Prostanoids and Resolution of Inflammation – Beyond the Lipid-Mediator Class Switch

Bioactive lipid mediators play a major role in regulating inflammatory processes. Herein, early pro-inflammatory phases are characterized and regulated by prostanoids and leukotrienes, whereas specialized pro-resolving mediators (SPM), including lipoxins, resolvins, protectins, and maresins, dominate during the resolution phase. While pro-inflammatory properties of prostanoids have been studied extensively, their impact on later phases of the inflammatory process has been attributed mainly to their ability to initiate the lipid-mediator class switch towards SPM. Yet, there is accumulating evidence that prostanoids directly contribute to the resolution of inflammation and return to homeostasis. In this mini review, we summarize the current knowledge of the resolution-regulatory properties of prostanoids and discuss potential implications for anti-inflammatory, prostanoid-targeted therapeutic interventions.


INTRODUCTION
Inflammation is an integral part of protective host responses against pathogens or injury (1). Herein, inflammatory processes usually follow a defined sequence of reactions characterized by the rapid induction of a pro-inflammatory response, which is closely followed by an anti-inflammatory response. To regain homeostasis resolution of inflammation represents an integral and crucial part of acute inflammation. In fact, failure to completely resolve inflammatory processes is associated with the emergence of chronic inflammatory conditions. While resolution was considered to merely represent the downregulation or inactivation of inflammatory mediators for a long time, it is now appreciated to be active and complex, involving the formation of pro-resolving mediators (2). Moreover, it is widely accepted that resolution processes are initiated early during inflammation, and classical pro-inflammatory mediators have been shown to directly impact on resolution as well (3). Therefore, pro-resolving therapeutic approaches are increasingly being considered rather than anti-inflammatory ones as the latter might diminish the host response against pathogenic challenges (4,5). Since lipid mediators play a crucial, yet often ambivalent role during the inflammatory process, and prostanoid synthesis represents a major target for anti-inflammatory therapies, the present mini review aims to recapitulate the current understanding of the role of prostanoids in the context of resolution of inflammation.

THE COURSE OF INFLAMMATION Cellular Mediators
Upon an insult (e.g., pathogen contact, injury) local resident immune cells are the first responders. Among these, tissue resident macrophages (MF) represent not only an important first line of defense but even more importantly they establish an environment prone to recruit neutrophils and monocytes from the circulation. Neutrophils are main executers in the acute inflammatory environment, i.e. they produce high amounts of reactive oxygen and nitrogen species to eliminate pathogens, and secrete a plethora of inflammatory mediators (6). Within the inflammatory environment neutrophils rapidly die by apoptotic processes and are phagocytosed by MF in a process termed efferocytosis (7,8). Importantly, uptake of apoptotic cells (AC) induces a shift in MF polarization from a pro-inflammatory to an anti-inflammatory and immunomodulatory phenotype (9)(10)(11)(12)(13)(14). MF and other antigen-presenting cells such as dendritic cells (DC) eventually activate adaptive immune responses, which ensures complete removal of the invading pathogens and enables accelerated responses upon anew contact to the triggering stimulus (15). The return to cellular homeostasis further requires successful emigration of infiltrated immune cells from the cleared site of inflammation (16). In contrast to the concept of a rapid return to homeostasis, recent reports provided evidence that resolution is followed by a longer lasting, immune-suppressive post-resolution phase characterized by infiltrating regulatory T cells (Treg) and myeloid-derived suppressor cells (MDSC), but also MF, which might be decisive for a successful adaptation vs. chronic inflammation and/or autoimmunity (17)(18)(19).

Lipid Mediators
Regulation and execution of inflammatory reactions is mediated by soluble mediators, such as cytokines and chemokines. In addition, bioactive lipids emerged as crucial factors during all phases of the inflammatory process (20, 21). Specifically, while leukotrienes and prostaglandins appear early during the onset of inflammation, specialized pro-resolving mediators (SPM), such as lipoxins, resolvins, and maresins, are produced later on, facilitating the resolution of inflammation (22). Prostanoids, like leukotrienes, belong to the eicosanoid family of lipid mediators (23). Both classes are synthesized from arachidonic acid (AA) after the latter is released from membrane phospholipids by phospholipase A 2 (24). While leukotrienes are synthesized by the lipoxygenases, prostanoid formation initially requires conversion of AA to the unstable prostaglandin H 2 (PGH 2 ) via the bifunctional cyclooxygenases 1 (Cox-1) and 2 (25,26). In line with the established pro-inflammatory function of the prostanoids, the cyclooxgenases became an important target in the therapy of inflammatory diseases, and as of today, non-steroidal antiinflammatory drugs (NSAIDs) rank amongst the most important anti-inflammatory drugs (27,28). Of note, while Cox-1 is constitutively expressed in most cells, Cox-2 often is inducible and emerged as the more important cyclooxygenase in the context of inflammation. Consequently Cox inhibitors (Coxibs) selectively targeting Cox-2 emerged as promising novel anti-inflammatory therapeutics (29,30). The short-lived Cox product PGH 2 is then substrate to specific synthases, which produce potent prostanoids including PGD 2 , PGE 2 , PGF 2a , and PGI 2 , as well as thromboxane A 2 (TXA 2 ) ( Figure 1).
Interestingly, the production of the different lipid mediators appears to be tightly connected across the course of inflammation. For example, the presumably pro-inflammatory PGE 2 was shown to attenuate the synthesis of leukotrienes and to induce the production of SPM, thus initiating the so-called lipidmediator class switch (31). While the SPM-regulatory impact of prostanoids on the resolution of inflammation is rather well characterized, there is accumulating evidence that prostanoids also directly impinge on other aspects of the resolution process. In the following sections, we will therefore summarize the current understanding of the role of the most prominent prostanoids in resolution of inflammation, aside from the aforementioned lipid-mediator class switch, with a special focus on PGE 2 .

PROSTANOIDS IN THE RESOLUTION OF INFLAMMATION
Thromboxane A 2 TXA 2 is predominantly produced by platelets, but also at appreciable amounts by MF (32). Interestingly, while thromboxane A synthase (TXAS) appears to be coupled to the activity of Cox-1 and Cox-2 constitutes the dominant Cox in inflammatory conditions, the activity of TXA 2 is largely proinflammatory (33,34). In fact, there is little evidence that TXA 2 might contribute to the resolution of inflammation. In line, Kupffer cell-derived TXA 2 was shown to contribute to liver fibrosis, a common outcome of ineffective resolution (35). The observation that TXA 2 synthesis is induced by pro-inflammatory stimuli, while it is suppressed upon inflammatory restimulation (36), further suggest that the prevention of TXA 2 , e.g. via redirection of Cox-1-provided PGH 2 towards PGE 2 synthesis, might be part of the immune-suppressive environment typical for the post-resolution phase (18).

Prostaglandin I 2 (Prostacyclin)
PGI 2 (prostacyclin) has been characterized as the counterpart of TXA 2 as it inhibits platelet aggregation and acts as a potent vasoand bronchodilator (37). It is produced primarily by vascular endothelial and smooth muscle cells, yet other cells such as fibroblasts and dendritic cells also synthesize PGI 2 (38). In the context of inflammation, PGI 2 was shown to inhibit LPS-induced expression of pro-inflammatory cytokines in MF, DC, CD4 + T cells, and endothelial cells (39)(40)(41)(42). PGI 2 further synergizes with anti-inflammatory cytokines interleukin-4 (IL-4) and IL-13 to suppress pro-inflammatory cytokines (43). Along the same lines, PGI 2 receptor (IP) deficient mice displayed stronger allergic inflammation, which was attributable to enhanced Th2 cell function (44). Thus, PGI 2 emerges as a predominant antiinflammatory mediator, positioning it also as a counterpart of TXA 2 in the context of inflammation.
Prostaglandin D 2 PGD 2 is produced by numerous immune cells including activated MF, DC, Th2 cells, eosinophils, platelets, but also endothelial cells (45). However, since its main source are mast cells (46,47), PGD 2 has been characterized extensively in the context of allergic reactions (48). PGD 2 can further be metabolized to the cyclopentenone, PGJ 2 -type prostanoids, including PGJ 2 and 15d-PGJ 2 , which also display biological activity in the context of resolution of inflammation (49). PGD 2 binds to the PGD 2 receptors 1 (DP1) and DP2 [also known as chemoattractant receptor-homologous molecule expressed on Th2 cells (CRTH2)] (50). Activation of DP2 is of specific importance in allergic inflammation, where mast cell-derived PGD 2 stimulates the recruitment of innate lymphoid type 2 cells (51) and Th2 cells (52), and causes activation of these as well as of basophils and eosinophils (53). While PGD 2 was shown to contribute to allergic inflammation, PGD 2 synthase (PGDS) decreases during the acute inflammatory phase, while it rises again at later stages corresponding to the resolution phase in animal models (54). PGD 2 and 15d-PGJ 2 both exert pro-inflammatory functions via CRTH2 (55). In contrast, PGD 2 -dependent activation of DP1 as well as 15d-PGJ 2 -mediated activation of peroxisome proliferatoractivated receptor g (PPARg) inhibit the production of inflammatory cytokines and chemokines by antigen-presenting cells including DC or MF by interfering with inflammatory transcription factors nuclear factor kappa B (NFkB), activator protein 1 (AP1), and signal transducer and activator of transcription 3 (STAT3) (56)(57)(58) and additionally by enhancing the activity of anti-inflammatory nuclear factor erythroid 2-like 2 (Nrf2) (59,60). Consequently, PGD 2 and 15d-PGJ 2 support emigration of MF to the draining lymph nodes and attenuate the recruitment of leukocytes (54). They further inhibit effector functions of and induce apoptosis in T lymphocytes (57). Thus, PGD 2 contributes to normalize the local environment, an important aspect of the resolution of inflammation (61).

Prostaglandin F 2a
PGF 2a is produced by the aldo-keto reductase (AKR) 1C3, also known as PGF 2a synthase (PGFS) using PGH 2 or PGD 2 (62). Alternatively, PGE 2 can be converted to PGF 2a by AKR1C1 or AKR1C2 (63). While PGF 2a is synthesized in most tissues (64), its prime site of production is the female reproductive system (65), where PGF 2a has been shown to be offunctional importance (66,67). Nevertheless, PGF 2a also appears to be involved in inflammatory processes (64,68). PGF 2a is elevated in rheumatoid arthritis (69) and was shown to contribute to and correlate with fibrosis (70, 71), which is characteristic for insufficient resolution. Interestingly though, PGFS expression and concomitantly PGF 2a levels decrease similar to PGD 2 during acute inflammation, only to increase again in the resolution phase (72), which might be indicative for an active role of PGF 2a in the resolution of inflammation. Yet, the exact role of PGF 2a in inflammation-resolving processes remains to be determined.
Upon inflammatory stimulation, both the expression of Cox-2 and microsomal PGE 2 synthase 1 (mPGES-1) are induced in MF (78), skewing the prostanoid spectrum towards PGE 2 production in the acute inflammatory phase (73). PGE 2 signals towards enhanced recruitment of neutrophils, MF, and mast cells (79)(80)(81) and enhances the expression and secretion of proinflammatory cytokines in DC, MF, and T cells (82)(83)(84). Its NFkB-activating properties further support neutrophil survival, thereby extending the pro-inflammatory impact of neutrophils in the inflammatory niche (85,86). Yet, this initial increase in PGE 2 is only moderate and transient in character, and PGE 2 levels rise again during the resolution phase, eventually increasing to much higher levels in the post-resolution phase (18,87,88). This seemingly biphasic regulation of PGE 2 might at least in part be due to a shift from transcriptional to posttranscriptional programs governing not only PGE 2 production, but more generally the course of inflammation (89). With respect to the regulation of PGE 2 synthesis, Cox-2 rapidly accumulates during early inflammation. In a negative feedback loop, elevated PGE 2 induces the expression of dual specificity phosphatase 1 (DUSP1), thereby enhancing the activity of the RNA-binding protein tristetraprolin (TTP), which destabilizes the mRNA of Cox-2, but also of pro-inflammatory tumor necrosis factor (TNF) (90). The second wave of PGE 2 production by MF appears to be facilitated by another increase of Cox-2 expression induced by sphingosine-1-phosphate released from apoptotic cells, which activates the RNA-stabilizing protein human antigen R (HuR) in MF to increase Cox-2 expression (91). As a side note, while PGE 2 is predominantly produced in a Cox-2/mPGES-1 dependent manner during the inflammatory and early resolution phase, PGE 2 levels during the postresolution phase are approx. 3-fold higher, which is due to enhanced Cox-1/mPGES-1 expression in MF (18). While these findings might explain the kinetics of PGE 2 production and even some of the inhibitory effects of PGE 2 on pro-inflammatory mediators, further evidence for an immunosuppressive function of MF-derived PGE 2 emerged in the tumor context, where PGE 2 inhibits CD80 expression via EP2, thereby attenuating activation of cytotoxic T cells (92,93). Similarly, PGE 2 suppresses cytolytic functions of NK cells (94,95) and inhibits phagocytic and bacteria killing activities of MF, thus preventing the establishment of appropriate inflammatory, anti-microbial responses largely via EP2-dependent cAMP induction (96)(97)(98). In line, EP2-and EP4-signaling limits secretion of TNF and IL-1b, and enhances expression of anti-inflammatory IL-10 in response to LPS by microglia (99), i.e. resident MF of the central nervous system (100). In general, EP2-and EP4activation by PGE 2 appears to be crucial to establish an antiinflammatory and resolving MF phenotype (101,102), which is characterized by further immune-modulatory factors such as transforming growth factor b (TGFb) (103). Nevertheless, there are contradictory reports regarding the exact impact of PGE 2 on T cell functions. On the one hand, PGE 2 appears to inhibit IL-2 production by T cells, thereby attenuating both T cell activation and activation-dependent apoptosis (104)(105)(106). On the other hand, while PGE 2 appears to contribute to sustained inflammation by differentiation and activation of Th1 and gd T cells, considered to support inflammation (107)(108)(109), other findings indicate that PGE 2 selectively inhibits Th1 cytokine production leaving Th2 cytokines, such as IL-4 and IL-5, unaffected (110,111), thus provoking a PGE 2 -dependent shift towards Th2 responses, which are associated with repair mechanisms instead (112)(113)(114). This notion is substantiated by the high levels of PGE 2 observed in Th2-driven diseases such as atopic allergy (115). Indeed, the intricate impact of PGE 2 on the balance between different T cell subtypes might be one of the key mechanisms how PGE 2 affects a self-limiting inflammation throughout resolution and post-resolution phases. Elevated PGE 2 impairs interferon g (IFN g) synthesis, thereby directly attenuating Th1 responses, while leaving Th2 responses unaltered (110,111). PGE 2 further favors Th17 responses via EP2 and EP4 by shifting the IL-12/IL-23 balance towards Th17supportive IL-23 (116)(117)(118). While Th17 cells are largely inflammatory in nature contributing to severe inflammatory diseases (107,119), they have been shown to be highly plastic, being able to trans-differentiate into Treg thereby contributing to resolution of inflammation (120). Yet, PGE 2 not only promotes differentiation of naïve T cells or Th17 cells towards Tregs (121,122), it also supports further expansion of differentiated Tregs (123). Conclusively, the concentration, source, and timing of PGE 2 appear critical to determine the exact T cell response in the context of inflammatory responses. Moreover, while the pro-resolving activity of cyclooxygenase metabolites has long been attributed predominantly to PGD 2 and 15d-PGJ 2 (124), PGE 2 emerged as an important facilitator of the lipid-mediator class switch inducing not only the production of PGD 2 and its derivatives, but also of the so-called specialized pro-resolving mediators (SPM) (20). SPM, synthesized by 15-lipoxygenase (ALOX15) (31,(125)(126)(127), are key players in the resolution of inflammation (114,128). PGE 2 induces the expression of the relevant lipoxygenases, thereby skewing the balance towards a pro-resolving lipid mediator profile (129,130). While SPM levels are mostly considered to reciprocally reflect PGE 2 levels during the course of inflammation, they in fact coexist (131). The exact temporal and spatial distribution of both PGE 2 and SPM might eventually determine the course of the resolution of inflammation. Of note, SPM also have been shown to affect the T cell balance (132)(133)(134). Yet, there is mounting evidence that PGE 2 also directly supports further resolution-associated functions. Along these lines, PGE 2 has been shown to inhibit pro-inflammatory cytokine production (99,135) and to stimulate the expression of anti-inflammatory cytokines (136,137), thereby contributing to the early steps of the resolution process (138). Furthermore, blocking PGE 2 synthesis attenuated efficient resolution in a peritonitis model by preventing the emigration of MF from the site of inflammation in a CX3CL1-dependent manner (87). Extending beyond its impact on the resolution phase, PGE 2 contributes to the immune suppressive postresolution phase, where PGE 2 suppresses innate immune responses, inhibits lymphocyte functions, and contributes to the generation and activity of Treg (121,139) and MDSC (140). While these immune suppressive effects appear negative in the context of novel infections, they lower the risk of autoimmune responses (18,19) (Figure 2). The complexity of the resolution of inflammation both at cellular as well as (lipid) mediator level, highlights the need for further studies unraveling details of the resolution phase to allow for the development of refined therapeutic intervention strategies, especially for chronic inflammatory diseases lacking proper resolution mechanisms.

THERAPEUTIC CONSIDERATIONS
Cox inhibitors are amongst the most widely used over the counter anti-inflammatory drugs worldwide (141). Despite the undisputed beneficial effects of the broad spectrum Cox inhibitors, specific Cox-2 inhibitors (Coxibs) were developed to more selectively interfere with the production of inflammationassociated prostanoids (142). Due to the prominent role of PGE 2 in the establishment of inflammatory processes, recent therapeutic approaches aimed at inhibiting the inflammationassociated terminal PGE 2 synthase mPGES-1 to selectively block the production of PGE 2 only (73,143,144). Yet, considering the major impact of PGE 2 on successful resolution of inflammation, therapeutic approaches targeting PGE 2 synthesis should be critically revisited as continued PGE 2 inhibition in inflammatory diseases might be expected to lead to chronic diseases due to insufficient resolution. In fact, even attenuating inflammation itself might interfere with successful resolution, since the process of resolution is initiated already very early during the inflammatory process (3). Thus, it might be warranted to focus on therapeutic approaches promoting resolution rather than on anti-inflammatory ones in the future (4,145). Indeed, there are numerous efforts to target SPM production or receptors (146). Considering promising effects of PGE 2 receptor antagonists (e.g. EP4) in the context of chronic inflammatory diseases (147), it will be interesting to see combinatorial approaches in the future. FIGURE 2 | PGE 2 in the context of inflammation. An inflammatory insult is recognized by resident immune cells such as resident MF. Upon inflammatory stimulation Cox-2 and mPGES-1 are induced in MF resulting in increased PGE 2 synthesis. PGE 2 then contributes to the MF-mediated recruitment of neutrophils, which act as a first line of defense to eliminate the pathogenic stimulus and incite further inflammatory responses. Already at this early stage of the inflammatory process PGE 2 initiates the lipid mediator class switch towards the production of specialized pro-resolving mediators (SPM) including lipoxins, resolvins, maresins, and protectins e.g. in neutrophils. Neutrophils are rapidly followed by monocytes again facilitated by PGE 2 , which upon infiltration into the affected tissue differentiate into pro-inflammatorily activated MF and release soluble mediators including cytokines and chemokines, as well as further PGE 2 . PGE 2 -triggered transcriptional programs eventually induce post-transcriptional feedback circuits, which reduce Cox-2 mRNA stability and thus protein expression to eventually inhibit PGE 2 production, and also attenuate the expression of pro-inflammatory cytokines such as TNF. In addition, anti-inflammatory mediators (including IL-10) are induced restricting the intensity of the inflammatory reactions. Within the inflammatory niche, activated neutrophils rapidly undergo apoptotic cell death and are phagocytosed by MF, which causes a shift in MF polarization towards an alternatively activated, immune-modulatory, resolution phenotype characterized by the secretion of e.g. TGFb. Apoptotic cells further release sphingosine-1phosphate (S1P), which enhances the mRNA stability of Cox-2 again, resulting in increasing PGE 2 production. Within the resolving environment PGE 2 supports a shift from Th1 T cells to the repair-associated Th2-type further supporting tissue normalization. In addition, PGE 2 attenuates expression of CX3CL1 in MF, thereby eventually allowing their emigration from the resolving tissue. PGE 2 further supports the recruitment of T cells and myeloid cells, which differentiate into regulatory T cells and myeloid-derived suppressor cells, respectively, thereby establishing an immune-suppressive post-resolution environment. Lower part: During inflammation, PGE 2 levels transiently increase in the acute inflammatory phase. After a decrease during the anti-inflammatory and the early resolution phase, PGE 2 tends to increase again during the progress of resolution, reaching highest levels in the post-resolution phase. AC, apoptotic cells; MF, macrophages (graynaïve, resident; redpro-inflammatory; blueanti-inflammatory; greenresolution phase); MDSC, myeloid-derived suppressor cells; MO, monocytes; PMN, neutrophils (graynaïve, redinflammatory); T cells (red -Th1; blue -Th2; green -Treg); ⇨, infiltration/emigration; !, development within the inflammatory niche.

Schmid and Brüne
Prostanoids and Resolution of Inflammation Frontiers in Immunology | www.frontiersin.org July 2021 | Volume 12 | Article 714042

AUTHOR CONTRIBUTIONS
TS and BB wrote and edited the manuscript. All authors contributed to the article and approved the submitted version.