As mentioned, ASA modifies COX-2 activity allowing 15-epi-LXA4 and AT-RvD1–4 production. Indeed, ASA-triggered resolving lipids are more stable than the endogenous molecules and could serve as anti-inflammatory drugs. Another fascinating group of drugs, statins, can also increase the production of pro-resolving lipids, in a way that is similar to the process mediated by ASA. Just as the acetylation of COX-2 by ASA causes a metabolic switch in COX-2, statins produce the same metabolic switch by nitrosylation. Recently, Serhan’s group reported the presence of another subtype of Rvs, 13-series Rvs, which provide protective activity against Escherichia coli infections. These lipids are produced by COX-2 in a process involving a neutrophil–endothelium interaction, and their production can also be triggered by nitrosylation induced by atorvastatin (Dalli et al., 2015). Thus, the change in COX-2 activity explains, at least in part, the anti-inflammatory properties of ASA and statins.

It is remarkable how the same enzymatic array participates in the generation of both inflammatory and resolving mediators. COX and LO activities, which are responsible for the production of prostaglandins (PGs) and leukotrienes, respectively, can switch to the production of LX, Rvs, PDs, and maresins. Indeed, AA derivatives, PGs E₂ (PGE₂), prostacyclin (PGs I2, PGI₂), and leukotriene B₄ (LTB₄), participate in leukocyte recruitment to the damaged site. However, as acute inflammation progresses, a metabolic switch occurs, and LX synthesis begins. The exact moment when this switch happens is unclear; however, the dampening of the inflammatory signals may be part of the input needed to promote the metabolic changes (Gilroy and De Maeyer, 2015). These switches are accomplished by transcriptional or posttranslational modifications, which involve PGE₂ and PGs D₂ (PGD₂) (Frolov et al., 2013; Croasdell et al., 2015). COX-mediated production of PGD₂ by human PGD₂ synthase (hPGD₂s) activates the PGD₂ receptor, DP₁ (a G-protein coupled receptor [GPCR]), which in turn stimulates the production of IL-10 (an anti-inflammatory cytokine), which then blocks the path to chronic inflammation. PGD₂ can also be converted to PGs J2 (PGJ₂) and 15-deoxy-Δ(12,14)-PGs J₂ (15-D-PGJ₂) products that activate peroxisome proliferator-activated receptor (PPAR)-γ to promote resolution (Croasdell et al., 2015).

The actions of LXA₄ are mediated by a GPCR called formyl peptide receptor 2/lipoxin A₄ receptor (FPR2/ALXR), via several signaling pathways, including the p38/mitogen-activated protein kinase (MAPK)-activated protein kinase (APK)/heat shock protein 27 (HSP27), c-Jun N-terminal kinase (JNK), and phosphatidylinositol 3-kinase (PI3K) pathways (Cooray et al., 2013), depending on the cell type. In monocytes and macrophages, LXA₄ triggers the synthesis of IL-10 (Souza et al., 2007), a cytokine responsible for driving the resolution of inflammation, and enhances non-phlogistic phagocytosis of apoptotic cells (Maderna et al., 2010). Also, some effects of LXA₄ are related to its ability to activate the cytosolic aryl hydrocarbon receptor (AhR), inducing the expression of suppressor of cytokine signaling (SOCS) (Machado et al., 2006).

The action of Resolvin E1 (RvE1) is intricate because it is an agonist of the seven-pass transmembrane GPCR ChemR23. Naturally, the activation of ChemR23 by low concentrations of chemerin, its natural agonist, favors chemoattraction of monocytes/macrophages and immature dendritic cells (DCs). Activation of ChemR23 increases intracellular calcium release and inhibits cAMP and MAPK extracellular signal-regulated protein kinases 1/2 (ERK1/2)-mediated signaling, by Gi/o protein (a G protein subtype) recruitment. Consequently, it leads to up-regulation of the PI3K/Akt signaling pathway and down-regulation of nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) (Mariani and Roncucci, 2015). RvE1, via a different set of G proteins, stimulates Akt phosphorylation and ribosomal protein S6 kinase (Ohira et al., 2010). Also, RvE1 is a partial agonist of the LTB₄ receptor 1 (BLT1) and competes with LTB₄ for binding (Arita et al., 2007).

RvD1 can activate FRP2/ALXR and the G protein-coupled receptor 32 (GPR32) (Krishnamoorthy et al., 2012). The activation of the FPR2/ALX receptor by RvD1 suppresses cytosolic calcium and decreases activation of the calcium-sensitive kinase calcium-calmodulin-dependent protein kinase II (CaMKII). CaMKII inhibition suppresses activation of p38 and the MAP kinase-APK 2 (MAPK-APK2), which reduces Ser²⁷¹ phosphorylation of 5-LO and shifts 5-LO from the nucleus to the cytoplasm (Fredman et al., 2014). Moreover, LXA₄ and RvE1 counter-regulate the LTB₄/LL-37 proinflammatory circuit, which is partially mediated by FPR2/ALX (Wan et al., 2011).

LXA₄ decreases polymorphonuclear leukocyte (PMN)-mediated tissue damage, angiogenesis, and PMN proliferation and adhesion, increasing non-phlogistic phagocytosis and IL-10 production (Chandrasekharan and Sharma-Walia, 2015). RvE1 increases PMN apoptosis, LXA₄ production, and microbial killing, and decreases IL-12 production, PMN transendothelial migration, and PMN infiltration; also, it inhibits NF-κB reporter gene activation and reduces organ fibrosis (Arita et al., 2007; Campbell et al., 2007; El Kebir et al., 2012). RvD1 decreases reactive oxygen species (ROS) generation and the production of pro-inflammatory cytokines. RvD1 also reduces organ fibrosis (Qu et al., 2012; Miyahara et al., 2013). RvD2 increases microbial killing and clearance, as well as the production of nitric oxide (NO•) and PGI₂ in endothelial cells (Spite et al., 2009; Chiang et al., 2012). PD1 decreases COX-2 expression, T-cell migration, and cytokine production, including that of TNF, IFN-γ, and microglial cell cytokines (Ariel et al., 2005; Qu et al., 2015). Also, PD1 has neuroprotective actions (Serhan et al., 2015). Maresins increase tissue regeneration, stimulating phagocytosis and the killing of oral pathogens by human leukocytes, and also promote the macrophage phenotype switch from M1 to M2 (Wang et al., 2015a). It is important to emphasize that both Rvs and LXs promote bacterial clearance; thus, they both constitute a pathway that impedes microorganism evasion of the immune response. Thus, the concert of the SPMs in the inflammatory broth give a change of helm heading toward a calmer environment, where leukocyte and macrophage hyperactivity decline, promoting a favorable environment for the repair.

Infections may progress toward chronicity due to dysregulation of the inflammatory process. Thus, extensive damage may occur involving irreversible structural alterations. Furthermore, it has been shown that in chronic inflammatory states such as Alzheimer’s disease (Wang et al., 2015b), periodontitis (Van Dyke, 2017), peripheral artery disease (Miyahara et al., 2013), or cystic fibrosis (Karp et al., 2004), the levels of pro-resolving factors such as LXA₄, or ATLs are decreased. Therefore, it is suggested that a deficit of resolutory ability promotes chronic inflammatory states. Evidence of the participation of lipid pro-resolving mediators in infectious diseases is abundant. For example, in experimental models of lung infections (e.g., disseminated influenza A), there is low expression of LX (Cilloniz et al., 2010), and exogenous administration of PD1 improves survival (Morita et al., 2013). Moreover, when inflammation persisted in a respiratory syncytial virus lung infection, and exogenous administration of LXs or Rvs reverses the inflammation (Shirey et al., 2014). In a self-resolving murine model of E. coli pneumonia, exogenous AT-RvD1 administration enhanced the clearance of E. coli and Pseudomonas aeruginosa in vivo, and lung macrophage phagocytosis ex vivo (Abdulnour et al., 2016). These findings were similar to the results of a lung coinfection model with Streptococcus pneumoniae and influenza A virus (Wang et al., 2017), which showed that AT-RvD1 decreased the inflammatory drive by acting on the FPR2/ALX receptor and antagonizing the effect of serum amyloid A, which is an agonist of this receptor. Thus, there is a clear relationship between bacterial or viral infection and SPMs generation.

However, the LXs-bacterial interplay in a chronic infection may be more complicated. In a mouse model of Mycobacterium tuberculosis infection, 5-LO knockout mice were more likely to survive compared with wild-type mice, which had a more protracted disease evolution. Administration of LXA₄ inhibited the development of a Th1 response, which is protective in the tuberculosis model (Bafica et al., 2005). Bacteria may also inhibit the resolution of infectious processes. A recent report demonstrated that chronic P. aeruginosa infections in cystic fibrosis patients promote the hydrolysis of an endogenous epoxide-containing eicosanoid 14,15-epoxyeicosatrienoic acid (14,15-EET) to its corresponding diol, thereby destroying the signal that triggers increased biosynthesis of 15-epi-LXA₄ and preventing the activation of resolution pathways that could help to eradicate the infection (Flitter et al., 2017).

Specialized pro-resolving mediators may contribute to eradicate infections and terminate the inflammatory input, as it is suggested by murine sepsis models, where there was reported that LXA₄ decreased plasma IL-6, chemokine (C–C motif) ligand 2 (CCL2), IL-10, and NF-κB activity in peritoneal macrophages, reduced neutrophil migration, and increased the clearance of bacteria by neutrophils without the production of excessive free radicals (Walker et al., 2011; Wu et al., 2015). Furthermore, LX may increase the production of antibacterial proteins. Bactericidal/permeability-increasing protein (BPI) is increased in an in vitro model of Salmonella typhimurium infection when LXs are administered (Canny et al., 2002). Thus, LX may have a role in the clearance of pathogenic microorganisms. Interestingly, LXA₄ decreases the release of the exotoxin pyocyanin by P. aeruginosa, reducing its pathogenicity (Wu et al., 2016). Thus, LXA₄ also affects infective agents. Furthermore, RvD1 and RvD5 were shown to decrease the dose of antibiotics needed to treat E. coli or Staphylococcus aureus infections (Chiang et al., 2012). Thus, RvD1 and RvD5 can help to clear bacteria and, most importantly, this research provides new insight into how to circumvent antibiotic resistance (Chiang et al., 2012).

In addition to its action in bacterial infections, LXA₄ plays a role in acute and chronic parasitic infections. However, it is controversial because in murine models of cerebral malaria, LXA₄ is associated with a lower parasite burden, less cerebral inflammation, and better survival (Shryock et al., 2013). However, in a murine model of Toxoplasma gondii infection, although LXA₄ production is increased and there is less cerebral inflammation (due to decreased IL-12 levels), the immune response against the parasite is dampened (Aliberti et al., 2002). Probably, this despair results are due to the distinct causative organism or by the inflammatory context where LXA₄ was produced.

Among the acute inflammatory mediators, AA derivatives have been described as essential drivers of the acute infection process and the chronic cardiac damage induced by the parasite. Experimental in vivo models of infection with T. cruzi showed an increase in the expression of COX-2 (Molina-Berrios et al., 2013a) and PGs E receptor 2 (EP₂) in cardiac tissue (Guerrero et al., 2015). These proteins are involved in the synthesis and activity of AA derivatives, such as PGE₂, TXA₂, PGF_2α, 6-oxo-PGF_1α, LTB₄, and other eicosanoids that have been observed increased after T. cruzi infection. Consequently, cells derived from T. cruzi-infected mice shows elevated levels of PGE₂ (Celentano et al., 1995; Borges et al., 1998; Freire-de-Lima et al., 2000; Guerrero et al., 2015). Also, PGE₂, TXA₂, PGI₂, PGF_2α, and LTB₄ levels have been found increased in plasma of mice infected with T. cruzi (Cardoni and Antunez, 2004; Molina-Berrios et al., 2013a,b; Sharma et al., 2013). Also, it is well-documented that T. cruzi itself can synthesize AA derivatives like TXA₂, PGE₂, and PGF_2α by the action of the T. cruzi old yellow enzyme (TcOYE) (Kubata et al., 2002; Ashton et al., 2007). There is evidence that both, host- and parasite-derived autacoids contribute importantly to cardiac remodeling. In fact, COX-2, PLA2γ, or 5-LO deficient mice, infected with T. cruzi, exhibited improved survival rate and reduced cardiac tissue inflammation when compared with wild-type mice (Celentano et al., 1995; Borges et al., 1998; Cardoni and Antunez, 2004). Furthermore, there is evidence indicating that T. cruzi induces the formation of lipid bodies, specialized organelles where PGs synthesis occurs, being an important strategy for parasite growth and survival (Melo et al., 2003; D’Avila et al., 2011). The effect of the host or parasite-derived AA derivatives in the acute infection onset has been reviewed extensively elsewhere (Machado et al., 2011).

An aspect of growing interest is the implication of the PPAR-γ pathway in the modulation of inflammatory processes in chronic infections (Kim et al., 2015). In this regard, this pathway is related to the decrease in the transcription of genes controlled by NF-κB and the activator protein-1 (AP-1). It is well known that 15-D-PGJ₂, which is a pro-resolving lipid derived from PGD₂, acts as an agonist of PPAR-γ (Paumi et al., 2004). Therefore, it may have a role in pro-resolving processes in conjunction with other SPMs. Consequently, 15-D-PGJ₂ has been tested as a modulator of acute and chronic heart inflammation in T. cruzi infection. 15-D-PGJ₂ (1 mg/kg) decreases the inflammatory infiltrate and amastigote nest count and significantly increases the IL-10 levels (Rodrigues et al., 2010). Also, 15-D-PGJ₂ attenuates acute liver damage induced by T. cruzi in mice. In this model, 15-D-PGJ₂ was able to decrease fibrosis and liver damage without influencing the course of the infection itself (Penas et al., 2016). Moreover, a preliminary report from the same group suggested that treatment with 15-D-PGJ₂ could regulate the number of intracellular amastigotes in the cardiac tissue via PPAR-γ-dependent and PPAR-γ-independent pathways (Penas et al., 2013).

Due to the role of AA derivatives in T. cruzi infection, inhibition of COX activity has been proposed as a strategy for controlling parasite-induced disease. Although several non-steroidal anti-inflammatory drugs (NSAIDs), such as indomethacin, meloxicam, and celecoxib, have been assayed in mouse models of T. cruzi infection (Freire-de-Lima et al., 2000; Michelin et al., 2005; Abdalla et al., 2008; Hideko Tatakihara et al., 2008), the most studied NSAID is ASA, also known as aspirin. Interestingly, the effect of ASA seems to be mouse-species dependent. When T. cruzi-resistant mice (C57BL/6 or CD-1) are treated with ASA, parasitemia and mortality increases (Michelin et al., 2005; Hideko Tatakihara et al., 2008; Mukherjee et al., 2011). On the other hand, Balb/c mice, which are sensitive to T. cruzi infection, become healthy when ASA is administrated (Freire-de-Lima et al., 2000; Paiva et al., 2007; Hideko Tatakihara et al., 2008). Importantly, C57BL/6 mice generate more NO• than Balb/c mice after T. cruzi infection, and ASA treatment induces NO• production in Balb/c mice infected with T. cruzi (Hideko Tatakihara et al., 2008). Furthermore, in murine macrophages infected with T. cruzi, the inhibition of NO• synthesis partially prevents the effect of ASA (Lopez-Munoz et al., 2010), suggesting an important role for this mediator in the ASA effect.

More importantly, the effect of ASA has an important relationship with its dose. Most studies investigating ASA have been performed using a fixed dose of 25–50 mg/kg. However, studies using higher doses (>75 mg/kg) have shown no effect, or they have shown that ASA aggravates the damage caused by intraperitoneal or oral T. cruzi infection (Molina-Berrios et al., 2013a; Cossentini et al., 2016). The fact that ASA has an antichagasic effect only at doses <50 mg/kg has been associated to the production of 15-epi-LXA₄, an ASA-triggered LX found in patients treated with low doses of ASA (Chiang et al., 2004). In Balb/c T. cruzi-infected mice, 25 and 50 mg/kg of ASA induced a significant increase in 15-epi-LXA₄ production without modification of LTB₄ levels. At these doses, the mice had prolonged survival, decreased mortality, and less cardiac inflammatory infiltrate. Additionally, the administration of 25 mg/kg exogenous 15-epi-LXA₄ significantly decreased the parasitemia peaks and cardiac parasite load, improved the survival of the infected mice, and partially reversed the detrimental effect of high-dose ASA (Molina-Berrios et al., 2013a). Low doses of ASA also improve the vascular reactivity of mice infected with T. cruzi. Molina-Berrios et al. (2013b) evaluated the effect of ASA at 2 and 40 mg/kg and found that both regimens decreased ECAM expression and the TXA₂ level. Also, 2 mg/kg/day ASA reduced the inflammatory infiltrate in mice hearts and improved the cardiac histology at 90 days post infection. Furthermore, in in vitro cardiac cells infected with T. cruzi, low doses of aspirin increased IL-1β and NO• release, and decreased transforming growth factor (TGF)-β release, and these effects disappeared when ASA concentrations were increased (Malvezi et al., 2014b).

Consistently, ATLs not only participate in the resolution of the damage produced by T. cruzi but also in the clearance of the pathogen. In an in vitro model of T. cruzi-infected macrophages, the addition of 0.3–1.25 mM of ASA significantly decreased the internalization of T. cruzi without altering macrophage viability (Carvalho de Freitas et al., 2017). Also, it has been reported that 0.625–2.5 mM of ASA decreased the internalization of T. cruzi and increased the release of IL-15 and NO•. However, co-administration of celecoxib (a COX-2 selective inhibitor) reverted the ASA effect and restored the invasive capacity of trypomastigotes, suggesting that functional COX-2 is necessary for the ASA effect. Furthermore, Boc-2 (a specific antagonist of the FRP2/ALX receptor) prevented the ASA effect, suggesting that the inhibition of invasion depends on the synthesis of 15-epi-LXA₄ (Malvezi et al., 2014a).

The effects of other SPMs triggered by ASA have been studied. Ogata et al. (2016) evaluated the effect of AT-RvD1 on peripheral mononuclear cells (PBMCs) from patients with Chagas heart disease at stage B1, that is, with few heart abnormalities. Stimulation of PBMCs with a T. cruzi-derived antigen increased the production of INF-γ, TNFα, IL-10, and IL-13, while the pre-treatment of PBMCs with AT-RvD1 (100 nM) significantly reduced the production of INF-γ, with no changes in TNFα, IL-10, and IL-13. As INF-γ polarizes the immune response to a type Th1 response, and a low level of IL-10 indicates loss of the Treg response, both phenomena could be associated with the development of heart damage in CD. Therefore, decreasing INF-γ using AT-RvD1 has a beneficial role in chagasic cardiomyopathy. Furthermore, AT-RvD1 was able to reduce the percentage of necrotic PBMCs and their proliferation after stimulation with a T. cruzi antigen (Ogata et al., 2016).

Recently, an elevated plasma level of RvD1 has been found in T. cruzi-infected CD-1 mice. Also, RvD1, RvD5, and RvE2 (but not LXs, maresins, or PDs) were found in the lysates of T. cruzi trypomastigotes. Interestingly, there were no Rvs found in the lysates of T. cruzi epimastigotes or other protozoan parasites such as T. gondii, suggesting that the trypomastigotes themselves synthesize these SPMs (Colas et al., 2018). However, there was no direct evidence for a metabolic switch involving TcOYE or a yet unknown enzymatic activity that could explain a T. cruzi origin of these SPMs. There is currently no evidence of a T. cruzi 5-LO enzyme or another enzyme for SPM synthesis.

15-epi-LXA₄ can also be produced by statins. Birnbaum et al. (2006) showed that statins stimulate 15-epi-LXA₄ release from myocardial tissue. This effect of statins over COX-2 could be mediated by the overexpression of inducible NO• synthase (iNOS), which in turns nitrosilates COX-2 at the Cys⁵²⁶ residue, giving COX-2 the ability to generate the 15R-HETE intermediate metabolite, which is cleaved by 5-LO to generate 15-epi-LXA₄ (Kim et al., 2005). This 15-epi-LXA₄ production also requires that 5-LO remains in the cytoplasm. Thus, 5-LO, phosphorylated by protein kinase A (PKA) is attached to the nuclear membrane, where it is committed to leukotriene synthesis (Ye et al., 2008).

In T. cruzi-infected human endothelial cells, simvastatin induces the synthesis of 15-epi-LXA₄ and decreases the ECAM expression induced by the parasite. This effect was reversed by the addition of AA-861 (a 5-LO inhibitor) and replicated when using exogenous 15-epi-LXA₄. Interestingly, 15-epi-LXA₄ inhibited NF-κB pathway activation, decreasing the phosphorylation of the NF-kappa-B inhibitor (IκB) and the IκB kinase (IKK), and preventing NF-κB p65 nuclear translocation. Thus, the action of 15-epi-LXA₄ in T. cruzi-infected endothelial cells involves the NF-κB signaling pathway (Campos-Estrada et al., 2015). In addition, in a murine model of CCC, 40 mg/kg/day simvastatin decreased endothelial activation, inflammatory infiltration, and fibrosis in heart tissue, an effect that persisted for a long time after treatment stopped. When simvastatin was administered to 5-LO-deficient mice, the anti-inflammatory effect was not observed unless exogenous 15-epi-LXA₄ was also administered. Thus, there is an association between simvastatin administration, 15-epi-LXA₄ production, and cardiac improvement. Furthermore, 15-epi-LXA₄ was still detectable 30 days post administration, suggesting that 15-epi-LXA₄ is stable in serum and is associated with the observed sustained effects (Gonzalez-Herrera et al., 2017).

Table 1 summarizes the experimental evidence showing the effect of pro-resolving mediators in T. cruzi infections in vivo. The findings support the idea that ASA and simvastatin have a positive impact on the resolution pathways, producing novel pro-resolving lipids such as 15-epi-LXA₄ or AT-RvD1, modulating the inflammatory response, and decreasing ECAMs, leukocyte recruitment, and inflammatory infiltration. Also, the SPMs increase the survival rate in animal models of CD.

Cutaneous leishmaniasis (CL) is the most common form of leishmaniasis around the world. CL is characterized by self-limiting skin lesions located in body areas where sandflies usually bite, such as the face, neck, and limbs. The disease progression varies between different world regions. The old-world CL (caused by Leishmania major or Leishmania tropica) progresses over weeks to months to form a dry ulcer, but healing occurs over several months or years, leaving a scar or depigmentation of the skin. On the other hand, the new-world CL (caused by Leishmania mexicana, Leishmania amazonensis, or several parasites from the Vianna subspecies, such as L. [V] braziliensis) produces a wet ulcer that is associated with lymphadenitis or lymphadenopathy and may involve mucosal manifestations. New-world leishmaniasis is also called tegumentary leishmaniasis (TL), which can be localized or disseminated (de Vries et al., 2015).

In patients with TL, there is a different pattern of expression of the genes involved in the AA cascade between patients with localized CL (LCL) and mucocutaneous leishmaniasis (MCL). Patients with MCL (the most inflammatory form, with a lower parasite burden per lesion) have decreased expression of the genes for PGE₂ synthesis enzymes (PTGS1 and PGES), whereas the expression of the gene for 5-LO (ALOX5) is increased. This divergent expression pattern correlates with the decreased levels of PGE₂ and increased levels of LTB₄ found in patients with MCL. Also, in this study, patients with MCL have increased expression of PTGER3, the gene coding for the PG receptor EP₃ (Franca-Costa et al., 2016). This finding suggests a role for LTB₄ in MCL-induced inflammation. However, studies on the role of the inflammatory AA derivatives (such as PGE₂ or LTB₄) at the molecular level show that this effect is highly Leishmania species-dependent (Figure 2).

In the 1980s, it was reported that higher levels of PGE₂ relate to reduced lymphocyte proliferation in the spleens of mice infected with L. major (Farrell and Kirkpatrick, 1987). Moreover, the increase in PGE₂ and LTB₄ levels induced a reduction in the Th1 cytokines TNF-α and IFN-γ, and an increase in the Th2 cytokine IL-4. This effect has been linked to the overexpression of the PG receptors EP₁ and EP₃ (Milano et al., 1996). Indeed, in macrophages from Balb/c mice infected with L. major, there is an increase in these receptors and downregulation of EP₂ and EP₄ receptors. Furthermore, agonists of EP₁ and EP₃ favor infection, whereas selective agonists of EP₂ and EP₄ decrease infection (Penke et al., 2013). PGE₂ also affects the role of B-1 cells. B-1 cells modulate the phagocytic activity of macrophages; this action is dependent on IL-10 release, which in turn modulates the PGE₂ levels in the media (Arcanjo et al., 2017). Also, PGE₂ (via IL-10 production) increases the phagocytic activity of Balb/c mice-derived B-1CDP cells (a particular B-1 cell type with phagocytic ability; Borrello and Phipps, 1999), but it impairs the ability of the mice to resist the infection (Arcanjo et al., 2015). Finally, in a model involving human PMNs co-stimulated with ionomycin or LPS/N-formyl-methionyl-leucyl-phenylalanine (LPS/fMLP), L. major was also able to increase LTB₄ release, which in turn induced a rapid and sustained decrease in LXA₄ production (Plagge and Laskay, 2017).

In contrast, in mouse models of L. amazonensis infection, higher levels of LTB₄ and PGE₂ enhance the parasite-killing ability of murine macrophages, an effect that is dependent of NO• production and IL-10 reduction (Guimaraes et al., 2006; Serezani et al., 2006). The high levels of LTB₄ from macrophages can also be induced by ATP, via P2X7 receptor activation (Chaves et al., 2014). LTB₄ also has a role in neutrophil-dependent L. amazonensis killing. LTB₄ activates its BLT₁ receptor, inducing neutrophil degranulation, the release of metalloproteinase 9 (MMP-9) and myeloperoxidase (MPO), ROS production, and overexpression of TLR2. All these phenomena may be associated with the activation of NF-κB, PI3K, ERK, and protein kinase C (PKC) signaling (Tavares et al., 2014).

Table 2 resumes all current available evidence regarding the role of pro-resolving lipids in Leishmania spp. infections. Of note, most of the researches have been focused in CL. Regarding the role of these pro-resolving lipids in CL, RvD1 increases in patients with diffuse CL (DCL), in comparison with those with LCL, who had lower levels of RvD1 and higher levels of RvD2. This RvD1 pattern in DCL patients correlates with higher levels of Arginase-I and TGF-β and lower levels of TNF-α (Malta-Santos et al., 2017). In an in vitro model of Leishmania infection based on human cells, RvD1 increased phagocytosis of the Leishmania parasites by human macrophages. The increase of RvD1 in L. amazonensis-infected macrophages is reversed by baicalein, an inhibitor of 15-LO (Malta-Santos et al., 2017).

RvD1 and LXA₄ also target AhR, a transcription factor activated by several tryptophan metabolites and some AA derivatives. AhR has been identified in Treg cells, Th17 cells, and DCs, and its activation induces the expression of IL-10 and TGF-β (Gutierrez-Vazquez and Quintana, 2018). In L. major infections, the deletion of this receptor accelerates the disease progression, increasing TNF-α production, and decreasing IL-12 and LXA₄ synthesis (Elizondo et al., 2011).

Annexin A1 is a 37-kDa protein that is expressed under the control of glucocorticoids and activates FPR2/ALXR, increasing IL-10 and IL-6 in experimental models of inflammation (reviewed by Perretti and D’Acquisto, 2009). In mice infected with Leishmania [V] braziliensis, Annexin A1 expression is correlated with the lesion size, being higher 2 weeks after infection. Annexin A1 is important to control the inflammatory response while not impairing the immune system’s parasite-killing ability. Indeed, Annexin A1-KO mice display more phosphorylation of NF-κB after L. braziliesis infection. Also, Annexin A1 is increased in patients with the mucosal form of the disease, but not in those with the localized cutaneous form (Oliveira et al., 2017). In addition, the synthesis of LXA₄, the primary agonist of FPR2/ALXR, is decreased in PMNs infected with L. major (Plagge and Laskay, 2017). LXA₄, via FPR2/ALXR activation, enhances the phagocytic activity of PMNs in a dose- and time-dependent manner (Wenzel and Van Zandbergen, 2009).

More than 30 years ago, it was reported that Leishmania donovani induces the synthesis of PGE₂ and LTB₄ in infected macrophages (Reiner and Malemud, 1984, 1985). The interaction of L. donovani with the TLR2 receptor in macrophages induces activity in the PI3K/Ca⁺² axis, activating the cytosolic phospholipase A2 enzyme (cPLA2) enzyme, which release AA from membranes. Also, the Ca⁺² increase activates the nuclear factor of activated T-cells (NFAT2), a transcription factor which translocates into the nucleus and induces the expression of COX-2. Both AA release and COX-2 overexpression are responsible for the resultant increased PGE₂ release (Bhattacharjee et al., 2016). PGE₂ is anti-inflammatory, leading to parasite survival via EP₂ receptor activation. This EP₂ activation triggers cAMP synthesis, with subsequent PKA activation. This second messenger cascade allows the release of IL-10 and TGF-β and reduces the levels of the inflammatory cytokines TNF-α, IL-12, and IL-17 (Figure 3). Consequently, the inhibition of COX-2 or EP₂-mediated activation of PKA enhances the antiparasitic ability of immune cells in vitro and in vivo (Saha et al., 2014).

15-D-PGJ₂, a pro-resolving lipid that acts as an activator of PPAR-γ receptors (Croasdell et al., 2015), decreases the parasite load in vitro and in vivo (Table 2). This antiparasitic activity correlates with a significant decrease in IL-10 and TGF-β, and a slight increase in the proinflammatory cytokines TNF-α and IL-12, suggesting that 15-D-PGJ₂ favors the Th1 response against L. donovani (Vishwakarma et al., 2016).

Nevertheless, as in CL, the effect of AA derivatives is also Leishmania strain-dependent (Figure 3). Leishmania infantum is a parasite that infects a broad variety of dog populations, inducing canine VL (CVL). This parasite can also infect humans, thus dogs are a reservoir of the disease in large endemic areas of South America (Romero and Boelaert, 2010). Dogs infected with L. infantum show lower levels of PGE₂ and LTB₄ compared with uninfected animals. Moreover, lower levels of PGE₂ and LTB₄ correlate with an increase in the severity of CVL presentation (Solca et al., 2016).

At the cellular and molecular level, PGE₂ increases parasite killing in macrophages infected with L. infantum; this effect is dependent on NO• release (Brandonisio et al., 2001; Panaro et al., 2001). Lymph node-derived leukocytes from dogs infected with L. infantum produce higher levels of NO•, TNF-α, and PGE₂. Consequently, the pharmacological inhibition of PGE₂ synthesis using indomethacin reduces TNF-α release from these cells (Venturin et al., 2016). This PGE₂ release is stimulated by the activation of TLR1/2 receptors by the lipophosphoglycans of L. infantum, which induces the PKC-ERK1/2 pathway, causing the release of the Th1 cytokines TNF-α, IL-12, and monocyte chemoattractant protein-1 (MCP-1). In a dog model, PPAR-γ agonists, such as rosiglitazone, inhibit the cytokine storm induction. The fact that Th1 cytokine overexpression is associated with the inhibition of PPAR-γ (Lima et al., 2017) concurs with the fact that NSAIDs not only inhibit COX activity but can also activate this transcription factor (Lehmann et al., 1997), adding another possible mechanism of action for the antiparasitic activity of NSAIDs in in vivo L. infantum infections (Figure 3) (Panaro et al., 2001; Venturin et al., 2016). Conversely, in an L. infantum-resistant mouse model, the animals became sensitive when the activity of 5-LO was prevented. Spleen cells obtained from these mice showed that the IL-17-producing CD4+ T cells were significantly impaired, with a consequent reduction of cytokine release related to the Th17 axis (Sacramento et al., 2014).

In humans presenting with VL, the pattern of cytokines and lipids that drives the inflammatory response is diverse. In a cohort of patients from an endemic area in Brazil, patients with VL had elevated levels of IL-10, IL-6, IL-8, IL-12, RvD1, LTB₄, and PGF_2α. Also, these patients had lower levels of TGF-β1 and TNF-α. Moreover, this pattern of immune modulators reverted after treatment with antimonial compounds, increasing the levels of TGF-β1 and decreasing the levels of LTB₄, RvD1, IL-6, IL-8, and IL-10. This study also showed that TNF-α levels were not modified by chemotherapy treatment, indicating that the inflammatory response to kill the parasite remained unaltered (Araujo-Santos et al., 2017).