The Role of Non-Phagocytic Cells in Mycobacterial Infections

Tuberculosis is one of the infectious diseases with great impact in the world. It affects mainly to young adults in the productive stage of the life and most of the deaths due to this disease occur in developing countries. All the cities are affected, but 85% of the cases occur in Africa (30%) and Asia (55%). In 2009, the WHO reported 9.4 million new cases, 14 million prevalent cases and 1.7 million deaths; in addition, there is an estimate that one third part of the world-wide population lodges the latent infection. These data place tuberculosis like the third cause of death at world-wide level, after HIV/AIDS and cardiac diseases (WHO, 2010).


Introduction
Tuberculosis is one of the infectious diseases with great impact in the world.It affects mainly to young adults in the productive stage of the life and most of the deaths due to this disease occur in developing countries.All the cities are affected, but 85% of the cases occur in Africa (30%) and Asia (55%).In 2009, the WHO reported 9.4 million new cases, 14 million prevalent cases and 1.7 million deaths; in addition, there is an estimate that one third part of the world-wide population lodges the latent infection.These data place tuberculosis like the third cause of death at world-wide level, after HIV/AIDS and cardiac diseases (WHO, 2010).
Mycobacterium tuberculosis, the causal agent of tuberculosis, is a slow growth, acid-fast resistant bacillus, has a complex cellular wall constituted by complex lipids and carbohydrates.It is a facultative intracellular pathogen able to infect and to survive within the hosts cells.M. tuberculosis has developed numerous strategies to evade the immune response, so it is considered one of the most successful pathogens.
M. tuberculosis reach the human body through small drops expel by individuals with active disease.Once within the respiratory tract M. tuberculosis reaches the alveolar space and bounds to specific receptors of alveolar macrophages, dendritic cells and monocytes.For years the paradigm has been that alveolar macrophages are the main cells responsible of endocytosis, recognition and handling of M. tuberculosis.However, actual evidence demonstrates that non-phagocytic cells interact, harbour and respond to M. tuberculosis.In this work we will present and discuss some of the most relevant information on the interaction of non-phagocytic cells-M.tuberculosis, but first a brief revision of the main receptors described on macrophages and their role in the recognition of M. tuberculosis will be presented.

Mycobacterium tuberculosis interaction with macrophages
The lung is a highly susceptible organ for the invasion and establishment of microorganisms that are transmitted by aerosols.To guarantee control and elimination of those microorganisms that skip physical and chemical barriers of respiratory system, the lung requires activation coordinated and regulated by cells from the immune system located at Dectin-1 a link between the innate and the adaptive immune response (Brown et al., 2003;Taylor et al., 2007).Most of the studies have been focused to establish the role of Dectin-1 on fungi infections; however, the importance of this receptor in bacterial infections is less known.Recently, some investigators have suggested that Dectin-1 has an important role in the proinflammatory response against mycobacterial infections (Rothfuchs et al., 2007;Yadav & Schorey, 2006).It has been reported that TNF-production by macrophages infected with virulent strains of M. avium and M. tuberculosis is Dectin-1 dependent, the opposite occurred in macrophage infections with avirulent or attenuated mycobacterial strains like M. smegmatis, M. phlei or M. bovis BCG; the minimal proinflammatory response elicited by macrophages infected with virulent mycobacteria has been attributed to Dectin-1 engagement (Yadav & Schorey, 2006).It has been demonstrated that during fungal pathogens recognition by macrophages, TLR2 and Dectin-1 have a synergic effect (Rothfuchs et al., 2007;Shin et al., 2008;Yadav & Schorey, 2006), a similar situation has not been reported for mycobacterial infections.On the contrary, a recent study showed that M. tuberculosis recognition by Dectin-1 triggers a Th1 and Th17 immune response against mycobacteria and is independent of TLR2 recognition (van de Veerdonk et al., 2010).The mycobacterial component recognized by Dectin-1 is unknown, the presence of -glucans in the mycobacteria has not been reported yet, some hypotheses suggest that -glucans presents in M. tuberculosis and BCG could be the molecules recognized by this receptor (Dinadayala et al., 2004;Lemassu & Daffe, 1994).
Other C-type lectin receptor involved in mycobacterial PAMPs recognition, include Mincle (macrophage inducible C-type lectin), which is expressed by macrophages expose to inflammatory stimuli like LPS, IL-6, etc. (Matsumoto et al., 1999).Ishikawa and collaborators demonstrated that Mincle bounds the mycobacterial glycolipid Trehalose-6-6´-dimycolate (TDM) also known as cord factor (Ishikawa et al., 2009), which is one of the most studied mycobacterial constituent with immunostimulatory activity.
Complement receptor 3 (CR3), also known as CD11b/CD18 or Mac-1, is a cell membrane heterodimeric receptor, belonging to the integrin superfamily, and in spite of this presents a lectin domain that interacts with M. tuberculosis components (Schäfer et al., 2009).Neutrophils, monocytes, natural killer cells and alveolar macrophages express this receptor, although it has been reported opsonic bound of M. tuberculosis to alveolar macrophages is inefficient because these cells express low amounts of CR3 (Stokes et al., 1998).It has been described that the CR3 is the main complement receptor in the opsonic mycobacteria phagocytosis, 80% of the opsonized mycobacteria is uptake trough this receptor.Interestingly, CR3 can bind a wide variety of ligands and since it presents multiple bounding sites, CR3 may promote opsonic and non-opsonic M. tuberculosis endocytosis (Velasco-Velazquez et al., 2003).Most of the evidence indicates that CR3 is responsible of mycobacterial bounding and internalization into the macrophage, however, lack of CR3 expression does not have a significant effect on the production of reactive oxygen species or nitric oxide by macrophages (Rooyakkers & Stokes, 2005;Velasco-Velazquez et al., 2003).

Scavenger receptors
The cell membrane glycoprotein CD36 (scavenger receptor class B) is found in macrophages, dendritic cells, endothelial cells and other cell types (Murphy et al., 2005).CD36 has been implied in several cellular functions like fatty acid transport, angiogenesis regulation, inflammation and as PRR participates in the innate immune response against several pathogens including mycobacteria (Febbraio et al., 2001).The role of CD36 in protection against mycobacteria is controversial, recently Hawkes and collaborators (2010) demonstrated that CD36-deficiency confers resistance to mycobacterial infection; macrophages from CD36-/-mice allowed a reduction in the mycobacteria intracellular survival, but the mechanisms responsible of mycobacterial growth containment were not clear.They suggest that the better response of the knock-out mice could be due to the impairment of mycobacterial immune evasion strategies that take advantage of CD36.Detailed structural studies of the mycobacterial cell wall lipomannans have demonstrated that diacylated lipomannans inhibit LPS-induced inflammation by murine macrophages (Doz et al., 2007).Intriguingly, CD36 is a sensor of diacylglycerides (Hoebe et al., 2005) and may be the host receptor through which diacylated mycobacterial lipomannans suppress macrophage function.Another possible explanation is that through its association with the TLR2/6 heterodimer, CD36 may participate in immune-suppressor responses dependent of TLR2 as was reported by Noss and collaborators (2001).The family of scavenger receptors also includes macrophage receptor with collagen structure (MARCO) and scavenger receptors class A type I and II (SR-A).They are expressed in alveolar macrophages and dendritic cells (Areschoug & Gordon, 2009;Pearson, 1996).MARCO and SR-A bound acetylated-low density lipoproteins and bacteria; they promote inhaled particles and bacteria uptake (Arredouani et al., 2005;Palecanda et al., 1999).A recent report describe that macrophages expressing MARCO and bounding TDM respond producing proinflammatory cytokines, but this response required the concomitant participation of TLR2 and CD14 (Bowdish et al., 2009).With respect to SR-A, this receptor can also bound TDM, nevertheless, this recognition diminishes TNF-/MIP-1-production by activated macrophages, suggesting that SR-A have an important role in the suppression of the excessive inflammatory response presented during mycobacterial infection (Ozeki et al., 2006).Recent evidence suggest that scavenger receptors along with C-type lectins, cooperate to maintain controlled the immune response established in vivo against the mycobacterial infection (Court et al., 2010).

Toll like-receptors and NOD2
Toll like-receptors (TLRs) are not directly implied in microbial uptake; however, bacterial recognition by TLRs triggers an innate and adaptive immune response through the activation of monocytes, macrophages and dendritic cells.TLRs engagement by M. tuberculosis or its components is one of the first events in the M. tuberculosis-host cell interactions, resulting in activation of signaling cascades that culminates in production of proinflammatory responses.TLRs are expressed on immune cells like macrophages, dendritic cells, B lymphocytes, some types of T cells and some non-immune cells also express TLRs, like endothelial, fibroblasts and epithelial cells (Akira et al., 2006).The TLRs involved in M. tuberculosis recognition are TLR2, TLR4, TLR9 and TLR1/TLR6 that forms a heterodimer with TLR2 (Jo, 2008;Jo et al., 2007;Ryffel et al., 2005).Several studies have reported a number of mycobacterial lipids and proteins implied in TLR recognition and signaling, Jo and colleagues in 2007 presented an excellent review of these findings.Most of the studies oriented to explain the role of TLRs in mycobacterial infection, have been made in TLR knock-out mice, in this way, the protective role of TLR2 and in lesser extend of TLR4 was demonstrated, especially in the acute phase of the infection (Tjärnlund et al., 2006).In addition, the finding that macrophages from mice TLR2-/-are not able to control mycobacterial growth, correlates with a reduction in TNF-production by this cells (Tjärnlund et al., 2006).TNF-is clue cytokine to contain mycobacterial infection (Roach et al., 2002); TLR2 engagement triggers an excellent TNF-production, making this receptor an important element against mycobacterial infection.The role of TLR4 in the protection against M. tuberculosis infection is not absolutely accepted, some evidences demonstrate that TLR4 plays a protective role in the defense against pulmonary tuberculosis, since mice expressing a non-functional TLR4 had higher mortality and higher bacillary loads in the lungs (Branger et al., 2004).However, other reports demonstrated that mice with deficient TLR4 receptors are equally susceptible to M. tuberculosis or M. avium infections that the wild type mice (Feng et al., 2003;Reiling et al., 2002).Jo and coworkers suggested that these differences may depend of the signaling route involved, whereas the molecules that engage TLR2 will involve only MyD88 (myeloid differentiation primary response protein 88), those molecules that engage TLR4 may generate signals MyD88 dependent and independent (Jo et al., 2007).Another TLR with an important role in the antimycobacterial response is TLR9; mice TLR9-deficient were more susceptible to M. tuberculosis and mice with a combined TLR2 and TLR9 deficiency, were much more susceptibility to the infection (Bafica et al., 2005), pointing towards the need of the combined presence of these receptors for a better protection against tuberculosis.Nevertheless, other studies demonstrated that none of these TLRs (2, 4 and 9) are required to induce an adaptive immune response cellular, but the single MyD88 deficiency favored the unrestricted growth of M. tuberculosis, indicating the crucial role of this molecule for the generation of efficient effectors mechanisms by the macrophage (Hölscher et al., 2008).The role of other TLRs like TLR1 or TLR6 is not clear yet, it has been reported that mice TLR6-deficient are resistant to M. tuberculosis infection at high doses (Sugawara et al., 2003) and some clinical studies looking for associations between the mycobacterial infection and polymorphisms of a single nucleotide (SNPs) in the TLRs genes demonstrated that TLR1 SNP 1602S regulates the innate immune response towards the triacyl-lipopeptide and towards mycobacterium extracts (Hawn et al., 2007).
TLRs and C-type lectin receptors are molecules present on the plasma or endosomal membranes that recognize microbial components, whereas NOD (nucleotide binding oligomerization domain) receptors are expressed in the cytoplasm.Several evidences indicate that NODs family has an important role in the recognition of intracellular bacteria like M. tuberculosis (Takeda & Akira, 2005).NOD2 is a member of the Nod like receptors (NLRs) that recognizes muramyl-dipeptide (MDP), a key component of the peptidoglycan of Gram negative and Gram positive bacteria; on the contrary NOD1 recognizes only peptidoglycan from Gram negative bacteria.The intracellular location of NOD2 and M. tuberculosis, and the high content of peptidoglycan in mycobacterium cell wall allowed to suggest NOD2 as candidate for mycobacterium recognition.A recent study reported that NOD2 deficient mice (-/-) had high bacterial load in their lungs after M. tuberculosis infection, and a deficient cytokine production, suggesting that NOD2 participates in the resistance against mycobacterial infection, favoring the innate and adaptive immunity (Divangahi et al., 2008).Ferwerda and coworkers reported that NOD2 along with TLR2 represent a non-redundant system for M. tuberculosis recognition, and demonstrated that mycobacterial ligands for TLR2 and NOD2 synergize for proinflammatory cytokine production (Ferwerda et al., 2005).
Numerous are the receptors expressed by the macrophage and numerous are the receptors described for M. tuberculosis recognition (Table 1), nevertheless is obvious that M. tuberculosis recognition and internalization will not depend solely of a single receptor-ligand interaction.Additionally, mycobacterium surface displays many molecules, which may bound diverse receptors and hence activate different signaling routes.For these reasons during mycobacteria-macrophage interaction the possibility of multiple receptor-ligands interaction is high, so the first question arose, which of the multiple interactions will be the predominant?Then, which of the multiple signaling pathways activated in the cell will be executed?Presented in this way it looks like a chaotic scenario, so it is probable that the ultimate response, will be the final result of the multiple individual responses, for instance if during the interaction, many TLRs are engaged, a robust intracellular signaling response leading towards an inflammatory response will be building up, but if at the same time some lectin C-type receptors like mannose-receptors are engaged, favoring an anti-inflammatory response, then the final response will be a weaker inflammatory response.On the other side, the mycobacteria will have an active role for unbalance the host-response for their benefit, for instance, virulent mycobacteria over-express and expose for their easy recognition, those ligands with anti-inflammatory activity like ManLAM (recognized mainly by MR), whereas less virulent mycobacteria expose phosphoinositide-capped lipoarabinomannans that are considered as proinflammatory molecules.In summary, the studies briefly described here, suggest that the final outcome of the intracellular mycobacteria (survival or elimination) into the phagocytic cell, will be depending in much by the first encounter with the host cell, the mycobacterial components recognized in the encounter and the type and amount of receptors involved in the recognition.

Mycobacterium tuberculosis interaction with non-phagocytic cells
As it was stated earlier, interaction between intracellular pathogens and their host cells is a highly complex process, conformed by mechanisms that during bacterial and host cell evolution have guaranteed their nowadays survival.For the host cell, diverse strategies are displayed to contain the infection.Most of the strategies elicited by phagocytic cells are known, but bacterial containment does not rely only on phagocytic cells, other cell types participate also, these cells in conjunction are known as non-phagocytic cells.From the host cells side, phagocytic or non-phagocytic cells employ diverse mechanisms for bacterial containment, beginning with bacterial uptake up to the establishment of an adequate cellular activation state, which includes the display of effective antimicrobial mechanisms, apoptosis induction, autophagy or bacterial Ag presentation for recognition and death induction by activated cytotoxic lymphocytes, among others.From the pathogen perspective, these have developed numerous strategies to overcome the defense mechanisms elicited by the host cells.Actual evidence of the ability of intracellular pathogens to induce their entrance into non-phagocytic cells, have supported the hypothesis that these cells may be reservoirs of intracellular pathogens favoring their survival and persistence.

Epithelial cells infection by M. tuberculosis
Among the cells that comprise the pulmonary epithelium, alveolar epithelial cells type II are the non-phagocytic cell most studied, their participation in the immune response is acknowledged by their capacity to produce surfactant proteins A, C and D and a great variety of cytokines and chemokines in response to certain stimulus (Lin et al., 1998).By their location and distribution, these cells are exposed to all type of pathogenic agents that can reach the alveolar space, being then susceptible to the infection.In fact, the possibility that a microorganism finds an epithelial cell is much greater than finding an alveolar macrophage, since comparatively the number of pneumocytes type II surpasses up to 30 times the number of macrophages (Bermudez et al., 2002).In addition, pneumocytes type II may participate as antigen presenting cells, since they express histocompatibility molecules class I and class II, they also express adhesion molecules like ICAM-1, VCAM, LFA-3 and B7 (Corbière et al., 2011;Cunningham et al., 1994).
The ability of M. tuberculosis to invade non-phagocytic cells was initially described by Shepard, who described the susceptibility of HeLa cells to M. tuberculosis infection (Shepard, 1955).Years later Mapother and Songer described the intestinal epithelium cell invasion by M. avium, emphasizing the active role of mycobacteria to induce its uptake (Maphoter & Songer, 1984).Later, Bermudez and Goodman demonstrated the invasion and intracellular replication of M. tuberculosis into epithelial cells, using a pneumocytes type II cell line as model (Bermudez & Goodman, 1996).The presence of M. tuberculosis DNA into pulmonary non-phagocytic cells including epithelial, endothelial and fibroblast from individuals that died by causes other than tuberculosis, confirmed the ability of M. tuberculosis to infect epithelial cells in vivo (Hernandez-Pando et al., 2000).After recognition and acceptance of the ability of M. tuberculosis to infect human pneumocytes type II, the next studies were carry out to determine the mechanism responsible of M. tuberculosis internalization into the host cell, as well as the mycobacteria ligands involved in the recognition and triggering of the endocytosis mechanisms.Nowadays the antimicrobial mechanisms exerted by the nonphagocytic cells to contain M. tuberculosis growth are been revealed.
The first studies achieved to explain the invasion mechanism, were made by Bermudez and Goodman, who reported the invasion and replication of M. tuberculosis into a cell line of human pneumocytes type II (A549 cells), they described that the internalization process was dependent of actin microfilaments and microtubules, and they suggested that vitronectin receptor and -1 integrin were the membrane receptors responsible of M. tuberculosis recognition by the epithelial cells (Bermudez & Goodman, 1996).Later, Reddy and Kumar reported cell membrane projections and membrane ruffling formation during M. avium interaction with cells from the respiratory epithelium (Hep-2), and identified two M. avium proteins (31 and 25 kD), as responsible of this interaction (Reddy & Kumar, 2000).Our group reported that M. tuberculosis entrance into the A549 cells was carried out by an endocytic process known as macropinocytosis, which involves the participation of actin filaments and membrane ruffling formation, in this study we suggested that the viability of the mycobacteria was an indispensable requirement for membrane projections formation and consequently for bacterial entry (Garcia-Perez, et al., 2003).Later, we also reported that entrance of the low-pathogenic M. smegmatis, into the A549 cells was induced by macropinocytosis as well, and we suggested that this endocytic route was not a virulence factor for the mycobacteria (Garcia-Perez et al., 2008).Diverse studies have postulated macropinocytosis as the endocytic pathway responsible of pathogenic microorganisms entrance into non-phagocytic cells (Kerr et al., 2009).It is also known that the establishment of the intracellular signaling for macropinocytosis induction requires an initial stimulation with growth factors, hormones, phorbol-esters and some bacterial products (Swanson, 1989;Swanson & Watts, 1995;Patel & Galan, 2006).Some microorganisms, like Salmonella enterica, presents a highly specialized organelle known as secretion system type III (SSTIII), through which proteins responsible of the pathogenicity of this microorganism are transferred into the host cell (Patel & Galan, 2006).Many of these proteins act imitating functions from the host cell especially those responsible of cytoskeleton rearrangements and macropinocytosis induction.Although M. tuberculosis does not count with a proper SSTIII, a related system known as ESX-1, modulates the early events occurred during mycobacterium infection (DiGiuseppe Champion & Cox, 2007).Different groups of investigators have reported the production of several molecules that can mediate M. tuberculosis interaction with nonphagocytic cells.Arruda and coworkers reported the production of an "invasive like" protein, although they did not show conclusive results on its role in tuberculosis infection (Arruda et al., 1993).The role of the Mce1 protein (one of the members of the Mce protein family) in the entrance of M. tuberculosis into the non-phagocytic cells, was demonstrated by using polystyrene microspheres covered with recombinant protein, Chitale and coworkers reported that Mce1 protein was able to induce the formation of "membrane disturbances" in non-phagocytic cells (Chitale et al., 2001).Recently, another group of investigators, reported that Mce3A protein bound to latex particles facilitated uptake by HeLa cells, suggesting that this protein also has a role in M. tuberculosis interaction with host cell (El-Shazly et al., 2007).Also, it has been reported that Mce4A protein facilitates the invasion of HeLa cells by a nonpathogenic E. coli strain expressing Mce4A.Since the Mce4A protein is expressed at the late stage of mycobacterial growth, it has been postulated that this protein besides favoring M. tuberculosis entrance, it can play an active role in mycobacterium survival and persistence (Saini et al., 2008).
Another protein of M. tuberculosis involved in the interaction with the non-phagocytic cells is the heparin binding hemagglutinin (HBHA), which has been reported that induces membrane projections and bacteria internalization into Hep-2 cells (respiratory epithelium) (Reddy and Hayworth, 2002).Additionally, HBHA has been implied in extrapulmonary dissemination of M. tuberculosis, hbha gene disruption has deep effects on M. tuberculosis interaction with epithelial cells, but does not affect interaction with macrophages.On the other hand, the mutant strain of HBHA expressed a reduced capacity to colonize spleen, although lung colonization was not affected.These data emphasize the importance of HBHA in extrapulmonary dissemination and remarks the importance of non-phagocytic cells-M.tuberculosis interaction in tuberculosis pathogenesis (Pethe et al., 2001).Menozzi and coworkers reported that HBHA induce endocytosis mediated by receptor through the recognition of proteoglycans containing heparan sulphate (Menozzi et al., 2006).Like other bacterial proteins, mycobacterial HBHA can mimic the function of some cellular proteins being responsible of the modification of diverse cellular activities, including entrance, survival and pathogen dissemination.Recent studies support this theory, Verbelen and collaborators reported that HBHA has a sequence similar to proteins that bound actin, like tropomiosin-1, ezrin-1 and the heavy chain of miosina-9, and their results demonstrated the specific and stable union of HBHA to actin (Verbelen et al., 2008).Another research group showed that HBHA was able to bind to G-actin without altering their nucleation but obstructing actin polymerization; these authors suggest that like profilin, HBHA can affect polymerizationdepolymerization dynamics of F-actin facilitating M. tuberculosis mobility into the cytoplasm (Esposito et al., 2011).This report emphasize the importance of HBHA in M. tuberculosis dissemination, since it has demonstrated that HBHA induces membrane protrusions formation, is possible to infer that HBHA may be one of the mycobacterial molecules that can act as "signals for internalization" hence inducing mycobacteria entrance by macropinocytosis, as previously we suggested (Garcia-Perez et al., 2008).Other molecules of M. tuberculosis responsible of the invasion of epithelial cells are the mycobacterial DNA-binding protein 1 (MDP1), which promotes A549 cells infection through hyaluronic acid (Aoki et al., 2004); and a group of high active bound peptides (HABPs) of different hypothetical proteins from M. tuberculosis (Rv2301, Rv0180c, Rv0679c, among others), which have shown that facilitates the bound and internalization of latex particles to A549 cells (Cáceres et al., 2011;Cifuentes et al., 2010;Ocampo et al., 2011).
The knowledge of the molecules that participate in M. tuberculosis i n v a s i o n t o n o nphagocytic cells is still scarce, basically two surface proteins, Mce and HBHA are the most studied.The great complexity of the mycobacterium cell wall and the external lipids "wall" composition allows us to suggest that most probably not only proteins are involved in mycobacterium invasion, but also some other constituents like mycobacterium-lipids may have an important role.We described that the morphological changes and the cytoskeleton modification of the A549 cells infected by M. tuberculosis were observed in cells with bound or internalized bacteria, but also were observed in cells without bacteria, this made us to suggest that those changes could be induced also by secretion molecules (Garcia-Perez et al, 2003;Garcia-Perez et al., 2008).Chopra and colleagues speculated that the nucleotide diphosphate kinase (Ndk), a M. tuberculosis secretion protein, which acts as a GTPases protein activator may contribute to internalization of M. tuberculosis into the host cell by inducing cytoskeleton reorganization (Chopra et al., 2004).Still there is a need to elucidate and explain the role of M. tuberculosis secretion molecules that could be related to invasion and survival of mycobacteria into the non-phagocytic cells.
M. tuberculosis entrance into the non-phagocytic cells is not a circumstantial event, it must be induce by mycobacterial product(s), the best candidates could be mycobacterial actively secreted products that could reach many cells in a short time, and these molecules can be recognized as the "first signal".This first signal initiates the intracellular signaling cascades that leads to the formation of membrane protrusions necessary to internalize the bacillus, but this event apparently is not enough to assure mycobacterial internalization.During the non-phagocytic cell infection event, only few cells will be finally infected; if we infect simultaneously, with the same mycobacterial cell suspension, during the same time, a monolayer of macrophages and a monolayer of non-phagocytic cells (for instance lung epithelial cells), up to 90% of the macrophages will be infected compared to 10% of epithelial cells (Garcia-Pérez et al., 2003).So, most probably another signal, the "second signal" is needed to assure that mycobacteria will be internalized by the non-phagocytic cell, and most probably this "second signal" will depend on the physical interaction of some bacterium molecule (PAMP) with a receptor molecule on the host cell (PRR).For the case of macrophages, already we described the variety of receptors involved in the interaction mycobacteria:phagocytic cell.With respect to the non-phagocytic cells, there are few studies of cell receptors involved in mycobacterium invasion.Given the frequency and ubiquitous distribution of glycosaminoglycans on the epithelial cell surface, it has been proposed that these molecules are good candidates for the union and penetration of pathogenic microorganisms in the tissues (Menozzi et al., 1996).The evidences indicate that cells from the respiratory epithelium express PRRs, specifically TLRs, making them able to contribute to the establishment of the pulmonary innate immune response.In the pulmonary epithelium, 11 TLRs are expressed, along with CD14, an important element for TLR4 function, these molecules have been involved in the recognition of pathogens by these epithelial cells (Bals & Hiemstra, 2004;Gribar et al., 2008).Interestingly, Lee and collaborators reported that M. tuberculosis actively induce Dectin-1 receptor expression by A549 cells; Dectin-1 is a classical receptor for fungi, and is expressed mainly by myeloid cells.They established that for Dectin-1 expression, the participation of TLR2 and ROS were crucial.In addition, the authors demonstrate d t h a t D e c t i n -1 a n d T L R 2 w e r e e q u a l l y responsible of the early activation of Src in the epithelial cells stimulated by M. tuberculosis, and suggest that Dectin-1 in the epithelial cell contribute to mycobacterial survival (Lee et al., 2009) (Fig. 1).

Endothelial cells infection by M. tuberculosis
The endothelium is composed of a single layer of thin flattened cells known as endothelial cells, that lines internal body cavities and the lumens of vessels, endothelial cells had a mesodermal origin.Endothelial cells participate in several cellular functions like vasoconstriction and vasodilation (blood pressure control), angiogenesis, inflammation, etc., so endothelial cells contribute to the whole homeostasis of the organism.In cases of bacteraemia or viraemia, endothelial cells are among the first cells to be in contact with microbial pathogens and are also among the first cells that respond to endogenous molecules released in events of tissue damage (Opitz et al., 2009).Different research groups have demonstrated M. tuberculosis interaction with the endothelial cells; in the first study, Birkness and coworkers used a bilayer model of epithelial and endothelial cells, and they demonstrated adhesion, internalization and passage of M. tuberculosis through the bilayer (Birkness et al., 1999).Later, the study of Bermudez and collaborators established that M. tuberculosis crosses the alveolar wall to reach the endothelial cell, hence suggesting the ability of M. tuberculosis to disseminate (Bermudez et al., 2002).The presence of M. tuberculosis DNA in lung endothelial cells of human samples from cadaveric donors from countries with high TB incidence, confirmed the infection of endothelial cells in in vivo situations (Hernandez-Pando et al., 2000).The first oriented study to establish some features of endothelial cell infection by M. tuberculosis was performed by Metha and coworkers; in this work, authors describe the interaction of M. tuberculosis with two endothelial cell lines, one from human lung (HULEC) and the other from human foreskin (HMEC-1).The authors determined that M. tuberculosis may bound, enter and survive better into the pulmonary cell line (HULEC), than in the skin cell line HMEC-1, suggesting that lung cells are a better niche for mycobacterial replication (Metha et al., 2006).Recently, our group reported the innate response of human umbilical vein endothelial cells (HUVEC) against M. tuberculosis, M. abscessus and M. smegmatis infection.The study demonstrated that although the three different mycobacterium species were internalized, the intracellular fate of each strain was different.The mechanism responsible of mycobacteria entrance into the endothelial cells is still unknown, but we found that during mycobacteria internalization the endothelial cells display cytoskeleton rearrangements and the magnitude of those changes was different for each mycobacterium strain, for instance, the more virulent mycobacteria induced the most dramatic rearrangements.It is necessary to determine if macropinocytosis is also the mechanisms responsible of mycobacterial uptake by endothelial cells, as it occurs for epithelial cells (Garcia-Perez et al., 2003).Contrary to other cell models, intracellular M. tuberculosis did not replicate, or was eliminated by HUVECs, resembling a "latency stage", that could be triggered by the high NO levels induced at early times of the infection (Garcia-Perez et al., 2011).
Up today there are not conclusive studies on the molecules responsible of M. tuberculosis recognition by endothelial cells, however is known that endothelial cells express a great variety of PRRs through which they interact with different pathogens (Opitz et al., 2009).The few reports on mycobacterial molecules involved in endothelial recognition indicate that phosphatidylinositide mannoside (PIMs) can act as adhesins that mediate M. tuberculosis bounding to non-phagocytic cells (Hoppe et al., 1997).Another study indicates that HBHA induces acting filament reorganization in a confluent endothelial cells monolayer, which suggests that this protein may also be involved in the recognition of M. tuberculosis by endothelial cells (Menozzi et al., 2006).

Interaction of M. tuberculosis with other non-phagocytic cells
M. tuberculosis interaction with other non-phagocytic cells like adipocytes or fibroblasts has been reported, although few reports exist on the matter.A recent study demonstrated that adipocytes are susceptible to M. tuberculosis invasion but in these cells, M. tuberculosis persists in a non-replicative state.Mycobacterial recognition by adipocytes was mediated by scavenger receptors.The authors suggest that due to the abundance and wide distribution of fatty tissues in the human body, adipocytes may represent an ideal reservoir for persistent tuberculosis bacilli (Neyrolles et al., 2006).
Fibroblasts and epithelial cells constitute two of the main structural cell lineages of the lung, in tuberculosis, fibroblasts are present in the periphery of granulomas; they synthesize extracellular matrix proteins that are required for the maintenance and repairing of the lung; in addition they produce cytokines and chemokines that allow leukocyte recruitment and activation (Suzuki et al., 2008).In spite of the physiological relevance of the fibroblasts, its interaction with pathogens, specifically with M. tuberculosis has been scarcely explored.The first study of the matter, performed by Rastogi and coworkers, demonstrated that M. tuberculosis and M. avium can infect a murine fibroblast cell line, and in response to the infection, fibroblasts secreted several chemical mediators that promote mycobacterium elimination by macrophages (Rastogi et al., 1992).Up today, two additional studies demonstrated the role of the fibroblasts in the mycobacterial infection.In the first study, Hernandez-Pando and coworkers reported the presence of mycobacterial DNA in lung fibroblasts (among other cells) from cadaveric donors that die for causes different to tuberculosis; these individuals lived in countries with high tuberculosis prevalence.These data suggest that a previous in vivo M. tuberculosis infection was established and controlled in those individuals and reveal the importance of lung fibroblast in the response (Hernandez-Pando et al., 2000).The second study confirms that fibroblast are susceptible of M. tuberculosis infection; and describes the multiplication features of several strains and mutants of M. tuberculosis into human lung fibroblast and rat lung fibroblasts, although they found that human fibroblast are more permissive to mycobacterial growth.Contrary to several studies with other cell lineages like epithelial or endothelial cells, these authors report that in fibroblast, M. tuberculosis internalization is actin and tubulin independent, and c o n c l u d e t h a t m y c o b a c t e r i a d o n o r e q u i r e s c e ll host cytoskeleton reorganization to be internalized (Ferrer et al., 2009).Preliminary results of our group demonstrate that infection of mouse lung fibroblasts (MLg cell line) by M. tuberculosis induce actin mobilization and membrane ruffling contrary to those observations described by Ferrer and coworkers (Fig. 2).The discrepancy in these results forces to make new studies to clarify and precise the mechanism responsible of M. tuberculosis internalization into fibroblasts, and to determine the mycobacterial molecules and the cell receptors involved in the recognition and triggering of this event.In addition, it is necessary determine the defense mechanisms elicited by the fibroblasts to contain the mycobacterial infections.

Innate immune response of non-phagocytic infected by M. tuberculosis
Phagocytic cells like macrophages, neutrophils and monocytes have a crucial role in the defense against infections; they are responsible of bacterial phagocytosis and their elimination.They count with well adapted systems for bacterial recognition and exhibit sophisticate and highly effective mechanisms to eliminate the pathogenic microorganisms.Production of reactive oxygen and reactive nitrogen species, cytokines and chemokines production, degradative enzymes production, and antimicrobial peptide production are some of the elements displayed by the phagocytic cells to contain and kill pathogens.On the other hand, pathogenic microorganisms have developed numerous mechanisms to evade the antimicrobial arsenal and to guarantee their survival and persistence.In particular, M. tuberculosis has developed several mechanisms to avoid the hostile environment of the macrophages and to persist within them: a) inhibition of the phago-lysosome fusion, b) inhibition of the phago-lysosome acidification, c) recruitment and retention of TACO protein (tryptophan-aspartate containing coat protein) on phagosomes to prevent fusion with lysosome and d) expression of proteins and lipids for protection against the oxidative stress (Meena & Rajni, 2010).Also the ability of mycobacterium to enter into non-phagocytic cells is another mechanism of immunological evasion, since mycobacteria "hided" in these cells could skip hostile environment of the macrophages and create an ideal niche for its replication and establishment.Nevertheless, recent evidences suggest that non-phagocytic cells have an important role in the immune response against mycobacteria and these cells have the potentially to exert potent antimycobacterial mechanisms (Garcia-Perez et al., 2011;Kwon et al., 1998;Sharma et al., 2007).Besides, we demonstrated that the intracellular destiny of M. tuberculosis will be dictated by the cell type infected, hence while in epithelial cells M. tuberculosis survives and replicates, in the endothelial cells tends to persist without showing replication (Garcia-Perez et al., 2011).We consider that non-phagocytic cells can actively respond to M. tuberculosis infection, but the magnitude and diversity of their response can be modulated by the bacillus to guarantee its survival, persistence and/or dissemination.Kwon and George demonstrated that epithelial cells are able to produce nitric oxide (NO) in response to cytokines like IL-1 and IFN- (Kwon & George, 1999) and also in response to M. tuberculosis (Kwon et al., 1998).Later, Roy and collaborators established that direct infection of epithelial cells by M. tuberculosis, stimulates the novo production of NO via iNOS, although they demonstrated that the low levels of NO produced during the direct-mycobacterial stimulation, were not sufficient to kill the bacilli, however NO production by cells infected and simultaneously stimulated with cytokines was sufficient to kill the mycobacteria and to reduce the bacterial load (Roy et al., 2004).We reported that endothelial cells produce NO in response to mycobacterial infection and we found that NO concentrations correlated with the capacity of infective bacteria to induce modifications in the actin cytoskeleton.We demonstrated that a high NO production does not correlate with the diminution of the bacterial load, on the contrary, the mycobacterium with the greater intracellular replication (M.abscessus), induced the greater amount of NO.In comparison, the mycobacterium that induced the lower NO levels was eliminated (M.smegmatis), but induced the higher ROS production, which allowed us to suggest that ROS have a more important role for mycobacterial elimination than NO.Interestingly, in the endothelial cells, M. tuberculosis exhibited a non-replicative state, resembling the latency state characteristic of tuberculosis.In these cells, M. tuberculosis triggered intermediate levels of NO compared with the other two mycobacteria, but at very early stages of infection, a "burst of NO" production was observed, and a rapid diminution of this metabolite proceed to this event (Garcia-Perez et al., 2011).Based on these results, we can suggest that the ability of the mycobacteria to modulate NO production by the endothelial cells will determine, in great extent, its final intracellular fate.
The importance of ROS in the innate immunity was recognized initially in the phagocytic cells, which under certain stimuli generate the "respiratory burst".ROS production is related to the activity of the NADPH oxidases, although ROS may be produced by other mechanisms (West et al., 2011).In the phagocytic cells, the NADPH oxidase responsible of ROS production is NOX2.In recent years, 7 isoforms of NOX/DUOX enzymes have been described, and their distribution and function in different cells or tissues is been recognized (Krause, 2004).ROS have an ample variety of biological activities, participating not only as antimicrobial agent (which is one of the least sophisticated activities).ROS participate in hormone biosynthesis, intracellular signaling, blood pressure control, etc. (Krause, 2004).At the moment it has been recognized that some non-phagocytic cells express some members of this oxidase family; one of the most studied cells are the pneumocytes type II A549 cells, which has been reported that express NOX1, NOX2, DUOX 1 and DUOX2 (Kolářová et al., 2010), and although in these cells the role of ROS for pathogens elimination has not been studied, there are reports that indicate the importance of ROS generation in the modulation of the immune response.In the infection of epithelial cells by M. tuberculosis, Lee and collaborators reported that ROS are produced via the activation of Src, after Dectin-1 engagement, Dectin-1 is a crucial receptor that contributes to mycobacterium internalization, the expression of pro-inflammatory mediators and the establishment of an effective antimicrobial environment, constituted in part by ROS (Lee et al., 2009).In the endothelial cells, Van Buul and coworkers, reported the differential expression and location of NOX2 and NOX4, whereas NOX4 was located in the endoplasmic reticulum, the regulating proteins of NOX2 were located and associated to actin present at membrane protrusions and ruffles; in this work the authors correlated ROS scavenging activity with an impair of cytoskeleton rearrangement and the formation of confluent monolayers (Van Buul et al., 2005).We reported that endothelial cells produce ROS in response to the mycobacterium infection, and we correlated a high ROS production with mycobacterium elimination (Garcia-Perez et al., 2011).The role of ROS as direct M. tuberculosis effectors is still controversial, however ROS indirectly, may contribute to M. tuberculosis elimination, for example the signaling route dependent of ROS and NOX2, is crucial for the achievement of the antimycobacterial effect of cathelicidin, an antimicrobial peptide induced by vitamin D3 in macrophages (Yang et al., 2009).
The participation of antimicrobial peptides or defensins in M. tuberculosis infection is a current issue.Defensin production by some non-phagocytic cells infected with mycobacteria has been reported; Rivas-Santiago and collaborators, found that in vitro infection of pneumocytes (A549 cells) by M. tuberculosis, induced human -defensin 2 (HBD-2) production, immunolocalization with colloidal gold demonstrated HBD-2 associated to damaged bacillus, suggesting the antimicrobial effect of this peptide (Rivas-Santiago et al., 2005).Later, the same group reported the expression and production of another antimicrobial peptide cathelicidin LL37, by epithelial cells infected by M. tuberculosis (Rivas-Santiago et al., 2008).In a mouse model of chronic pulmonary tuberculosis and latent tuberculosis, defensin expression was analyzed; the results demonstrated that -defensin-3 (MBD-3), -defensin-4 (MBD4) and CRAMP (the mouse equivalent of the cathelicidin LL37) were expressed by different cell types including lung epithelial cells.In the progressive tuberculosis model, initial production of MBD3 and MBD4 correlated with the transitory control of mycobacterial growth, although their expression diminished at the late stage of the disease.In the latent infection, MBD3 and MBD4 were expressed continuously, but after disease reactivation, their production was abrogated.CRAMP expression was high in the progressive phase of the disease and the protein was observed in different structures of the mycobacteria.The authors of these studies conclude that both defensins and cathelicidin participate in mycobacterial containment, not only by the direct activity against the bacteria, but by their recognized chemotactic activity (Rivas-Santiago et al., 2006;Castañeda-Delgado et al., 2010).Another antimicrobial peptide involved in the defense against M. tuberculosis is hepcidin.Recently, the production of this antimicrobial peptide by the epithelial cells stimulated with ManLAM and PIMs from M. tuberculosis was reported, and proposed that the antimicrobial activity was mediated by a reduction in the iron available at the alveolar space (Sow et al., 2011).We recently reported that M. tuberculosis infection of endothelial cells stimulated HBD-1 production, and we proposed that this defensin could have an important role in the mycobacterial growth control observed in this model (Garcia-Perez et al., 2011).The direct antimycobacterial role of HBD1 is not determined so far.
One of the actual hypotheses is that the environment surrounding the non-phagocytic cells is crucial for their activation and preparation for bacterial elimination.Greco and collaborators reported that the enriched lysophospholipid pulmonary microenvironment protects pneumocytes type II from cytotoxicity induced by M. tuberculosis, and increases the antimycobacterium response by promoting acidification of endosomal compartments (Greco et al., 2010).Desvignes and Ernst demonstrated that epithelial and endothelial cells after IFN-stimulation express a better response for M. tuberculosis growth control.These authors established that one of the mechanisms involved in the response induced by IFN-is expression of indoleamine-2,3-dioxygenase (IDO) and regulation of the expression of IL-17 (Desvignes & Ernst, 2009).The capacity of epithelial cells to improve their response after IFN-exposure is supported by the observation that cells infected by M. tuberculosis express elevated levels of the receptor specific for this cytokine (Sharma et al., 2007) Vol.197, No.9, (May 2003), pp. 1119-1124, ISSN 1540-9538 Bulut, Y.;Faure, E.;Thomas, L.;Equils, O. & Arditi, M. (2001) Cellular microbiology, Vol.9, No.6, (June 2007) pp.1376-1384, ISSN 1462-5822 Dinadayala, P.; Lemassu, A.;Granovski, P.;Cerantola, S.;Winter, N. & Daffe, M. (2004).
Revisiting the structure of the anti-neoplastic glucans of Mycobacterium bovis

Fig. 2 .
Fig. 2. Interaction of M. tuberculosis with fibroblasts.Cells were infected during 1 h with M. tuberculosis.Before the infection, bacilli were stained with fluorescein isothiocyanate (FITC, green color).After infection, actin filaments were labeled with rhodamine-phalloidin (red).A) Uninfected cell showing actin filaments longitudinally distributed.B) Infected cell showing membrane protrusions and membrane ruffling in contact with mycobacteria.