Role of apoptosis and autophagy in tuberculosis

Abstract

Tuberculosis (TB) is one of the oldest known human diseases and is transmitted by the bacteria Mycobacterium tuberculosis (Mtb). TB has a rich history with evidence of TB infections dating back to 5,800 b.c. TB is unique in its ability to remain latent in an individual for decades, with the possibility of later reactivation, causing widespread systemic symptoms. Currently, it is estimated that more than one-third of the world’s population (~2 billion people) are infected with Mtb. Prolonged periods of therapy and complexity of treatment regimens, especially in active infection, have led to poor compliance in patients being treated for TB. Therefore, it is vitally important to have a thorough knowledge of the pathophysiology of Mtb to understand the disease progression, as well as to develop novel diagnostic tests and treatments. Alveolar macrophages represent both the primary host cell and the first line of defense against the Mtb infection. Apoptosis and autophagy of macrophages play a vital role in the pathogenesis and also in the host defense against Mtb. This review will outline the role of these two cellular processes in defense against Mtb with particular emphasis on innate immunity and explore developing therapies aimed at altering host responses to the disease.

tuberculosis (TB), transmitted by the bacteria Mycobacterium tuberculosis (Mtb), is a global epidemic and 1 of the top 10 causes of mortality worldwide. In 2015, the disease accounted for 1.4 million deaths, surpassing the mortality of other epidemic diseases, such as human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) and malaria (144). The disease burden of TB extends beyond mortality, also contributing to high levels of disability-adjusted life years (DALYs) in many regions (38). In fact, estimates show that Mtb surpasses malaria for DALYS due to illness (65 million DALYs for TB vs. 50 million DALYs for malaria; 144). Commonly regarded as a disease of the poor and disenfranchised, TB is more common in developing countries, with up to 40% of some populations being carriers of the bacteria (101). In 2015, there were 10.4 million new cases of TB, with 6 countries accounting for 60% of these cases: India, Indonesia, China, Nigeria, Pakistan, and South Africa (144).

Individuals infected with Mtb commonly present with pulmonary or generalized complaints, including cough, chest pain, fatigue, and unintentional weight loss (65). Though the lungs are the most common organs affected, the bacteria can disseminate intravascularly and seed multiple organs and structures within the body. Immunocompromised individuals are especially susceptible to disseminated disease and have higher rates of mortality associated with infection (85). Furthermore, disseminated disease can present as a diagnostic challenge, as disease manifestation can vary among different individuals (104, 110, 135).

The prolonged incubation of the bacteria, along with their ability to remain latent in an individual for extended periods of time, makes disease eradication almost impossible. In fact, it is estimated that >90% of TB infections are latent (152). Because of this, advances in clinical diagnosis are of paramount importance in controlling the spread of disease. Latent infections are commonly tested by the purified protein derivative test or the IFN-γ release assay (3). Both assays test the cell-mediated immune response and have comparable sensitivities (86). In active infection, diagnostic recommendations are chest radiograph followed by three sputum specimens for both acid-fast staining and culture (23). In addition, the Centers for Disease Control and Prevention now recommends nucleic acid amplification of respiratory specimens for all individuals in which the diagnosis of Mtb is considered. However, if clinical suspicion for infection is high, empiric therapy may be initiated even before the presence of a positive test.

Although the presence of TB dates back thousands of years (20, 151), efficacious treatments have only been present for the last half-century. Whereas initial treatments were crude and involved isolating patients in sanatoriums (42), management now consists of long-term antibiotics. Streptomycin, the first effective treatment for Mtb, was developed in 1946, and many other antibiotics were developed shortly thereafter (42). Isoniazid and pyrazinamide, discovered in the 1950s, are still in use today. Isoniazid therapy for 9 mo is the current preferred therapy for latent Mtb infection (83). For uncomplicated active infections, the treatment of choice consists of using four drugs: rifampin, isoniazid, pyrazinamide, and ethambutol (RIPE therapy) for 2 mo, followed by rifampicin and isoniazid for another 4 mo (144).

Prolonged periods of therapy and complicated treatment regimens have led to poor compliance in patients being treated for TB. The World Health Organization estimates that ~4% of new cases of Mtb are already multidrug resistant and only half of patients who receive treatment for Mtb are considered cured on follow-up (144). Poor compliance, along with inappropriate use of antimicrobials, has effectively paved the way for the development of drug-resistant, multidrug-resistant (MDR-TB), and extensively drug-resistant tuberculosis (10). Treatments for these cases vary depending on patterns of drug resistance and, in some cases, may require surgical resection (29). With these limitations in diagnosis and treatment, a fundamental understanding of the pathophysiology of TB is necessary to develop both novel diagnostic tools as well as therapeutics.

General Pathogenesis of Tuberculosis

M. tuberculosis (Mtb) is a nonmotile, facultative intracellular bacteria of macrophages. Entry into resident alveolar macrophages following exposure is an absolute prerequisite for Mtb to cause lung infection in a susceptible host. Experiments have demonstrated that the depletion of resident alveolar macrophages in mice using liposome-encapsulated, dichloromethylene diphosphonate protects mice from Mtb infection (68) but increases the susceptibility of mice to other infections, such as Streptococcus pneumoniae (60). These results suggest that although alveolar macrophages protect the host against typical extracellular bacterial pathogens, they facilitate the establishment of Mtb at least during the initial stages of infection. Once Mtb comes in contact with the alveolar macrophages, they are readily engulfed by the granulocytes using a variety of phagocytic receptors. The complement receptors (C receptors), mannose receptors, surfactant protein-A (SP-A), cluster of differentiation 14 (CD14) receptor, and scavenger receptors are some of the important receptors involved in the adhesion and entry of the Mtb in the macrophages (27, 106, 139). C receptors are important for bacterial opsonization (a process by which a pathogen is marked for ingestion and eliminated by phagocytes; 117). Mannose receptors assist the macrophages in phagocytosing unopsonized bacteria (117). SP-A acts as an opsonin and may also modulate the activity of other receptors to enhance macrophage binding and uptake of Mtb (32). This initial uptake of bacteria and subsequent intracellular growth inside alveolar macrophages establish the infection within the host. However, late in the infection, Mtb can exist as an extracellular microbe in the necrotic cavities of lung parenchyma that connect to airways, which provide an oxygen-rich environment and facilitate aerosol transmission between hosts. Although resident alveolar macrophages are the primary cells involved in the initial uptake of Mtb, during the later stages of infection, dendritic cells and monocyte-derived macrophages also participate in the phagocytic process (41, 131). The interaction between Mtb and these antigen-presenting cells ultimately activates the macrophages and stimulates the production of cytokines and chemokines, such as tumor necrosis factor-α (TNF-α), interleukin (IL)-1B, IL-6, IL-12, IL-15, IL-18, and interferon (IFN)-γ (132). These cytokines further stimulate macrophages and play a crucial role in the inflammatory response and the outcome of mycobacterial infection.

The pathogenesis of TB is dependent on the delicate interplay between the mechanisms involved in host defense and the various survival strategies employed by Mtb (115). Alveolar macrophages can facilitate both innate and adaptive defense mechanisms to combat Mtb infection. The mechanisms involved in the development of adaptive immunity to Mtb require a close interaction between CD4⁺ T cells and dendritic cells. First, the major histocompatibility complex (MHC) class II molecules on dendritic cells present processed Mtb antigens to CD4⁺ T cells. Subsequent macrophage activation by type 1 T helper (Th1) cells involves the production of IFN-γ by Th1 cells and the interaction of its T-cell receptor and CD4 molecule with Mtb peptide-MHC class II complex on macrophages in the presence of costimulation by CD40 ligand (CD40L) on the Th1 cell and CD40 on the macrophage. The dendritic cells, which are able to uptake the infected apoptotic macrophages and extracellular Mtb, stimulate CD8⁺ T cells by cross-presenting Mtb antigen via MHC class I molecules. Then, the MHC class I molecules, expressed on all nucleated cells, present mycobacterial proteins to activated, antigen-specific CD8⁺ T cells and are eliminated. The stimulation of T cells is facilitated by the costimulatory signals B7.1 (CD80) and B7.2 (CD86), which are expressed on macrophages and dendritic cells and bind to CD28 on naive T cells. In addition, several cytokines produced by the macrophages and dendritic cells, such as IL-2, IL-12, IL-18, and IL-23, help in the stimulation of T lymphocytes (98). The strategies employed by Mtb to circumvent adaptive immunity include 1) increasing the production of anti-inflammatory cytokines or decreasing the expression of proinflammatory cytokines, leading to the dampening of T-cell stimulation and reducing the expression of antigen-presenting molecules in macrophages (33, 103); and 2) attenuating the expression of the costimulatory molecule, CD80 (111), and impairing adequate costimulatory signaling leading to T-cell anergy and apoptosis (43, 44). These considerations, taken together, undoubtedly demonstrate that the interactions between Mtb and host macrophages play a central role in the pathogenesis of TB.

Macrophages can also destroy the microbes directly, without requiring the activation of the adaptive T-cell response. In this regard, the active metabolite of vitamin D, 1,25-dihydroxyvitamin D, has been shown to assist macrophages in attenuating the growth of Mtb (22, 107). Interestingly, vitamin D deficiency has been reported in some patients with TB (21) and in susceptible populations at higher risk of being infected with TB (141). Inside the activated macrophages, the putative mechanisms involved in the killing of Mtb include the generation of reactive oxygen and nitrogen species. In an experimental model of Mtb infection, it was determined that mice that lack the ability to produce superoxide suffered from increased early outgrowth of mycobacteria (19). However, the increased susceptibility to TB is not seen in patients with chronic granulomatous disease, a disease associated with the defective production of reactive oxygen species (143). The precise role of reactive nitrogen species in TB is also debatable and remains to be elucidated, although the production of nitric oxide by the macrophages in response to mycobacterial infection has been shown to be bactericidal (108). In addition, natural resistance-associated macrophage protein (Nramp1), an integral membrane protein expressed exclusively in macrophage/monocytes and polymorphonuclear leukocytes, influences the rate of intracellular replication of Mtb in macrophages (9). Nramp1 belongs to a family of transporters that facilitate the cellular entry of metal ions such as Fe²⁺, which ultimately induce macrophages to produce antimicrobial toxic radicals (153). Experiments have shown that the addition of small quantities of iron to resident alveolar macrophages stimulates their antimicrobial activity and that this effect is abrogated by the addition of antioxidants, catalase, or mannitol, suggesting that the Fe(II)-mediated inhibition of mycobacterial growth is mediated by the Fenton/Haber-Weiss reaction and free hydroxyl radicals (153). Interestingly, studies from West Africa have shown that a functional polymorphism in the promoter region of Nramp1 is associated with reduced gene expression and increased susceptibility to TB (8, 11). However, further mechanistic studies in animal models and humans are required to clearly delineate the role of this protein in the pathogenesis of TB. Apoptosis and autophagy are other effector mechanisms that the infected host uses to limit the outgrowth of Mtb and will be discussed in detail below.

Apoptosis and Endoplasmic Reticulum Stress in Tuberculosis

The endoplasmic reticulum (ER) participates in the modification and proper folding of newly synthesized proteins. ER is also the major storage site of intracellular Ca²⁺ and synthesis of lipids. Cellular infection or stress induces a loss of Ca²⁺ from the ER and an increase in the intracellular redox state, resulting in impaired folding and subsequent accumulation of misfolded proteins in the ER. The aggregation of misfolded proteins ultimately results in ER stress (ERS; 15). This pathological event triggers an essential cellular response called the unfolded protein response (UPR; 94), which halts early protein synthesis to attenuate the accumulation of unfolded or misfolded proteins in the ER, restoring normal cellular function. However, prolonged or uncontrolled ERS activates downstream signaling pathways that push the cell toward apoptosis. There are three major ER-localized signaling pathways involved in ERS: inositol-requiring-1α (IRE1α), double-stranded RNA-dependent protein kinase (PKR)-like ER kinase (PERK), and activating transcription factor 6 (ATF6; 148). The pathway for ERS-induced apoptosis also includes CCAAT/enhancer-binding protein (C/EBP) homologous protein (CHOP); glucose-regulated protein-78 (GRP78/BiP), a major UPR target protein that targets misfolded proteins for degradation (100); and phosphorylated alpha subunit of eukaryotic initiation factor 2 (eIF2α), which inhibits protein synthesis. The role of ERS in promoting the apoptosis (90, 129) of macrophages infected with Mtb, and thereby limiting the spread of infection, is discussed below.

Necrosis and apoptosis are the two known mechanisms by which cells die. Necrosis is caused by accidental and irreversible damage to the plasma membrane, which destroys the integrity of the cell, whereas apoptosis is a programmed cell death associated with changes in nuclear organization and chromatin fragmentation. In apoptosis, the cytoplasm and other cellular components never leave the dying cell, and hence apoptosis is immunologically silent (125). Although apoptosis in microbial infections causes tissue damage, the initiation of apoptosis may be beneficial to the host, as it promotes the removal of the microorganisms (125). Apoptosis plays a vital role in host defenses against intracellular pathogens, including Mtb, by preventing the release of intracellular pathogens and the spread of mycobacterial infection. The apoptosis of macrophages can limit mycobacterial infection by activating both innate and adaptive immune response (79, 102). Apoptotic bodies containing cell cytoplasm, other cellular organelles, and microorganisms are taken up by macrophages and dendritic cells via receptor-mediated phagocytosis and then undergo degradation and are presented via MHC class II complexes. This process is defined as efferocytosis and may promote both innate and adaptive immunity. There is currently intense investigation and debate over the mechanisms and functional consequences of macrophage apoptosis in TB. Recent studies have shown that alveolar macrophages undergo apoptosis as an innate defense response against Mtb (7). In an experiment conducted to show the role of apoptosis and necrosis in the viability of mycobacteria, researchers exposed peripheral blood monocytes infected with replicating and viable M. bovis bacillus Calmette-Guérin (BCG) to the cellular inducers of either necrosis (hydrogen peroxide) or apoptosis (adenosine triphosphate). The treatment killed the infected monocytes. However, the necrosis of monocytes had no effect on the microbial viability, but the apoptosis of monocytes reduced the mycobacterial viability (88). Therefore, it is evident that macrophage death by either apoptosis or necrosis appears to have drastically different outcomes for the course of infection. Apoptotic cell death of Mtb-infected macrophages is associated with mycobacterial killing (64, 77, 88, 95) and stimulation of T-cell responses via antigen presentation (116, 142). In contrast, necrotic cell death of macrophages aggravates infection by allowing the release of viable mycobacteria for subsequent reinfection (67, 88, 126). Although some recent work has demonstrated that necrosis, too, can follow a programmed series of events, a beneficial role of necrosis in Mtb infection has yet to be found (13). In TB, necrosis may also be the primary pathological event in granuloma formation, which causes liquefaction, cavitation, and tissue damage, resulting in massive expansion of bacillary numbers (78).

Almost all mycobacteria can induce apoptosis. Mildly virulent mycobacterial strains, such as BCG, and the avirulent strain, Mtb H37Ra, are better inducers of macrophage apoptosis compared with the more virulent mycobacteria, such as Mtb H37Rv (31, 55, 56, 63, 105). In fact, studies have shown that virulent Mtb may not elicit macrophage apoptosis, but rather induce macrophage necrosis to avoid bacterial killing (66). These virulent mycobacteria have effective mechanisms to evade apoptosis and survive. To escape apoptosis, virulent Mtb manipulate the ultrastructural architecture of the macrophage ER. Macrophages infected with the virulent strain, H37Rv, have predominantly rough ER (RER) compared with macrophages infected with the avirulent strain, H37Ra, which have a smooth ER phenotype. The functional consequences of this phenotype change result in increased cytosolic Ca²⁺ levels and the simultaneous induction of phosphatidyl choline/phosphatidyl ethanolamine (PC/PE) expression in the macrophages infected with the H37Ra strain, which facilitates apoptosis. However, in macrophages infected with the H37Rv strain, cholesterol homeostasis is disturbed, resulting in the inhibition of apoptosis and sustained infection (114).

Furthermore, the secretion of TNF-α, a proinflammatory cytokine, in an autocrine/paracrine manner has been shown to play a significant role in the apoptosis of macrophages (87). Interestingly, TNF-α levels are significantly higher at the site of disease in patients with TB (5). TNF-knockout mice, when infected with Mtb, display increased lung infection, which can be ameliorated by the administration of recombinant TNF-α (52). TNF-α also plays an important role in the containment of latent infection in granulomata (87). The administration of mice with MP6-XT22, a monoclonal antibody that neutralizes TNF-α, results in the fatal reactivation of TB, increased tissue bacillary burden, squamous metaplasia, and fluid accumulation in the alveolar spaces (87). TNF-α mediates its proapoptotic activity through its receptors 1 and 2 (TNFR1 and TNFR2). TNFR1 (55), the FAS receptor (FasR), also known as apoptosis antigen 1 (APO-1 or APT) or tumor necrosis factor receptor superfamily member 6 (TNFRSF6), and the FAS ligand (FasL) play important roles in the induction of apoptosis. The binding of FasL to FasR results in signal transduction, leading to the apoptosis (92) of human macrophages infected with Mtb (95). Strangely, the level of TNF-α production induced by the Mtb avirulent strain, H37Ra, is comparable to that induced by the virulent strain, H37Rv, even though apoptosis is higher in H37Ra (4, 55). It was, however, observed that virulent Mtb employs tactics to combat TNF-α-induced apoptosis. First, H37Rv induces the release of the soluble TNF-α receptor (TNFR2), which forms inactive TNF-α-TNFR2 complexes that account for the decrease in TNF-α bioactivity (4). The release of soluble TNFR2 by H37Rv-infected macrophages is dependent on IL-10 production (4). Thus the pathogenic strains of Mtb may selectively induce IL-10, leading to decreased TNF-α activity and subdued macrophage apoptosis. Second, virulently infected macrophages also show reduced susceptibility to FasL-induced apoptosis, correlating with a reduced level of FasR expression (95). Finally, the virulent Mtb strains may interfere with TNF-α signaling by upregulating the expression of the antiapoptotic MCL1 gene, a member of the B-cell lymphoma (BCL2) gene family (123).

ER stress response plays a significant role in limiting the survival of Mtb by inducing apoptosis (62). Mtb induces ER stress in macrophages by disrupting Ca²⁺ homeostasis and affecting macrophage polarization. Depending on the stimuli, macrophages can acquire a distinct phenotype during the process of polarization. M1 and M2 are the well-described phenotypes, also referred to as classically or alternatively activated macrophages, respectively. Classical activation is stimulated by microbial products and results in M1 macrophages characterized by high antigen presentation and high production of IL-1, IL-6, IL-12, IL-23, TNF-α, nitric oxide, and reactive oxygen intermediates (133). These M1 macrophages increase inflammatory response and mediate resistance against intracellular microbes (75). In contrast, alternative/M2 activation restrains inflammation and upregulates surface molecules, such as dectin-1, mannose receptor, and scavenger receptors A and B1 (35, 81). M2 macrophages also produce high levels of IL-10 (89). Interestingly, macrophages infected with attenuated Mtb strain, H37Ra, display a M1 phenotype, whereas the M2 phenotype is dominant in macrophages infected with the virulent Mtb, H37Rv (74). The mannose receptor and scavenger receptors on M2 macrophages further facilitate H37Rv entry, and IL-10 inhibits macrophage apoptosis, as mentioned earlier. ER stress stimulates M1 polarization of macrophages and facilitates Mtb removal, suggesting that ER stress may be an important component of the host immune response to Mtb (74).

Mycobacteria are composed of several virulence factors, such as the 38-kDa antigen (Ag; Rv0934) (113), 19-kDa Ag (p19; Rv3763) (17), early secreted antigenic target of 6 kDa (ESAT-6; Rv3875) (14), heparin-binding hemagglutinin Ag (HBHA; Rv0475) (15), and PE_PGRS33 (Rv1818c) (6). All of these factors have been shown to cause ER stress and affect major apoptogenic factors on host cells. The 19-kDa Mtb glycolipoprotein, p19, shown to be both cell wall associated and secreted, is capable of inducing apoptosis in macrophages in both a dose- and time-dependent manner (77). This effect of p19 is mediated by Toll-like receptor 2 (TLR2), as the incubation of cells with anti-TLR2 monoclonal antibody inhibited p-19-induced apoptosis. López et al. also established that the viability of Mtb was significantly reduced in cells undergoing apoptosis induced by p19 (77). Additionally, the early secreted mycobacterial antigen, ESAT-6, increases intracellular Ca²⁺ concentration, reactive species accumulation, and ER stress-induced apoptosis in macrophages (14). Furthermore, the 38-kDa Ag has been demonstrated to increase the expression of ER molecular chaperones, including CHOP and BiP, and phosphorylated eIF2α in bone marrow-derived macrophages and to induce apoptosis via the activation of caspase-12, caspase-9, and caspase-3 (73). Recently, Sohn et al. demonstrated that the Mtb HBHA, a virulence factor and diagnostic antigen involved in extrapulmonary dissemination of Mtb, effectively induces apoptosis in macrophages. HBHA increased reactive oxygen species production, DNA fragmentation, nuclear condensation, caspase activation, and poly (ADP-ribose) polymerase cleavage in apoptotic macrophages (124). Furthermore, Choi et al. demonstrated that HBHA stimulation induced the ER stress sensor molecule, CHOP, in a caspase-dependent manner. In addition, enhanced reactive oxygen species production and elevated cytosolic Ca²⁺ levels are essential for HBHA-induced ER stress responses (15). In fact, Lim et al. showed that infection with M. kansasii also induced ER stress-mediated apoptosis of macrophages. They found that the apoptosis was associated with calpain activation, as the inhibition of calpain prevented the induction of CHOP and BiP in infected macrophages (71). The exogenous silencing of CHOP expression enhanced the survival of Mtb, whereas increasing ER stress reduced the rate of Mtb survival (72). In fact, calpain activation has also been associated with the development of pleural fibrosis during Mtb infection (145). It is, however, important to note that although initially Mtb increases ERS, the bacteria gradually decrease ERS over time in an attempt to survive within macrophages (72), suggesting that ERS plays an important role in limiting Mtb infection (Fig. 1).

Fig. 1.

Role of apoptosis in mycobacterial tuberculosis (Mtb) infection. The major pathophysiological events that lead to either induction or inhibition of apoptosis in macrophages by the avirulent Mtb strain (H37Ra) and virulent Mtb strain (H37Rv) are summarized. Macrophages infected with H37Ra strain attain M1 phenotype, have predominantly smooth endoplasmic reticulum (ER), and have high levels of ER stress, resulting in increased apoptosis and Mtb death. However, macrophages infected with H37Rv strain attain M2 phenotype, have predominantly rough ER, and have comparatively lower levels of ER stress, resulting in the attenuation of apoptosis and improved Mtb survival.

The role of ERS and apoptosis is particularly evident at the site of a pulmonary tubercle or granuloma formation in Mtb infection. In immunocompetent hosts, the control of TB infection occurs during granuloma formation. Mtb granulomata are highly cellular and are characterized by differentiated myeloid cells surrounded by lymphocytes (13). These macrophage-rich areas of the lung that are infected with virulent Mtb show increased levels of ERS markers, such as CHOP, ATF3, and phosphorylated Ire1α and eIF2α (119). Tubercles, characteristic of active Mtb disease, are distinguished from other granulomata by the extent of necrosis, which leads to liquefaction and cavitation at the necrotic center. This necrotic center causes the hallmark symptomatology of tissue destruction and bloody sputum in pulmonary TB. Interestingly, ERS markers and apoptotic cells are more abundant in the granulomatous tissue surrounding the centralized areas of caseation (119). Therefore, ERS-induced macrophage apoptosis is beneficial in protecting the host from Mtb infection and may limit the dissemination of the bacteria in advanced granulomata (140).

Summary of apoptosis and endoplasmic reticulum stress in tuberculosis.

First, macrophage apoptosis is beneficial to the host as it clears the intracellular bacteria and activates the host innate and adaptive immune response; however, macrophage necrosis is detrimental to the host as it allows the expansion of Mtb numbers and subsequent tissue damage. Second, macrophages can differentiate into M1 and M2 phenotypes: M1 macrophages promote inflammatory responses and increase apoptosis, whereas M2 macrophages attenuate inflammation and decrease apoptosis. Third, ER stress pushes macrophages to the M1 subtype and apoptosis and thereby facilitates Mtb clearance. Fourth, virulent Mtb selectively induce the M2 phenotype in macrophages to suppress apoptosis and killing by the host. Fifth, strategies employed by virulent strains of Mtb to evade macrophage apoptosis include 1) abundance of RER and impaired cholesterol homeostasis, 2) induction of IL-10 and subsequent decrease in TNF-α-mediated apoptotosis, 3) reduced FasR expression and susceptibility to FasL-induced apoptosis, and 4) upregulation of the antiapoptotic MCL1 gene. Finally, several virulence factors of Mtb (38-kDa antigen, p19, ESAT-6, HBHA, and PE_PGRS33) are known to induce macrophage apoptosis in an ER stress-dependent manner.

Autophagy in Tuberculosis

Autophagy is upregulated in several lung pathologies (16, 45, 70, 149, 150). The major difference between apoptosis and autophagy is that during apoptosis the characteristic cell changes (blebbing, cell shrinkage, nuclear fragmentation, chromatin condensation, and chromosomal DNA fragmentation) finally lead to cell death. However, during autophagy, cells degrade dysfunctional and unnecessary cellular components, a process that may or may not lead to cell death. Oxidative stress and other environmental stressors commonly induce autophagy (2, 50). Autophagy prolongs cell survival by turning over cellular constituents and preserving cellular homeostasis (50). During autophagy, these cellular constituents, such as proteins, lipids, and organelles, are sequestered into double-membrane vesicles called autophagosomes that are transported to endosomes or lysosomes to become autophagolysosomes. Within the autophagolysosomes, hydrolase enzymes digest the autophagic components to their basic elements (i.e., amino acids and fatty acids) to be reused for cell building and energy generation. Autophagy maintains organelle quality control by disposing of dysfunctional or damaged cellular organelles. Autophagy of ribosomes, peroxisomes, and mitochondria is termed “ribophagy,” “pexophagy,” and “mitophagy,” respectively.

Autophagy is upregulated by the activation of adenosine 5′-monophosphate (AMP)-activated protein kinase (AMPK) and by the inhibition of the mechanistic target of rapamycin (mTOR) pathway (109). Two ubiquitin-like conjugation systems, microtubule-associated protein-1 light chain 3 (LC3) and autophagy protein 5 (ATG5)-ATG12, are required for the formation of the double-membrane autophagosomes (28, 51, 128). The autophagosome formation can be detected by measuring the conversion of LC3-I (unconjugated cytosolic form) to LC3-II (autophagosomal membrane-associated phosphatidylethanolamine-conjugated form; 28, 51, 128). Additionally, the SQSTM1/p62 protein, also known as sequestosome-1, recognizes cellular components marked for degradation and targets them for autophagy by the virtue of its ubiquitin association domain (UBA) and a LC3-interacting region (LIR; 46). Impairment of autophagy results in the buildup of p62 and cellular dysfunction (47, 61). Autophagy is also upregulated by the association of an autophagy-related protein, beclin 1, and VPS34, a class III phosphatidylinositol-3-kinase (PI3KC3; 40, 53). Beclin 1 can also influence apoptosis and necrosis of cells through its interactions with the B-cell lymphoma-2 (BCL2) family of proteins, BCL2-associated X protein (BAX) and BCL2-antagonistic/killer (BAK; 53, 54, 138).

Xenophagy, the process in which autophagy eliminates bacteria by degradation, was first reported in Mtb (39). It was shown that the physiological stimulation of autophagy in macrophages caused mycobacterial phagosomes to mature into phagolysosomes and led to the suppression of intracellular survival of Mtb (39). During Mtb infection, the increased production of IFN-γ activates macrophages to induce autophagy and traffic Mtb to lysosomes for degradation (25). Induction of autophagy by IFN-γ is associated with protective immunity against TB (25). In addition, studies have shown the role of several autophagy factors, such as ATG5, ATG12, ATG16L1, p62, nuclear dot protein 52 (NDP52), beclin 1 (BECN1), and LC3 (25, 48, 112, 120, 136), in the pathogenesis of infection with Mtb and other infections, where stimulation of autophagy has been shown to increase bacterial killing (39, 112, 134, 146), whereas the suppression of autophagy increases Mtb survival (25, 39, 112, 136).

Transcriptional upregulation of autophagy by Mtb is a complex phenomenon requiring the involvement of numerous genes in multiple steps of autophagy, including the biogenesis of the phagophore, autophagosome and lysosome fusion, and the targeting/degradation of substrate (121). Transcription factor EB (TFEB) is a critical regulator of genes involved in autophagy activation, assembly, and target degradation (121). Recently, Kim et al. identified the role of the nuclear receptor peroxisome proliferator-activated receptor-α (PPARα) in host defense against Mtb and BCG infections via the upregulation of TFEB (59). PPARα binds to the PPAR response elements in the promoter region of numerous target genes (80) and regulates key cellular activities, such as mitochondrial and peroxisomal function and energy metabolism (57, 118). The authors demonstrated that PPARα is essential for host antimycobacterial responses, as the deficiency of PPARα increased the bacterial load and exaggerated the inflammatory responses to mycobacterial infection (59). Interestingly, this immune modulator role of PPARα was mediated by the upregulation of Lamp2, Rab7, and Tfeb, genes involved in autophagy and lysosomal biogenesis (59). These results suggest that further studies are required to explore the possibility of using PPARα agonists in anti-TB management.

The process of elimination of intracellular bacteria by autophagy can involve LC3-associated phagocytosis, the sequestration of phagosomes within autophagosomes, but is largely dependent on targeting cytosolic bacteria for ubiquitination. Mtb evades its destruction in phagolysosomes and can translocate from the phagosome into the cytosol of macrophages and dendritic cells, where it is susceptible to ubiquitination. During autophagy, ubiquitination of intracellular bacteria recruits several key proteins that mediate bacterial delivery to autophagosomes for degradation. Although the process of Mtb ubiquitination remains largely unclear, several key proteins have been shown in recent years to be involved in the likely mechanisms of ubiquitin-dependent autophagic clearance. First, the autophagy adaptors, such as p62, NDP52, NBR1, and optineurin (OPTN), are recruited to the ubiquitin-associated bacteria and then are attached to LC3 to target the intracellular bacteria to autophagosomes for degradation (34). Parkin, an E3 ubiquitin ligase, targets Mtb and Mtb-associated structures to autophagosomes and promotes autophagy-mediated host resistance to TB (82). Ubiquilin-1 (UBQLN1), a member of a protein family that contains an ubiquitin-like domain, an ubiquitin-associated domain, and stress-inducible protein-1 (STI1) motifs, recruits ubiquitin, p62, and LC3 to Mtb-containing vacuoles and hampers bacterial growth and replication (112). Another such protein, Smurf1, which is an E3 ubiquitin ligase, catalyzes the ubiquitination of substrates for subsequent proteasomal degradation. Recently, Franco et al. demonstrated that macrophages lacking Smurf1 are unable to recruit polyubiquitin, the proteasome, ubiquitin-binding autophagy adaptor NBR1, autophagy protein LC3, and lysosomal marker lysosomal associated membrane protein-1 (LAMP1) to Mtb-associated structures (30). They further found that the mice lacking the Smurf1 gene had higher loads of Mtb, exaggerated lung inflammation, and poorer survival than the control mice (30).

Nevertheless, mycobacteria have evolved several mechanisms to inhibit, modulate, or exploit the autophagy response of the host. Certain mycobacteria, such as M. avium complex (MAC), have evolved mechanisms to evade both apoptosis and autophagy, leaving the bacteria free to infect nearby macrophages in the process of spreading (26). Furthermore, the virulent Mtb H37Rv has been known to strategically upregulate IL-6 production to combat innate immunity (25). The production of IL-6 by the Mtb-infected macrophages selectively inhibits IFN-γ-induced autophagosome biogenesis by attenuating the ATG12-ATG5 complex (91). In fact, infection with Mtb time-dependently increases IL-6 production, and the neutralization of IL-6 by an anti-IL-6 antibody significantly enhances the IFN-γ-mediated killing of the intracellular bacteria (25).

In addition, the virulent Mtb H37Rv possesses a unique gene called the “enhanced intracellular survival” (Eis) gene. The protein product of Eis is a unique 42-kDa protein that has been shown to enhance the survival of Mtb during repeated passages through human macrophages (137). Shin et al. demonstrated that the Mtb EIS modulates macrophage autophagy, inflammation, and cell death (122). They found that the macrophages infected with an Mtb Eis deletion-mutant H37Rv (Mtb-Δeis) had markedly increased autophagic vacuoles and autophagosome formation (122). Duan et al. showed that EIS inhibits autophagy in macrophages by upregulating IL-10 gene expression and increasing Akt/mTOR/p70S6K pathway activity (24). They further found that EIS increased the acetylation of H3 histones in the IL-10 gene promoter sequence, thereby increasing IL-10 expression (24). Furthermore, Mtb also inhibits Ras-related protein Rab-7a, an endosomal marker that is also involved in the maturation of Mtb-containing autophagosomes into autolysosomes; thus Mtb selectively modulates autophagy flux in macrophages (12).

Mtb can live in macrophages and evade immune attack by also regulating microRNAs (miRNAs). miRNAs are a class of noncoding small single-strand RNA molecules (~22 nucleotides in length) that play a critical role in macrophage function. miRNAs bind to the 3′ untranslated region of targeted mRNAs and regulate gene expression posttranscriptionally (127). It is well known that miRNAs regulate immune responses through TLR signaling (97), play a critical role in autophagy (147), and regulate host immunity during Mtb infection (84). In a recent publication, Gu et al. demonstrated that the infection of macrophages by Mtb upregulated the expression of miR-23a-5p in a time- and dose-dependent manner (36). Furthermore, they found that the increased expression of miR-23a-5p increased Mtb survival, whereas the inhibition of miR-23a-5p attenuated Mtb survival in the cells (36). Interestingly, miR-23a-5p overexpression prevented the activation of autophagy by the Mtb through TLR2-mediated signaling, suggesting an important role of miR-23a-5p in influencing the innate immune response by inhibiting autophagy and enhancing Mtb survival (36). In another study, Guo et al. showed that BCG infection of macrophages increased the expression of miR-20a, which inhibits autophagic processes by targeting ATG7 and ATG16L1 mRNAs (37). miR-20a also decreased the levels of LC3-II in macrophages and promoted BCG survival (37). In addition, it was found that by inducing miR-33, Mtb inhibited integrated pathways involving autophagy, lysosomal function, and fatty acid oxidation to support bacterial survival and replication (99).

The clinical relevance of autophagy in Mtb infection has been investigated by Li et al. in a recent publication. The authors examined the relationship of autophagy induction by clinically isolated Mtb strains with the disease outcomes of patients with TB (69). They collected sputum, urine, or cerebrospinal fluid samples from 185 patients and found that most of the clinical isolates of Mtb were able to induce autophagosome formation in macrophages; however, the autophagy-inducing ability varied significantly among different isolates (69). They further deduced that the patients infected by Mtb with poor autophagy-inducing ability had more severe disease and worse outcomes (69). The study also suggests that through intimate and persistent interactions with its human host, Mtb has evolved counterstrategies to negate the host antibacterial effects of autophagy. Collectively, it is evident that autophagy represents an important immunologically regulated process in Mtb infection (Fig. 2), and a better understanding of the process will have greater implications in the management of TB infection.

Fig. 2.

Strategies to inhibit autophagy by mycobacterial tuberculosis (Mtb). The major strategies adopted by Mtb to combat bacterial death by autophagy are summarized. Mtb has evolved techniques to evade killing by autophagy, allowing unchecked bacterial growth. Macrophages infected with Mtb have upregulated IL-6 production, which selectively inhibits IFN-γ-induced autophagosome biogenesis. In addition, Mtb possess the “enhanced intracellular survival” (Eis) gene, which attenuates autophagy and improves Mtb survival. Finally, Mtb upregulates the microRNAs miR-23a-5p, miR-20a, and miR-33, which prevent the activation of autophagy by Mtb. Currently, there is not enough data in the literature to compare the role of autophagy in avirulent and virulent Mtb strains, as mentioned in Fig. 1 for apoptosis.

Summary of autophagy in tuberculosis.

First, autophagy is a homeostatic lysosomal process that involves the degradation of cellular components to their basic elements. Second, stimulation of autophagy increases Mtb killing, whereas suppression of autophagy increases bacterial survival in macrophages. Third, impaired autophagy is associated with severe disease and poor outcomes in humans infected with Mtb. Fourth, the upregulation of Lamp2, Rab7, and Tfeb genes by PPARα plays a significant role in autophagy and lysosomal biogenesis and host antimycobacterial responses. Fifth, increased production of IFN-γ by Mtb activates macrophages to induce autophagy and traffic Mtb to lysosomes for degradation. This process requires a family of proteins including LC3, LAMP1, and Smurf1. Finally, the strategies incorporated by Mtb to evade autophagy include 1) upregulation of the Eis gene, 2) inhibition of IFN-γ by IL-6 produced by Mtb, 3) inhibition of Rab-7a-dependent maturation of Mtb-containing autophagosomes into autolysosomes, and 4) upregulation of specific microRNAs that attenuate TLR signaling and autophagy.

Potential Therapeutics and Concluding Remarks

A nonnaturally occurring type I IFN, called IFN alfacon-1 or Infergen (IFG), has been recently approved against the hepatitis C virus (18, 130). Importantly, IFG has been recently shown to enhance the maturation and activation of macrophages (58). IFG improved bactericidal activity through autophagy and attenuated the survival of Mtb in human macrophages (58). Therefore, immunotherapy with IFG may provide a novel adjuvant treatment strategy in targeting autophagy in Mtb infection (58). Similarly, PPARα agonists, which regulate energy homeostasis and inflammation, are currently approved to be used therapeutically for cholesterol and triglyceride disorders. As mentioned in the review above, these PPARα agonists promote autophagy, lysosomal biogenesis, phagosomal maturation, and antimicrobial defense against Mtb and BCG (59) and therefore can be used in the future to manage TB. Furthermore, loperamide, an antidiarrheal drug, which increases the degradation of long-lived cellular proteins (76), has been shown to induce autophagy in macrophages and reduce Mtb growth and burden (49). Additionally, it was shown that vitamin D supplementation in patients with pulmonary TB without cavitation could enhance autophagy in macrophages and improve innate immune function and therefore may help to control the intracellular growth of mycobacteria (1). Interestingly, it was also noticed that adjunct therapy employing CD40 and TLR4 agonists (C40.T4) significantly enhanced the bactericidal potency of anti-TB drugs by increasing autophagy (58), thus providing a novel therapeutic avenue to target drug resistance in anti-TB therapy.

In conclusion, despite advances in understanding the disease, TB remains a global health challenge and a major cause of morbidity and mortality throughout the world. Rising rates of drug resistance coupled with frequently inappropriate treatment regimens have impeded efforts to control TB worldwide. Therefore, a better understanding of Mtb pathology and the development of novel therapies based on that knowledge are urgently needed. In this regard, targeting molecular pathways controlling apoptosis and autophagy (93, 96) may play a significant role in combating TB.

GRANTS

This work was supported by the Countermeasures Against Chemical Threats (CounterACT) program, National Institutes of Health Office of the Director, National Institute of Neurological Disorders and Stroke, and National Institute of Environmental Health Sciences, Grants 5U01-ES-026458-02 and 1U01-ES-027697-01.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

S.A. prepared figures; A.L., R.P., C.M.G., V.S., and S.A. drafted manuscript; A.L., L.A.R., V.S., and S.A. edited and revised manuscript; V.S. and S.A. approved final version of manuscript; S.A. conceived and designed research.