Introduction

Toll-like receptors (TLRs) are transmembrane receptors which recognize various pathogen-associated molecular patterns (PAMPs) [1]. The myeloid differentiation primary response gene 88 (MyD88) is critical for TLR signaling [2], and except for TLR3, all TLRs require MyD88 for their downstream signaling [1, 3]. First identified as a myeloid differentiation gene [4], MyD88 was later on shown to be an essential adaptor protein in the signaling pathway, required for NF-κB activation in response to IL-1R1 signaling [5–7]. MyD88 adaptor protein mediates numerous biologically important signal transduction pathways in innate immunity [8]. Also, MyD88 signaling is important in the regulation of inflammation during bacterial infection and cancer progression [9].

MyD88 mutations are associated with immunodeficiencies that predispose patients to recurrent life-threatening bacterial infections, similar to what is observed in MyD88-deficient mice which are susceptible to various pathogens [10, 11]. MyD88-deficient macrophages are able to express proinflammatory cytokines, but are defective in TNF, IL-12, and NO production in response to mycobacterial stimulation [12]. They are also able to expand IFN-γ-producing antigen-specific T cells, but in a delayed fashion [13]. Infection with Mtb in these mice is lethal, as MyD88 signaling is necessary to raise an adequate innate and acquired immune response to Mtb [12–14]. The functions of MyD88 appear to go beyond its signaling pathway, as we will be describing in the following sections.

MyD88 in TB phagocytosis

The phagocytic process starts with recognition of the pathogen. This occurs through receptors present on the cell surface, such as scavenger or mannose receptors, which bind ligands present on the surface of the pathogen [15]. Upon ligation, these receptors cluster on the phagocyte surface and induce rearrangement of the actin cytoskeleton. TLR stimulation upregulates the Fc receptor, complement receptor, scavenger receptors MARCO and SR-A, and many other genes regulating phagocytosis. Upregulation of these genes is dependent on p38 activation, downstream of MyD88-TLR signaling [16]. Mtb interacts with MARCO, which tethers cell wall glycolipid, trehalose 6,6′-dimycolate (TDM/cord factor) from Mtb and activates a TLR2/CD14 signaling pathway [17]. CD36 and MARCO can interact with TLR2 and CD14, regulating NFκ-B cytokine response, and are all required for macrophage cytokine responses to mycobacterial TDM and Mtb [17] (Fig. 1a).

Fig. 1
figure 1

The distinct roles of MyD88 in Mtb infection. a PRR-mediated phagocytosis of Mtb. Scavenger receptors like MARCO interact with TLR2 and CD14. MARCO tethers the cell wall from Mtb and activates a TLR2/CD14 signaling pathway. Upregulation of scavenger receptors is in turn dependent on activation of downstream MyD88-TLR signaling. b Phagosomal presence of Mtb and/or Mtb structural components liberated in the phagosome lead to PRR activation. TLR activation by Mtb ligands leads to downstream NF-κB- signaling and IRF-mediated responses. c Expression of inflammatory and anti-inflammatory cytokines ensues. d MyD88 also regulates the expression of various negative regulators (e.g., IRAK-M, SOCS-1 and SOCS-3, etc.) in addition to IL-10 and type I IFNs e which suppress host protective inflammatory mediators. f The Myddosome may also have a functional role in stabilizing “cytosolic” receptors when recruited to the phagosome

Full size image

Many genes involved in the subsequent actin rearrangement that occurs during the phagocytosis process are dependent on MyD88 [18], so it is no surprise that phagocytosis of bacteria is impaired in the absence of MyD88 [19].

Phagocytosis, in turn, is essential for an optimal MyD88-dependent response to bacteria [20]. In addition to playing a central role in phagocytosis, MyD88 is crucial to the downstream TLR activation and signaling occurring after bacterial degradation in the phagosome of human monocytes [21] and murine macrophages [20]. Bacteria must be phagocytized and it is within the acidic environment of the phagosome that activated host enzymes and endosomal TLR receptors can recognize their bacterial ligands. Once these bacterial-derived ligands have been liberated after internalization and degradation, a TLR-dependent response can be initiated from the phagosome [20, 22, 23].

Besides being necessary for mounting an adequate innate and adaptive immune response [12, 14], MyD88 signaling in DCs and macrophages is essential to control the mycobacterial burden and prevent pathogen persistence [14, 24]. In the case of Mtb phagocytosis, MyD88 is necessary for confinement of Mtb in the phagosome [25].

PRR recognition of Mtb

Several pattern recognition receptors (PRRs) on phagocytic cells have been identified in the recognition of conserved (PAMPs) of mycobacteria [26]. Mtb can mask macrophage PRRs via cell surface-associated phthiocerol dimycocerosate (PDIM) lipids. These lipids also promote the recruitment of Mtb-permissive macrophages, while impairing reactive oxygen species (ROS)-response of microbicidal macrophages [27].

Mtb is a bacterium that mainly resides in the phagosome [28], and has evolved mechanisms to evade its destruction in phagolysosomes to successfully survive and replicates within phagocytes [29]. Suboptimal recognition by phagosomal receptors seems important in deciding the fate of this organism and active disease in the host. This could explain the association of different endosomal TLR polymorphisms associated with pulmonary TB in different populations [30–33].

Various pathogenic mycobacterial species might be able to invade the cytosol as it has been able to previously circumvent phagosomal maturation [34]. This process of phagosomal escape may occur at later stages of infection [35], and appears to rely on the region of deletion 1 (RD1)-dependent secretion of early secreted antigenic target 6 (ESAT-6) [36], otherwise reported to possess pore-forming activity [37]. Virulent strains of Mtb reported to translocate from the phagosome into the cytosol of macrophages and dendritic cells (DC) [25, 36] must inhibit innate immune signals to facilitate their translocation to the cytosol. This is accomplished by down-regulation of key genes in the TLR2-MyD88 pathway [25].

Some studies proposing that Mtb might be able to escape into the cytosol of macrophages, alluding to the inflammasome activation and cell death after infection by Mtb and other mycobacteria [38, 39], do not clarify if these events are due to phagosomal escape or a consequence of phagosomal membrane rupture which has previously been found to occur [40]. Furthermore, rupture of the phagosomal membrane by Mtb results in necrotic cell death of the infected macrophages [41].

Cytosolic receptor activation by recognition of their respective Mtb ligand may not necessarily require cytosolic escape of the bacillus. Intraphagosomal Mtb, is able to perforate the phagosomal membrane, allowing for the transfer of phagosomal contents in to the cytosol [42, 43]. NLRP3 is activated after recognition of Mtb’s ESAT-6 [44]. ESAT-6 is known to possess pore-forming activity, and in fact, NLRP3 activation and necrosis are dependent on ESAT-6-damage to the phagosomal membrane [38, 41] .

NOD2, a member of NOD-like receptors, recognizes mycobacterial N-glycolyl muramyl dipeptide (MDP) within the cytoplasm [45], and along with MyD88, is also required for optimal inflammatory response and protective immunity to Mtb.

DC-SIGN and dectin-1, both members of C-type lectin receptors, are able to selectively recognize mannose caps on lipoarabinomannan as well as b-glucan present in Mycobacteria [46, 47]. Mtb-induced DC-SIGN signaling affects TLR signaling at the transcriptional level [48]. In summary, Mtb-derived ligands are being sensed via endosomal receptors or recruitment of cytosolic receptors to the phagosome, or by cytosolic transfer due to loss of phagosomal integrity.

The Myddosome and cytosolic PRR phagosomal recruitment

Although MyD88 appears to be essential for an effective innate immune response against Mtb [14, 49], controversy exists if TLRs may or may not be essential. These discrepancies may be resolved if we consider the distinct immune functions of the TLR/MyD88 system in TB, beyond the function of MyD88 as a TLR adaptor. The Myddosome is a multiprotein complex involving MyD88 that forms after TLR activation [50, 51]. Myddosome stabilizes the weakly associated receptor–adaptor Toll-IL-1 receptor resistance (TIR) complexes [51], and could serve stabilizing NOD2 and NLRP3 when these “cytosolic” receptors are recruited to the phagosome [52, 53]. This could help explain macrophage activation observed in response to non-pathogenic mycobacteria through TLR2 and Nod2 occurring in an MyD88-dependent manner [54] or the synergistic inflammatory response by these two receptors [55].

Mtb possesses a system to prevent inflammasome [56] and IL-1β activation [57]. Although NLRP3 activation by Mtb does exist [58], NALP3 does not appear to be necessary for protection against Mtb [59]. NLRP3 activation could occur at the phagosomal level, a phenomenon described in Staphylococcus aureus infection [60]. All this suggests an interaction of MyD88 with other “cytosolic” receptors at the phagosome, where the MyD88 could exert a regulatory function on other pathways like that of the inflammasome.

MyD88 in the inflammatory response to Mtb

The importance of MyD88 in the protection against Mtb infection is underlined by the fact that MyD88 polymorphisms have been associated with pulmonary TB [61]. Early post-infection, Mtb elicits inflammation-dampening signals like NF-κB dampening molecule TNFAIP3 (A20) in macrophages which would favor its survival [62]. ESAT-6 can down regulate MyD88-dependent TLR signaling [63]. Proteins like Rv0625, which are heavily secreted by virulent strains like the Beijing strain can induce a strong TNF and MCP-1 response which occurs in an MyD88-dependent manner [64]. Mycobacterial TLR-signaling crosstalks with other intracellular antimicrobial innate pathways, like the autophagy process and functional vitamin D receptor (VDR) signaling [62]. Infection with Mtb induces differential expression (DE) of a series of genes in MyD88-independent pathways (i.e., iNOS, COX-2, IP10, MIG, RANTES, IRG-1, argininosuccinate synthetase 1, and chemokines JE and KC) [65].

Macrophages in the lung are the primary cells encountering Mtb, and along with neutrophils, they are crucial in mounting an early innate immune response. Macrophages can be activated by cytokines and microbial products [66]. MyD88 primes macrophages for IFN-γ mediated activation [65]. In fact full-scale activation of macrophages by IFN-γ can only be achieved when MyD88 is present. Mtb inhibits macrophage responses to IFN-γ through MyD8-dependent and MyD88-independent mechanisms [67]. When MyD88 k.o. mice are infected with Mtb, rather than being highly susceptible to the infection they develop granulomatous pulmonary lesions with larger neutrophil infiltration compared to wild-type (WT) mice [68]. These massive cellular infiltrations accompanied by extensive necrosis of granulomas are also described in a later study that reported high susceptibility of MyD88 k.o. mice to Mtb infection [13]. Both studies are also concordant in a higher bacterial burden in the lungs and spleens of MyD88 k.o. mice at 5 weeks post-infection [13, 68]. Lifespan of neutrophils is short, and removal of apoptotic cells by resident macrophages is key in the resolution of inflammation. Dying neutrophils also enhance the capacity of infected macrophages to control intracellular growth of Mtb [69]. NLRP3 inflammasome activation and IL-1β signaling by neutrophils, however, appear to enhance macrophage proinflammatory activation. IL-1R-mediated signaling is MyD88-dependent and is crucial in the early control of Mtb infection [70]. IL-1 and type I IFNs represent two major counter-regulatory mechanisms that control the outcome of Mtb infection [71]. IL-1 confers host resistance through the induction of eicosanoids (like PGE2) that limit excessive type I interferon (IFN) production keeping bacteria growth in check [71, 72]. Type I IFNs promote disease by inducing IL-10 and IL1 receptor antagonist [73]. The type I interferon, IFN-β, is induced after Mtb infection and can also suppress NLRP3-inflammasome activation [74]. All this evidence connects Mtb phagosomal recognition and phagocyte MyD88-regulated signaling, with subsequent neutrophil recruitment, that ultimately is associated with pulmonary disease [75].

“Losing control” in the absence of MyD88

Progressive cellular destruction in the lung is associated with massive neutrophilic infiltration suggesting that inflammation may be a key factor in progression towards active tuberculosis [76]. A down-regulation of the NF-κB response is a necessary mechanism to prevent excessive inflammation and return to homeostasis [77]. MyD88s acts as a negative signaling factor to shut down LPS-induced NF-κB activation [78]. Regulation of the inflammatory response by MyD88 might be an indirect process (e.g., through MyD88-dependent expression of long non-coding RNA lincRNA-Cox2) [79]. MyD88s-mediated inhibition of TLR signaling is selective for the NF-κB pathway and does not alter the activity of other pathways, like the one of transcription factor AP-1 [80]. Resolution of inflammation also involves apoptosis and subsequent clearance of activated inflammatory cells, in order to prevent a chronic inflammatory condition [81]. An aberrant inflammatory response ensues when MyD88 is absent, possibly due to a failure in the mechanism of regulation of inflammation. Mice lacking MyD88 are hyper-susceptible to dextran sodium sulfate (DSS)-induced colitis [82–84], and present an increased and more persistent inflammation after Borrelia burgdorferi infection [85–87]. In fact, in vivo infection of Mtb has shown higher levels of proinflammatory cytokines in lung tissue and sera of MyD88 k.o. mice [68]. Absence of MyD88 can lead to a Th10-mediated autoimmunity when mice are immunized with complete Freund’s adjuvant [88], composed of heat killed and dried avirulent Mtb strain H37Ra.

MicroRNAs (miRNAs) are important regulators of TLR signaling. miR-149 regulates MyD88 expression in murine macrophages [89]. In alveolar epithelial cells, immunomodulatory miR-124 is induced upon infection with BCG in an MyD88-dependent manner [90]. Although miR-124 and miR-147 target MyD88 and TRAF6, and are highly induced by Mtb to promote Mtb growth [91], they may also reflect an MyD88 mechanism of regulation of inflammation.

BCG, the attenuated M. bovis strain bacilli Calmette–Guerin, is the only currently used vaccine that has variable protective efficacy against Mtb infection [92]. Future vaccine models utilizing recombinant non-pathogenic M. smegmatis bearing Mtb genes (IKEPLUS strain) appear to be able to induce bactericidal immunity against challenge with virulent Mtb [93]. This response can only occur in the presence of MyD88, since infection with this recombinant M. smegmatis is lethal in MyD88-deficient mice [93]. The functions of MyD88 in Mtb infection are multiple and should be taken into account when assessing an individual’s genetic risk, severe disease predisposition, and designing strategies to control Mtb infection.