Peter F. Wright, Dartmouth Medical School, Division of Infectious Disease and International Health, 1 Medical Center Drive, Lebanon, NH, 03756, USA. E-mail: email@example.com
Citation Wright PF. Inductive/effector mechanisms for humoral immunity at mucosal sites. Am J Reprod Immunol 2011; 65: 248–252
Problem Inductive and effector functions in the human mucosal immune system in relationship to protective immunity to HIV are poorly defined.
Method of Study A broader review of vaccine-induced immunity in mucosal systems was undertaken.
Results A series of questions are posed. The limited answers to the questions indicate our profound lack of knowledge of some of the broader issues of the induction of mucosal immunity. The questions posed include the following: Is there a common mucosal immune system? Is IgA the critical immunoglobulin class in mucosal protection? Will systemic administration of antigen stimulate mucosal immunity? Is there true sterilizing mucosal immunity? What is the duration of mucosal immunity?
Conclusion The focus on mucosal immunity in HIV of this conference highlights recent remarkable advances in our understanding of the early events in HIV pathogenesis at the mucosal surface. This review identifies areas for further research.
The vast majority of human immunodeficiency virus (HIV) infections start on mucosal sites at the epithelial border. This is also true for many other viral illnesses; indeed, some viruses restrict their growth entirely to mucosal epithelial cells, e.g. influenza. For the latter viruses, epithelial cells play an underappreciated role in cytokine expression and antigen processing that directs the mucosally initiated immune response. For other viruses, including HIV, the epithelial barrier must be crossed to reach target cells for effective replication. Therefore, there is a strong presumption that immune events at the mucosal surface will influence the acquisition, the clinical course, and the potential for transmission of many viral diseases. In discussing this broad topic in HIV biology, the author has drawn on analogies with the induction of immunity to other viral agents. Knowledge of human biology in this area is fragmentary, 1 and identification of lacuna in our understanding of human mucosal immunology may be helpful to the field. Thus, I have posed the sections of this review on the afferent and efferent arms of the mucosal system as a series of questions.
Is there a common mucosal immune system?
Substantial differences exist in the mucosally associated lymphoid systems. There are areas of the mucosal epithelium that are heavily colonized with microbes and exposed to exogenous proteins, including the nasopharynx, the lower gastrointestinal tract, and the vaginal vault. In contrast, there are other areas that are largely sterile including the lung, the duodenum, the uterus, the renal collecting system, the breast, and the male genital tract. These differences would appear to dictate major differences in how potential foreign proteins and pathogens are recognized both in terms of innate pathogen-associated molecular pattern recognition and in terms of induction of cellular and humoral immunity. The role of tolerance versus allergy is extensively studied in the gastrointestinal tract in relationship to foods but has not been widely applied to understanding responses to vaccines.2
Some mucosal sites have well-developed mucosally associated lymphoid aggregates with defined relationships to the overlying epithelium, e.g. the presence of microfold or M cells. Such sites include the Peyer’s patches in the gastrointestinal tract, the nasopharyngeal Waldeyer’s ring, and the bronchial-associated lymphoid tissue. In other sites, notably the vagino-cervical area, associated lymphoid tissue is minimal.
A basic assumption about the mucosal immune system is that the effector arm acts through IgA. In this regard, a generalization that does seem valid is that IgA and IgM are actively produced in mucosally associated lymphoid tissue and are actively transported across mucosal cells, achieving relatively high concentrations in mucosal secretions. Further, there are homing mechanisms through expression of α-4, β-7 integrins that traffic CD4 cells to mucosal-associated lymphoid tissue.3 The optimal way to stimulate such homing with systemic immunization remains to be defined.
Conversely, when IgG is detected at a mucosal site, it is likely to be there as a result of passive leakage rather than active transport across the mucosal barrier. However, there are sites in which IgG is the predominant antibody including the lung and cervicovaginal tract. Immunity at the latter site is obviously of special significance for protection from the acquisition of HIV.
Although there is some quite compelling evidence of a functional common mucosal immune system from animal models, there is less evidence of such in humans. The best example may be that of the live-attenuated adenovirus types 4 and 7 vaccines, which were widely used in the US military.4 The adenovirus vaccines were given in an enteric capsule, which degraded in the small intestine leading to viral replication in the gut that was highly effective in preventing adenovirus-associated lower respiratory disease. We still do not know whether this protection is the result of the induction of serum antibody from an effective gastrointestinal immunization or whether the mucosal delivery and replication of adenovirus stimulated generalized local mucosal immune responses. We do know from animal models that protection against influenza lower respiratory tract disease can be conferred by passively administered IgG antibody.5 We are currently trying to address some of those issues with a live adenovirus vector expressing the influenza hemagglutinin with hopes that we may demonstrate antibody that has spread beyond the enteric site of delivery to the respiratory tract and female genital tract. Preliminary evidence suggests that the enteric delivery has lead to the induction of some IgA responses at upper respiratory tract mucosal surfaces.
The key operational significance of understanding the commonality of the mucosal immune system is to define our potential ability to deliver vaccines to a distant mucosal site, e.g. enteric or nasal administration site, to stimulate mucosal immunity at sites of HIV acquisition.
Is IgA the critical mucosal antibody?
IgA is regarded as the primary mucosal antibody and has the ability, shared with IgM, to be actively transported across the epithelial barrier. However, the relative concentrations of IgA and IgG and their subclasses vary greatly by mucosal sites, Table I. As a highly relevant example for HIV, the dominate immunoglobulin in the female genital tract is IgG.6
Table I. The relative concentrations of IgA and IgG vary by mucosal site
aTotal IgG and IgA in ugm/ml.
Under what conditions does systemic immunization stimulate mucosal antibody?
Early work on HIV vaccines was marked by extensive efforts to demonstrate mucosal immunity after systemically delivered HIV protein-based vaccines. Virtually, no induction of mucosal immunity was detected. We are not certain of the optimal collection methods for mucosal samples or optimal approaches to measuring both functional and binding antibodies. There may be degradation of antibody either in vivo or after collection. It is very important for IgG in particular, but also for IgA, to inhibit proteolytic degradation of the antibody. While the inability to date to detect a mucosal response to HIV vaccines may reflect the low level of immunogenicity of the antigens or the relative insensitivity of assays, this also can be indicative that systemic delivery of protein-based antigens is not an optimal way to stimulate mucosal immunity.7
There is potent immunity against a locally invasive viral infection, human papilloma virus (HPV), provided by a systemically administered virus-like particle vaccine that appears to act primarily through neutralizing antibody, although clearance of infection may be mediated by T-cell immunity.7 The observations that HPV replicates at the basal cell layer of the endocervical epithelium, that IgG is the dominant local antibody at the site of replication and that HPV often initiates infections at abrasions or other breaks in the epithelial layer may explain the effectiveness of IgG antibody stimulated by this vaccine and suggest that it is of systemic rather than local origin.
Can there be sterilizing immunity at mucosal surfaces?
Live, viral vaccines delivered systemically obviously provide against systemic viral infections that start on the mucosal surface, such as measles, rubella, varicella, and mumps. That these vaccines inhibit mucosal replication seems certain from their blockage of transmission, but whether that immunity is sterilizing to the extent that might be protective against HIV infection is not known. It is worth noting that there is experience with the delivery of measles vaccine by an aerosol. It appears more effective than systemic immunization. Measles vaccine is most effective when delivered into the lung by a small particle aerosol as opposed to when it is delivered primarily into the upper respiratory tract by a large particle spray.8
With HIV, the Holy Grail is the induction of sterilizing immunity – a complete block of replication at the site of infection. We do not know with HIV whether true sterilizing immunity can be induced and sustained at the mucosal surface. We also know relatively little about the ways in which IgA can block the entry of a virus like HIV. Experimental models are extending well beyond virus neutralization to aggregation in mucus, blockage of transcytosis, and antibody-mediated cytotoxicity. With HIV, many of these mechanisms are becoming somewhat clearer as discussed in other presentations in this symposium.
How effective is mucosal immunization?
We have clear examples of effective mucosal immunization and their stimulation of IgA antibody. These include licensed, live, attenuated rotavirus, influenza, and polio vaccines. With polio, the comparison of mucosal IgA induction with live, attenuated oral polio vaccine and its lack after inactivated parenteral polio vaccine remains, 50 years after being published, the clearest example of the differing patterns of immunity between the two approaches.9 In this article, recipients of live Sabin vaccine had detectable serum, nasal, and duodenal IgA responses, while recipients of inactivated Salk vaccine had serum IgA but no detectable IgA at the mucosal sites. Further, there are examples with live and inactivated poliovirus and influenza vaccines that reduction in virus shedding is minimal when challenged with live vaccine after receipt of inactivated vaccines and approaches sterilizing immunity with previous live vaccines.10,11
Can local HIV-specific IgA immunity be induced by exposure or abortive infection in the female genital tract?
The presence and degree of protection in HIV highly exposed but uninfected woman remains a controversial area. Recent analyses of 41 shared samples between multiple laboratories have failed to show such responses (Fig. 1). Where low-level responses were seen, they were not replicable and almost always seen by only a single laboratory, green bar.
In contrast, in 26 HIV-infected individuals, all laboratories detected HIV-specific IgG in serum and most in cervical specimens, and dark purple bar in Fig. 2 represents all laboratories detecting antibodies. This study12 also confirmed an earlier observation6 that there is a very limited HIV-specific IgA response in infected individuals. It is unknown whether the lack of HIV-specific antibody in infected individuals is unique to HIV or may be pattern of IgA suppression in other chronic viral infections.
Can a targeted IgA response be elicited on mucosal surfaces?
We do not yet know how to safely induce an immune response that is targeted to produce IgA in humans or indeed whether this is a good idea. Experience with delivery of antigens with cholera toxin B or other enterotoxins and their dramatic stimulation of IgA responses has not been extended to man with the exception of one experience in which an adverse reaction was seen when an influenza vaccine was delivered with enterotoxin with resultant facial paralysis.13 We do not know whether inactivated protein antigens delivered to a mucosal surface are sufficient to induce good mucosal immunity with the exception of oral cholera vaccines which appear to induce effective antitoxin antibody that is protective.14
How durable is an IgA response?
We have relatively few clues to the duration of the mucosal immune response; it is generally felt to be of short duration in months to a small number of years period. Where it has been looked at it indeed does seem to decline. In several instances in which it has been examined, there appears to be an anamnestic response with a secondary rise occurring within 3 days. With rotavirus challenge, a significant rise had occurred within 3 days.15
What happens to the IgA response with chronic infection?
Remarkably, little is known about the mucosal immune response in chronic viral infection such as HIV. We have had the experience of demonstrating that in a chronically HIV-infected individual, there is a very limited IgA response as shown in Fig. 2 and in reference6. There is an early IgA response in acute infection that appears to be largely disappear as shown by the work of Tomaras.16 A productive line of research would be to understand the IgA response in other chronic viral infections such as cytomegalovirus or hepatitis B. The lack of a sustained HIV-specific IgA response in established HIV infection might represent antigen–antibody complexes, destruction of IgA producing cells, or the induction of tolerance.
This article draws analogies from induced immunity with other viruses that replicate on mucosal surfaces to determine the functionality of afferent and efferent arms of the mucosal immune system. It points out areas that are conducive to further research and helps put in perspective the aims of this symposium to more directly understand mucosal events in relationship to HIV infection and to stimulate the field in terms of scientific endeavor.