Immunity in the Gut: Mechanisms and Functions

1. Introduction

Mammalian host defense has successfully handled environmental confrontations over millions of years. To this end, numerous genes involved in innate and adaptive immunity have been subjected to evolutionary modifications, thus being shaped according to microbial pressure and other exogenous impacts. This modulation has been influenced by various ways of living such as hunting, fishing, gathering, agriculture, and animal husbandry.

The major arena for the complex interactions between genes and the exogenous impact is the gut. This is therefore the most crucial organ for communication between the environment and the body (Fig. 1.2.1). In an adult human being, the intestinal epithelium covers a surface area of perhaps 300 m² when villi, microvilli, crypts, and folds are taken into account. This barrier has generally only one cell layer and is therefore vulnerable but normally well protected by numerous chemical and physical innate defense mechanisms which cooperate intimately with a local adaptive immune system (Brandtzaeg, 2013a). The dominating component of the latter is an immunoglobulin A (IgA)-generating mucosal B-cell population which basically provides an antiinflammatory first-line defense by giving rise to secretory IgA (SIgA) antibodies performing “immune exclusion”. This term is coined for low- and high-affinity antibody functions at the mucosal surface, aiming to control both microbial colonization and penetration of noxious antigens through the epithelial barrier (Brandtzaeg, 2013b).

Figure 1.2.1 Influence of exogenous variables on gut–host communication and immunological development.
The gut represents a large contact organ that transmits biological signals from the environment to the host. These signals are essential for the development of adaptive immunity and may determine the balance between tolerance and proinflammatory responses. The interaction of exogenous factors with host genes (A, B, C, etc.) provides opportunities not only for programming but also for reprogramming of the immune system and may be influenced by epigenetic variables. The immunophenotype of the host may therefore be determined at various levels, partly through epigenetic regulation. (Source: Based on information discussed in Renz et al., 2011.)

Acute or protracted diarrhea is a sign of barrier break in the gut and an increasing variety of viral agents are suspected in the etiology (Desselberger, 2000; Glass et al., 2001). Both DNA and RNA viruses are obligate intracellular microorganisms, so those infecting via the gut must in one way or another affect the epithelial lining. However, the inconsistently observed damage of enterocytes, villous atrophy, and crypt hyperplasia should not necessarily be ascribed to cytopathic effects of the actual viruses but may reflect immunopathology resulting from an immune attack on the infected host cell or from immune-mediated inflammation. Due to ethical considerations, there are only few studies of the pathogenesis of mucosal injury caused by gastroenteritis in humans (Davidson and Barnes, 1979; Philips, 1989). Overt or obscure cellular damage could result from complement- or T-cell-mediated cytotoxicity, antibody-dependent cell-mediated cytotoxicity (ADCC), complement-induced inflammation, or activation of Th1 or Th17 lymphocytes which secrete proinflammatory cytokines. However, similar immune reactions are involved in virus elimination and mucosal healing; the clinical outcome most likely reflects the balance that is continuously evolving in the fine-tuning of the respective defense mechanisms of the pathogen and its host. An inherent biological problem in this field is that host defense against pathogenic microorganisms and immunopathology basically go hand-in-hand (Shacklett, 2010).

The main focus of this chapter is the adaptive defense mechanisms exerted by the intestinal mucosal immune system. Secretory immunity is of special interest because asymptomatic virus infections, or prevention of infections, seems to be related to a relatively high local antibody level—at least for poliovirus (Ogra and Karzon, 1969) and apparently also for rotavirus (Clarke and Desselberger, 2015). Experiments in mice and gnotobiotic pigs support this notion by showing that improved protection correlates with an enhanced specific B-cell response in the gut after oral rotavirus inoculation (Moser and Offit, 2001).

2. Neonatal adaptive mucosal immunity

The enterocytes play a vital role in the defense of the neonate, not only by forming a mechanical barrier but also by transferring breast milk-derived maternal antibodies from the gut lumen, thus providing passive systemic immunity in the newborn period. This enterocytic Ig transmission differs remarkably among species. In the ungulate (horse, cattle, sheep, pig) the whole length of the intestine is involved in a nonselective protein uptake, including all Ig isotypes in a poorly defined pinocytotic process. Since colostrum of these animals is particularly rich in IgG, this antibody class will preferentially reach the circulation of the neonate via its gut epithelium during the two first postnatal days, after which so-called “gut closure” takes place (Mackenzie, 1990). Rodents, on the other hand, express an Fc receptor specific for IgG apically on neonatal enterocytes in the proximal small intestine. This receptor, FcRn, which disappears at weaning, has been particularly well characterized on enterocytes of the neonatal rat; it is a major histocompatibility complex (MHC) class I-related molecule associated with β2-microglobulin (Simister and Mostov, 1989). Complexes of FcRn and IgG are internalized in clathrin-coated pits at the base of the microvilli; binding of the ligand takes place in the acidic luminal environment, and IgG release occurs at physiological pH at the basolateral face of the enterocyte, after which the receptor is recycled.

Contrary to the previously mentioned species, the human fetus acquires maternal IgG via the placenta (Mackenzie, 1990), and perhaps to some extent from swallowed amniotic fluid via FcRn expressed by fetal enterocytes (Israel et al., 1993). Indeed, a bidirectional transport mechanism for IgG has been demonstrated in a human intestinal epithelial cell line (Dickinson et al., 1999), but the functional significance of FcRn on enterocytes in the human newborn remains unknown. Intestinal uptake of SIgA antibodies after breast-feeding appears of little or no importance in the support of systemic immunity (Klemola et al., 1986), except perhaps in the preterm infant (Weaver et al., 1991). Although gut closure in humans normally seems to occur mainly before birth, a patent mucosal barrier function may not be established until after 2 years of age; the different variables involved in this process are poorly defined.

Although breastfeeding initially may reduce the induction of SIgA by antibody shielding of the infant’s intestinal immune system, it appears later on in infancy (up to 8 months) to boost secretory immunity (Brandtzaeg, 2013c). One possibility is that SIgA antibodies in mother’s milk guide the uptake of cognate luminal antigens via receptors for IgA on M cells of follicle-associated epithelium (FAE) covering the infant’s Peyer’s patches. Mouse experiments suggest that the antigens may further be targeted to dendritic cells (DCs) which migrate to mesenteric lymph nodes where they induce a homeostatic immune response (Mathias et al., 2014). One possibility is that SIgA interact with DCs through the specific intercellular adhesion molecule (ICAM)-3 grabbing nonintegrin receptor 1 (SIGNR1)—a mouse homolog of the human C-type lectin receptor DC-SIGN. DC-SIGNR1 interaction has been shown to induce Treg cells and antiinflammatory IL-10 (Monteiro, 2014). Therefore, the remarkable output of SIgA during feeding represents an optimally targeted passive immunization of the breast-fed infant’s gut, which might serve as a positive homeostatic feed-back loop.

3. Postnatal adaptive mucosal immunity

Most babies growing up under privileged conditions show remarkably good resistance to infections if their innate nonspecific mucosal defense mechanisms are adequately developed. This can be explained by the fact that immune protection of their mucosae is additionally provided by maternal IgG antibodies, which are distributed in interstitial tissue fluid at a concentration 50–60% of the intravascular level (Offit et al., 2002). In the first postnatal period, only occasional traces of SIgA and secretory IgM (SIgM) normally occur in intestinal juice, whereas some IgG is more often present—either as a result of epithelial FcRn-mediated translocation or, perhaps more likely, mainly passive paracellular leakage from the highly vascularized lamina propria (Persson et al., 1998), which particularly after 34 weeks of gestation contains readily detectable maternal IgG (Brandtzaeg et al., 1991). However, an optimal mucosal barrier function in the neonatal period depends on an appropriate supply of breast milk, as highlighted in relation to mucosal infections, especially in the developing countries (Brandtzaeg, 2013c). Exclusively breast-fed infants are better protected against a variety of infections, and epidemiological data suggest that the risk of dying from diarrhea is reduced 14–24 times in nursed children. In the westernized part of the world, the protective value of breast-feeding is clinically evident in relation to mucosal infections, being most apparent in preterm infants. The role of secretory antibodies for mucosal homeostasis is furthermore supported by the fact that knock-out mice lacking SIgA and SIgM show increased mucosal leakiness and overstimulation of systemic (IgG) immunity to Escherichia coli (Johansen et al., 1999) and also to other commensal bacteria (Sait et al., 2007).

After the peak of passive immunity mediated by systemic maternal IgG antibodies and SIgA from breast milk, the survival of the infant will to an increasing extent depend on its own adaptive immune responses (Offit et al., 2002). At mucosal surfaces such responses are largely expressed by local IgA antibody production (Brandtzaeg, 2013b, 2015a). In addition, hypersensitivity due to penetration of exogenous antigens into the mucosa is normally counteracted by adaptive hyporesponsiveness to innocuous agents (Fig. 1.2.2). This phenomenon is traditionally referred to as “oral tolerance” when induced via the gut; it inhibits in particular overreaction to dietary proteins and components of the commensal microbiota as long as the epithelial barrier remains fairly intact (Brandtzaeg, 2013a).

Figure 1.2.2 Two levels of intestinal antiinflammatory immune defense aiming at preserved integrity of the epithelial barrier.
(1) Secretory immunity providing immune exclusion to limit microbial colonization and penetration of harmful agents. This first line of defense is principally mediated by polymeric antibodies of the IgA (and IgM) class in cooperation with various innate defense mechanisms (not shown; for rotavirus see Chapter 2.8). Polymeric antibodies are exported by transcytosis after interaction with, and apical clevage of, the epithelial polymeric Ig receptor (pIgR), also called membrane secretory component (SC). Secretory immunity is preferentially stimulated by pathogens and other particulate antigens taken up through specialized M cells (M) located in the dome epithelium covering inductive mucosa-associated lymphoid tissue (Fig. 1.2.4). (2) Innocuous soluble antigens (eg, food proteins; magnitude of uptake after a meal is indicated) and the commensal microbiota are also stimulatory for secretory immunity, but induce additionally suppression of pro-inflammatory Th2-dependent responses (IgE antibodies), Th1-dependent delayed-type hypersensitivity (DTH), IgG antibodies, and Th17-dependent neutrophilic reactions (graded arrows indicate presumed importance of stimulatory pathways). This Th-cell balance is regulated by a complex mucosally induced phenomenon called “oral tolerance” in the gut, in which regulatory T (Treg) cells are important. Their suppressive effects can be observed both locally and in the periphery.

The cellular basis for the SIgA-mediated defense is the fact that exocrine glands and secretory mucosae contain most of the body’s activated B cells—terminally differentiated to Ig-producing plasmablasts and plasma cells (collectively referred to as PCs). Most of these mucosal PCs (70–90%) produce dimers and some larger polymers of IgA (collectively referred to as pIgA) which, along with pentameric IgM, can be transcytosed through serous-type of secretory epithelia to act as SIgA and SIgM in surface defense (Fig. 1.2.3). This function depends on the polymeric Ig receptor (pIgR)—a 110-kDa epithelial glycoprotein also known as membrane SC—and the presence of a small (∼15 kDa) joining (J) peptide in the Ig polymers (Brandtzaeg and Prydz, 1984).

Figure 1.2.3 Model for external transport of J chain-containing polymeric IgA (pIgA, mainly dimers) and pentameric IgM by the polymeric Ig receptor (pIgR), expressed basolaterally on glandular epithelial cells.
This receptor was previously called membrane secretory component (mSC). The polymeric Ig molecules are produced with incorporated J chain (IgA + J and IgM + J) by mucosal plasma cells. The resulting secretory Ig molecules (SIgA and SIgM) act in a first line of defense by performing immune exclusion of antigens at the mucus layer on the epithelial surface. Although J chain is often (70–90%) produced by mucosal IgG plasma cells (Brandtzaeg, 2015a), it does not combine with this Ig class but is degraded intracellularly as denoted by (±J) in the figure. Locally produced and serum-derived IgG is not subjected to active external transport, but can be transmitted paracellularly to the lumen as indicated. Free SC (depicted in mucus) is generated when unoccupied pIgR (top symbol) is cleaved at the apical face of the epithelial cell in the same manner as bound SC in SIgA and SIgM. Most commensal bacteria become coated with SIgA in vivo (right panel) which presumably contains them for mutualism with the host without inhibiting their growth.

IgA⁺ PCs are normally undetectable in human intestinal mucosa before 10 days of age but thereafter a rapid increase takes place, although IgM⁺ PCs usually remain predominant up to 1 month after birth (Brandtzaeg et al., 1991). Adult salivary IgA levels are reached quite late in childhood, but only a small increase of IgA⁺ PCs has been reported in intestinal mucosa after the first year (Brandtzaeg, 2013d). These observations have notably been made in industrialized countries; a faster development of the mucosal IgA system is usually seen in children from developing countries, reflecting the adaptability of local immunity according to the environmental antigenic load (Hoque et al., 2000).

4. Mucosa-associated lymphoid tissue

4.1. Induction of Gut Immunity

Intestinal lymphoid cells occur in three distinct tissue compartments: organized gut-associated lymphoid tissue (GALT), the mucosal lamina propria, and the surface epithelium (Fig. 1.2.4). GALT comprises the Peyer’s patches, the appendix and numerous isolated (solitary) lymphoid follicles, especially in the large bowel. All these lymphoid structures are believed to represent inductive sites for gastrointestinal immune responses (Brandtzaeg, 2015a). The lamina propria and the epithelial compartment constitute effector sites but are nevertheless important in terms of cellular expansion and differentiation within the mucosal immune system. GALT and other mucosa-associated lymphoid tissue (MALT) structures (see later) are covered by a characteristic FAE, containing specialized M cells (Fig. 1.2.4) which are effective in the uptake especially of live and dead particulate antigens from the gut lumen. However, many enteropathogenic infectious bacterial (eg, Salmonella spp., Vibrio cholerae) and viral [eg, poliovirus, reovirus, human immunodeficiency virus (HIV)-1] agents may use the M cells as portals of entry (Hathaway and Kraehenbuhl, 2000).

Figure 1.2.4 Antigen-sampling and B cell-switching sites for induction of intestinal IgA responses.
The classical inductive sites are constituted by gut-associated lymphoid tissue (GALT), here represented by Peyer’s patches and isolated (solitary) lymphoid follicles (ILFs). GALT structures are equipped with antigen (Ag)-sampling M cells, T-cell areas (T), B-cell follicles (B), and antigen-presenting cells (APCs). Class switching from surface sIgM to sIgA occurs in GALT and mesenteric lymph nodes; from here primed B and T cells home to the intestinal lamina propria (LP) via lymph and blood. T cells mainly end up in the epithelium (EP). Primed B cells may also migrate from ILFs directly into LP where IgG-producing cells occur with downregulated J chain (±J). The sIgA⁺ cells differentiate to plasma cells that produce dimeric IgA with J chain (IgA + J) which becomes secretory IgA (SIgA). Bone marrow-derived sIgM⁺ B2 cells (CD5^–) may also give rise to pentameric IgM (IgM + J) and secretory IgM (SIgM). B1 cells (CD5⁺) from the peritoneal cavity reach the LP by an unknown route (?), perhaps via mesenteric lymph nodes. These sIgM⁺ cells are particularly abundant in mice and may switch to sIgA within the LP under the influence of the cytokines BAFF and APRIL when dendritic cells have sampled luminal antigen within the epithelium. The sIgA⁺ cells then differentiate to plasma cells that provide SIgA mainly directed against the commensal gut flora. Red dots denote antigen.

MALT structures resemble lymph nodes with B-cell follicles, intervening T-cell areas, and a variety of antigen-presenting cells (APCs), but there are no afferent lymphatics supplying antigens (Brandtzaeg, 2015a). Therefore, exogenous stimuli must come directly from the gut lumen via the M cells, although DCs in FAE may participate in the antigen uptake. Among the T cells, the CD4⁺ helper subset predominates—the ratio between CD4 and CD8 cells being similar to that of other peripheral T-cell populations. In addition, B cells aggregate together with T cells in the M-cell pockets, which thus represent the first contact site between immune cells and luminal antigens (Yamanaka et al., 2001). The B cells may perform important antigen-presenting functions in this compartment, perhaps promoting antibody diversification and immunological memory. Other types of professional APCs, macrophages and DCs, are located below the FAE and between the follicles.

Pioneer studies, performed in animals almost 30 years ago, demonstrated that immune cells primed in GALT are functionally linked to mucosal effector sites by integrated migration or “homing” (Brandtzaeg, 2015a). T cells activated in GALT preferentially differentiate to CD4⁺ helper cells which—aided by DCs and secretion of cytokines such as transforming growth factor (TGF)-β and interleukin (IL)-10—induce the differentiation of antigen-specific B cells to predominantly IgA expression. Induction of rotavirus-specific B-cell responses, moreover, seems to be promoted by type I interferon derived from plasmacytoid DCs (Deal et al., 2013).

The activated B-cell blasts proliferate and differentiate further on their route through mucosa-draining lymph nodes and the thoracic duct into the blood stream (Fig. 1.2.4). They thereafter home preferentially to the gut mucosa and complete their terminal differentiation to IgA⁺ PCs locally—most likely under the influence of DCs which may pick up luminal antigens at the secretory effector site. The migration of lymphoid cells into the mucosal lamina propria is facilitated by “homing receptors” that interact with ligands on the microvascular endothelium at the effector site (addressins)—with an additional fine-tuned level of navigation conducted by local chemoattractant cytokines (chemokines). Under normal conditions, therefore, the lamina propria microvasculature exerts a “gatekeeper” function to allow selective extravasation of primed lymphoid cells belonging to the mucosal immune system, but this restriction is not maintained during inflammation (Table 1.2.1). Thus, it has been shown in mice that the selectivity of chemokine-dependent attraction of IgA⁺ PC to the small intestinal lamina propria becomes redundant during rotavirus infection (Feng et al., 2006). Also, the inductive part of MALT exerts an important impact as shown by the virtual lack of B-cell homing to the small intestinal lamina propria from human nasopharynx-associated lymphoid tissue (NALT) (Johansen et al., 2005), similar to the lack of homing of rotavirus-specific B cells after NALT immunization in mice (Ogier et al., 2005).

Table 1.2.1

Characteristics of the Systemic Versus the Mucosal Immune System

	Systemic immunity	Shared features	Mucosal immunity
Inductive sites
Antigen uptake and transport	Ordinary surface epithelia		Epithelia with membrane (M) cells
		Dendritic cells (DCs)
	Blood circulation
	Peripheral lymph nodes, spleen and bone marrow		Mucosa-associated lymphoid tissue (MALT): Peyer’s patches, appendix and isolated (solitary) lymphoid follicles (GALT)
		Tonsils and adenoids Local (regional) lymph nodes
Influx of circulating lymphoid cells: adhesion molecules and chemokines/chemokine receptors		Postcapillary high endothelial venules (HEVs) PNAd/L-selectin (CD62L) SLC (CCL21), ELC (CCL19)/CCR7
			GALT: MAdCAM-1/α4β7
Effector sites
Homing of memory and effector B and T cells	Peripheral (lymphoid) tissues and sites of chronic inflammation: a variety of adhesion molecules and chemokines/chemokine receptors		Mucosal lamina propria and exocrine glands: MAdCAM-1/α4β7 (gut), other adhesion molecules (? extraintestinal), TECK (CCL25)/CCR9 (small intestine), MEC (CCL28)/CCR10 (? elsewhere)
		Tonsils and adenoids
Antibody production	IgG > monomeric IgA > pIgA > pentameric IgM		pIgA > pentameric IgM > >IgG

Abbreviations: GALT, gut-associated lymphoid tissue; PNAd, peripheral lymph node addressin; SLC, secondary lymphoid-tissue chemokine; ELC, Epstein Barr virus-induced molecule 1 ligand chemokine; CCR, CC chemokine receptor, MAdCAM-1, mucosal addressin cell adhesion molecule 1; TECK, thymus-expressed chemokine; MEC, mucosae-associated epithelial chemokine; pIgA, polymeric IgA, mainly dimeric.

The immunological immaturity of the newborn period (Offit et al., 2002) causes very few IgA⁺ B cells (presumably GALT-derived) to appear in peripheral blood of newborns (<8/10⁶ lymphocytes), but after 1 month of age this number increases remarkably (∼600/10⁶ lymphocytes) in parallel with the progressive environmental stimulation of GALT (Stoll et al., 1993). Later on in childhood, B-cell trafficking between inductive and effector sites is reflected by a significant correlation between the number of IgA⁺ cells in the circulation and in the gut lamina propria, as shown for the small intestine of patients with rotavirus infection (Brown et al., 2000).

4.2. Additional Sources of Intestinal B Cells

GALT constitutes the major part of human MALT, but induction of mucosal immunity may also take place in lymphoepithelial structures of Waldeyer’s pharyngeal ring, particularly the palatine tonsils and the adenoids (Brandtzaeg, 2015b), and probably also bronchus-associated lymphoid tissue at an early age (Heier et al., 2008). Regionalization exists in the mucosal immune system, especially a dichotomy between the gut and the upper aerodigestive tract with regard to homing properties and terminal differentiation of B cells (Brandtzaeg, 2015a). This disparity may be explained by microenvironmental differences in the antigenic repertoire as well as adhesion molecules and chemokines involved in preferential local leucocyte extravasation (Table 1.2.1). It appears that primed immune cells selectively home to effector sites corresponding to the inductive sites where they initially were triggered by antigens. Such compartmentalization has to be taken into account in the development of local vaccines. However, there is interaction among the various immune compartments, and even the mucosal and systemic cell systems are not completely segregated. Thus, it is possible to obtain a detectable SIgA response also by the parenteral route of immunization, as shown with inactivated poliovirus vaccine when GALT has been previously primed by poliovirus infection (Herremans et al., 1997).

The peritoneal cavity is an additional source of intestinal lamina propria B cells (CD5⁺ B1 cells) in the mouse. The switching to IgA expression may take place in the intestinal lamina propria and give rise to so-called natural SIgA antibodies (Fig. 1.2.4), which react particularly with commensal bacteria (Macpherson et al., 2000). However, there is no evidence that B1 cells are significantly involved in intestinal IgA production in humans (Brandtzaeg, 2015a), although human secretions contain considerable levels of polyreactive SIgA antibodies recognizing both self and microbial antigens (Bouvet and Fischetti, 1999). Interestingly, however, even in mice the traditional B2 cells, but not the B1 cells, apparently contribute along with CD4⁺ T cells to the clearance of rotavirus infection by an intestinal IgA response (Kushnir et al., 2001).

5. Intestinal immune-effector compartments

5.1. Lamina Propria B Cells and Their Epithelial Cooperation

The intestinal lamina propria of adults contains approximately 10¹⁰ PCs per meter of intestine, so at least 80% of all antibody-producing cells are located in the gut. Some 80–90% are IgA⁺ PCs and a relatively large fraction consists of the IgA2 subclass (17–64%) compared with the proportion (7–25%) seen in peripheral lymphoid tissue, tonsils, and airway mucosae (Brandtzaeg, 2015a,b). A predominance of IgA2⁺ PCs is regularly seen only in the large bowel but this cannot be explained by isotype switching outside of GALT structures (Lin et al., 2014) as previously proposed (He et al., 2007). The concentration ratios of the two SIgA subclasses in various exocrine secretions are quite similar to the relative distribution of IgA1⁺ and IgA2⁺ PCs at the corresponding effector sites, attesting to the fact that pIgA of both isotypes are equally well transported externally (Brandtzaeg, 2013b). The relative increase of the IgA2 subclass in secretions compared with serum may be important for the stability of secretory antibodies because SIgA2, in contrast to SIgA1, is resistant to several IgA1-specific proteases which are produced by a variety of potentially pathogenic bacterial species (Plaut et al., 1974).

More than 90% of the mucosal IgA⁺ PCs synthesize a small polypeptide called the joining (J) chain (Brandtzaeg, 2015a). The J chain is essential for correct polymerization of pIgA (and also pentameric IgM) and for the subsequent interaction of the Ig polymers with the pIgR which is expressed basolaterally on the secretory epithelial cells (Brandtzaeg, 1985, 2013b; Brandtzaeg and Prydz, 1984). After transcytosis the pIg-receptor complexes are released into the gut lumen by apical pIgR cleavage (Fig. 1.2.3). The extracellular portion of the receptor (∼80 kDa) remains as so-called bound SC in SIgA and SIgM, thereby stabilizing the secretory antibodies. Particularly the covalent bonding between SC and one α-chain of pIgA makes SIgA the most stable antibody functioning in external secretions.

5.2. Secretory Immunity

The function of SIg antibodies is principally to perform immune exclusion by complexing with soluble antigens and by binding to the surface of microorganisms, thus preventing their adherence to epithelial cells and penetration of the surface barrier (Fig. 1.2.5, far left panel). Interestingly, during transcytosis of pIgA and pentameric IgM these antibodies may even inactivate infectious agents, for example, rotavirus, influenza virus, and HIV—inside secretory epithelial cells and carry the pathogens and their products back to the lumen, thus avoiding cytolytic damage to the epithelium (Fig. 1.2.5, next to left panel). It has also been proposed that pIgA during transcytosis may neutralize endotoxin from Gram-negative bacteria and thereby inhibit the proinflammatory gene repertoire of the epithelial cell (Fig. 1.2.5, next to right panel). Finally, as previously reviewed (Johansen and Brandtzaeg, 2004), stromal clearance of penetrating antigens has been experimentally documented for pIgA-containing immune complexes (Fig. 1.2.5, far right panel).

Figure 1.2.5 Schematic representation of four proposed mechanisms for intestinal surface defense performed by dimeric IgA (pIgA) after being produced with J chain (IgA + J) by plasma cells in the lamina propria.
Far left panel: pIgA is transcytosed in vescicles by the polymeric Ig receptor (pIgR) across secretory epithelial cells and released into the lumen as secretory IgA (SIgA) antibodies which perform immune exclusion by interaction with luminal antigens (▬). Next to left panel: During pIgR-mediated transport pIgA antibodies may interact with viral antigens within apical epithelial endosomes, thereby performing intracellular virus neutralization and removal of viral products. Next to right panel: Intracellular neutralization of endotoxins (lipopolysaccharide, LPS) from Gram-negative (Gr−) bacteria may inhibit potentially harmful activation of the proinflammatory NFκB pathway in the epithelial cell. Far right panel: Dimeric IgA antibodies interact with paracellularly penetrating antigens reaching the lamina propria and shuttle them back to the lumen by pIgR-mediated export. (Source: References to primary articles on which the cartoon is partially based, can be found in Johansen and Brandtzaeg, 2004.)

Both the agglutinating and virus-neutralizing antibody effect of pIgA is superior compared with monomeric antibodies (Brandtzaeg, 2013b), and SIgA antibodies may block microbial invasion quite efficiently. This has been particularly well documented in relation to HIV (Mazzoli et al., 1997); specific SIgA antibodies isolated from human colostrum were shown to be more efficient in this respect than comparable IgG antibodies (Hocini and Bomsel, 1999). Moreover, serum IgG antibodies mainly depend on a passive paracellular transfer to reach mucosal surfaces (Fig. 1.2.3), and the same is true for the locally produced IgG that is detectable in various secretions (Bouvet and Fischetti, 1999). Also notably, less than 5% of the human intestinal PCs normally produce IgG—a figure that in selective IgA deficiency, however, is increased to ∼25% together with ∼75% IgM⁺ PCs (Brandtzaeg, 2015a). The proportion of IgG⁺ PCs is also strikingly increased in inflamed mucosae.

The protective role of SIgA has been questioned by observations in IgA knockout (IgA^−/−) mice which remain healthy under ordinary laboratory conditions (Mbawuike et al., 1999). When challenged with influenza virus, they show similar pulmonary virus levels and mortality as control wild-type (IgA^+/+) mice. However, leakage of serum IgG antibodies through an irritated mucosal surface lining (Fig. 1.2.3), according to the classical principle of “pathotopic potentiation of local immunity” (Fazekas de St. Groth, 1951), probably plays a greater protective role in the airways than in the gut (Persson et al., 1998). In addition, the IgA^−/− mice usually show a compensatory SIgM response. Interestingly, IgA^−/− mice on exclusive parenteral nutrition have reduced IgA antiinfluenza virus titers in the upper respiratory tract and no compensatory SIgM, which most likely explains that they show impaired mucosal immunity (Renegar et al., 2001). The fact that individuals with selective IgA deficiency do not suffer significantly more than others from intestinal virus infections, may largely be ascribed to their consistently enhanced SIgM and IgG1 response in the gut (Brandtzaeg, 2015a). Altogether, a secretory antibody response appears to be essential for adequate mucosal protection. This notion is supported by the finding that systemic IgG antibody production against E. coli and other commensals are triggered in pIgR knockout mice lacking both SIgA and SIgM (Johansen et al., 1999; Sait et al., 2007).

The complexity of local immunity is emphasized by the initial finding that clearance of rotavirus after infection, as well as long-term resistance to reinfection, was not impaired in IgA^−/− mice (O’Neal et al., 2000). Careful analysis of different variables suggested that in the absence of IgA, the protective effect was primarily mediated by IgG antibodies. Notably, however, both clinical data and animal experiments have suggested that when IgA is present, rotavirus infection is primarily cleared and subsequently prevented by intestinal SIgA antibodies (O’Neal et al., 2000). Rotavirus clearance and protection therefore seems to be critically dependent on IgA (Blutt et al., 2012).

In a similar study, the role of IgA versus IgG in gut immunity to reovirus was investigated (Silvey et al., 2001). Although this virus does not normally cause disease in humans, it constitutes a useful rodent model for enteric virus infections. Both wild-type and IgA^−/− mice were competent at clearing a primary infection, but only the wild-type mice were fully protected against entry of reovirus through Peyer’s patch M cells upon subsequent infection; partial protection shown by the IgA^−/− mice was ascribed to a compensatory SIgM response (Silvey et al., 2001). IgG antibodies to the σ1 protein could protect when preincubated with the virus prior to viral challenge, but only the corresponding monoclonal pIgA antibodies protected as SIgA against virus entry when administered via the hybridoma “back-pack” model (Hutchings et al., 2004). Together, these data imply that SIgA antibodies provide the best protection of Peyer’s patches against this enteric virus. In addition to local antibodies, mucosal cytotoxic T lymphocytes are likely important to prevent systemic seeding of virus, particularly when the pathogen such as HIV persistently tends to replicate in gut mucosa (Berzofsky et al., 2001).

5.3. Complexity of Mucosal Antigen Clearance

Unlike IgG, antibodies of the IgA class do not cause complement activation by the classical pathway. External transport of pIgA-containing immune complexes has therefore been suggested as an efficient, noninflammatory antigen clearance mechanism (Fig. 1.2.5, far right panel). This possibility is supported by in vivo experiments (Robinson et al., 2001). Pentameric IgM (in contrast to hexameric IgM without J chain) also appears to have poor complement-activating properties and may therefore support the noninflammatory functions of pIgA in competition with corresponding proinflammatory IgG antibodies. Interestingly, monomeric IgA or IgG antibodies, when cross-linked to pIgA via the same antigen, may contribute to this pIgR-mediated stromal immune clearance. Conversely, IgG antibodies against infectious agents and dietary proteins can on their own increase mucosal penetrability of exogenous bystander antigens as suggested by in vivo (Lim and Rowley, 1982) and ex vitro experiments (Brandtzaeg and Tolo, 1977). The mucosal integrity is apparently damaged by lysosomal enzymes released from polymorphonuclear granulocytes which are attracted when complement-activating immune complexes are formed locally.

The proinflammatory potential of IgG is probably less important in the gut of infants who are breast-fed because milk SIgA antibodies will exert a noninflammatory blocking effect (Brandtzaeg, 2013c). Moreover, complement regulatory proteins are expressed by the gastrointestinal epithelium and may counteract immune complex-mediated damage (type III hypersensitivity) of the epithelial lining (Berstad and Brandtzaeg, 1998).

Experimental evidence further suggests that IgA may influence mucosal homeostasis through its binding to the FcαI receptor (CD89) when present on lamina propria leucocytes. In the normal state CD89 is not detectable on human intestinal macrophages (Hamre et al., 2003), but this situation may differ in the inflamed mucosa with extravasation of leucocytes. Both activating and inhibitory signals may then be induced by IgA. Thus, cross-linking of CD89 during infection with IgA-opsonized pathogens causes proinflammatory responses, whereas naturally occurring IgA (not complexed) induces inhibitory signals through CD89, thereby dampening excessive reactions (Monteiro, 2014). Thus, IgA can downregulate the secretion of the proinflammatory cytokine tumour necrosis factor (TNF)-α from activated monocytes and inhibit activation-dependent generation of reactive oxygen intermediates in neutrophils and monocytes (Wolf et al., 1994a,b). On the other hand, perhaps aggregated monomeric IgA and pIgA can trigger monocytes to secrete TNF-α and upregulate the B7 (CD68) costimulatory molecules (Geissmann et al., 2001).

Eosinophil degranulation can also be induced by aggregated IgA (Abu-Ghazaleh et al., 1989). This proinflammatory potential of serum IgA probably reflects the need for reinforcement of mucosal antigen elimination when immune exclusion fails. Moreover, IgA can induce ADCC by interaction with CD89 (Black et al., 1996), and particulary pIgA can activate complement via the mannan-binding lectin pathway (Roos et al., 2001). Therefore, it still remains an open question whether IgA plays a good or bad role in, for instance, HIV infection in humans although a protective effect of mucosally administered pIgA antibodies against HIV challenge has been documented (Zhou and Ruprecht, 2014).

6. Difficulties in evaluating the protective effect of secretory immunity

As emphasized earlier, the protective effect of SIgA and SIgM antibodies during intestinal infection is blurred by concurrent systemic immunity, both after natural infection (or enteric live vaccines) and when nonproliferating virus-like particles are mucosally applied together with an adjuvant (O’Neal et al., 1998). The effect of serum antibodies, both locally in the gut and systemically such as in chronic HIV infection (Haas et al., 2011), may be to inhibit further spread of the infectious agent by neutralization and immune elimination. In the gut lumen, however, IgG (and probably also monomeric IgA) is less stable than SIgA and may thus be of little protective value, although binding of antibody fragments to the biliary protein Fv (Fv fragment-binding protein) has been suggested to reinforce immune exclusion (Bouvet and Fischetti, 1999).

Methodological problems render evaluation of intestinal antibody induction difficult. Important variables beyond current control are the impact on microbial mucosal challenge caused by differences in mucus layers and mucins produced by goblet cells and enterocytes (Pelaseyed et al., 2014). To maintain mucosal homeostasis, the mucus coat collaborates with SIgA in keeping the microbiota away from the epithelial cells (Rogier et al., 2014a,b). Moreover, mucus contains antimicrobial peptides and can bind DC-SIGN, probably through its content of lactoferrin (Stax et al., 2015). Thus, it has been shown that the capture of HIV-1 via DC-SIGN by dendrites from mucosal DCs, and thereby the transfer of virus to mucosal CD4⁺ T lymphocytes (so-called trans-infection) may be inhibited.

Analysis of intestinal fluid or fecal extracts for secretory antibodies is further jeopardized by leakage of serum IgA antibodies, and even SIgA may be considerably degraded in such samples (Johansen et al., 1999). Thus, it is often possible to measure intestinal IgA antibodies directed against, for instance, rotavirus, whereas a corresponding SIgA activity may be undetectable—depending on whether the employed immunoenzyme assay reveals the α-chain of IgA or bound SC. Another problem is the fact that an unknown fraction of SIgA may remain adsorbed to microbes in feces after the extraction procedure.

In rodent models a significant proportion of intestinal SIgA is derived from bile and therefore does not represent mucosal immunity, which is in important contrast to the human situation where the hepatocytes do not express pIgR/SC (Brandtzaeg, 1985). In addition, it has recently been reported that microbially driven dichotomous fecal IgA levels in mice within the same facility mimic the effect of chromosomal mutations (Moon et al., 2015). Bacteria from mice with low fecal IgA levels degraded bound SC in SIgA as well as IgA itself.

Finally, it is controversial whether serum IgA antibody levels are helpful in the diagnosis of rotavirus infection (Angel et al., 2012). Some studies have suggested that SIgA appearing in plasma to a certain degree may be a correlate for protection of rotavirus infection or attenuated live vaccine (Clarke and Desselberger, 2015). However, it was shown several decades ago that conditions that increase the permeability of glandular and mucosal tissue may raise the “spill-over” or absorption of SIgA into the circulation (Brandtzaeg, 1971). The presence of SIgA in plasma can therefore not be taken as a reliable proxy of mucosal immunity. SIgA antibody levels in sublingual/submandibular secretion, on the other hand, seem to be a promising noninvasive correlate of immune induction in the gut (Aase et al., 2015).

7. Mucosal T cells and their putative protective roles

7.1. Phenotypic Heterogeneity

The protective effect of live attenuated rotavirus vaccines is apparently mediated not only by IgA antibodies but also by cytotoxic CD8⁺ T cells (Azegami et al., 2014). The human intestinal lamina propria contains, in addition to IgA⁺ PCs and scattered B lymphocytes, an abundance of T cells which resemble other peripheral T lymphocytes in that they virtually always express the T-cell receptor (TCR)-α/β and show a predominance of the CD4 over the CD8 phenotype (Brandtzaeg et al., 1998). This picture is becoming even more complex by the discovery of mucosal natural killer (NK) T cells (Zeissig and Blumberg, 2014) and innate lymphoid cells (Sonnenberg et al., 2012).

Although the CD4⁺ lamina propria T cells exhibit a memory/effector phenotype, they show little proliferative activity and low expression of CD25 (the high-affinity receptor for IL-2). Also, they are refractory to CD3-triggered activation in vitro, but can be activated through CD2–CD28 costimulation (Abreu-Martin and Targan, 1996). Rotavirus-specific CD4⁺ T cells with intestinal homing molecules have been shown in peripheral blood of healthy individuals (Parra et al., 2014). Similarly, rotavirus-specific IgA⁺ PC are detectable in duodenal mucosa of subjects with no signs of ongoing infection (Di Niro et al., 2010).

Intraepithelial lymphocytes (IELs) are strikingly dominated by the CD8 phenotype (80–90%), and normally amount to 5–10 per 100 epithelial cells (Fig. 1.2.4). B lymphocytes are virtually absent from the intestinal epithelium outside of GALT (Brandtzaeg et al., 1998). In addition, the epithelium contains a small population of non-T non-B lymphocytes of obscure origin and function. Such peculiar subsets consist of T cells expressing the alternative TCR-γ/δ, the αα homodimer of the CD8 molecule (both the TCR-α/β+ and TCR-γ/δ+ subsets contain CD8αα+ subpopulations), as well as CD4 + CD8+ “double positive” and CD4−CD8− “double negative” T cells. These subsets contribute substantially to the IEL pool in rodents, and coexpression of NK-cell markers have been reported also in humans (Cheroutre et al., 2011). It is difficult to decide when IELs are important for mucosal barrier protection or when they are involved in epithelial damage.

Consistent with the expression of CD8 and cytoplasmic granules, human and murine IELs show vivid cytotoxicity in vitro, and this can be elicited through both the CD3 complex, TCR-α/β and TCR-γ/δ⋅ Interestingly, the IELs are potent producers of IFN-γ after in vitro stimulation (Lundqvist et al., 1996), and this capacity may be important for protection of neighbouring epithelial cells against intracellular infectious agents (Chardes et al., 1994).

7.2. Possible Antimicrobial Functions

It is generally believed that IELs play a role in a first line of microbial defense but only few studies have directly demonstrated such a function (Cheroutre et al., 2011). Experiments in mice have suggested that certain infectious intracellular pathogens elicit a dynamic disseminated IEL response which may be quite persistent. Thus, enteric infection with reovirus, rotavirus and Toxoplasma gondii resulted in rapid accumulation of cytotoxic TCR-α/β+ IELs that lyzed infected intestinal epithelial cells. The effector cells belonged to the CD8αβ+ subpopulation and recognized intracellular peptides presented on classical MHC class I molecules, thus being similar to the conventional CD8αβ+ cells of systemic lymphoid organs (Helgeland and Brandtzaeg, 2000).

Transfer of germ-free animals to a specific pathogen-free environment converts the composition and functional profile of IELs to that of conventional animals, suggesting that the response of these cells is primarily directed against the commensal microbiota. However, attempts to identify the stimulatory bacteria have been inconclusive. To elucidate the nature of the antigens recognized by IELs, their TCR repertoire has been subjected to detailed investigation. Surprisingly, it has been found in humans, mice and rats that TCR-α/β+ IELs express an oligoclonal antigen receptor repertoire and the same limited number of T-cell clones can be identified over long distances in the gut (Helgeland and Brandtzaeg, 2000). This result is difficult to reconcile with the possibility that IELs recognize diverse microbial antigens. Furthermore, inbred mice and rats express individually distinct oligoclonal repertoires, indicating that TCR-α/β+ IELs recognize different antigens in different animals. Microbial colonization of previously germ-free rats induces random oligoclonal expansion of CD8α/β+ and CD8γ/δ+ IELs (Helgeland et al., 2004). This may exclude specific recognition of indigenous microorganisms, which are believed to be quite comparable but not phenotypically identical among animals of the same species. Moreover, TCR-α/β+ IELs in germ-free mice also express oligoclonal repertoires, further speaking against a role of the intestinal microflora in the clonal T-cell selection process (Regnault et al., 1996).

The possibility remains that different T-cell clones recognize different epitopes of the same microbial species, and certain features of the TCR repertoire suggest that it is shaped by the intestinal flora. Thus, the widely disseminated clones were found to vary markedly in size at different levels of the gut and some clones occurred only locally (Dogan et al., 1996). Also, it has been shown that microbial colonization of rats leads to selective expansion of TCR-α/β+ clones bearing certain variable β gene segments (Helgeland and Brandtzaeg, 2000). Finally, data from chicken suggest that the first T cells to arrive in the intestinal epithelium during ontogeny express a polyclonal repertoire (Dunon et al., 1994), which then might be subjected to oligoclonal selection by naturally occurring microbial antigens in postnatal life. Altogether, it is tempting to speculate that IELs recognize conserved microbial antigens.

The γ/δ IEL subset is associated with certain TCR Vγ and Vδ gene segments in defined anatomical sites. Thus, in the human gut, Vδ1 in expressed by some 70% of the TCR-γ/δ⁺ IELs, apparently in the main together with Vγ8 (Brandtzaeg et al., 1998). This may reflect a preferential specificity directed against selected gut antigens, and molecular analyses of Vδ1 sequences have suggested an individually imprinted oligoclonality along extensive gut segments for considerable time periods. Interestingly, Groh et al. (1998) showed that Vδ1 TCR-γ/δ⁺ IELs could recognize stress-induced nonclassical MHC class I (or class Ib) molecules on epithelial cells, mainly MICA and MICB. This observation might hint to the existence of an immune surveillance mechanism for detection of damaged, infected or transformed intestinal epithelium, and/or the possibility for stimulated secretion of immunoregulatory cytokines from the TCR-γ/δ⁺ IELs.

Interestingly, the number of intestinal TCR-γ/δ⁺ IELs is generally increased in AIDS patients, although within a wide range, probably reflecting individual responses to opportunistic infections (Nilssen et al., 1996). Conversely, the almost total depletion of mucosal CD4⁺ cells is not compensated for by a numerical increase of duodenal CD8⁺ cells. In AIDS patients with particularly short life expectancy (<7 months), γ/δ IELs are decreased to virtually normal levels. Longitudinal studies in a few patients supported this observation, suggesting that γ/δ IELs might be involved in prolonging the life of patients with AIDS.

8. Conclusions

Secretory immunity is desirable in the defense against intestinal virus infections because it can operate both at the luminal face of the epithelium and intracellularly in infected epithelial cells without causing mucosal damage. In addition, IELs may act in front-line defense by eliminating infected epithelial cells. Several virological studies have shown that natural gut infection and enteric vaccination are more efficient in giving rise to SIgA antibodies than parenteral vaccination. Also, the intraluminal vaccines have been much more efficient when consisting of live than killed viruses. Like natural infections, live vaccines give rise not only to SIgA antibodies, but also to longstanding serum IgG and IgA responses. SIgM and serum IgM antibodies often appear in the acute infection phase but decline after covalescence. In infants and subjects with selective IgA deficiency, however, SIgM may play a major protective role against infectious agents (Brandtzaeg, 2015a).

Despite advanced mechanistic understanding of the protective effects of mucosal immune responses against gut viruses, it is generally difficult to determine unequivocally the importance of SIgA versus serum antibodies. The degree of immune protection may range from complete inhibition of reinfection (poliovirus and murine reovirus) to reduction of symptoms (rotavirus). It is furthermore not clear to what extent T-cell-mediated mechanisms or ADCC are involved in this effect. This complexity is an inherent difficulty in design of vaccines to protect against mucosal infections (Azegami et al., 2014) and makes it difficult to understand how the current live oral rotavirus vaccines function in various parts of the world (Glass et al., 2014).

Acknowledgments

Supported by the Research Council of Norway through its Centers of Excellence funding scheme (Project No. 179573/40) and by Oslo University Hospital Rikshospitalet. Hege Eliassen is thanked for her excellent assistance with the manuscript.

Basicmedical Key

Fastest Basicmedical Insight Engine

Immunity in the Gut: Mechanisms and Functions

Immunity in the Gut: Mechanisms and Functions

Abstract

Keywords

1. Introduction

2. Neonatal adaptive mucosal immunity

3. Postnatal adaptive mucosal immunity

4. Mucosa-associated lymphoid tissue

4.1. Induction of Gut Immunity

4.2. Additional Sources of Intestinal B Cells

5. Intestinal immune-effector compartments

5.1. Lamina Propria B Cells and Their Epithelial Cooperation

5.2. Secretory Immunity

5.3. Complexity of Mucosal Antigen Clearance

6. Difficulties in evaluating the protective effect of secretory immunity

7. Mucosal T cells and their putative protective roles

7.1. Phenotypic Heterogeneity

7.2. Possible Antimicrobial Functions

8. Conclusions

Acknowledgments

Full access? Get Clinical Tree