Secret Operations: Intestinal Epithelial Cells, Macrophages and MAP

Sun Tzu writes that ‘Secret operations are essen­tial in war; upon them the army relies to make its every move’ (Tzu, 1971, p. 149). The ability of MAP to infiltrate and enter host cells with­out overtly alarming the immune system may explain why it is so well adapted to its ruminant host (Fig.

Fig. 9.1. Intestinal cell wall and macrophage invasion tactics used by Mycobacterium avium subsp. paratuberculosis (MAP). MAP preferentially invades M cells by creating a fibronectin bridge and causes subsequent invasion of subepithelial macrophages. Entry into the macrophage is accomplished by ManLAM binding to CR3 and mannose receptors. MAP invasion into the lamina propria may also be gained through intestinal epithelial cells by an unknown mechanism. Dendritic cells may also transport MAP inside the lamina propria during sampling through tight junctions. MAP interaction with the dendritic cell receptor, DC-SIGN, may prime and promote a Th2 response. TLR, Toll-like receptor.

due to the lack of lysosomes and hydrolytic en­zymes present in these cells (Miller et al., 2007). Therefore, many antigenic properties of MAP would remain unaltered after passing through M cells. It is well established that fibronectin (FN) at­tachment proteins present on MAP facilitate FN binding of the bacterium, which in turn forms a FN bridge with β1 integrins located on intes­tinal epithelial cells (SigurSardottir et al., 1999; Pieters, 2001; Secott et al., 2001, 2004).

MAP primarily resides within macrophage cells; however, some studies suggest that epithe­lial cell processing prior to macrophage exposure may enhance efficiency of invasion into mac­rophages. MAP exposed to Mac-T cells, a mam­mary epithelial cell line, displayed increased invasion efficiency in subsequent infections of MDBK cells. Increased invasion efficiency by prior exposure may be due to upregulation of MAP3464, encoding an oxidoreductase that activates host Cdc42 and Rho internalization pathways (Patel et al., 2006; Alonso-Hearn et al., 2008). DNA microarray analysis of MAP 24h post-infection of Mac-T cells revealed up­regulation of 20 MAP genes related to regulato­ry, metabolic and virulence-associated functions compared with MAP grown in Middlebrook 7H9 broth cultures. From this gene set, Patel et al. (2006) hypothesized that a 35-kDa MAP pro­tein (major membrane protein, or MMP), which previously was shown to enhance invasion in epithelial cells, may be upregulated in response to prior exposure to Mac-T cells (Bannantine et al., 2003). Following up, Abdellrazeq et al. (2018) observed cytotoxic T lymphocyte (CTL) responses to a MAP ∆relA strain and found this response targeted MMP. This CTL-mediated re­sponse was recapitulated by stimulating bovine dendritic cells ex vivo with MMP alone, showing killing of intracellular MAP and signifying the importance of MMP in infection.

Following M-cell invasion into the sub- epithelial dome, MAP may encounter dendritic cells and/or macrophages. It is well established that MAP interacts with intestinal dendritic cells through its cell wall glycolipid mannosylated lipoarabinomannan (ManLAM) and the den­dritic cell receptor DC-SIGN (Jozefowski et al., 2008). MAP may use intestinal dendritic cell invasion as a strategic manoeuvre, since the primary function of intestinal dendritic cells is to sample and present commensal bacteria through tight junctions to the gut-associated lymphoid tissue. Thus, MAP would be able to overcome tight junction barriers and be directly transported to the lamina propria to interact with subepithelial macrophages. Furthermore, ManLAM-DC-SIGN interaction may prime MAP to direct a Th2 response, which would lead to immune subversion, as suggested for M. tubercu­losis (Jozefowski et al., 2008).

Although murine monocyte-derived den­dritic cells (MoDCs) are capable of phagocytizing MAP and secrete IL-10 after doing so (Basler et al., 2013), not much is known about their in­teraction. Inside dendritic cells (DCs), M. tubercu­losis vacuoles cannot access exogenous material or biosynthetic pathways and fail to replicate (Tailleux et al., 2003), which is also seen with Mycobacterium avium (Salte et al., 2011). Like macrophages however, the vacuole is arrested and does not acidify (Tailleux et al., 2003). MAP- infected DCs also fail to mature properly when cultured in media previously used for MAP/ MDM culture (Basler et al., 2013). DC's inability to mature in the presence of MAP was also ob­served in tissues (Lei et al., 2008).

As with dendritic cells, ManLAM from MAP is capable of interacting with macrophage cell surface receptors. The best-documented interaction is between MAP and the mannose receptor, which enhances macrophage phago­cytosis of MAP (Pieters, 2001; Gatfield and Pieters, 2003; Rowe and Grant, 2006; Souza et al., 2007a).

The efficiency with which MAP enters macrophage cells appears to differ with respect to MAP genotype, such that species-specific vari­ation is observed (Gollnick et al., 2007). Despite variation in invasion efficiency, MAP strains all seem to employ a similar infection protocol or modus operandi. Zhu et al. (2008) determined that expression patterns from three MAP strains of different types based on short sequence re­peats showed upregulation of 2 7, 22 and 35 genes, respectively, when isolated from infected bovine monocyte-derived macrophages (MDMs) at 48 and 120h post-infection.

Previous studies have investigated the abil­ity of MAP strains to regulate expression of MHC molecules during macrophage infection. Although MAP infection of J774 murine mac­rophages did not affect expression of MHC class II molecules, antigen presentation decreased, which may be due to MAP limitation of anti­gen processing (Kuehnel et al., 2001). However, these results conflict with those demonstrating downregulation of MHC class I and class II mol­ecules in bovine macrophages infected with live and killed MAP (Weiss etal., 2001). It is exciting to speculate on the potential that one or a num­ber of hypothetical genes identified by Zhu et al. (2008) may be responsible for MAP regulation of MHC class I and class II molecules and control of antigen processing.

While immediate activation of MAPK sig­nalling by MAP is of obvious importance in the response of macrophages to infection, studies focused on this cannot tell us what effect MAP might be having on the ability of macrophages to respond to T cells and to activate these cells to respond to infection. One of the most criti­cal components of macrophage-T-cell interac­tions is engagement of CD40 on macrophages by CD154 (CD40 ligand) on activated T cells.

Studies on CD40 signal transduction have resulted in a complex picture of different me­diators and pathways involved. Two major sig­nalling pathways are activated downstream of CD40, which both involve activation of latent transcription factors. One pathway involves ac­tivation of the inhibitor of nuclear factor kappa B kinase complex, leading to nuclear translo­cation of active nuclear factor kappa B. The second mechanism is activation of the MAPK pathway, a cascade of phosphorylation events that primarily results in post-transcriptional activation of transcription factors like cyclic adenosine monophosphate (cAMP)-response element binding protein, activating transcrip­tion factor, Ets, and AP-1 (van Kooten and Banchereau, 1996, 2000). Both pathways synergize in inflammatory gene expression, including expression of IL-12p40, iNOS, IL-6, IL-8 and TNFα. Not surprisingly, CD40-CD154 signalling is a target for many intracellular pathogens. For example, Mathur et al. (2004) demonstrated that Leishmania major, an intra­cellular parasite causing leishmaniasis in hu­mans, is able to inhibit CD154-CD40-mediated IL-12p40 and iNOS gene expression in murine peritoneal macrophages. This blockade ap­pears to involve interference with activation of two main members of the MAPK pathway, p38 and ERK1/2 (Awasthi et al., 2003; Mathur et al., 2004).

MAP-infected macrophages are defective in some aspects of CD40 signalling (Sommer et al., 2009). In uninfected macrophages, CD40-CD154 binding results in large in­creases in TNFα, IL-6, IL-10, IL-8, IL-12p40 and iNOS gene expression within 6 h. In MAP- infected macrophages, TNFα and IL6 gene expression following CD40-CD154 binding is relatively unaffected. In contrast, MAP- infected macrophages fail to activate expres­sion of IL-12p40 and iNOS gene expression. This is a critical difference, since IL-12 is a major driving force for development of an ap­propriate Th1-like response and production of IFNγ by T cells. In macrophages, iNOS activity and production of reactive nitrogen species is a major mechanism used to kill phagocytosed bacteria. For an intracellular bacterium such as MAP, limiting production of IL-12 and iNOS would ensure survival and development of an inappropriate immune response, particularly in the face of enhanced IL-10 production. It has also been suggested that failure to prop­erly activate and/or engage T cells could lead to development of regulatory T cells, which would further reduce Th1-like immune ac­tivity against MAP (de Almeida et al., 2008). Results of Sommer et al. (2009) are also con­sistent with observations in vivo (Coussens et al., 2004), where MAP-infected intestinal tissues contained elevated levels of IL-10 but not IL-12p40 or IL-12p35. These tissues also contain elevated levels of TGFβ (Coussens et al., 2004; Khalifeh and Stabel, 2004). TGFβ can be produced by regulatory Th3 cells and by regulatory γδ T cells, among others. These de­fects may be more important in early responses to MAP infection however, since despite the presence of IL-10, T regulatory cells seem to be largely absent in later stage MAP-infected tis­sues (Roussey et al., 2016).

9.4