The Mammary Gland as an Integral Component of the Common Mucosal Immune System

relevant mucosal niches. It appears that in addition to specific antibody activity, abundant SIgA-associated glycan side chains on SC and heavy chains effectively
bind to the corresponding bacterial glycan structures, thus preventing their adherence to epithelial receptors [12]. However, SIgA in concert with humoral factors of innate immunity enhances (or focuses) their antimicrobial activities [6].

The inhibition of absorption of soluble but biologically inert antigens from food by SIgA has been well documented as a means to prevent the overstimulation of the entire immune system by the increased absorption of such antigens in the absence of specific antibodies [6]. Furthermore, it should be stressed that due to the multivalency of milk SIgA (4 antigen binding sites in dimers and 8 in tetramers), the biological effectiveness for viral neutralization is enormously enhanced due to the bonus effect of multivalency [6, 12]. Specificity of Antibodies in Milk Reflects the Site of Antigenic Stimulation
at Various Inductive Sites The mammary gland as the effector site is populated by precursors of IgA-producing cells from remote inductive sites [7, 8, 10]. Consequently, the specificity of SIgA in milk depends on the encounter with antigens, dominantly in the gastrointestinal and respiratory tracts. As shown in Table 2, human colostrum and milk contain antibodies of the IgA isotype to a broad spectrum of environmental antigens of microbial and food origin. In response to the antigens encountered by the mother at the time of pregnancy and after the delivery, such antibodies provide the most relevant passive SIgA-mediated immunity [1–4, 13]. Although there is considerable individual variability in total and antigen-specific SIgA antibody levels among lactating mothers, antibodies to gram-positive and gramnegative bacteria, viruses, and food antigens are commonly found in all samples [4, 13]. Not surprisingly and in harmony with earlier results, antibodies to food antigens and antigens I/II of the oral bacterium Streptococcus mutans are associated with the IgA1 subclass, while those against lipopolysaccharides of gramnegative bacteria present in the large intestine; bacterial polysaccharides are
mostly IgA2 [9, 13]. Interestingly, the distribution of total IgA1 and IgA2 was ∼53 and 47%, respectively, with marked individual variability [13]. In comparison to other external secretions, human milk is in this respect reminiscent of the IgA subclass distribution in the lower intestinal tract and markedly differs from the secretions of the upper respiratory and upper digestive tracts [5]. There are, however, several interesting observations which remain unexplained. External secretions, including milk and sera obtained from HIV-infected individuals, display extremely low levels of HIV-specific antibodies of the IgA isotypes irrespective of the route of HIV infection [14]. It appears that HIV, specifically its negative factor (nef), selectively suppresses IgA responses. Furthermore, milk collected from lactating mothers contains high titers of antisperm antibodies of the IgA isotype [15]. Because the female genital tract is a poor inductive site, and various antigens administered intravaginally stimulate only minimal or no local
responses [16], it is possible to speculate that these antisperm antibodies are induced due to the exposure to sperm by alternative inductive sites, specifically through oral or anal receptive sexual encounter, both of which are effective in the induction of generalized mucosal immune responses [17].

The presence of antibody-secreting cells as well as T cells in secretory glands, including the mammary gland [18], was exploited in several attempts to induce local responses in animal models. Thus, the injection of microbial antigens or hapten carrier conjugates into the mammary glands of experimental animals has been used for the induction of antibody responses in milk [1, 19]. Importantly, such immunization induces the concomitant systemic responses dominated by IgG rather than IgA in sera of immunized animals. Furthermore, adjuvants causing local inflammatory reactions in the immunized gland were usually required for boosting of local responses. Although retrograde instillation of antigens into the duct of mammary glands has been explored with some success in animal experiments [1], the acceptance of such immunization attempts in humans is highly unlikely. The possibility that antibodies from the circulation
could contribute significant quantities of IgG or IgA to external secretions, including the milk, has been explored in early studies [1]. In animals whose milk contains IgG as the dominant isotype, systemic immunization is effective, and IgG of plasma origin is present in their milk [1]. In contrast, in human milk, low levels of total IgG are present. Nevertheless, subcutaneous immunization with 3 selected rubella virus vaccines resulted in the induction of IgG-, IgM-, and IgAspecific antibodies in sera, but only minimal levels of IgG or IgM virus-specific antibodies were detectable in milk [20]. However, IgA antibodies were detectable in the milk of all systemically immunized women 2–4 weeks after immunization, with peak responses at the 4th week. Although primary immunization with previously unencountered antigen does not stimulate mucosal immune responses, in women who had been previously sensitized by the mucosal route, systemic immunization could evince an SIgA response in milk, as demonstrated with cholera vaccine [19]. In previously unexposed lactating Swedish women, systemically administered cholera vaccine did not induce SIgA antibodies in
milk. In contrast, in lactating Pakistani women, presumably naturally exposed
to Vibrio cholerae, such an immunization induced SIgA antibodies in milk [19]. Thus, the initial mucosal exposure profoundly influenced the outcome of subsequent systemic immunization and may favor the induction of milk SIgA antibodies to some antigens.
Induction of Generalized Mucosal Responses and the Common Mucosal Immune System Pioneering studies performed in animals and later in humans helped to discover
the origin of antibody-forming cells in anatomically remote mucosal tissues and associated secretory glands, including the mammary, salivary, and tear glands, and had an enormous impact on the feasible approaches and design of mucosally administered vaccines [19]. Oral administration of dinitrophenylated pneumococci to lactating rabbits led to the appearance of specific antibodies of the IgA isotype in milk [21]. Subsequent extensive experiments with oral administration of various particulate or soluble antigens confirmed earlier observations and demonstrated that, in addition to milk, SIgA antibodies were present in secretions of other mucosal tissues and glands [19]. The fundamental explanation for these observations of such importance was provided by earlier studies 


concerning the tissue origin of IgA-producing cells found in mucosal tissues and glands [22]. Several investigators demonstrated that lymphocytes from the intestinal Peyer’s patches, bronchus-associated lymphoid tissue, and perhaps Waldeyer’s ring of the oropharynx are an enriched source of cells that express the potential to populate remote mucosal tissues and glands [19, 23]. Experiments performed primarily in mice led to the extended phenotyping of IgA precursor cells as surface IgA-positive cells that express surface receptors, later termed homing receptors, that are involved in interactions with specific ligands present on the surface of endothelial cells of postcapillary venules in mucosal tissues [24]. Importantly, for immunization studies and their physiologic interpretation, cells from such inductive sites, including Peyer’s patches, bronchusassociated lymphoid tissue, Waldeyer’s ring, rectal tonsils, and sublingual tissue, differ in their expression of homing receptors, which interact with corresponding ligands and, therefore, lead to the tissue-selective distribution of cells from various inductive site to tissue-elective effector sites (Table 3) [19, 23, 24]. It
should be stressed, however, that information relevant to the human mammary gland with respect to the origin of IgA precursor cells from various inductive sites has not been, for obvious ethical reasons, explored and remains controversial. Based on immunohistochemical staining of 2 samples of lactating mammary glands and spectra of antibodies, some authors speculate that the IgA precursors originate in the upper respiratory tract and Waldeyer’s ring, while others concerning the tissue origin of IgA-producing cells found in mucosal tissues and glands [22]. Several investigators demonstrated that lymphocytes from the intestinal Peyer’s patches, bronchus-associated lymphoid tissue, and perhaps Waldeyer’s ring of the oropharynx are an enriched source of cells that express the potential to populate remote mucosal tissues and glands [19, 23]. Experiments performed primarily in mice led to the extended phenotyping of IgA precursor cells as surface IgA-positive cells that express surface receptors, later termed homing receptors, that are involved in interactions with specific ligands present on the surface of endothelial cells of postcapillary venules in mucosal
tissues [24]. Importantly, for immunization studies and their physiologic interpretation,
cells from such inductive sites, including Peyer’s patches, bronchusassociated lymphoid tissue, Waldeyer’s ring, rectal tonsils, and sublingual tissue, differ in their expression of homing receptors, which interact with corresponding ligands and, therefore, lead to the tissue-selective distribution of cells from various inductive site to tissue-elective effector sites (Table 3) [19, 23, 24]. It should be stressed, however, that information relevant to the human mammary gland with respect to the origin of IgA precursor cells from various inductive sites has not been, for obvious ethical reasons, explored and remains controversial. Based on immunohistochemical staining of 2 samples of lactating mammary glands and spectra of antibodies, some authors speculate that the IgA precursors originate in the upper respiratory tract and Waldeyer’s ring, while others 

1. Which immunization route is effective in the induction of antibodies of desired specificity in human milk? The intranasal, sublingual, oral, and perhaps rectal routes of immunization should be compared.
2. Determine the phenotypes of specific antibody-secreting cells in the mammary gland, as compared to the peripheral blood, with respect to the immunoglobulin isotypes and expression of homing receptors.
3. Determine the levels and duration of humoral immune responses of specific antibodies with regard to the immunization route, immunoglobulin isotypes, and the antigen delivery systems in human milk.

Acknowledgment