The complex role of prostaglandin E₂-EP receptor signaling in wound healing

Abstract

Prostaglandins are critical lipid mediators involved in the wound healing response, with prostaglandin E₂ (PGE₂) being the most complex and exhibiting the most diverse physiological outputs. PGE₂ signals via four G protein-coupled receptors, termed EP-receptors 1-4 that induce distinct signaling pathways upon activation and lead to an array of different outputs. Recent studies examining the role of PGE₂ and EP receptor signaling in wound healing following various forms of tissue damage are discussed in this review.

INTRODUCTION

Wound healing is a highly complex and coordinated process that has three sequential but overlapping phases: 1) hemostasis/inflammation, consisting of vascular changes, immune infiltration, and inflammatory signaling; 2) proliferation, focusing on re-epithelialization and angiogenesis; and 3) remodeling, aiding in maturation and strengthening of the newly generated tissue (1). Following a wounding event, these processes must be tightly regulated to ensure that proper healing occurs. Dysregulation of any stage of the response can lead to either insufficient repair or chronic inflammation and excessive tissue damage. Therefore, coordinated inflammation is critical for a wound to heal properly, which requires a number of different cell types and inflammatory signals. Here, we focus on the role of prostaglandin E₂ (PGE₂) in the response to tissue injury and the important roles of this signaling mediator to induce proper wound healing. In addition, we discuss recent discoveries related to PGE₂ and wound repair.

PROSTAGLANDIN E₂ SYNTHESIS AND SIGNALING

Prostaglandins (PGs) are a family of bioactive lipids that exert a wide range of physiological effects throughout the body and are commonly involved in vasodilation, tissue repair/homeostasis, and inflammation. PGs are enzymatically synthesized from plasma membrane-derived phospholipids. These phospholipids are cleaved by the action of phospholipase A₂ (PLA₂), which releases fatty acids into the cytosol to be acted upon by a variety of enzymes (2). Arachidonic acid (AA) is the predominant fatty acid that is released, which is converted to prostaglandin G₂ (PGG₂) and PGH₂ via a two-step reaction by cyclooxygenase (COX) enzymes, including the constitutively active isoform, COX-1, and the inflammation-inducible isoform, COX-2. Prostaglandin H₂ (PGH₂) is then subsequently acted upon by PGE synthase enzymes to generate PGE₂. Although some prostaglandins are preferentially synthesized from COX-1-specific or COX-2-specific products, PGE₂ can be generated from both COX-1- and COX-2-derived PGH₂ (3)_. PGH₂ is acted on by one of three isoforms of PGE synthase enzymes: microsomal PGE synthase-1 (mPGES-1), mPGES-2, and cytosolic PGE synthase (cPGES). Generally, mPGES-1 is considered the inflammation-activated isoform, whereas mPGES-2 and cPGES are both constitutively expressed in most cell types (4). PGE₂ is secreted by most cells throughout the body and is inactivated by the enzyme 15-hydroxyprostaglandin dehydrogenase (15-PGDH), which oxidizes PGE₂ to form 15-keto-PGE₂ (5).

PGE₂ exerts varying effects depending on the concentration of PGE₂ present, the cell/tissue type, and the receptor isoform (EP 1–4) of which it interacts with (2). For these reasons, PGE₂-EP receptor signaling is highly complex yet crucially important for an appropriate inflammatory response (Fig. 1). EP receptors are PGE₂-specific G protein-coupled receptors (GPCRs) that are widely expressed throughout the body and induce distinct intracellular signaling cascades via activation of their respective G proteins: EP1 utilizes G_q, EP3 utilizes G_i, and EP2 and EP4 utilize G_s (6). These G proteins activate different downstream signaling pathways that lead to either pro- or anti-inflammatory responses (7). EP1 activates phospholipase C (PLC), leading to increased diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP3), increasing intracellular calcium levels, and subsequent protein kinase C (PKC) activation (2). EP2 and EP4 receptors both activate two identical signaling cascades: The first entails activation of adenylate cyclase causing increased adenosine 3′,5′-cyclic monophosphate (cAMP) production and subsequent protein kinase A (PKA) and cAMP response element binding protein (CREB) activity; the second includes activation of phosphoinositol-3-kinase (PI3K) and Akt via the β-arrestin pathway following EGFR transactivation (8). EP3 receptor activation leads to inhibition of adenylate cyclase and reduced cAMP levels, contradicting EP2/4 signaling. In general, EP1-induced signaling and EP3-induced signaling are considered anti-inflammatory, whereas EP2 and EP4 are considered pro-inflammatory (2). EP receptor distribution varies across tissues with EP3 and EP4 being the most commonly expressed subtypes found almost ubiquitously throughout cell types and EP1 and EP2 being present in only select tissues (9). Tissue mRNA and protein content of EP receptors can be evaluated via the Human Protein Atlas (6), and a more detailed description of arachidonic acid metabolism and prostaglandin signaling can be found in previously published reviews (2, 4, 10).

Figure 1.

EP receptor signaling and PGE₂-mediated wound healing phenotypes. EP1 (blue) associates with G_q G protein, which activates phospholipase C (PLC), leading to cleavage of phosphatidylinositol 4,5-bisphosphate (PIP₂), increasing intracellular levels of diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP₃). IP₃ binding to IP₃-gated calcium channels on the endoplasmic reticulum induces release of calcium into the cytosol, which acts in concert with DAG to activate protein kinase C (PKC), leading to phosphorylation of various proteins. EP1-regulated wound healing outputs are listed in the blue box. EP2 (yellow) and EP4 (red) activate two identical signaling pathways. First, G_s G protein activates adenylate cyclase (AC, orange), which increases intracellular adenosine 3′,5′-cyclic monophosphate (cAMP), activating protein kinase A (PKA), leading to phosphorylation of downstream targets including the transcription factor cAMP response element binding protein (CREB). The second pathway (green) occurs via β-arrestin/Src-induced EGFR transactivation. This activates IP₃ kinase (IP3K) to phosphorylate PIP₂ to become phosphatidylinositol 3,4,5-trisphosphate (PIP₃), which recruits Akt to the plasma membrane to be phosphorylated and activated. Activated Akt phosphorylates various downstream targets and often induces NF-κB nuclear translocation. EP2-regulated wound healing outputs are listed in the yellow box. EP4-regulated wound healing outputs are listed in the red box. EP3 (purple) associates with the inhibitory G protein (G_i) blocking the activity of AC and reducing intracellular cAMP levels. EP3-regulated wound healing outputs are listed in the purple box. PGE₂, prostaglandin E₂.

PGE₂ IS PRODUCED FOLLOWING TISSUE INJURY

PGE₂ is the most complex prostaglandin due to the multiple receptor isoforms that each induce distinct signaling cascades and outputs. The opposing signaling pathways initiated by PGE₂ make it difficult to determine whether PGE₂ is beneficial or detrimental to the healing response. Many types of tissue injury are described below, with the commonality of each of the studies described being that PGE₂ is produced following injury. For example, in DSS-induced colitis, PGE₂ levels were elevated during the induction phase, at day 4 post-DSS initiation, that coincided with increased PLA₂ and COX-2 expression, which was sustained for 2 wk (11). PGE₂ levels were also increased following dermal excisional wounding, with the highest concentration observed days 1–3 following injury with levels remaining elevated above baseline through day 12 (12). Although these studies suggest PGE₂ is important for the wound healing process, more details describing how PGE₂ exerts its effects are necessary to fully understand the potentially beneficial and/or detrimental pathways induced by PGE₂-EP signaling.

PGE₂ MODULATES CELL PHYSIOLOGY

EP1 signaling and EP3 signaling have been shown to induce vasoconstriction, whereas EP2 signaling and EP4 signaling lead to vasodilation (13). In general, vascular alterations are necessary for tissue repair and these observations alone are not indicative of promoting or impeding wound healing. In the murine renal vasculature system, EP3 activation by PGE₂ led to constriction of the vasculature and EP3-deficient mice showed a reduced ability to vasoconstrict (14). Interestingly, varying levels of PGE₂ have been shown to regulate cerebral blood flow in a dichotomous fashion; low-level PGE₂-induced EP4 activation led to vasodilation, whereas high-level PGE₂-induced EP1 activation led to vasoconstriction in cerebral parenchymal arterioles, with both phenotypes inhibited by antagonism of EP4 and EP1, respectively (15). Further, in vascular smooth muscle cells, deletion or inhibition of EP4 led to increased vascular constriction and increased blood pressure following stimulation with angiotensin II, although EP4 agonism inhibited this phenotype (16). PGE₂-EPR regulation of vascular flux is therefore necessary for the beginning stages of wound healing to prevent blood loss upon immediate injury and to allow for increased blood flow and immune infiltration at later time points. Pharmacological modulation of PGE₂ concentration could provide a critical therapeutic benefit for individuals needing alterations in vascular pressure following injury.

PGE₂ is also able to induce differentiation of cell subtypes. Tendon-derived stem cells isolated from rats that were treated with PGE₂ display osteogenic differentiation by increasing protein content of bone-specific proteins, bone morphogenetic protein-2 (BMP2), runt-related transcription factor 2 (Runx2), osterix (OSX), and osteocalcin (OCN) along with increased alkaline phosphatase activity and calcium deposition (17). Additionally, in zebrafish intestinal tissue and human-derived intestinal organoids, PGE₂ treatment increased mucus production, indicative of intestinal differentiation toward a goblet cell phenotype (18). PGE₂ can also influence macrophage polarization and metabolism. In bone marrow-derived macrophages (BMDM), IL-4 and PGE₂ cotreatment induced M2 macrophage polarization, an anti-inflammatory macrophage phenotype that is represented an increase by M2 markers, arginase (Arg1), mannose receptor C type 1 (Mrc1), resistin-like α (Fizz1), and Chitinase type protein 3 (Ym1) due to CREB-Kruppel-like factor 4 (KLF4)-mediated transcription (19). In addition, PGE₂ treatment following IL-4-induced M2 macrophage polarization modulated metabolic pathways by reducing mitochondrial membrane potential and subsequent ATP production due to reduced transcription of malate-aspartate shuttle components that led to an overall decrease in intracellular reactive oxygen species (20). Further, in mice with ear hole punch injuries, EP3 and EP4 were shown to influence M2 macrophage recruitment and polarization, with EP3- and EP4-deficient mice exhibiting altered ratios of M1 to M2 macrophages (21). These studies illustrate how PGE₂ treatment can alter specific cell physiology in a manner that could be either beneficial or detrimental to an injured site, depending on the context.

PGE₂ IN PROMOTING WOUND HEALING

EP Receptor-Mediated Phenotypes

PGE₂-EP receptor signaling positively influences wound healing by altering immune migration, increasing proliferation, and promoting survival. In a rodent model of myocardial infarction (MI), EP2 activation in macrophages was critical for macrophage recruitment to the injured site with EP2-deficient mice exhibiting increased infarct size, reduced immune cell mobilization, and poorer tissue regeneration than wild-type counterparts (22). In addition, mice deficient in EP3 or EP4 exhibit reduced wound closure following ear hole punch injury that correlated with reduced macrophage recruitment to the injured site, an overall reduction in M2 macrophages and a reduction in new lymph and blood vessel formation that coincided with reduced mRNA of vascular endothelial growth factor receptor-3 (VEGFR3) and lymphatic vessel endothelial hyaluronan receptor-1 (LYVE-1) in granulation tissue compared to wild-type mice (21). Corroborating these results, mice receiving bone marrow transplants exhibited increased hematopoiesis, chemokine production, colony formation, and immune cell homing posttransplant that correlated with increased survival, which was greatly reduced with either EP2 or EP4 antagonism (23). Interestingly, inhibiting EP4 signaling has also been shown to increase immune infiltration at sites of injury. Mice lacking EP4 on vascular smooth muscle cells (VSMCs) present with increased immune infiltration, aortic size, and oxidative stress following angiotensin II (ANG II)-induced damage to the vasculature (16). Similarly, in angioplasty wire-injured mice, EP4-specific deletion on endothelial cells increased immune infiltration and impeded tissue repair, whereas EP4 agonism improved endothelial repair and reduced neointima formation in wild-type mice (24). Likewise, zebrafish with amputated tail fins required EP4 signaling to eliminate neutrophils from the injured site to allow for continued tissue repair postinjury (25). These studies suggest a dual role for EP receptor signaling, especially EP4, in mediating immune infiltration and egress during tissue repair.

EP2 and EP4 also induce proliferation postinjury to aid in tissue regeneration. EP4 signaling following angioplasty wire injury was shown to be important for re-endothelialization because of EP4-cAMP-PKA-induced proliferation, which occurred to a lesser extent following EP2 activation (24). In mice with DSS-induced colitis, PGE₂ treatment increased expression of EP4 that led to β-arrestin-PI3K-Akt signaling, which increased proliferation and reduced disease severity and apoptosis levels (26). Interestingly, samples from human colitis patients or DSS-treated mice show a reduction in EP4, COX-1, and PGE₂ content but increased COX-2 levels compared to healthy controls (26). Together, these studies show the importance of EP4, and to a lesser extent EP2 signaling, in proliferation induction to allow for re-epithelialization at the injured site to enhance tissue repair.

Modulation of PGE₂ Concentration

Modifying PGE₂ levels via treatment with PGE₂, PGE analogs, or by impeding PGE₂ inactivation can improve wound healing phenotypes. Supporting this, mice deficient in the PGE₂ inactivating enzyme, 15-PGDH, have elevated PGE₂ levels in bone marrow and liver and have improved regenerative abilities and reduced markers of tissue damage following bone marrow transplantation, or hepatectomy-induced liver regeneration, respectively, that was also observed following pharmacological 15-PGDH blockade (23). In a mouse model of DSS-induced colitis, PGE₂ treatment reduced symptom severity and apoptosis levels while increasing colonic repair and proliferation via upregulation of EP4-dependent β-arrestin-PI3K-Akt signaling (26). Corroborating these results, genetic or pharmacological inhibition of 15-PGDH led to increased colonic PGE₂ content that correlated with reduced histological markers of colitis and increased proliferation of intestinal tissue 7 days following DSS treatment in mice (23). Further, in a zebrafish model of DSS-induced intestinal injury, PGE₂ treatment increased mucus production, barrier integrity, autophagy and cell viability, reduced Escherichia coli uptake, and improved overall survival following injury (18). Interestingly, treatment of cutaneous wounds with PGE₂-incorporated hydrogels improved speed of recovery by enhancing angiogenesis, reducing inflammation and immune infiltration, and shortening the time of wound closure (27). In addition, in mice with bleomycin-induced pulmonary fibrosis, 15-PGDH inhibition reduced markers of fibrosis, inflammatory cytokine expression in lung and serum, and improved airway function that correlated with increased survival (28). Finally, following angioplasty wire injury, treatment of mice with a PGE analog, misoprostol, increased re-endothelialization and improved healing, whereas loss of mPGES-1 blocked tissue repair (24). Combined, these studies illustrate the essential role of PGE₂-induced reparative tissue responses following injury and suggest that increasing PGE₂ concentration at sites of damage may improve wound resolution.

PGE₂ treatment can also alter macrophage polarization and metabolism, leading to an M2, anti-inflammatory macrophage phenotype (20), which has been observed in white adipose tissue in an obesity model (19), in skin following cutaneous wounding (22) and in microglia following traumatic brain injury (29). As mentioned previously, macrophages treated with PGE₂ present markers of M2 macrophages, including Arg1, Mrc1, Fizz1, and Ym1 that is dependent on CREB-KLF4-induced transcription (19) and can modify mitochondrial membrane potential and transcription of malate-aspartate shuttle components to reduce ATP production (20). Interestingly, macrophage proliferation is influenced by PGE₂ signaling in a dose-dependent manner, with the highest rate observed following 1 µM PGE₂ treatment in vitro, that leads to M2 macrophage polarization and increased expression of arginase, interleukin (IL)-1 receptor a, IL-10, CD68, and CD206 with reduced M1 markers, IL-1β, IL-6, and TNFα (27). Mice with skin wounds that are treated with PGE₂ incorporated hydrogels have increased M2 macrophages present at the site of injury that correlated with increased angiogenesis, faster wound closure, and reduced scarring (27). These signaling pathways are critical during a healing response, as the transition from a pro-inflammatory to an anti-inflammatory immune cell phenotype will directly dictate a tissue’s ability to begin the remodeling phase of the response.

Inhibiting PGE₂ production with COX inhibitors can be detrimental to wound healing. After mouse ear hole punch injury, treatment with the selective COX-2 inhibitor, celecoxib, reduced wound closure due to decreased macrophage recruitment and M2 macrophage numbers, and reduced blood and lymph vessel formation at the injured site due to decreased transcriptional content of LYVE-1, VEGFR3, VEGF-C, and VEGF-D (21). In a rodent model of bone injury, treatment with the nonsteroidal anti-inflammatory drug, naproxen, led to a reduction in serum PGE₂ levels, which correlated with poorer woven bone formation following mechanical loading, indicating a need for PGE₂ signaling in bone remodeling postinjury (30). Together, these studies underline the importance of PGE₂-mediated signaling following injury, but more studies defining the role of EP receptors in these responses would be highly beneficial.

PGE₂ IN IMPAIRED WOUND HEALING

EP Receptor-Mediated Phenotypes

Regulating EP-mediated signaling is essential to allow for adequate repair following injury. Excessive EP2 signaling can increase cytokine production and lead to dysregulated inflammation. PGE₂-EP2 signaling has been shown to impede blood-brain barrier integrity and exacerbate hippocampal injury induced by seizures by increasing oxidative stress and inflammatory cytokine and chemokine production, which was reversed by EP2 antagonism (31). EP2 blockade also reduced markers of gliosis and neuronal injury that correlated with improvements in behavior scores postinjury (31). Furthermore, following spinal cord injury, there is increased COX-2, mPGES-1, and PGE₂ content in astrocytes and astrocyte-mediated PGE₂ secretion increased production of pro-inflammatory cytokines IL-1β, and IL-6 in macrophages via EP2, which was inhibited by antagonizing EP2 or blocking COX-2 enzyme function (32). Conversely, in human epidermal keratinocytes, EP3 agonism reduced cytokine production following Toll-like receptor stimulation, which is thought to influence symptom severity in patients with toxic epidermal necrolysis, as there is typically reduced EP3 expression and concomitant chronic inflammation in conjunctival epithelial tissue of patients (33). Although cytokine production is generally necessary during an injury response, having excessive or dysregulated inflammation can exacerbate tissue damage and lead to inadequate repair. Modifying EP signaling to alter cytokine production may prove to be a valuable therapeutic option for highly inflamed tissues.

Modulation of PGE₂ Concentration

Although PGE₂ signaling has been found to be critical for adequate wound healing to occur in some injury models, there have also been situations where PGE₂ signaling was detrimental to the healing response. In radiation-induced salivary gland damage, elevated PGE₂ levels correlated with salivary gland dysfunction, whereas reduced PGE₂ levels correlated with improved salivary gland function postdamage (34). In an impaired wound healing model utilizing LIGHT−/− mice, PGE₂ levels were greatly elevated days 1–7 postdermal excisional wounding when compared to wild-type mice, which correlated with increased elastase activity, platelet aggregation, and reduced hemostasis following injury (12). In a rodent model of tendinopathy, PGE₂ treatment increased the activity of the osteoblast marker alkaline phosphatase, leading to increased ossification of tendon-derived stem cells, which is indicative of poorer recovery. Interestingly, treatment with the COX-2-selective inhibitor, celecoxib, attenuated this phenotype and improved tendon repair post injury (17). Furthermore, reducing PGE₂ levels via celecoxib treatment improved pressure ulcer healing by reducing iNOS, hydroperoxidase, matrix metalloprotease-1, and TNF-α levels. In addition, celecoxib treatment improved re-endothelialization, shortened time to wound closure, and strengthened scar formation postinjury (3). These studies suggest that fine-tuning of PGE₂ signaling is necessary for sufficient healing to occur and that dysregulation of PGE₂-mediated inflammation can impair healing following an injury. More detailed descriptions of the studies outlined above can be found in Table 1.

Table 1.

PGE₂-EP receptor-mediated wound healing phenotypes in different damage models

Model	EP Receptor	[PGE2]	Major Findings	Ref
PGE2 synthesis or signaling in humans
Ex vivo human cerebral parenchymal arterioles	EP1EP4	1–10 µM PGE21–100 nM PGE2	Low PGE2 doses (1–100 nM) induce vasodilation while high PGE2 doses (1–10 µM) induce vasoconstriction. Vasodilation was inhibited by EP4 antagonism (BGC 20-1531, 1 µM). Vasoconstriction was inhibited by EP1 antagonism (SC-51322, 1 µM).	(15)
In vitro human acute epidermal inflammation	EP3	100 µg/mL PGE2	PGE2 treatment (100 µg/mL) decreased CCL5, CXCL10, and IL-6 transcription and secretion induced by Toll-like receptor activation with Poly(I:C). EP3 agonism (ONO-AE-248, 10 µg/mL) decreased CCL5, CXCL10, and IL-6 transcription and secretion induced by Toll-like receptor activation with Poly(I:C).	(33)
Effects of PGE2 signaling on cell physiology
Ex vivo and in vivo mouse renal arterial pressure	EP3EP2/EP4	0.3 µM PGE2	Administration of PGE2 (0.3 µM) induced vasoconstriction. EP3-deficient mice exhibited vasodilation following PGE2 treatment. EP2 (PF-04418948, 10 µM) or EP4 antagonism (L-161982, 1 µM) induced vasoconstriction following PGE2 treatment. Lower PGE2 doses (0.001–0.3 µM) minimally altered arterial pressure and higher PGE2 doses (10–30 µM) did not exacerbate pressure changes.	(14)
In vitro rat tendon-derived stem cells (TDSCs)	Not evaluated	50 ng/mL PGE2	PGE2 treatment (50 ng/mL) induced osteogenic differentiation, measured by alkaline phosphatase activity, calcium deposition and increased content of osteogenic proteins BMP2, Runx2, OSX, and OCN.	(17)
In vitro human-derived enteroid monolayers	Not evaluated	1 µM PGE2	PGE2 treatment (1 µM) increased intestinal mucus production.	(18)
In vivo Zebrafish DSS-induced colitis	Not evaluated	0.1-10 µM PGE2	PGE2 treatment (0.1–10 µM) increased intestinal mucus production in a dose-dependent fashion, indicative of intestinal differentiation and improved barrier integrity and survival	(18)
In vitro mouse bone marrow-derived macrophage (BMDM) activation	Not evaluated	10 µM PGE2	PGE2 treatment (10 µM) to IL-4-activated M2 macrophages reduced cytoplasmic and mitochondrial reactive oxygen species generation and mitochondrial membrane potential. PGE2 treatment (10 µM) reduced metabolism and ATP levels due to reduced transcription of malate-aspartate shuttle components.	(20)
In vitro and in vivo mouse BMDM activation	Not evaluated	10 nM PGE2	PGE2 (10 nM) and IL-4 cotreatment enhanced M2 macrophage polarization, indicated by increased M2 gene markers, Arg1, Mrc1, Fizz1, and Ym1 that was mediated by CREB-induced upregulation of KLF4.	(19)
In vivo mouse ear-hole punch injury and macrophage polarization	EP3/EP4	Not quantified	EP3 (ONO-AE-248, 50 nmol/day) or EP4 (ONO-AE-329, 50 nmol/day) agonism increased LYVE-1 and VEGFR3 mRNA in wounds. EP3- or EP4-deficient mice exhibited slower wound healing and had reduced LYVE-1, VEGFR3, VEGF-C, and VEGF-D mRNA content along with a reduction in macrophage recruitment and reduced M2 markers	(21)
PGE2 signaling leading to promotion of wound healing
In vivo mouse myocardial infarction (MI) induced injury	EP2	Not quantified	EP2-deficient mice exhibit poorer heart function and increased infarct size post-MI that correlates with reduced IL-1β, IL-17A, and IL-18 content and reduced macrophage recruitment to the damaged site. EP2 signaling led to macrophage migration through downregulation of erythroid differentiation regulator 1 (Erdr1); macrophage migration defect was rescued via Erdr1 knockdown in EP2−/− macrophages.	(22)
In vivo mouse in vitro human DSS-induced colitis	EP4	20 µM PGE2	Colitis tissue from humans and mice have increased COX-2, but reduced COX-1, and EP4 mRNA and reduced PGE2 content compared with healthy controls. PGE2 treatment (20 µM) to mice reduced disease index and apoptosis while increasing proliferation of colon tissue. PGE2 treatment (20 µM) upregulated EP4 and increased β-arrestin-PI3K-Akt signaling.	(26)
In vivo zebrafish tailfin amputation	EP4	0.001-1 µM PGE2	Reducing endogenous PGE2 production in zebrafish elevated neutrophil counts at the site of tailfin injury 8–24 h postdamage. Treating ptges-deficient zebrafish with PGE2 (0.001-1 µM) reduced neutrophil numbers Concomitant EP4 antagonism (AH23848, 1 µM) and PGE2 treatment to ptges-deficient zebrafish elevated neutrophil counts at the injured site	(25)
In vivo mouse angiotensin II (ANG II)-induced vascular damage	EP4	Not quantified	Mice deficient in EP4 on vascular smooth muscle cells (VSMC-EP4−/−) are significantly more susceptible to ANG II-induced aortic dissection. ANG II-treated VSMC-EP4−/− mice have increased aortic size, elastin degradation, oxidative stress, immune infiltration, and matrix metalloprotease-2 and matrix metalloprotease-9 activity. VSMC-EP4−/− mice had increased vasoconstriction and higher blood pressure compared with WT mice following ANG II-infusion.	(16)
In vivo mouse angioplasty wire injury-induced vascular damage	EP2/EP4	100 µg/kg misoprostol (a PGE analog)	Mice deficient in microsomal PGE synthase-1 have elevated neointimal formation and increased immune infiltration 28 days postinjury that correlates with reduced PGE2 content in urine. Global loss of mPGES-1 reduced re-endothelialization that was rescued by treatment with misoprostol, a PGE analog (100 µg/kg), or by EP4 agonism (ONO-AE1-329, 1 µM). Endothelial cell proliferation was mediated by EP4-cAMP-PKA signaling. Endothelial specific deletion of EP4 blocked tissue repair and increased immune infiltration postinjury. Misoprostol or EP4 agonism following injury improved endothelial repair and reduced neointima formation.	(24)
In vivo mouse bone marrow alterations/efficiency of bone marrow transplantation	EP2/EP4	2–10 ng/mg protein PGE2	15-PGDH-deficient mice or mice treated with a 15-PGDH inhibitor (SW033291, 10 mg/kg) have elevated PGE2 content in bone marrow that was significantly higher than other prostaglandins measured. 15-PGDH inhibition enhanced hematopoiesis, colony formation, and immune cell homing following bone marrow transplantation that could be impeded by antagonizing EP2 (10 µg/mouse) or EP4 (10 µg/mouse).	(23)
In vivo mouse DSS-induced colitis	Not evaluated	10–30 ng/mg protein PGE2	15-PGDH-deficient mice or mice treated with a 15-PGDH inhibitor (SW033291, 10 mg/kg) have elevated PGE2 content in colon. 15-PGDH blockade (15-PGDH−/− mice or SW033291 at 5 or 10 mg/kg/twice per day) reversed histological markers of colitis and increased intestinal proliferation after 7 days of DSS treatment.	(23)
In vivo mouse hepatectomy-induced liver regeneration	Not evaluated	5–20 ng/mg protein PGE2	15-PGDH-deficient mice or mice treated with a 15-PGDH inhibitor (SW033291, 10 mg/kg) have elevated PGE2 content in liver. Inhibiting 15-PGDH (15-PGDH−/− or SW033291, 5 mg/kg/twice/day) increased liver regeneration by increasing proliferation at day 2 postsurgery that increased liver weight through day 7.	(23)
In vivo mouse ear-hole punch injury and macrophage polarization	EP3/EP4	Not quantified	Ear-hole punch injury induced COX-2 and mPGES-1 expression that correlated with lymphangiogenesis and angiogenesis of wounds. Celecoxib (100 mg/kg/day) treatment reduced lymph vessel formation and decreased mRNA content of LYVE-1, VEGFR3, VEGF-C, and VEGF-D at day 5. Celecoxib treatment decreased macrophage recruitment and increased the ratio of M1:M2 polarized macrophages at injured sites. EP3 (ONO-AE-248, 50 nmol/day) or EP4 (ONO-AE-329, 50 nmol/day) agonism increased LYVE-1 and VEGFR3 mRNA in wounds. EP3- or EP4-deficient mice exhibited slower wound healing and had reduced LYVE-1, VEGFR3, VEGF-C, and VEGF-D mRNA content along with reduced macrophage recruitment and M2 markers.	(21)
In vivo mouse mechanical loading and stress fractured bone	Not evaluated	0.75–1.25 pg/mL PGE2 in serum	Treatment with the nonsteroidal anti-inflammatory drug naproxen (10.9 mg/kg/day) led to worsened stress fracture healing, poorer bone remodeling, and increased bones fractures due to reduced toughness. Naproxen treatment altered fibril formation, reduced 4-hydroxyproline content, and led to reduced cortical bone formation that correlated with reduced serum PGE2 levels.	(30)
In vivo zebrafish DSS-induced colitis	Not evaluated	0.1–10 µM PGE2	Concomitant or post-DSS PGE2 treatment (0.1–10 µM) improved survival and reduced intestinal Escherichia coli uptake in zebrafish. PGE2 increased cell viability and autophagy induction in the zebrafish intestine 5-h post-DSS removal.	(18)
In vivo mouse skin excision wounding In vitro macrophage polarization	Not evaluated	0.5–2 µM; 0-1 ng/mL PGE2	Hydrogels incorporated with PGE2 improved closure of skin wounds in mice. The proliferation rate of macrophages was altered by PGE2 dose, with the highest rate observed following 1 µM PGE2 treatment. Macrophages cultured in PGE2 containing hydrogels expressed M2 macrophage markers CD68, CD206, IL-10, IL-1Ra, and arginase and had increased mRNA content of angiogenic factors and reduced M1 macrophages markers IL-1β, IL-6, and TNF-α. Mice receiving PGE2-incorporated hydrogels had improved wound closure and reduced skin fibrosis at day 14 postinjury.	(27)
In vivo mouse bleomycin-induced pulmonary fibrosis	Not evaluated	Not quantified	Pulmonary fibrosis was reversed with 15-PGDH inhibition (SW033291, 5 mg/kg/twice per day). Inflammatory markers in serum and lung were reduced by 15-PGDH blockade at day 7 postlung injury that correlated with reduced fibrosis, improved airway function, and increased survival at day 35.	(28)
In vitro mouse BMDM activation	Not evaluated	10 µM PGE2	PGE2 treatment (10 µM) to IL-4-activated M2 macrophages reduced cytoplasmic and mitochondrial reactive oxygen species generation and mitochondrial membrane potential, leading to a reduction in metabolism and ATP levels due to reduced transcription of malate-aspartate shuttle components.	(20)
In vitro and in vivo mouse BMDM activation	Not evaluated	10 nM PGE2	PGE2 (10 nM) and IL-4 cotreatment enhanced M2 macrophage polarization, indicated by increased M2 gene markers Arg1, Mrc1, Fizz1, and Ym1, which was mediated by CREB-induced upregulation of KLF4.	(19)
PGE2 signaling leading to inhibition of wound healing
In vivo mouse seizure-induced neuronal inflammation	EP2	Not quantified	Seizure-induced injury increased inflammatory cytokines and chemokines (IL-1β, IL-6, TNF-α, CCL2, CCL3, CCL4) that were reduced by EP2 antagonism (TG6-10-1, 5 mg/kg). EP2 blockade dampened gliosis and neuronal injury, preserved blood-brain barrier integrity, and improved behavioral scores.	(31)
In vivo rat spinal cord contusion injury	EP2	200–1,000 pg/mg protein; 100–6,000 pg/mL media; 0.1–10 µM PGE2	Astrocytes have increased levels of COX-2, mPGES-1, and PGE2 due to macrophage migration inhibitory factor activation of JNK-ERK-p38 signaling. PGE2 (2.5 µM) or astrocyte-conditioned media with simultaneous LPS treatment to primary macrophages increased IL-1β, IL-6, and decreased TNF-α content in supernatant and cell lysates. EP2 antagonism (PF-04418948, 10 µM) or COX-2 inhibition (NS389, 30 µM) modulated IL-1β, IL-6, and TNF-α content.	(32)
In vivo and in vitro rat Achilles tendinopathy	Not evaluated	50 ng/mL PGE2	12 wk postinjury, Achilles tendons exhibit ossification and increased OSX and Runx2 staining. Treatment with COX-2 inhibitor, celecoxib (10 mg/kg) immediately after tenotomy reduced ossification that correlated with reduced Akt phosphorylation.	(17)
In vivo mouse cutaneous pressure ulcer wounding	Not evaluated	200–500 pg/mg tissue PGE2	Ulceration of cutaneous tissue increased COX-1, COX-2, PGE2, TNF-α, iNOS, and lipid hydroperoxide content and induced immune cell recruitment to the wound days 3–7 postinjury. Treatment with a COX-2 inhibitor, celecoxib (5 mg/kg/day), beginning 1 day before injury decreased iNOS, TNF-α lipid hydroxyperoxide, COX-2, and PGE2; improved re-epithelialization markers; and reduced scar formation by day 14.	(3)
In vivo mouse DSS-induced colitis	Not evaluated	35–75 pg/mg tissue PGE2	PGE2 content increased ∼2-fold from 4 to 7 days postinitiation of DSS treatment. PLA2 and COX-2 expression levels were elevated 16.6-fold and 4 to 7-fold, respectively, at day 4 of DSS treatment. Elevated PGE2 levels correlated with reduced colon length and worsened fecal and histological scores.	(11)
In vitro and In vivo mouse radiation-induced salivary gland damage	Not evaluated	200–16,000 pg/mL media PGE2	Primary parotid gland cells treated with radiation significantly increase PGE2 secretion, which correlates with reduced saliva secretion days 3–30 following radiation. Mice deficient in the extracellular ATP-gated P2X7 receptor secrete significantly less PGE2 from primary parotid gland cells which correlates with improved saliva secretion days 3–30 following radiation. Parotid gland tissues treated with radiation increased COX-1 and COX-2 and decreased mPGES-2 and cPGES expression. COX-1 and COX-2 activity remain unchanged. Parotid gland tissues from P2X7R−/− mice treated with radiation decreased COX-1, mPGES-1, mPGES-2, and cPGES expression. COX-1 activity was increased with treatment.	(34)
In vivo mouse impaired dermal excisional wound healing	Not evaluated	Relative values; twofold to sixfold increase in WT, 1.3- to 1.8-fold increase in LIGHT−/−	WT mice exhibit a twofold to sixfold increase in PGE2 days 1–3 postinjury. LIGHT−/− mice (impaired wound healing model) exhibit 1.3-fold to 1.8-fold increase in PGE2 content days 1–3, with sustained elevated levels at day 7 compared with WT mice. Elevated PGE2 levels in LIGHT−/− mice correlated with elevated elastase activity, platelet aggregation, and reduced hemostasis.	(12)
In vitro and in vivo mouse BMDM activation and high-fat feeding-induced insulin resistance	Not evaluated	10 nM PGE2	Myeloid-specific CREB knockdown increased M1 macrophage markers and macrophage infiltration in white adipose tissue and induced systemic insulin resistance during high-fat feeding in mice.	(19)

CONCLUSIONS

Taken together, these studies illustrate that PGE₂ is a crucial eicosanoid in all three phases of the wound healing response. Regulated PGE₂ signaling is important for proper repair to occur and treatment with nonsteroidal anti-inflammatory drugs may be detrimental to a healing tissue. More clearly deciphering the roles of upstream mediators of PGE₂ production, including PLA₂, COX-1, COX-2, mPGES-1, mPGES-2, and cPGES with future studies will help delinate the inflammatory pathways leading to elevated PGE₂ levels and provide more direct therapeutic targets for modulating PGE₂ production. Additionally, defining the regulation and activity of the PGE₂ inactivating enzyme 15-PGDH will more precisely outline the timeframe that PGE₂ is able to induce EP-mediated signaling and may unveil opportunities to activate or inhibit 15-PGDH to improve wound resolution. A better understanding of these components of the PGE₂ pathway will allow for more precise drug targeting to modify PGE₂-induced signaling.

EP receptor signaling plays a major role in dictating the transition from pro- to anti-inflammatory state, either allowing for or inhibiting completion of wound healing. The physiological outputs of EP receptor activation have been more clearly defined for EP2 and EP4 receptors, whereas the importance of EP1 and EP3 signaling is insufficiently studied in the context of wound healing (Fig. 1). Uncovering different components that regulate EP receptor distribution, such as signaling components that modulate transcription and/or translation, would be highly useful to better understand the importance of each EP receptor subtype. Finally, a clearer understanding how EP receptor activation is PGE₂ concentration-dependent across isoforms and tissue types could help guide the use of NSAIDs or 15-PGDH inhibitors to modulate PGE₂-mediated outputs. Further evaluation of EP-specific signaling and generation of EP-selective agonists and antagonists may provide novel therapeutic treatment options to enhance wound recovery following an injury.

GRANTS

The authors are supported by the National Institute of Dental and Craniofacial Research Grants R01DE029166 (to K. H. Limesand) and F31DE028737 (to K. E. Gilman).

DISCLAIMER

The funding agency had no role in preparation of this manuscript.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

K.E.G. drafted manuscript; K.E.G. and K.H.L. edited and revised manuscript; K.E.G. and K.H.L. approved final version of manuscript.