PPARs are activated by endogenous and exogenous compounds. For instance, eicosanoids and long-chain fatty acids (LCFAs) are the endogenous ligands for PPAR-α and PPAR-β/δ [10,11]. PPAR-γ is activated by metabolites of arachidonic acid, such as 5-oxo-eicosatetraenoic acid (5-oxo-ETE) and 5-oxo-15(S)-hydroxyeicosatetraenoic acids (5-oxo-15(S)-HETE) [12,13], in addition to oxidized low-density lipoprotein (oxLDL) derivatives [14]. Several exogenous compounds are highly specific activators and modulators for mammalian PPAR subtypes: PPAR-α by the hypolipidemic drugs clofibrate and fenofibrate and the synthetic ligand Wy-14643 and PPAR-γ by the thiazolidinedione (TZD) group of antidiabetic drugs (including rosiglitazone, ciglitazone, troglitazone, and pioglitazone) [15]. GW501516, GW0742, and bezafibrate are highly selective PPAR-β/δ agonists, while GW1929 and GW2090 are specific PPAR-γ activators [16].

PPARs are also activated by mycobacterial infection; however, the bacterial component(s) responsible are not well understood. Organisms that naturally produce unsaturated fatty acids at the C10 position are relatively rare in nature, while several mycobacteria species, including M. vaccae, are able to accomplish this desaturation [17,18,19,20]. The mycobacteria-derived 10 (Z)-hexadecenoic acid upregulates genes in the PPAR signaling pathway and represses the proinflammatory cytokines in macrophages [21]. Furthermore, 10 (Z)-hexadecenoic acid and monoacylglycerol (MAG), which contains 10 (Z)-hexadecenoic acid, both activate PPAR-α but have no effect on PPAR-γ or PPAR-δ. The observed effects are blocked by PPAR-α antagonists and absent in PPAR-α-deficient mice. Recently, we found that PPAR-γ and PPAR-δ are activated in M. leprae-infected macrophages [22]. Infection with a recombinant strain of M. bovis BCG that produces phenolic glycolipid-1 (PGL-1) of M. leprae activates PPAR-γ in primary cultures of human Schwann cells [23].

Mannose-capped lipoarabinomannan (ManLAM) is present in the members of the M. tuberculosis complex, which interact with the mannose receptor (MR) in alveolar macrophages (AMs). High levels of PPAR-γ are expressed in activated AMs and macrophage-derived foam cells [24,25]. ManLAM upregulates PPAR-γ expression in human macrophages, consistent with M. tuberculosis infection. Furthermore, activation by ManLAM is suppressed by MR siRNA. These results indicate that the activation of PPAR-γ by M. tuberculosis is due to the interaction between its cell wall component ManLAM and host MRs.

Several molecules are known to bind to PPARs, including polyunsaturated fatty acids (PUFAs), such as certain ω3-PUFAs (e.g., docosahexaenoic acid with C22:6 and α-linolenic acid with C18:3) and certain ω6-PUFAs (e.g., arachidonic acid with C20:4 and linoleic acid with C18:2) [26,27]. Saturated fatty acids, such as stearic acid with C18:0 and myristic acid with C14:0, also bind to PPAR-α. M. leprae cell wall lipids also contain mycolic acids, other types of LCFAs typical for mycobacteria, such as alpha-mycolic acids and ketomycolic acids [28]. However, whether or not this lipid could be a ligand for PPARs is not known.

Mycobacterial infection induces lipid droplet formation in macrophages. These lipids are essential for mycobacterial survival and are presumed to be a carbon source. In several different models, M. tuberculosis has been shown to use accumulated lipids as a carbon source at various stages of the infectious process [49,50,51,52]. M. tuberculosis-induced lipid droplets in macrophages primarily contain cholesterol esters and TAG. The cholesterol is transported through the bacterial cell membrane by Mce4, a bacterial lipid transporter required for cholesterol import and its utilization [53,54]. Many of the active compounds that limit M. tuberculosis growth in macrophages have been found to inhibit cholesterol-related processes, indicating that cholesterol is central to M. tuberculosis infection [55]. Fatty acids are also an abundant lipid in human granulomas [56]. Although it has been thought that M. tuberculosis assimilates and metabolizes fatty acids, recent genome sequencing has identified many putative fatty acid β-oxidation genes [57].

Since M. leprae has lost the mce4 operon, M. leprae seems to use cholesterol oxidase (ML1492) in order to convert cholesterol to cholestenone for survival [58]. In leprosy skin tissue sections, M. leprae-containing histiocytes and Schwann cells are filled with cholesterol [59,60]. This has been confirmed with the observation of cholesterol accumulation in M. leprae-infected primary macrophage [60,61]. Furthermore, the expression of cholesterol synthase, HMG-CoA reductase, was increased following infection, and when de novo cholesterol synthesis was inhibited by lovastatin, viability of M. leprae was reduced [61].

Conversely, high-performance thin-layer chromatography (HPTLC) analysis demonstrates that TAG is the main component of the lipid in M. leprae-infected human monocytic THP-1 cells [62]. It has been reported in Schwann cells that M. leprae infection enhances glucose uptake and stimulates the pentose phosphate pathway, which is required for TAG synthesis [63]. The accumulated TAGs are maintained by the enhanced expression of adipose differentiation-related protein (ADRP) and perilipin and by the reduced expression of hormone-sensitive lipase (HSL), which contributes to lipid degradation [64,65]. Glycerol-3-phosphate acyltransferase 3 (GPAT3) is an important rate-limiting enzyme for TAG synthesis [66]; accordingly, the internalization and viability of bacilli are lower in GPAT3 knockout cells [62]. Furthermore, clofazimine, a therapeutic agent for leprosy, reduces the accumulation of lipid in M. leprae-infected THP-1 cells and promotes the production of interferon (IFN)-β and IFN-γ [67]. Therefore, mycobacterial viability is hypothesized to be closely related to lipid metabolism in host cells, especially the accumulation of TAG and cholesterol.

A recent study demonstrated that PPAR-mediated lipid metabolism is a key process in foamy cell formation following M. leprae infection. Among PPARs, the involvement of PPAR-γ in mycobacterial infections has been studied. Infection with M. tuberculosis modulates homeostasis of host lipid and induces foamy macrophages, which is necessary for intracellular parasitization and growth [68,69]. The virulent H37Rv strain of M. tuberculosis induces PPAR-γ expression [25], while attenuated M. bovis BCG slightly upregulates PPAR-γ [25,70]. In vitro interference with PPAR-γ signaling in M. tuberculosis-infected macrophages decreases intracellular lipid accumulation and increases mycobacterium killing [71]. Pretreatment with a PPAR-γ antagonist significantly suppressed mycobacterial (M. bovis BCG and M. tuberculosis) induction of intracellular lipid droplet accumulation [70,71,72]. In addition, M. tuberculosis growth was attenuated in human lung macrophages after PPAR-γ deletion or isolation from PPAR-γ-deficient mice. Taken together, these data suggest that PPAR-γ is required for foam cell formation in tuberculous granulomas, which is related to bacilli survival.

Recently, in M. leprae-infected THP-1 cells, we reported that the increased expression of PPAR-γ and PPAR-δ coincided with the induction of intracellular lipid droplet formation [22]. Further, the expression of the PPAR-γ target genes ADRP, scavenger receptor CD36, fatty acid-binding protein 4 (FABP4), and apolipoprotein C-1 (APOC1) were significantly increased. Activation of the PPAR-γ signaling pathway is responsible for the upregulation of Gpat3 expression during adipocyte differentiation [73,74,75]. We also found that GPAT3 expression is induced in THP-1 cells infected with M. leprae, suggesting that the mechanism of intracellular TAG accumulation is triggered by PPAR-γ activation [62].

The expression of CD36, an essential receptor for LDL-C incorporation, is also induced by M. tuberculosis through PPAR-γ in THP-1 macrophages [71]. CD36 can interact with surfactant in the lungs and promote the proliferation of M. tuberculosis in human macrophages in vitro [76]. CD36 directly interacts with TLR2 in macrophages infected with M. bovis BCG, as demonstrated by co-immunoprecipitation [77]. The neutralization of CD36 subsequently decreases PPAR-γ expression and lipid droplet formation and prostaglandin E2 (PGE2) secretion. In addition, M. tuberculosis upregulates the expression of GLUT1 and GLUT3 on the cell membrane by PPAR-γ activation of glucose metabolism. Its activation is suppressed by the PPAR-γ inhibitor T0070907 but enhanced by the agonist pioglitazone [78]. These data suggest that the activation of PPAR-γ promotes cholesterol and TAG uptake, both of which are components of the lipid droplets in mycobacteria-infected macrophages. Cholesterol accumulation in infected macrophages reduces cell wall permeability to rifampin, one of the first-line antituberculosis drugs, and masks surface antigens of mycobacteria [79]. Thus, lipids also play a role in drug resistance.

On the other hand, PPAR-α is known to promote the metabolism of lipids accumulated in M. tuberculosis-infected macrophages and suppress lipid droplet formation. Following infection with M. tuberculosis, PPAR-α^-/- bone marrow-derived macrophages decrease the activation of the transcription factor EB (TFEB), a responsible factor for the regulation of autophagy, and increase lipid droplet formation. Conversely, PPAR-α activation significantly reduces the amount of lipid droplets in mycobacteria-infected macrophages, suggesting that PPAR-α promotes lipid catabolism in mycobacterial infection [80]. Thus, PPAR-α and PPAR-γ may have opposed roles in the host defense during mycobacterial infection.