Canine IgA and IgA deficiency: Implications for immunization against respiratory pathogens

Abstract

Immunoglobulin A (IgA) is widely recognized as the important antibody isotype involved in protective responses on mucosal surfaces, where it acts primarily by effectuating immune exclusion of foreign material. Selective IgA deficiency (SIgAD) is the most common immunodeficiency disease in dogs and humans and has consequences for mucosal immunity. This review is a comparative look at the biology of IgA and SIgAD with a focus on how this branch of immunology relates to vaccine selection and efficacy for canine infectious respiratory disease.

Introduction

Most veterinarians, regardless of their occupational stripes and level of interest in immunology would associate immunoglobulin A (IgA) with protection of mucosal surfaces. Many would summon a related homily that it is the (only) important isotype stimulated when intranasal or oral vaccines are administered. Beyond that, for most, details about IgA are somewhere back, fuzzy, in the academic past.

Dogs have been domesticated for about 15 000 y (1). Even before recognized breeds emerged a few hundred years ago, humans have been selectively breeding their best friend, the dog, to have certain traits. Initially, most of this matchmaking aimed to improve dogs’ abilities at a variety of working tasks to aid the dog owner. More recently, as the unemployment rate of dogs, overall, has increased, and most dogs are mere pets, the focus of dog breeding has largely devolved to esthetics; pleasing judges in the show ring. Together, these canine eugenic endeavors have resulted in unintended consequences, or, as geneticists would say, linkage disequilibrium (1). This refers to the association between a desired trait and some seemingly, but not random, undesired, often pathologic, trait, running the gamut from dysplastic hips to a spectrum of immune deficiencies. Selective deficiency in IgA (SIgAD) is the most common of the latter, and its expression can be insidious (2,3).

As in so many areas of immunology, the IgA story is mostly one of mice and men. But, as is often overlooked in overly extrapolative medicine, species differences matter, as in the case of IgA. The purpose of the following is to review the comparative biology of IgA and its role in the respiratory tract in the context of some continuing controversies concerning the choice and efficacy of vaccines for canine infectious respiratory disease (CIRD).

A brief history of IgA

The time between von Behring and Kitasato’s performance of their Nobel Prize-winning experiments on passive immunity to tetanus in the horse in the late 1890’s and the mid 1950’s could be viewed as the “Dark Ages” of antibody history. This was a period of benightedness regarding the nature and naming of globulins, including antibodies, in blood; progress being inhibited by the relatively low specificity of techniques available to dissect complex mixtures of proteins. Indeed, despite considerable phenomenological data documenting the antimicrobial effects of antibodies in serum, by 1956 only 2 antibodies were recognized, “7S” and “19S” (4). This designation was based on their approximate molecular weight or sedimentation coefficient, which is where the molecules end up in an ultracentrifuge tube (4). Meanwhile, in 1953, Grabar and Williams (5) working at the Pasteur Institute in Paris invented immuno-electrophoresis, which used antisera to specifically identify electrophoretically separated proteins, by virtue of the formation of precipitin lines, and, as part of this work, provided the first evidence for the existence of IgA. This ushered in a renaissance of sorts in the study of antibodies. Amongst the leaders of this renaissance were JF Heremans et al (6) at the Medical School in Louvain, Belgium. In 1959, it was this group that first definitively isolated and characterized IgA from human serum, then called “β2_A-globulin” based on its original electrophoretic migratory pattern (6). To do this, they used a modified salt precipitation strategy to deal with the relatively high carbohydrate content of IgA, and immuno-electrophoresis. Heremans then introduced the term “immunoglobulin” to include globulins with antibody activity, which his laboratory demonstrated in the case of IgA in 1963 (7,8). After a brief stint being called “γ_1A,” IgA finally became “IgA” by international agreement in 1964 (9,10). Moving beyond work in humans, Heremans’ group proceeded to identify IgA in the sera of a variety of other mammals, including dogs, in the late 1960’s, using cross-reactive antisera raised against human IgA and components thereof (11). They were, however, scooped on the first report of canine IgA by Johnson and Vaughan (12) at the medical school in Rochester, New York, who were better characterizing the dog immune system because dogs were, at the time, a primary experimental model for human transplantation. They used various sizing and electrophorectic techniques and called the putative IgA candidate “intermediate Sγ1.” This seminal work was then “confirmed” by the laboratory in Belgium primarily using antisera against human IgA (13).

Coincident with the early work on IgA in serum was the discovery of what came to be known as “secretory IgA (SIgA).” The identification of IgA in human milk by laboratories in Switzerland and Sweden in the late 1950’s was the first evidence that IgA was present in secretions (14,15). Shortly thereafter, in 1963, Tomasi et al (16) in the medical school at the University of Vermont first reported the presence of γ₁A, or IgA, initially in saliva, then in other human secretions, during the course of extensive investigations of the mucosal immune system (17). From a canine perspective, Johnson and Vaughan’s seminal work with canine serum immunoglobulins also provided the first evidence that IgA was contained in canine secretions including colostrum, saliva, and bronchial secretions (12). These observations were again confirmed and extended by Heremans’ group in Belgium (18).

The structural features of IgA

Biophysically, aside from its relatively higher carbohydrate content which made early isolation attempts from serum difficult (9), IgA is not markedly dissimilar from the other immunoglobulin isotypes, IgD, IgG, IgM, and IgE (19). Sophomorically, we all learned these molecules are glycoproteins having the shape of a Y, with the top of the Y containing the hypervariable regions (Fab), that are predestined to interact with specific epitopes, and a more conserved base or constant region containing the “Fc” fragment that interacts with cells (19). Each part of the Y, top and bottom, comprises 2 chains, and the top and bottom of the Y are joined by a flexible “hinge” region. Each side of the Y is made up of a heavy chain with a constant region at the base and a variable region that extends to the inside top of the Y. The outside of each arm of the top of the Y is a “light chain,” that similarly has a constant region and a hypervariable region. These bits are held together by disulphide bonds or “bridges.” What defines the different isotypes is the biochemistry of the “heavy” chains in the base; IgA has so-called α heavy chains.

Beyond this electron micrograph-inspired cartoon, that can be found in any immunology textbook (19,20), the plot thickens. The first complexity is the difference among species in the number of genes coding for α chains; rabbits have 13, humans have 2, and dogs only 1, each with variable numbers of alleles, or forms of the genes (9,19). In humans these 2 genes code for 2 subclasses of IgA: IgA1 found in serum and IgA2 found in secretions, a dichotomy absent in dogs. And, where the simple Y cartoon works for immunoglobulins D, G, and E which exist as single copies of the Y, or “monomers,” it does not work for IgA, or IgM, which often exist as more complicated molecules; depending on the species and where the immunoglobulins are found (9,17,19). Early in the studies of human SIgA it was noted that the IgA in secretions was bigger than that in serum, approximately 2× bigger, as it existed as an 11S molecule, versus a 7S (16). The same size disparity was noted in early examinations of IgA in canine secretions (18,21). As well, in preliminary biochemical analyses of SIgA, specifically those examining SIgA that had been chemically reduced to break the disulphide bonds, it was noted that there was a lot of other “stuff ” in the mixture than heavy and light chains; more lines or bands in separating gels (9,17). This suggested that SIgA was not the simple Y. One of these extra bands was what became known as the ”J chain” because it joined 2 monomers of IgA together using another disulphide bond to make a dimer. That was definitively identified in human SIgA in 1970 (22) and in canine SIgA in 1972 (23) and has long been recognized to be added to IgA monomers during their transit through plasma cells (24). On mucosal surfaces IgA exists as a dimer or uncommonly another polymer, regardless of the species (9,17). However, importantly, in dogs it also exists as a dimer in plasma, whereas in humans and mice it is a monomer (9,17,25).

Sites of IgA production and transport

The observation that IgA predominated at mucosal surfaces begged the questions, where does it come from, and how does it get there? The Ockham’s razor that it derives from serum was essentially discounted by simple logic in the first studies on SIgA (16). In other words, the idea that SIgA on mucosal surfaces simply results from transudation was not compatible with observation that the globulin:albumin ratio in saliva, a representative secretion, was 6 to 8 times higher than in serum, and albumin is a considerably smaller molecule. This of course does not discount the role of inflammation-induced capillary leakage, exudation, as a contributor to the presence of IgA and other immunoglobulins at sites of inflammation. But, the first clue regarding sites of IgA production came, again, from Heremans’ laboratory with the observation that the lamina propria of the human small intestine was the major site at which IgA-producing plasma cells are found (26). They then documented that the same was true in dogs (27). Around the same time, Per Brandtzaeg, a Norwegian dentist and nascent giant in mucosal immunology, together with colleagues, reported a similar predominance of IgA-producing plasma cells in the mucosa of the respiratory tract (nose) of humans (28). It took almost 40 y before similar studies were conducted in dogs (29); IgA-producing plasma cells predominate with decreasing prevalence, from the nose to the bronchioles, while in the pulmonary parenchyma IgG-producing plasma cells predominate. From a practical standpoint all of this implies that measurement of IgA in serum is mostly a direct reflection of what is happening immunologically on mucosal surfaces, at least in dogs and other species in which the dimeric form of IgA predominates in plasma. Evidence for this possibility was found in a recent study comparing intranasal and oral vaccination of dogs for Bordetella bronchiseptica (Bb; 30). Dogs that had only mucosal exposure to that antigen(s) by vaccination and aerosol challenge developed high titers of Bb-specific IgA in their sera after challenge; there had been no systemic exposure to the bacterium by parenteral vaccination. Advantageously, the ability to use serum as a surrogate for SIgA precludes the difficulty in sampling nasal cavities in dogs, and the difficult to control variation attendant to that process.

As with the dimerization, understanding IgA’s odyssey from plasma cell to luminal surface began with investigators noticing other molecules in disulfide bond-broken preparations of SIgA besides heavy and light chains of immunoglobulin and J chain (9,31). This other important glycoprotein is now known as the polyIg receptor (pIgR; 31). During the late 1960’s and 1970’s there was a sweepstakes concerning the precise mechanism of trans-epithelial transport of SIgA, with at least 6 different models championed (32). The winner was first proposed by Brandtzaeg in 1974, and was based, largely, on just looking at immuno-fluorescently-stained human nasal mucosae: pIgR is synthesized and expressed on the base of epithelial cells lining mucosal surfaces; it binds, “lock and key” dimeric IgA via the J chain; the resulting complexes are internalized, traverse the cell, and are extruded on the mucosal surface (33). Part of the pIgR, is cleaved from the complex, and part, called secretory component, remains, which protects SIgA from proteolytic digestion (31,33). While the focus of SIgA biology has been mostly gut, this means of transport, also occurs in the nose (33), and in (human) bronchioles (34), explaining the long-documented presence of SIgA in respiratory secretions in both humans and dogs (35,36). An important variation on this theme, and consistent with the immuno-histology in dogs (29) is the differential presence of SIgA and IgG in the respiratory tract; in the upper airways SIgA predominates, while in the lower, IgG rules (36). This has implications for vaccine choice and efficacy.

Immune effector functions of IgA

From a functional standpoint, early on, IgA was caricatured as “antiseptic paint” or “immunological flypaper” that worked in collaboration with mucus to entrap and remove or “exclude” microbes and other extraneous antigenic material from mucosal surfaces (9). Coinage of the term “immune exclusion” and evidence supporting it was provided by Stokes et al in 1975, originally in the respiratory tract (37). They identified a mouse IgA myeloma antibody that bound the classical hapten dinitrophenol (DNP), passively inoculated rat tracheas with serum containing the antibody, then injected the tracheas with DNP conjugated to radioactive albumin. Compared to rats that were inoculated with normal mouse serum, rats that received the DNP-reactive myeloma IgA had negligible radioactivity in their sera, indicating that the IgA specifically immunologically excluded entry of the target antigen into the blood. Since then, immune exclusion is often viewed as the cornerstone of IgA activity; it is the stuff of figures in chapters on mucosal immunity in textbooks (38). The dimeric nature of SIgA together with the wider opening of the its hinge versus in other isotypes, enhances this agglutinating activity (9). In addition, from a protective standpoint, IgA is thought to function in 2 other important ways. Based on extrapolation from extensive studies in other species, primarily humans (39), compared to IgG and IgM, both monomeric and dimeric SIgA are, at best, poor activators of complement, at least via the classical pathway by binding of Cq1 to its Fc portion. This means that it cannot effectuate punching fatal holes in viral and bacterial membranes via the commonly enlisted membrane attack complex, the final common pathway of complement activation. However, by virtue of its incredible journey through epithelial cells lining mucosal surfaces, it is thought that dimeric IgA can interact with viruses while they are replicating and neutralize them, or prevent the assembly of new virions, thereby maintaining the integrity of the epithelial barrier. The actual evidence for this in the respiratory tract is somewhat limited, but, experimentally, this modus operandi has been demonstrated in cultured dog kidney epithelial cells infected with either of 2 important representative respiratory viral agents, influenza virus and Sendai virus, and is postulated to occur in vivo (40). Relatedly, it is also thought, again based on in vitro studies using a variety of cell lines, that IgA can effectively neutralize bacterial toxins, notably LPS, by binding them intracellularly and preventing them from inducing pro-inflammatory cytokine production, thereby serving an anti-inflammatory function (41). A third commonly acknowledged function of IgA is to excrete foreign antigens that have penetrated into submucosal spaces. So, dimeric IgA can agglutinate the foreign material, forming immune complexes which bind to the pIgR at the base of an epithelial cell, then undergo “transcytosis” and extrusion onto lumen at the apex of the cell (19,31,39). IgA is therefore unique, functionally, among immunoglobulins since it can act in 3 sites around mucosae; in the lumen, in the lining cells, and in the subjacent tissue fluid; all being achieved without significant inflammation, thereby sparing the mucosae of self-inflicted tissue damage.

There has been considerable work in humans, mice, and less in some other species examining the activity of IgA in serum (39). This has focused on the interaction and inflammatory consequences of the binding of the Fc region of monomeric IgA immune complexes to a variety of effector cells via an Fc receptor (FcαR), which binds poorly if at all to SIgA dimers (39). As noted, dogs have minimal circulating monomeric IgA and, to date, a canine FcαR has not been described; any consequences of this apparent species difference are currently unknown.

Selective IgA deficiency: What is it? How common is it? How does it present?

Ironically, things are usually best appreciated in their absence, and so it is with IgA. Selective IgAD was first reported in humans in the early 1960’s (2). In early reports, some with the condition were clinically normal, others were diseased, but there were no consistent signs or symptoms (2). Twenty-odd years later, SIgAD was first reported in beagle dogs; the notable history in that colony being recurrent respiratory disease etiologically associated with B. bronchiseptica despite routine intranasal vaccination for Bb (and canine parainfluenza virus) (42).

Selective IgAD is recognized as the most common primary immunodeficiency in humans, being immunologically defined by a serum IgA concentration of < 0.07 g/L, together with normal concentrations of IgG and IgM with an estimated overall prevalence of 1 in 600 (2). Selective IgAD is also thought to be the most common canine immunodeficiency, but the situation in dogs is complicated, by being less well-documented, still lacking an agreed-upon definitive “cutoff ” value for serum IgA concentrations and exhibiting an apparent breed dependence (3). In early, mostly breed-centric, studies of canine SIgAD, a range of serum IgA concentrations was reported, and a range of cut-off values for IgAD was suggested (3). In the most recent and extensive study of canine SIgAD comprising 1267 dogs representing 22 breeds, a range of 0.01 to 3.0 g/L serum IgA concentrations was reported, but the data were not normally distributed (3). The mean serum IgA concentration in the test population was 0.27 g/L, whereas the more relevant, given the distribution, median was 0.18 g/L. The authors suggested a cutoff value of 0.15 g/L, much higher than the condition-defining one in humans. What this cut-off (value) confusion results in is the potential for an underestimation of SIgAD across breeds in the case the cutoff is, in fact, significantly higher than the human value, and an overestimation of breed susceptibility due to the small data set within breeds; a conundrum that can only be addressed with more data, including a more equal representation among breeds in any further study populations. Although data are somewhat conflicting, at least there is apparently no gender, or castrated or not, difference in canine IgA concentrations, avoiding further confounding variables Acknowledging these limitations, the most recent and extensive study identified 10 breeds, including, notably, the commonly owned golden and Labrador retrievers, in which > 10% of tested individuals had low serum IgA concentrations when applying the human cutoff value of 0.07 g/L (3). Consistent with previous work, the Shar-Pei stood out as the breed at highest risk for SIgAD, having the lowest median IgA concentration (0.08 g/L) and 45% individuals deficient in IgA. But, unfortunately, this most recent study muddies the waters with regard to German shepherd dogs, which had previously been identified in several studies to be a high-risk breed for SIgAD. In this study they were not amongst the dogs with the lowest IgA concentrations. However, consistent with previous work, low IgA concentrations were associated with canine atopic dermatitis (CAD). Further complicating the matter, and also relating to the cutoff value, is the likelihood that intra-breed differences, essentially those based on pedigree, are most apropos to the condition, rather than indicting an entire breed as IgA deficient which has been the tendency; it makes things (overly) simple that way.

In humans and dogs, symptomatic patients with SIgAD suffer most prominently from recurrent respiratory and gastrointestinal infection, which is not surprising given the central role of IgA in protecting mucosal surfaces (2,43). Clinical correlates of SIgAD in some human and canine patients also include various forms of autoimmunity and allergy, in dogs, notably CAD, as first recounted in the seminal description of SIgAD in beagles (2,3,43). Why exactly the latter diseases usually associated with an overactive immune system are associated with decreased IgA remains puzzling. Many are searching for answers to this in the microbiome, also known as the “normal flora.” In a nutshell, not surprisingly, recent application of “deep sequencing,” in other words sequencing essentially every fragment of genetic material in a sample, in this case the metagenomes of mucosal surfaces, has revealed microbial communities considerably more diverse than detected previously on culture plates. What this means for the immune system is a lot to respond to; IgA is central in this, but various other effectors such as regulatory T-cells and innate responses are also involved. But it’s a mutual relationship; as the microbiome shapes the immune response, the immune response shapes the microbiome maintaining a homeostasis that is perceived as “health.” Microbiomes have co-evolved with hosts over millennia. Perturbations to microbiomes resulting from environmental changes, changes in diets, use of antibiotics, to name a few, can have profound effects on IgA responses, including contributing to SIgAD, and other maladies associated with immune dysregulation (2,44).

Hopping on Ockham’s razor again, one would probably predict that SIgAD results from mutations in the genes encoding for the IgA molecule, specifically the α chains. Simple, yet profound, evidence that this is not the case was provided almost 40 y ago by looking under the microscope at lymphocytes from SIgAD human patients that had been immunofluorescently stained for IgA; it was there (45). This allowed the investigators to conclude that the genes for IgA were intact and that the molecular mechanisms involved in the recombination and class-switching necessary to get from IgM to IgA apparently take place in SIgAD-effected individuals (45). These data also suggested that the defect in SIgAD related to the overall production of IgA and/or its secretion. Mice, the still commonly used model for human diseases, do not naturally have a SIgAD that is clinically or immunologically similar to that found in humans, or dogs, and, in fact, the biology of IgA in mice differs significantly from that of dogs and humans (2,39). Nevertheless, several different gene “knockout” (GMO) mice have been engineered, yielding excruciatingly detailed molecular data, shedding light on the genes and molecules involved in the synthesis of IgA (2). However, these GMO mice fall well short of providing exact phenocopies of the human or canine malady, and at best provide only a piecemeal mechanistic explanation of human or canine SIgAD. Beyond that, these resource and intellect-intensive efforts underscore the caveat that a multifactorial disease is unlikely ever to be explained or mimicked by knocking out 1 or a few genes. To wit, 35 genes, the putative products of which have a variety of functions, have been associated with SIgAD in humans (46). A recent study employing whole genomic sequencing of DNA from individuals from 4 “high-risk IgAD breeds,” including Shar Pei, German shepherd, and Labrador and golden retrievers revealed 35 loci (meaning at least different alleles of the same gene) that could be associated with concentration of IgA (47). Loci in 4 genomic regions were significantly associated with IgA levels in the dogs, and the authors concluded that the candidate gene products were involved with early B-cell development and proliferation, while other nominally associated regions are thought to be responsible for inflammation, suggesting that, in the dog, the etiology of SIgAD is not simply restricted to the antibody response, but also the inflammatory response, including the innate immune response (47); again consistent, overall, with the current thoughts on the molecular basis of SIgAD in humans (2). In interesting contrast with humans, none of the implicated canine loci were within the Major Histocompatibility Complex (MHC) region, which has many associations with immune responses and has been repeatedly associated with SIgAD in humans (2,47).

Given the apparent multi-genic basis of SIgAD it is predictable that the heritability of the malady may not be particularly high; in humans approximately 20% of cases are thought to be inherited (2). The heritability of SIgAD in dogs is currently unknown, although it is probably considerably higher, given their genetic makeup. Mostly by virtue of iatrogenic genetic bottlenecks resulting from selective breeding (relative inbreeding) dogs have long haplotype blocks and long linkage disequilibrium (1). What this means is that there are long segments of chromosomes in which there is relatively little if any recombination or variation, reading from single nucleotide polymorphisms (SNPs). Relatedly there is a tendency for nonrandom association of, often unrelated, traits, for example, a desired coat color and an undesired abnormality such as SIgAD (1,3). This decreased heterogeneity and its consequences contrast with humans in whom inbreeding is generally frowned upon, royal families aside. Interestingly, genetic bottlenecks in nature resulting from human activities such as predation and habitat destruction have resulted in a higher prevalence of SIgAD in Swedish versus Canadian wolves; the latter apparently benefiting from a larger gene pool, but still having the genes (for SIgAD) long ago transmitted to their domesticated descendants (48).

Recent investigations in mice and men of the interaction between the enteric microbiome and IgA suggest a link between the macro- and micro-environments and IgA function and dysfunction (2,44). This together with the multi-genic basis and wide range of clinical severity attendant to a diagnosis of SIgAD, has led to the suggestion that SIgAD is actually a collection of different diseases with a convergent immunological phenotype, reduced or absent IgA (2); a wholism that rings true.

IgA and SIgAD in the choice and efficacy of vaccines for respiratory pathogens

Intranasal vaccination for respiratory pathogens has a 40-year history in canine medicine; in fact, canine medicine has led all others in its application, except perhaps, now with the recent upsurge in interest in intranasal vaccination in cattle. The simple justification being, for a local infection one wants a local response, that induced by intranasal, and now oral, delivery. In that logic, IgA is considered of paramount, or exclusive, importance. However, a consideration of the biology of IgA and SIgAD calls that exclusivity of importance into question; at best it is perhaps an over-simplification of a more complex reality.

As discussed, effector functions of IgA are sub-lethal (to pathogens) and diminished in the lower respiratory tract. Human neonates do not produce IgA at all, and IgA is the last immunoglobulin isotype to mature, taking 10 to 15 y before reaching adult levels (2). Available data in dogs mirror these phenomena (3); puppies have low IgA and fluctuating isotypes in nasal secretions early in life (49), and there is a highly significant age dependency with serum IgA concentrations positively correlated with age, being thought to stabilize around 1 y of age (3). In humans, a distinctive feature of SIgAD, compared to other immunodeficiencies, is that over 50% of patients who meet the diagnostic criteria do not have clinical symptoms (2); apparently other innate and adaptive effector mechanisms take up the immunological slack in protecting mucosal surfaces (2). The same scenario has been proposed in dogs (3); an insidious likelihood that a relative IgA deficiency is underappreciated in dogs. How variable “penetrance” and subclinical degrees of SIgAD may contribute, overall, to mucosal responses to natural and vaccinal antigen exposure is an open question. All of these liabilities should summon insecurity about depending exclusively on IgA in a disease-sparing response, mucosal or otherwise, to vaccination.

There are few published data concerning the kinetics or half-life of IgA responses in dogs. In one paper that lionizes the importance of IgA (versus IgG), puppies vaccinated intranasally against Bb did not have a nasal IgA response that was different than placebo until 28 days after vaccination. The response peaked at day 42 and began rapidly declining thereafter, being less than half peak by day 56 (50). Concerning different approaches to stimulating antigen-specific IgA responses in the canine respiratory tract, it has been nearly a decade since the introduction of oral vaccines for Bb, which are now probably the most commonly used vaccines for this pathogen in small animal practices, mostly because of their relative ease of administration. From a tacit, if uncommonly stated theoretical standpoint, presumably this relies on the concept of a “common mucosal immune system,” with IgA as its chief enforcer (51). This concept was nascent in the early 1960’s with documentation of the universal presence of secretory IgA on the mucosae of all organ systems examined, from stem to stern. Experimental evidence for this putatively unifying immunological principle emerged in the late 1970’s, based on the selective trafficking of murine IgA (and IgG) expressing lymphocytes derived from various tissues; cells induced at mucosal sites eventually came home again (51). Since then it has become apparent that this concept was an oversimplification; immune responses on mucosal surfaces are compartmentalized, GI versus respiratory, and induction sites matter (52). Indeed, in that seminal paper the investigators documented that lymphocytes from the bronchial lymph nodes localized in the lungs rather than the intestine, although this was not reflected in the title (51). Now, there is an expanding database clarifying this important variation on the overall theme of mucosal immunity involving IgA (52). From the practical standpoint of vaccination, many studies of prototype bacterial and viral vaccines in laboratory animals and humans demonstrate that oral vaccination is not the route of choice for inducing immune responses in the respiratory tract; intranasal delivery is clearly better, and better at stimulating adjunct T-cell and systemic antibody responses as well (52). A recent comparative study of intranasal and oral vaccines for Bb (30) is consistent with these current concepts and supportive data; another in-house study conducted by a manufacturer of oral vaccine is not (53).

Conclusions

If the immune response to CIRD were a movie, most veterinarians would give the Oscar to IgA as best actor in a leading role, not undeservedly. However, there is a big supporting cast comprising players from innate and adaptive immunity; the show can go on without IgA. The attributes of IgA dictate that immunity in CIRD is not a one-effector performance. Although IgA is good at housekeeping, principally by immune exclusion, and maintains homeostasis, it is incapable of performing a complement-mediated coup de grâce on pathogens; IgG, and to a lesser extent IgM, play that role. IgA is speechless when it comes to eliminating virus-infected cells; T-cells steal the show there. As well, IgA is not center stage in every scene; it is a bit player in the lower respiratory tract. And, behind the scenes there are pleotropic gene products interacting with the resident microbiome that produce and direct IgA in individual dogs. How various canine “specialty” diets and lifestyle co-factors, such as whether the dog gets to be a dog and interact extensively with other dogs and the natural world, or whether it lives in isolation on a couch, can affect the physiology of IgA, and, by extension, how this may affect the response to vaccines, are open questions. All this to say, that a vaccine protocol in which the spotlight doesn’t shift from IgA, is unlikely to be the critic’s choice, at least for a hip immunologist. Indeed, the final soliloquy in the 40-year old seminal paper on the “common mucosal immune system” (51) and much data since then (52), proclaim some combination of mucosal and systemic exposure to vaccines as optimal for establishing lasting protective responses to respiratory pathogens, including those with a starring role for IgA. CVJ