virus glycoproteins in micelles and immunostimulatory complexes (ISCOMs). Scandinavian Journal of Immunology 27:645-652, 1988.

Kamat BR, PG Isaacson. The immunocytochemical distribution of leukocytic subpopulations in human endometrium. American Journal of Pathology 127:66-73, 1986.

Kaminski JM, NA Nuzzo, L Bauer, et al. Vaginal contraceptive activity of aryl 4-guanidinobenzoates (acrosin inhibitors) in rabbits. Contraception 32:183-189, 1985.

Katz DH, JF Marceletti, MH Khalil, et al. Antiviral activity of 1-docosanol, an inhibitor of lipidenveloped viruses including herpes simplex. Proceedings of the National Academy of Sciences, USA 88:10825-10829, 1991.

Katz DH, JF Marceletti, LE Pope, et al. n-Docosanol: Broad spectrum antiviral activity against lipid-enveloped viruses. Annals of the New York Academy of Sciences 724:472-488, 1994.

Kilshaw PJ, SJ Murant. Expression and regulation of ß7(ß) integrins on mouse lymphocytes: Relevance to the mucosal immune system. European Journal of Immunology 21:2591-2597, 1991.

Kilshaw PJ, SJ Murant. A new surface antigen on intraepithelial lymphocytes in the intestine. European Journal of Immunology 20:2201-2207, 1990.

Klebanoff SJ. Effects of the spermicidal agent nonoxynol-9 on vaginal microbial flora. Journal of Infectious Diseases 165:19-25, 1992.

Klebanoff SJ, RW Coombs. Viricidal effects of Lactobacillus acidophilus on human immunodeficiency virus type 1: Possible role in heterosexual transmission. Journal of Experimental Medicine 174: 289-292, 1991.

Koito A, T Hattori, T Murakami, et al. A neutralizing epitope of human immunodeficiency virus type 1 has homologous amino acid sequences with the active site of inter-a-trypsin inhibitor . International Immunology 1:613-618, 1989.

Kojima H, S-P Wang, C-C Kuo, et al. Local antibody in semen for rapid diagnosis of Chlamydia trachomatis epididymitis. Journal of Immunology 140:528-531, 1988.

Kreiss J, E Ngugi, K Holmes, et al. Efficacy of nonoxynol-9 contraceptive sponge use in preventing heterosexual acquisition of HIV in Nairobi prostitutes. Journal of the American Medical Association 268:477-482, 1992.

Krieger JN, MF Rein. Canine prostatic secretions kill Trichomonas vaginalis. Infectious Immunology 37:77-81, 1982.

Kroese FGM, EC Butcher, AM Stall, et al. Many of the IgA producing plasma cells in the murine gut are derived from self-replenishing precursors in the peritoneal cavity. International Immunology 1:75-84, 1988.

Kutteh WH, RP Edwards, AC Menge, et al. IgA immunity in female reproductive tract secretions. IN Local Immunity in Reproductive Tract Tissues. PD Griffin, PM Johnson, eds. Oxford, UK: Oxford University Press, 1993.

Kutteh WH, C Kutteh, RE Blackwell, et al. Secretory immune system of the female reproductive tract. II. Local immune system in normal and infected fallopian tube. Fertility and Sterility 54:51-55, 1990.

Kutteh WH, J Mestecky. Secretory immunity in the female reproductive tract. American Journal of Reproductive Immunology 31:40-46, 1994.

Lehner T, LA Bergmeier, C Panagiotidi, et al. Induction of mucosal and systemic immunity to a recombinant simian immunodeficiency viral protein. Science 258:1365-1369, 1992.

Lehner T, R Brookes, C Panagiotidi, et al. T- and B-cell functions and epitope expression in nonhuman primates immunized with simian immunodeficiency virus antigen by the rectal route. Proceedings of the National Academy of Sciences, USA 90:8638-8642, 1993.

Lehrer RI. Defensins: Antimicrobial and cytotoxic peptides of mammalian cells. Annual Review of Immunology 11:105-128, 1993.

Lenaerts V, R Gurny. Bioadhesive Drug Delivery Systems. Boca Raton, FL: CRC Press. 1990.

London SD, JJ Cebra, DH Rubin. Intraepithelial lymphocytes contain virus-specific MHC-restricted precursors after gut mucosal immunization with reovirus serotype 1/Lang. Regional Immunology 2:99-105, 1989.

Ma JK-C, A Hiatt, M Hein, et al. Generation and assembly of secretory antibodies in plants. Science 268:716-719, 1995.

Mann J, DJM Tarantola, TW Netter, eds. A Global Report: AIDS in the World. Cambridge, MA: Harvard University Press. 1992.

Mardh P-A, S Colleen, J Sylwan. Inhibitory effect on the formation of chlamydial inclusions in McCoy cells by seminal fluid and some of its components. Investigations in Urology 17:510-513, 1980.

Masters WH, VE Johnson. Human Sexual Response. New York: Bantam Books. 1980.

Mauck CK, M Cordeo, HL Gabelnick, et al., eds. Barrier Contraceptives: Current Status and Future Prospects. New York: Wiley-Liss. 1994.

Mayer KH, DJ Anderson. Issue of the day: Heterosexual HIV transmission. Infectious Agents and Disease 4:273-284, 1995.

Mazanec MB, JG Nedrug, et al. A three-tiered view of the role of IgA in mucosal defense. Immunology Today 14:430-435, 1993.

McAnulty PA, DB Morton. The immune response of the genital tract of the female rabbit following local and systemic immunization. Clinical Laboratory Immunology 1:255-260, 1978.

McCune JM, LB Rabin, MB Feinberg, et al. Endoproteolytic cleavage of gp160 is required for the activation of human immunodeficiency virus. Cell 53:5-57, 1988.

McDermott MR, J Bienenstock. Evidence for a common mucosal system. 1. Migration of B immunoblasts into intestinal, respiratory and genital tissues. Journal of Immunology 122:1892-1898, 1979.

McDermott MR, CH Goldsmith, KL Rosenthal, L Brais. T lymphocytes in genital lymph nodes protect mice from intravaginal infection with Herpes Simplex type 2. Journal of Infectious Diseases 159:460-466, 1989.

McGhee JR, KW Beagley, K Fujihashi, et al. Regulatory mechanisms in mucosal immunity: Roles for T helper cell subsets and derived cytokines in the induction of IgA responses. IN Local Immunity in Reproductive Tract Tissues. PD Griffin, PM Johnson, eds. Oxford, UK: Oxford University Press. 1993.

McGhee JR, J Mestecky, MT Dertzbaugh, et al. The mucosal immune system: From fundamental concepts to vaccine development. Vaccine 10:75-88, 1992.

McMillan A, G McNeillage, H Young. Antibodies to Neisseria gonorrhoeae: A study of the urethral exudates of 232 men. Journal of Infectious Diseases 140:89-95, 1979.

McShane PM, I Schiff, DE Trentham. Cellular immunity to sperm in infertile women. Journal of the American Medical Association 253:3555-3558, 1985.

Mestecky J. The common mucosal immune system and current strategies for induction of immune responses in external secretions. Journal of Clinical Immunology 7:265-276, 1987.

Mestecky J, JR McGhee. The secretory IgA system. IN Local Immunity in Reproductive Tract Tissues. PF Griffin, PM Johnson, eds. Oxford, UK: Oxford University Press. 1993.

Mestecky J, JR McGhee. Oral immunization: Past and present. IN Current Topics in Microbiology and Immunology: New Strategies of Oral Immunization. J Mestecky, JR McGhee, eds. Berlin: Springer Verlag. 1989.

Mestecky J, JR McGhee, RR Arnold et al. Selective induction of immune responses in external secretions. Journal of Clinical Investigation 61:731-737, 1978.

Mestecky J, J Zikan, W Butler. Immunoglobulin M and secretory immunoglobulin A: Evidence for a common polypeptide chain different from light chains. Science 171:1163-1165, 1971.

Metter L, D Schirwani. Macrophage migration inhibitory factor in female sterility. American Journal of Obstetrics and Gynecology 121:117-120, 1975.

Michaels EB, EC Hahn, AJ Kenyon. Effect of C31G, an antimicrobial surfactant, on healing of incised guinea pig wounds. American Journal of Veterinary Research 44:1378-1381. 1983.

Michalek SM, I Morisaki, IL Gregory, et al. Oral adjuvants enhance salivary IgA responses to purified Streptococcus mutans antigens. Protides Biology of Fluids 32:47-52, 1985.

Miller CJ, JR McGhee, MB Gardner. Biology of disease: Mucosal immunity, HIV transmission and AIDS. Laboratory Investigation 68:129-145, 1992.

Miller CJ, P Vogel, NJ Alexander, et al. Pathology and localization of SIV in the reproductive tract of chronically infected male rhesus macaques. Laboratory Investigation 70:255-262, 1994.

Mims CA. The Pathogenesis of Infectious Disease. London: Academic Press. 1988.

Montgomery PC, J Cohn, ET Lally. The induction and characterization of secretory IgA molecules. Advances in Experimental Medicine and Biology 45:453-462, 1976.

Moore KS, S Wehrli, H Roder, et al. Squalamine: An aminosterol antibiotic from the shark. Proceedings of the National Academy of Sciences, USA 90:1354-1358, 1993.

Mosmann TR, RL Coffman. Thl and Th2 cells: Different patterns of lymphokine secretion lead to different functional properties. Annual Review of Immunology 145-173, 1989.

Mowat AM. The regulation of immune responses to dietary protein antigens. Immunology Today 8:93-98, 1987.

Mowat AM, AM Donachie, G Reid, et al. Immune-stimulating complexes containing Quil A and protein antigen prime class 1 MHC-restricted T lymphocytes in vivo are immunogenic by the oral route. Immunology 72:317-322, 1991.

Murray PD, DT McKenzie, SL Swain, et al. Interleukin 5 and interleukin 4 produced by Peyer's patch T cells selectively enhance immunoglobulin A expression. Journal of Immunology 139:2669-2674, 1987.

National Research Council. Neem: A Tree for Solving Global Problems. Washington, DC: National Academy Press. 1992.

Naz RK, K Metha. Cell-mediated immune responses to sperm antigens: Effect on mouse sperm and embryos. Biology of Reproduction 41:533-542, 1989.

Neurath AR, S Jiang, N Strick, et al. Bovine beta-lactoglobulin modified by 3-hydroxyphatalic anhydride blocks the CD4 cell receptor for HIV. Nature Medicine 2:230-234, 1996.

Niruthisard S, RE Roddy, S Chutivongse. Use of nonoxynol-9 and reduction in rate of gonococcal and chlamydial cervical infections. Lancet 339:1371-1375, 1992.

Niruthisard S, RE Roddy, S Chutivongse. The effects of frequent nonoxynol-9 use on the vaginal and cervical mucosa. Sexually Transmitted Diseases 18:176-179, 1991.

Norman LR, J Cornett. Discussing nonpenetrative sexual activities as safer sex alternatives for HIV prevention with HIV-infected clients. Presentation at workshop sponsored by the Centers for Disease Control, Washington, DC, 1995.

Norsigian J. Feminist perspective on barrier use. IN Barrier Contraceptives: Current Status and Future Prospects. CK Mauck, M Cordero, HL Gabelnick, et al., eds. New York: Wiley-Liss. 1994.

O'Hagan DT, JP McGhee, J Holmgren, et al. Biodegradable microparticles for oral immunization. Vaccine 11:149-154, 1992.

O'Hagan DT, D Rafferty, S Wharton. Intravaginal immunization in sheep using a bioadhesive microsphere antigen delivery system. Vaccine 11:660-664, 1993.

O'Reilly RJ, L Lee, BG Welch. Secretory IgA antibody responses to Neisseria gonorrhoeae in the genital tract secretions of infected females. Journal of Infectious Diseases 133:113-125, 1976.

Ogra PL, DT Karzon. Distribution of poliovirus antibody in serum, nasopharynx and alimentary tract following segmental immunization of lower alimentary tract with poliovirus. Journal of Immunology 102:1423-1427, 1969.

Ogra PL, SS Ogra. Local antibody response to polio vaccine in the human genital tract. Journal of Immunology 110:1307-1311, 1973.

Owen RL, AL Jones. Epithelial cell specialization within Peyer's patches: An ultrastructural study of intestinal lymphoid follicles. Gastroenterology 66:189-203, 1974.

Parker CM, KL Cepek, GJ Russell, et al. A family of ß7 integrins on human mucosal lymphocytes. Proceedings of the National Academy of Sciences, USA 89:1924-1928, 1992.

Parr MB, EL Parr. A comparison of antibody titres in mouse uterine fluid after immunization by several routes, and the effect of the uterus on antibody titres in vaginal fluid. Journal of Reproduction and Fertility 89:619-625, 1989a.

Parr MB, EL Parr. Immunohistochemical localization of secretory component and immunoglobulin A in the urogenital tract of the male rodent. Journal of Reproduction and Fertility 85:115-124, 1989b.

Parr MB, HP Ren, LD Russell, et al. Urethral glands of the male mouse contain secretory component and immunoglobulin A plasma cells and are targets of testosterone. Biology of Reproduction 47:1031-1039, 1992.

Patton DL, SK Wang, CC Kuo. In vitro activity of nonoxynol-9 on HeLa 229 cells and primary monkey cervical epithelial cells infected with Chlamydia trachomatis. Antimicrobial Agents and Chemotherapy 36:1478-1482, 1992.

Pierce NF, JL Gowans. Cellular kinetics of the intestinal immune response to cholera toxoid in rats. Journal of Experimental Medicine 142:1550, 1975.

Potts M. The urgent need for a vaginal microbicide in the prevention of HIV transmission. American Journal of Public Health 84:890-891, 1994.

Pudney J, DJ Anderson. Immunology of the human male urethra. American Journal of Pathology 147:155-165, 1995.

Quayle AJ, J Pudney, D Muñoz, et al. Characterization of T lymphocytes and antigen presenting cells in the murine male urethra. Biology of Reproduction 51:809-820, 1994.

Quigg JM. Development and evaluation of pH sensitive bioerodible polymers for the controlled release of vaginal contraceptive agents. PhD dissertation, University of Illinois at Chicago, 1991.

Quigg JM, IF Miller, SR Mack, et al. Development of polyurethane sponge as a delivery system for aryl 4-guanidinobenzoates. Contraception 38:487-497, 1988.

Raff HV, C Bradley, W Donaldson, et al. Comparison of functional activities between IgG and IgM class-switched human monoclonal antibodies reactive with group b streptococci or Escherichia coli K1. Journal of Infectious Diseases 163:346-354, 1991.

Romagnani S. Induction of Th and Th2 responses: A key role for the "natural" immune response? Immunology Today 13:379-381, 1992.

Rosenberg MJ, AJ Davison, JH Chen et al. Barrier contraceptives and sexually transmitted diseases in women: A comparison of female-dependent methods and condoms. American Journal of Public Health 82:669-674, 1992.

Rosenberg MJ, W Rojanapithayakorn, PJ Feldblum, et al. Effect of the contraceptive sponge on chlamydial infection, gonorrhea, and candidiasis: A comparative clinical trial. Journal of the American Medical Association 257:2308-2312, 1987.

Rush CM, LM Hafner, P Timms. Lactobacilli: Vehicles for antigen delivery to the female urogenital tract. Advances in Experimental Medicine and Biology 371B:1547-1552, 1995.

Sadownik A, G Deng, V Janout, et al. Rapid construction of a squalamine mimic. Journal of the American Chemical Society 117:6138-6139, 1995.

Schonwetter BS, ED Stolzenberg, MA Zasloff. Epithelial antibiotics induced at sites of inflammation. Science 267:1645-1648, 1995.

Sharman D, E Chantler, M Dukes, et al. Comparison of the action of nonoxynol-9 and chlorhexidine on sperm. Fertility and Sterility 45:259-264, 1986.

Shedlovsky L, D Belcher, I Levenstein. Titrations of human seminal fluid with acids and alkalis and their effects on the survival of sperm motility. American Journal of Physiology 136:535-541, 1942.

Shelton JA, E Goldberg. Local reproductive tract immunity to sperm-specific lactate dehydrogenase-C4. Biology of Reproduction 35:873-876, 1986.

Sherwood JK, L Zeitlin, KJ Whaley, et al. Controlled release of antibodies for long-term topical passive immuno-protection of female mice against genital herpes. Nature Biotechnology 14:468-471, 1996.

Sobrero AG. Use and effectiveness of condoms, diaphragms, cervical cap, vaginal sponge, and spermicides. IN Gynecology and Obstetrics, Vol. 6. GI Zatuchni, JJ Laferla, JJ Sciarra, eds. Philadelphia: J.B. Lippincott. 1989.

Stein Z. HIV prevention: An update on the status of methods women can use. American Journal of Public Health 83(10):1379-1382, 1993.

Stein Z. HIV prevention: The need for methods women can use. American Journal of Public Health 80:460-462, 1990.

Stone KM. HIV, other STDs, and barriers. IN Barrier Contraceptives: Current Status and Future Prospects. CK Mauck, M Cordero, HL Gabelnick, et al., eds. New York: Wiley-Liss. 1994.

Stratton P, NJ Alexander. Prevention of sexually transmitted infections. Infectious Disease Clinics of North America 7(4):841-859, 1993.

Sullivan DA, CR Wira. Hormonal regulation of immunoglobulins in the uterus: Uterine response to multiple estradiol treatments. Endocrinology 114:650-658, 1984.

Svennerholm A-M, LA Hanson, J Holmgren, et al. Different secretory immunoglobulin A antibody responses to cholera vaccination in Swedish and Pakistani women. Infectious Immunology 30:427-430, 1980.

Takahashi H, T Takeshita, B Morein, et al. Induction of CD8+ cytotoxic T cells by immunization with purified HIV- envelope protein in ISCOMS. Nature 344:873-875, 1990.

Talwar GP, S Garg, R Singh, et al. Praneem polyherbal cream and suppositories. IN Barrier Contraceptives: Current Status and Future Prospects. CK Mauck, M Cordero, HL Gabelnick, et al. New York: Wiley-Liss. 1994.

Thapar MA, EL Parr, JJ Bozzola, et al. Secretory immune responses in the mouse vagina after parenteral or intravaginal immunization with an immunostimulating complex (ISCOM) . Vaccine 9:129-133, 1991.

Thapar MA, EL Parr EL, MB Parr. Secretory immune response in mouse vaginal fluid after pelvic, parenteral or vaginal immunization. Immunology 70:121-125, 1990.

Thormar H, CE Isaacs, HR Brown, et al. Inactivation of enveloped viruses and killing of cells by fatty acids and monoglycerides. Antimicrobial Agents in Chemotherapy 31:27-31, 1987.

Titus RG, JM Chiller. Orally induced tolerance. International Archives of Allergy and Applied Immunology 65:323-338, 1981.

Tjokronegoro A, S Sirisinha. Quantitative analysis of immunoglobulins and albumin in secretion of female reproductive tract. Fertility and Sterility 26:413-417, 1975.

Tomasi TB, EM Tan, A Soloman, et al. Characteristics of an immune system common to external secretions. Journal of Experimental Medicine 121:101, 1965.

Tomasi TB, S Ziegelbaum. The selective occurence of gamma 1A globulins in certain body fluids. Journal of Clinical Investigation 42:1552-1560, 1963.

Trussell J. Contraceptive efficacy of barrier contraceptives. IN Barrier Contraceptives: Current Status and Future Prospects. CK Mauck, M Cordero, HL Gabelnick, et al., eds. New York: Wiley-Liss. 1994.

Tsai C-C, KE Follis, A Sabo et al. Prevention of SIV infection in macaques by (R)-9-(2 phosphonylmethoxypropyl) adenine. Science 270:1197-1199, 1995.

Tung KSK, EH Goldberg, E Goldberg. Immunobiological consequence of immunization of female mice with homologous spermatozoa: Induction of infertility. Journal of Reproductive Immunology 1:145-158, 1979.

Voeller B, DJ Anderson. Heterosexual transmission of HIV. Journal of the American Medical Association 31:27-31, 1992a.

Voeller B, DJ Anderson. pH and related factors in the urogenital tract and rectum that affect HIV- I transmission. Mariposa Occasional Paper No. 16, Topanga, CA, 1992b.

Wachsmann D, JP Klein, M Scholler, et al. Local and systemic immune response to orally administered liposome-associated soluble S. mutans cell wall antigens . Immunology 54:189-193, 1983.

Walter BA, GA Digenis et al. High-performance liquid chromatographic (HPLC) analysis of oligomeric components of the spermicide nonoxynol-9. Pharmacology Research 8:409-411, 1991a.

Walter BA, AA Hawi et al. Solubilization and in vitro spermicidal assessment of nonoxynol-9 and selected fraction using rabbit spermatozoa. Pharmacology Research 8:403-408, 1991b.

Wang YE, A-F Holstein. Intraepithelial lymphocytes in the human epididymis. Cell and Tissue Research 233:517- 521, 1983.

Wasserheit JN. Epidemiological synergy: Interrelationships between human immunodeficiency virus infection and other sexually transmitted diseases. Sexually Transmitted Diseases 19:61-77, 1992.

Wiley JA, SJ Herschkorn, N Padian. Heterogeneity in the probability of HIV transmission per sexual contact: The case of male-to-female transmission in penile-vaginal intercourse. Statistics in Medicine 8:93-102, 1988.

Wira CR, CP Sandoe. Effect of uterine immunization and estradiol on specific IgA and IgG antibodies in uterine, vaginal and salivary secretions. Immunology 68:24-36, 1989.

Wira CR, DA Sullivan. Estradiol and progesterone regulation of IgA, IgG and secretory component in cervicovaginal secretions of the rat. Biology of Reproduction 32:90-95, 1985.

Wolff H, DJ Anderson. Immunologic characterization and quantization of leukocyte subpopulations in human semen. Fertility and Sterility 49:497-504, 1988a.

Wolff H, DJ Anderson. Male genital tract inflammation associated with increased numbers of potential human immunodeficiency virus host cells in semen. Andrologia 20:404-410, 1988b.

Yang SL, GFB Schumacher. Immune response after vaginal application of antigens in the rhesus monkey. Fertility and Sterility 32:588-598, 1979.

Zaneveld LJD. Vaginal contraception since 1984: Chemical agents and barrier devices. IN Contraceptive Research and Development, 1984 to 1994: The Road from Mexico City to Cairo and Beyond. PFA Van Look, G Pérez-Palacios, eds. Delhi: Oxford University Press. 1994a.

Zaneveld LJD. Vaginal contraceptive efficacy: Animal models. IN Barrier Contraceptives: Current Status and Future Prospects. CK Mauck, M Cordero, HL Gabelnick, et al., eds. New York: Wiley-Liss. 1994b.

Zaneveld LJD, AK Bhattacharyya, DS Kim, et al. Primate model for the evaluation of vaginal contraceptives. American Journal of Obstetrics and Gynecology 129:368-372, 1977.

Zaneveld LJD, RT Robertson, WL Williams. Synthetic enzyme inhibitors as antifertility agents. Federation of European Biochemical Societies Letters 11:345-347, 1970.

Zheng H-Y, TM Alcorn, MS Cohen. Effects of H202-producing lactobacilli on Neisseria gonorrhoeae growth and catalase activity. Journal of Infectious Diseases 170:1209-1215, 1994.