Autophagy as a Target for Drug Development Of Skin Infection Caused by Mycobacteria

Pathogenic mycobacteria species may subvert the innate immune mechanisms and can modulate the activation of cells that cause disease in the skin. Cutaneous mycobacterial infection may present different clinical presentations and it is associated with stigma, deformity, and disability. The understanding of the immunopathogenic mechanisms related to mycobacterial infection in human skin is of pivotal importance to identify targets for new therapeutic strategies. The occurrence of reactional episodes and relapse in leprosy patients, the emergence of resistant mycobacteria strains, and the absence of effective drugs to treat mycobacterial cutaneous infection increased the interest in the development of therapies based on repurposed drugs against mycobacteria. The mechanism of action of many of these therapies evaluated is linked to the activation of autophagy. Autophagy is an evolutionary conserved lysosomal degradation pathway that has been associated with the control of the mycobacterial bacillary load. Here, we review the role of autophagy in the pathogenesis of cutaneous mycobacterial infection and discuss the perspectives of autophagy as a target for drug development and repurposing against cutaneous mycobacterial infection.


INTRODUCTION
Pathogenic mycobacteria species subvert the innate immune system barriers and modulate the activation of phagocytes to cause disease not only in the respiratory tract but also in soft tissues and skin, sometimes resulting in disseminated infection (1). Cutaneous mycobacterial infections may cause different clinical manifestations, such as cutaneous manifestations of Mycobacterium tuberculosis (M. tuberculosis) infection, Buruli ulcer caused by M. ulcerans and other related slowly growing mycobacteria, leprosy caused by M. leprae and M. lepromatosis, and cutaneous infections caused by rapidly growing mycobacteria such as M. abscessus subsp. abscessus, M. abscessus subsp. bolletti, M. abscessus subsp. massiliense, M. chelonae and M. fortuitum (1)(2)(3)(4)(5)(6)(7)(8)(9). Among patients with advanced immunosuppression, M. avium-intracellulare complex, the M. haemophilum, and M. kansasii may cause cutaneous or disseminated disease. Mycobacterial infections of the skin and subcutaneous tissue are associated with important stigma, deformity, and disability. The treatment for cutaneous mycobacterial infections depends on the specific pathogen, whereas for rapidly growing mycobacteria, the official treatment guidelines recommend carrying out susceptibility tests for antibacterial drugs of different classes (10,11). Management often includes use of multiple antibiotics for several months (12). Treatment options for cutaneous tuberculosis follow the same recommendations for the treatment of other forms of TB, being limited to conventional oral therapy and surgical intervention for severe forms, such as lupus vulgaris (13,14). The therapeutic regimen is based on the combination of isoniazid, rifampicin, pyrazinamide, ethambutol and streptomycin according to the needs of each individual. In most cases, skin manifestations result from hematogenous dissemination or are a direct extension from the focus of the infection (14,15). In addition, treatment of leprosy is performed with multidrug therapy (MDT) and comprises 6 or 12 doses, depending on the clinical form. There is not a consensus for the treatment of cutaneous infections caused by non-tuberculous mycobacteria. Recently, much effort has been made to develop more effective therapies by modulating host responses to mycobacteria (i.e., host-directed therapy).
After recognition by skin cells, mycobacteria may use a wide range of strategies to escape the microbicidal activity of skin host cells. Some of these immune escape mechanisms are the inhibition of the maturation of phagolysosomes, inhibition of the acidification of phagolysosomes, bacterial escape to reside in the cytosol, modulation of host cell metabolism, inhibition of oxidative stress, and inhibition of apoptosis and autophagy associated with increased type 1 interferon (IFN) expression and inflammasome activation (16)(17)(18)(19)(20)(21)(22)(23).
Autophagy is an intracellular catabolic process that may contribute to the removal of invading pathogens via a lysosomal degradation pathway. The activation of autophagy by diverse drugs or agents may represent a promising treatment strategy against mycobacterial diseases. In this review, we discuss the current knowledge of, advances and perspectives on new therapeutic strategies targeting autophagy against mycobacterial infections in the skin.

OVERVIEW OF AUTOPHAGY MACHINERY ON SKIN CELLS
Autophagy is a homeostatic mechanism highly conserved evolutionarily and dependent on the lysosome action (24). It is responsible for the cellular catabolism of dysfunctional organelles, components of the cytoplasm and, more recently, invading pathogens, thus determining the maintenance of homeostasis and adaptation of the cell to stress (25,26). Autophagy has been described as having a primary role in physiological cellular processes such as development and growth, in the senescence process, and immune defense (25,(27)(28)(29). Based on the way the autophagy target is taken to the lysosome, its final destination of degradation, autophagy was didactically classified into three forms: macroautophagy, microautophagy, and chaperonemediated autophagy. In this review, we will exclusively address the action and manipulation of the macroautophagy pathway.
Only a small amount of research has considered the impact of autophagy on the pathogenesis of skin diseases, including diseases caused by mycobacteria. Skin is the largest organ of the body and it is not only the first line of defense against numerous insults but it is also the site whereas some infectious, including mycobacterial diseases, may manifest.
Autophagy is considered an effector tool of the immune system since it is a relevant pathway of elimination and recognition of pathogens by the immune system (30). As well to cellular homeostasis, autophagy works to eliminate intracellular pathogens, including some pathogens associated with skin diseases, such as Streptococcus pyogenes from group A (31,32), Staphylococcus aureus (33,34), M. leprae (35,36), M. marinum (37,38), and M. tuberculosis (39)(40)(41)(42). Through a process called xenophagy, which plays a principal role in innate immune defense, intracellular pathogens are directed to the autophagosome and then to the lysosomal degradation pathway (43,44). Xenophagy is the process of eliminating intracellular pathogens through autophagic machinery, being a unique type of macroautophagy/selective autophagy that targets invasive pathogens, being an important defense mechanism against infectious diseases (45,46).
Few studies have focused on deciphering autophagy machinery in skin cells, such as: keratinocytes, skin fibroblasts, melanocytes, Langerhans cells, dendritic cells, mast cells, neutrophils, NK and B cells. The current knowledge regarding skin cell autophagy during mycobacterial diseases is based mainly in studies with cell lineage and dermal macrophages.
Briefly, after pathogen recognition by host cells, the first step is the formation of the isolation membrane, which starts to grow and expand in size until sequestration and the surrounding of the target and finally closure to form the autophagosome. Subsequently, autophagosomes fuse with lysosomes to generate autolysosomes through elimination and recycling the sequestered charges via the lysosomal proteases ( Figure 1) (28). A large number of proteins have been identified as highly relevant in different stages of control and action in autophagic flow. Several cell types have autophagy as an effector mechanism for homeostatic/immune functions as skin cells like keratinocytes and macrophages ( Figure 1) (47).
A wide variety of signals regulates the activation of autophagy. The induction of autophagy can occur through the recognition of microbial factors that are ubiquitinated and recognized by autophagy cargo adaptor proteins (these include p62 (sequestosome 1), NBR1 (neighbor of BRCA1 gene 1 protein), NDP52 (calcium binding and coiled-coil domain 2), optineurin and galectin) or can occur by the production of reactive oxygen radicals and IFN-g-mediated proteolysis, and autophagosome formation (43,(48)(49)(50)(51)(52). The autophagy pathway may be negatively regulated by PI3K (phosphoinositide 3-kinase)/Akt (protein kinase B)/mTOR (target of rapamycin FIGURE 1 | Different steps of the autophagic pathway targeted by autophagy-modulating drugs. A schematic view of the different cell types populating the skin. Vertebrate skin is comprised of two major compartments: the epidermis and the dermis. The superficial part of the epidermis, known as the stratum corneum, is composed of dead keratinocytes and acts as a barrier. The epidermis is composed mainly of keratinocytes with few melanocytes. The major immune cells in this compartment include Langerhans cells (LCs) and CD8 T-cells. The dermis is composed of fibroblasts, NK cells, T-cells (CD4 ab, and gd), B cells, dermal dendritic cells, macrophages, mast cells, and neutrophils (non-exhaustive list). The knowledge of skin cell autophagy is mainly based in studies with dermal macrophages. Briefly, (1) autophagy is inhibited by mTOR and activated by AMPK. mTOR is inhibited by the autophagy-initiation signals as metabolic stress, ROS, infection and drugs, and leads to the activation of AMPK. After AMPK activation, the ULK1 complex (ATG13, ULK1/2, FIP200) initiates the phagophore formation (2), involving the targets (pathogens, dead cells, cellular components and organelles, protein aggregates), which in turn activates the Class III PI3K complex (Beclin 1, VPS34, VPS15, ATG14) (3). This complex completes the autophagosome maturation and elongation by forming PI3P in the omegasome membrane and recruiting downstream ubiquitin-like conjugation systems that convert LC3-I to LC3-II (4). Fully formed autophagosomes then fuse with lysosomes (autolysosomes), degrade the sequestered cargo via lysosomal hydrolases and recycle macromolecule components (5). Several drugs can interfere with the autophagic pathway by inhibiting or activating different parts of the process (see also Table 1). Drugs as rapamycin, resveratrol and nitazoxanide, that inhibit mTOR, or carbamazepine, metformin and pyrazinamide, that activate AMPK, induce autophagy. Bedaquiline, ambroxol and linezolid increase the formation of autophagosomes. Loperamide and valproic acid increase the colocalization of LC3-decorated autophagosomes with M. tuberculosis. Ibrutinib and isoniazid facilitate the fusion of phagosome and lysosome. Vitamin D3 (1,25D3) induces the expression of antimicrobial peptides as cathelicidin and upregulates the expression of Beclin 1 and ATG5, that are pivotal for the autophagosome formation. On the other hand, azithromycin was demonstrated to inhibit the acidification of the autolysosome impairing M. abscessus degradation.

AUTOPHAGY AS AN INNATE IMMUNE MECHANISM AGAINST MYCOBACTERIAL DISEASES
There is a strong relationship between autophagy signals and pattern recognition receptors, such as TLR (Toll-Like Receptors) that include TLR3, TLR4, TLR5, TLR6, TLR9, and the heterodimers TLR1/2, TLR7/8 that are capable of activating autophagy in macrophages, dendritic cells, and neutrophils (56)(57)(58). This activation occurs via signaling of the adaptor proteins MyD88 (myeloid differentiation factor 88) and TRIF (TIR-domain-containing adapter-inducing interferon-b). Xu and colleagues (59) demonstrated that after the stimulation of TLR4, positive LC3 (microtubule-associated protein 1A/1B-light chain 3) aggregates form in the macrophage cytoplasm and increase mycobacterial elimination through autophagy. Interestingly, for the LC3-aggregates induction, via TLR4 induction, it is necessary to activate the protein TRIF, as well as other proteins like RIP1 (receptor-interacting protein 1) and p38 for autophagic induction (56,59). TLR4 acts as a proautophagic receptor in TRIF-dependent pathways. TLR4 induces the production of TNF (tumor necrosis factor) by a mechanism that is mediated both by reactive oxygen species (ROS) and nitrogen intermediates (i.e. nitric oxide), and by p38 and MAPK and the inhibition of these components may lead to total autophagy inactivity (60)(61)(62). Studies have shown that in LPS (lipopolysaccharide)-TLR4-mediated autophagy, activation of the transcription factor Nrf2 (nuclear factor erythroid 2related factor 2) occurs, which leads to increased p62 transcription and formation of aggresome-like induced structures (ALIS) with subsequent autophagic degradation (63,64), showing the ability of this receptor to link innate immunity with cellular oxidative response or adaptive immunity.
It is known that TLR receptors are of great importance for the activation of dendritic cells (DCs) and their subsequent maturation, some of these receptors such as TLR4 and TLR2 are already described as inducing an innate response against M. tuberculosis (65-67). Khan and colleagues (68) observed that the co-stimulus of CD40 and TLR4 leads to the production of pro-inflammatory cytokines such as IL-6, IL-12 and TNF, autophagy and death of mycobacteria. Interestingly, when they evaluated this co-stimulus as an adjunct to anti-TB therapy, they observed an increase in vivo and in vitro of the deadly potential of anti-TB drugs. Shin and colleagues (69) showed that stimulation of TLR2/1/CD14 by mycobacterial lipoprotein LpqH can activate antibacterial autophagy by activating vitamin D receptor signaling and inducing cathelicidin. They suggested that the TLR2/1/CD14-Ca 2+ -AMPK (Adenosine monophosphateactivated protein kinase)-p38 MAPK pathways contribute to cathelicidin-dependent expression, which played an important role in LpqH-induced autophagy. A study comparing the induction of autophagy by different species of mycobacteria found that non-pathogenic mycobacteria, such as M. smegmatis, induce a more robust autophagy response than M. tuberculosis (strain H37Rv) (70). The group observed a decrease in LC3-II protein expression when the TLR2 receptor was blocked, as well as a reduction in the colocalization of LC3 with M. smegmatis DpmmB (lipoglycan deficient mutant), suggesting the participation of TLR2 in the activation of autophagy during infection with M. smegmatis (70). M. smegmatis can also be recognized by NOD2 (nucleotide-binding oligomerization domain-containing protein 2) and dectin-2 receptors (71).
In addition to the TLR receptors, another group of innate receptors was the nucleotide-binding oligomerization domain (NLRs). It has already been described that the presence of the NOD2 receptor is capable of synergistically amplify the production of pro-inflammatory cytokines and their bactericidal activity (72). In previous studies, Khan and colleagues (73) have demonstrated that after the induction of both receptors, an increase in the bactericidal capacity of DCs in vitro was observed and they required a much lower dose of the drug to kill M. tuberculosis, in addition, activated DCs induced a more effective T cell response in vivo with an increase in autophagy (73,74). Since pathogenic mycobacteria can modulate the autophagy machinery in skin cells, we hypothesize that autophagy may be a target for new therapeutic strategies against mycobacterial infections in the skin.

AUTOPHAGY-TARGETING THERAPEUTICS UPON MYCOBACTERIAL INFECTION
Despite the efficacy of anti-TB treatment based on classic isoniazid and rifampicin, limitations in terms of drug resistance, duration of treatment, associated with the use of a complex treatment regimen (75), made the researchers use another strategy in the treatment of different bacterial disease. Besides, unlike infections caused by M. tuberculosis and M. leprae for which there is a well-established therapeutic regimen, there are no standardized and effective regimens for the treatment of non-tuberculosis mycobacteria (NTMs) (10). A promising strategy in the treatment of infectious diseases is the use of host-directed therapy. It works as an adjuvant therapy, which aims to enhance the main components of the host's antimycobacterial effector mechanisms (76)(77)(78)(79). Several studies on immunity, host-pathogen interactions, and host-directed interventions have shown that the antimycobacterial action of anti-TB drugs (standardized scheme) is associated with the induction of autophagy (40). Thus, several drugs used in the clinical area to treat infectious diseases may have their action through the autophagic process.
We previously showed that xenophagy is a crucial mechanism in the leprosy outcome. A functional autophagy pathway driven by IFN-g and Beclin 1 in skin lesion macrophages was associated with the self-healing paucibacillary tuberculoid form of the disease, whereas a BCL2 (apoptosis regulator Bcl-2)-mediated block of Beclin 1 autophagy axis was linked to the progressive multibacillary lepromatous pole (35). While macrophages patrol the dermis, the human epidermis is enriched for Langerhans cells (LC). Langerhans cells restricted human immunodeficiency virus (HIV)-1 infection through the capture of viral particles by langerin and subsequent internalization into Birbeck granules and targeting of HIV-1 for destruction in the TRIM5 (tripartite motif-containing protein 5) auto lysosomal pathway (80), which in turn is induced by IFN-g (81). In M. leprae-infected LC, the antimicrobial activity induced by IFN-g treatment is achieved through autophagy, which improves the degradation of M. leprae-containing phagolysosomes and fine-tunes LC's power to present antigens for T cells in a CD1a-restricted manner (82). Thus, IFN-g therapy or a drug targeting autophagy on skin cells could be favorable to the clinical management of leprosy and other skin-related mycobacteriosis such as fish-tank granuloma (83) and Buruli ulcer (1), as well as outbreak associated postsurgical and tattoo ink infections caused by rapidly growing mycobacteria (2,4). Indeed, the acid-fast bacilli clearance in the skin of multibacillary leprosy patients is accelerated when multidrug therapy is used along with an intradermal treatment with recombinant human IFN-g (84).
Cell-based studies in leprosy have predominantly focused on dermal cells such as macrophages, neutrophils and T cells. In the dermis, macrophages are an important cell type that promote Th1-type responses, but there is evidence about the involvement of the epidermis in the development of reactional episodes (85) which are acute inflammatory episodes that can occur before, during or after the release of multidrug therapy, being responsible for the cases of disability caused by the disease (86). The relevance of autophagy as a drug target is not only restricted to the control of M. leprae infection but also to its potential to regulate the exacerbated inflammation associated with leprosy reactional episodes, as autophagy tempers inflammation by hijacking active inflammasomes for destruction (87). The downregulation of autophagy observed in skin lesion macrophages of multibacillary leprosy patients also predicts the reversal reaction onset. This impairment of the autophagic pathway correlates with the activation of NLRP3 (NALP3; NACHT, LRR and PYD domains-containing protein 3) inflammasome and IL-1b production, which drive the inflammatory status found in multibacillary patients when undergoing reversal reaction (36). On the other hand, due to Th2!Th1 shift and increased IFN-g production, autophagy levels are restored in lepromatous patients when the reversal reaction episode is established, which in turn help to reduce the bacillary load in skin cells (35). Therefore, leprosy lesion skin cells can earn a dual benefit from the use of autophagy as a platform for drug development; both inflammasome and antimicrobial optimal activities can be reached by modulating autophagy to a certain level. However, some bacterial pathogens inhibit autophagosome maturation and promote bacterial replication, such as M. tuberculosis (88,89). Given the background, Silva and colleagues (35) demonstrated that live but not dead M. leprae can inhibit the autophagic flux in macrophages, which indicates a requirement for an active mycobacterial ESX-1 secretion system. The ESX-1 secretion system is also involved in the targeting of M. marinum by LC3; however, ubiquitination does not seem to be necessary for this process (83). Legionella pneumophila and Coxiella burnetii also developed strategies to explore or subvert autophagy (88). Kim and colleagues (42) demonstrated that M. abscessus (UC22rough variant) induces autophagy and inhibits autophagic flow in murine macrophages. Also as observed, the lipid components of the clinical isolate UC22, which is highly virulent, play a critical role in the formation of the autophagosome. These data suggest that virulent M. abscessus can survive and grow within autophagosomes, preventing autophagosome-lysosome fusion and clearance from cells (42). A study demonstrates the role of lactoferrin, an antimicrobial peptide, in the autophagy of macrophages infected with M. avium. D-lactoferrin inhibits intracellular growth of M. avium and, at the same time, leads to structural changes in infected macrophages leading to increased lysosomal content and increased numbers of autophagic vesicles (90).
P-aminosalicylate, one of oldest drugs used against tuberculosis, inhibits the assimilation of iron (91). Depletion of iron is strongly associated with increased expression and accumulation of regulated in DNA damage and development 1 (REDD1), which inhibits mTOR activation, decrease phosphorylation of Akt and TSC2 (tuberous sclerosis complex 2) (92, 93). Iron depletion was also shown to increase the activation of HIF-1a (hypoxia-inducible factor) and AMPK and induce autophagy (92,94).
Zinc has been shown to be a positive regulator of autophagy in several different cell types and conditions, increasing the production of ROS, the formation and turnover of autophagosomes and cellular clearance (95)(96)(97)(98)(99)(100)(101). Nevertheless, zinc depletion was found to induce non-selective autophagy in yeast to release zinc recycled from zinc-rich proteins (91,102,103), demonstrating the key role of autophagy on zinc homeostasis. Zinc chelation was found to arrest autophagy and impair lysosomal acidification (95,104). Phosphorylation of ERK1/2 is necessary for the regulation of zinc-induced autophagy by either activating the Beclin 1-PI3K complex or by promoting disassembly of mTOR complex but the mechanisms in which zinc modulates autophagy are still not completely understood (95,99,105). Uncoupling of autophagy and zinc homeostasis in the airway epithelial cells was demonstrated to be a fundamental mechanism in the pathogenesis of chronic obstructive pulmonary disease (106). In TB, previous studies have shown that zinc levels in the peripheral blood decrease with age and during active disease but are improved after the beginning of treatment with anti-TB drugs (107)(108)(109)(110)(111). Oral zinc supplementation in Brazilian children exposed to adults with pulmonary TB was demonstrated to increase the positivity of tuberculin test (PPD) and induration size, decreasing false negative results (112). It is postulated that zinc supplementation could correct asymptomatic zinc deficiencies, improve the effect of autophagymediated therapy in TB, as well as giving a booster to immunity (109,111,112). There are currently several studies associating autophagy and infection by bacteria, including studies showing the different strategies developed by bacteria to inhibit the host's autophagic responses (113)(114)(115)(116)(117), as well as studies that show that the activation of autophagy by starvation or by treatment with rapamycin restricts bacterial growth and is capable of improving cell resistance to infection (39,40,(118)(119)(120). The therapeutic benefit of pharmacological agents that can modulate autophagy must be considered since a diverse variety of pathogens using autophagic machinery has been described in their favor. It is primary to understand whether the pathogen exploits this pathway as a whole (systemically) or just part of components to increase its intracellular replication and/or survival. Besides, it is necessary to consider whether the drug will act on all autophagic pathways or only on a specific component, which may, or may not, be used to replicate for the pathogen. For example, intracellular Brucella abortus (B. abortus) survives by promoting the formation of vacuoles containing B. abortus, which requires the activity of the autophagy initiation proteins PIK3C3 (phosphatidylinositol 3-kinase catalytic subunit type 3), ULK1 (serine/threonine-protein kinase ULK1), ATG (autophagy-related protein) 14L (Barkor; Beclin 1-associated autophagy-related key regulator), and Beclin 1, but not the autophagy activity stretching proteins ATG16L1, ATG4B, ATG5, ATG7 and LC3-II (121). In this condition, the use of inhibitors of the autophagy protein conjugation systems or inhibitors of autophagosome maturation would not have a protective effect against the survival of this bacterium. Still in this context, it is important to consider those patients who are affected by infections (for example, TB) that can be eliminated if autophagy is regulated positively, but who are co-infected with pathogens that use the autophagic pathway in their favor, such as concomitant infections with the Hepatitis B virus and HIV (122). Under other conditions, the co-infected patient is favored by autophagic activation, as is the case of patients with cystic fibrosis (CF) who are treated with cysteamine. The autophagic stimulus mediated by cysteamine in macrophages of cystic fibrosis (with the CFTRdel506 mutation) patients favors the clearance of Pseudomonas aeruginosa, a bacterium that frequently infects the lungs of CF patients (123). Therefore, it is primary to understand the differences between each stimulus, pathogen, and the type of cell under study so that the use of this route as a target for the development of antimycobacterial drugs can be advanced.

TREATMENTS INDUCING AUTOPHAGY DURING TUBERCULOUS MYCOBACTERIAL INFECTION
When autophagy studies were started, the only drug that was able to chronically induce this pathway was rapamycin. There is evidence of its antimycobacterial activity, where it has been observed that it significantly inhibits infection by M. kansasii, M. avium, Bacillus Calmette-Gueŕin (BCG), and virulent strains of M. tuberculosis (124,125). However, the adverse effects of rapamycin (which were not associated with autophagy induction) made this drug unattractive for use. Several drugs are capable of inducing autophagy and treating mycobacterial diseases, some examples are summarized in Table 1 and their activities are illustrated in Figure 1.
Among the various drugs described in the literature with proautophagic properties, ambroxol (126), metformin (127), verapamil (143), carbamazepine (128,129), valproic acid (129,130), and loperamide (130) are already approved for clinical use in different pathologies. The strategy of using drugs with a known safety profile for new indications related to autophagy is attractive because they do not need to undergo a complete toxicological assessment (18,147,148).
Regarding the pro-autophagic property of ambroxol, it has been shown to potentiate the antimicrobial activity of rifampicin in the murine model in trials for TB (126). The antidiabetic drug metformin reduced the intracellular growth of M. tuberculosis in a manner dependent on AMPK. Also, metformin was able to induce reactive mitochondrial oxygen species and facilitate phagosome-lysosome fusion (127). However, a more recent study failed to show the improvement in the bacterial activity of antituberculosis drugs by metformin in the murine model (149). This data makes us reflect on the importance of considering whether the anti-TB drug may or may not alter the pharmacokinetics of the repositioning drug. The use of rifampicin in this more recent study (149) may have altered the pharmacokinetics of metformin. Besides, it is also prudent to pay attention to the differences in the experimental design carried out to assess the effectiveness of the therapy, which can be combined (149) or used alone (monotherapy) (127).
Initial studies that evaluated the effect of verapamil and its analogs on macrophages infected with M. tuberculosis showed that the structural analog KSV21 had an additive effect on the inhibitory antimicrobial activity of Isoniazid and Rifampicin (143). In addition, the antibiotics isoniazid and pyrazinamide, two first-line cocktail drugs used to treat TB, exert their antimycobacterial activity through autophagy (40).
Recently, the impact of linezolid and bedaquiline on the intramacrophagic behavior of M. tuberculosis has been reported. It was observed that the anti-Mtb effect of these new drugs occurred via activation of autophagy and increased formation of autolysosomes in infected macrophages (131). Bedaquiline induces metabolic stress in M. tuberculosis, which results in the accumulation of NADH (nicotinamide adenine dinucleotide), followed by the generation of ROS (subsequently generating ROS by the bacteria) (150). Although not directly proven, ROS can trigger autophagy activation and be responsible for antibioticinduced death of M. tuberculosis (151).
Resveratrol has also been studied for its antioxidant effect and its role in inducing autophagy. Regarding the antioxidant effect, resveratrol can increase the activity of antioxidant enzymes and works by eliminating free radicals (152,153). Resveratrol has inhibitory activity on the mTOR molecule (133,154). Other studies have shown antibacterial properties, mainly activity against gram-positive bacteria, flavonoid, and resveratrol (132). Still, on drugs capable of stimulating the autophagic death of M. tuberculosis, the anticonvulsant drug carbamazepine was able  to induce autophagy in mice infected with the multidrugresistant M. tuberculosis strain, resulting in a decrease in their bacterial load and improvement in pulmonary pathology (129). It was observed that carbamazepine induces antimicrobial autophagy due to decreased levels of Myoinositol (by blocking myoinositol uptake) into a pathway independent of mTOR. Furthermore, it was seen that this drug also activates AMPK (128). In that same study, the group described the induction of autophagy by the drug valproic acid, another anticonvulsant drug (129), which favored the increase in the co-localization of LC3 with M. tuberculosis, an effect similar to that observed after treatment with anti-diarrhea medication loperamide (130). Unlike carbamazepine, which activates AMPK, the induction of autophagy by baicalin in macrophages infected by M. tuberculosis occurred through inhibition of the PI3K/Akt/ mTOR pathway. Additionally, baicalin showed a suppressive effect on the activation of the NLRP3 inflammasome via PI3K/ Akt/NF-kB (nuclear factor-kB), as well as reduced the levels of the pro-inflammatory cytokine IL-1b (134). Both the induction of autophagy and the inhibition of NF-kB contribute to limit the activation of the NLRP3 inflammasome. Autophagy can limit the activation of the inflammasome indirectly or directly. Indirectly, it can reduce endogenous stimuli that favor the activation of the inflammasome (155,156) and can directly inhibit the autophagic degradation of inflammasome components (87,157).
Fluvastatin is a statin class drug currently used to treat hypercholesteromia and prevent cardiovascular disease, by blocking the enzyme hydroxy-methyl-glutaryl-CoA (HMG-CoA) reductase, which catalyzes a key step in cholesterol synthesis. Fluvastatin was demonstrated to be effective in targeting not only the mycobacteria but also increasing the ability of the host cells to eliminate M. tuberculosis infection (158). Other statins, including simvastatin and rosuvastatin were also demonstrated to control M. tuberculosis infection by promoting phagosomal maturation and autophagy (145).
Some studies demonstrated the protective role of autophagy in excessive inflammation during M. tuberculosis infection (159). Based on these studies, we conclude that autophagy plays an important role in the fight against TB, by direct killing of the pathogen, while also avoiding excessive inflammatory damage. This makes an antimycobacterial agent that has autophagy as a pharmacological target, a promising candidate to assist in therapy directed at the host.

ROLE OF AUTOPHAGY IN THERAPEUTIC APPROACHES FOR NTMS AND SKIN DISEASES
The treatment of nontuberculous mycobacteriosis is not very rewarding. Currently, the proposed therapeutic regimen for infection with NTMs is based on the use of macrolides (clarithromycin or azithromycin), ethambutol, and rifamycins (160). Azithromycin is a potent antibiotic and is often prescribed for prophylaxis and treatment regimens of mycobacterial infections (10). However, one study reported that long-term use of azithromycin by adults with CF increased the risk of infection with M. abscessus. That was observed because the therapeutic dosage of azithromycin impaired autophagic degradation (135). That is, these data emphasize the importance of autophagy in the host's response to infection by NTMs.
The challenge of treating lung diseases caused by M. abscessus is related to antibiotic resistance, including all firstline drugs for anti-TB treatment (161,162). Even rifampicin, which has bactericidal activity against M. tuberculosis and M. leprae, has low potency against M. abscessus. Although rifampicin is part of the treatment regimens established for M. kansasii and Mycobacterium avium complex infections, it is not recommended against M. abscessus (163,164). Recently, rifabutin (of the rifamycin group) was shown, through its bactericidal activity, to be effective against strains of clinical isolates from the three subspecies of the M. abscessus complex (subsp. abscessus, subsp. massiliense, and subsp. bolletii) (136). Recently, the in vitro activity of omadacycline and tigecycline against clinical isolates of M. abscessus, M. chelonae and M. fortuitum was evaluated (146). Omadacycline, a new tetracycline analog, approved for the treatment of acute bacterial skin and skin structure infections (ABSSSI) (165) showed activity against the three clinical isolates (146). There are reports that these microorganisms have been identified in postoperative infections caused by mycobacteria, including the three opportunistic pathogens: M. fortuitum (166), M. abscessus (167) and M. chelonae (168). Postoperative infections have been reported after orthopedic, laparoscopic, ophthalmic procedures and cosmetic operations (mainly liposuction, abdominoplasty, rhinoplasty) (169,170). M. chelonae can cause localized skin infection after being accidentally inoculated from the environment (pedicure beds, water heaters, and tattoo parlors) (171,172). In immunocompromised patients, the infection caused by this mycobacterium can manifest itself as a disseminated skin disease. A case report demonstrated M. chelonae skin and soft tissue infection in a patient with chronic lymphocytic leukemia (LLC) who was using ibrutinib, an oral drug, which acts by inhibiting Bruton tyrosine kinase (BTK) for the treatment of various malignant B-cell diseases (173,174). After 6 months of therapy with ibrutinib, the 85-year-old man developed skin lesions on his arms and legs (175). Fiocari and colleagues (176) showed that ibrutinib promotes an M2 phenotype by modifying the function of macrophages/ monocytes in the LLC. Taken together, these results showed that ibrutinib can have detrimental consequences on the microbicidal response in patients treated with ibrutinib. On the other hand, a more current study reported the impact of the drug ibrutinib on the intra-macrophagic behavior of M. tuberculosis. It was observed that the anti-TB effect of this medication occurred via activation of autophagy and facilitates phagosome-lysosome fusion in infected macrophages (142).
Nitazoxanide has also been studied for its role in inducing autophagy. The use of nitazoxanide in C57BL/6 mice infected with M. leprae showed a bactericidal action similar to that of rifampicin, an antibiotic used in the therapeutic regimen against leprosy (137). Based on this study, nitazoxanide (NTZ) may be an effective option for the treatment of leprosy (137).
The epidermis is composed mainly by keratinocytes, which contributes to the defense responses against various stimuli in the environment (177). Numerous findings indicate that autophagy plays an important role in the biology and pathology of keratinocytes (177). It has already been seen that calcipotriol, a vitamin D analog, has the ability to induce autophagy in keratinocytes (178). Analogous vitamin D molecules have been used to treat different skin diseases, such as psoriasis, lamellar ichthyosis and epidermolytic hyperkeratosis (179). The autophagic pathway converges with the vitamin D3cathelicidin pathway, which is preferably seen in the paucibacillary form of leprosy (140,141). Vitamin D3 induces autophagy via cathelicidin in macrophages infected with M. tuberculosis, with cathelicidin being required for IFNgmediated antimicrobial activity (180,181). Also  (182). Vitamin D3 has been used successfully in the treatment of patients with TB (183). Vitamin D3 could be one of the components for the treatment of leprosy and other chronic infectious diseases in which the cellular immune response is unregulated (184,185). Vitamin D prevents tissue damage through the negative regulation of perforin, granzyme B and granulisine in cytotoxic T lymphocytes (186).
Many species of mycobacteria that cause skin infections are considered to have a natural ability to acquire resistance to antibiotics and to have a significant reduction in sensitivity to antibiotics, which makes treatment efficacy more difficult by increasing failure rates (187,188). Thus, using therapies directed at the host, such as those that induce autophagy, to inhibit bacterial cell release and form biofilms or bacterial media can increase the effectiveness of currently available antibiotics, i.e. azithromycin (135) and verapamil (143,189) already mentioned in the text, as well as, Carvacrol (190)(191)(192)(193), Tetracycline (146,194,195), Thioridazine (196)(197)(198)(199) and, Mefloquine (200,201).

CONCLUSION
This review describes the potential of host cell autophagy as a target for the development of new strategies against mycobacterial diseases. There are few studies focusing on skin cell autophagy during mycobacterial infections but in this review we summarized autophagy mechanisms in some cells most relevant to skin mycobacterial diseases. In addition, drug repurposing presents itself as a promising perspective in the control of infections caused by mycobacteria, being used in isolation or complementary to existing treatments. Some challenges still need to be faced, such as the understanding of the mechanisms used by different species of mycobacteria to induce autophagy, the evaluation of host cell autophagy by different clinical isolates, including resistant strains, the impact of a therapy directed at the host cell in cases where there is coinfection and, finally, if the use of a drug in combination with current therapeutic regimens will have a beneficial effect on bacillary load.