Inhibiting the PGE₂ Receptor EP2 Mitigates Excitotoxicity and Ischemic Injury

Abstract

Prostaglandin E2 (PGE₂) is elevated in the brain by excitotoxic insults and, in turn, aggravates the neurotoxicity mainly through acting on its Gα_s-coupled receptor EP2, inspiring a therapeutic strategy of targeting this key proinflammatory pathway. Herein, we investigated the effects of several highly potent and selective small-molecule antagonists of the EP2 receptor on neuronal excitotoxicity both in vitro and in vivo. EP2 inhibition by these novel compounds largely decreased the neuronal injury in rat primary hippocampal cultures containing both neurons and glia that were treated with N-methyl-d-aspartate and glycine. Using a bioavailable and brain-permeant analogue TG6-10-1 that we recently developed to target the central EP2 receptor, we found that the poststroke EP2 inhibition in mice decreased the neurological deficits and infarct volumes as well as downregulated the prototypic inflammatory cytokines in the brain after a transient ischemia. Our preclinical findings together reinforced the notion that targeting the EP2 receptor represents an emerging therapeutic strategy to prevent the neuronal injury and inflammation following ischemic stroke.

Stroke is one of the leading causes of death and disability among adults, as the prevalence of stroke continues to rise worldwide. Currently, intravenous thrombolysis with a recombinant tissue-type plasminogen activator is the only medication approved by the U.S. FDA for acute ischemic stroke. However, its risks often outweigh the benefits if administered 4.5 h after the incidents.¹ With a slightly extended time window (6 h), the endovascular therapy by intra-arterial mechanical thrombectomy has been emerging as an option for a proportion of patients.² However, the overall narrow time window, its limitation to the occlusion in large arteries, and the potential risks including intracerebral hemorrhage largely limit the patient eligibility. Owing to the urgent demand for newer, safer, and more effective therapies for acute cerebral ischemia, hundreds of potential candidate agents have been developed; none has succeeded in the subsequent clinical trials.³ Acute brain injury, such as ischemia, can initiate complex neuroinflammatory processes that often persist and can further compromise the brain functions in the long term,⁴ suggesting that targeting inflammatory and immune pathways may provide a longer therapeutic time window and greater efficacy in ameliorating reperfusion injury.^3,5 Over the past few decades, intervening with acute neuroinflammatory processes for ischemic stroke has been mainly focused on reducing microglial activation, inhibiting neutrophil migration, and blocking interleukin-1 (IL-1) receptors.⁴ However, targeting these proinflammatory components of the innate immune system did not lead to any new treatment for cerebral ischemia,⁶ attesting to the cellular and molecular complexities of the neuroinflammatory pathways triggered by ischemic injuries.

The inducible cyclooxygenase (COX) isozyme, COX-2, sits atop another large inflammatory signaling axis associated with ischemic stroke. COX-2 is highly regulated by neuronal activities and is often rapidly and robustly induced within the brain by cerebral ischemia.^7,8 A large number of preclinical studies using various rodent stroke models have shown that the induced COX-2 by cerebral ischemia might contribute to the neuronal injury, whereas the administration of selective COX-2 inhibitors or genetic ablation of COX-2 reduced infarct volumes.⁷⁻¹⁰ As a major product of COX-2 enzymatic activity in the brain, prostaglandin E2 (PGE₂) plays a central role in mediating the detrimental effects of induced COX-2 in various neurological conditions, such as ischemic stroke and several neurodegenerative diseases.¹¹ PGE₂ exerts its physiological and pathological effects via signaling through a group of four distinct G-protein-coupled EP receptors (EP1–EP4) that have divergent engagement in cAMP and phosphoinositol turnover.¹² These membrane-bound receptors are differentially expressed on neuronal and glial cells throughout the central nervous system (CNS). Given that targeting PGE₂ downstream signaling might provide more therapeutic specificity than targeting COX-2 for blocking PGE₂ synthesis, the four EP receptors—particularly EP2—have been under intense investigation as potential new therapeutic targets for ischemic stroke in animal models.¹¹

Congenital global ablation of the EP2 receptor increased the cerebral infarction in both transient and permanent middle cerebral artery occlusion (MCAO) models of focal forebrain ischemia.^13,14 In line, the EP2 receptor activation by a selective agonist, ONO-AE1-259-01, decreased the infarct volumes and alleviated neurological deficits.¹⁵ Conversely, a more recent study revealed the selective ablation of neuronal EP2 or the induced postnatal deletion of the receptor-reduced cerebral ischemic injury in mice.¹⁶ Poststroke treatment with a selective EP2 antagonist—benzoxazepine 52—improved the neurological scores and decreased the infarct volumes in mice after transient MCAO.¹⁶ Considering the discrepancy in these previous findings, it is important and necessary to validate the pharmacological results using other selective EP2 antagonists with different chemical structures, providing preclinical evidence to further support the pharmacological inhibition of the EP2 receptor as a novel strategy to treat ischemic stroke. We recently developed a series of novel small-molecule compounds as selective EP2 antagonists,¹⁷ among which TG6-10-1 is highly potent, bioavailable, and brain-permeable and showed marked anti-inflammatory effects, neuroprotection, and many other benefits following prolonged seizures in mice and rats.^18,19 The present study was aimed to validate the role of the EP2 receptor in neuronal injuries following excitotoxic insults in vitro and in vivo. We first evaluated the effects of several potent selective EP2 small-molecule antagonists on neuronal death in a rat primary culture model of excitotoxicity induced by N-methyl-d-aspartate (NMDA) and glycine. The therapeutic effects of our current lead EP2 antagonist, TG6-10-1, in a mouse transient MCAO model of cerebral ischemia were also assessed.

Results

Inhibition of EP2 Receptor Is Neuroprotective in a Neuron–Glia Mixed Setting

Activation of the EP2 receptor by its selective agonists such as PGE₂ and butaprost was neuroprotective against glutamate receptor-mediated excitotoxicity in neuron-enriched cortical and hippocampal cultures (95%), seemingly via activating protein kinase A (PKA).¹³ In contrast, EP2 activation by PGE₂ or butaprost exacerbated NMDA-induced neurotoxicity in glia-containing cortical cultures through a cAMP- but not PKA-dependent pathway.²⁰ Although the reason for these conflicting outcomes remains uncertain, the levels of glial presence in the culture settings might make a difference in the overall effect of EP2 activation, as the receptor in neurons and microglia can exert opposing effects.^12,21 To explore this possibility, we first examined the effects of EP2 receptor inhibition on NMDA/glycine-induced excitotoxicity in rat primary neuron–glia mixed cultures (DIV 13–15) using several EP2-selective small-molecule antagonists that we recently developed (Figure 1A).¹⁷ The primary mix cultures typically consisted of about 55% neurons, 22% microglia, and 23% astrocytes (Figure 1B,C), which approximately mimic the cellular constitutions of the human brain, where there are roughly 40–130 glial cells for every 100 neurons.²² We used the lactate dehydrogenase (LDH) release from the cells to indicate cellular injury. Indeed, the neuronal excitotoxicity can be induced in these mixed hippocampal cells by NMDA/glycine treatment, indicated by a substantial increase of LDH release into the culture medium (Figure 1D). However, the NMDA/glycine-promoted cell injury was attenuated by co-incubating the cells with these EP2 antagonist compounds in a concentration-dependent manner, as the NMDA/glycine-induced LDH release was reduced to 94, 88, 74, and 55% by 0.01, 0.1, 1, and 10 μM of compound TG4-155, respectively (Figure 1D). TG4-155 is our first-generation EP2 antagonist that was originally identified by high-throughput screening.¹⁷ In line, LDH release by NMDA/glycine treatment in these primary hippocampal cultures was also reduced to 78, 42, 51, 78, and 72% by treatment with several active TG4-155 analogues, TG4-166, TG4-290-1, TG4-292-1, TG4-294-2, and TG6-10-1, respectively, at 10 μM (P < 0.001, Figure 1D). In contrast, the analogue TG6-109-1 that was previously tested to be EP2-inactive¹⁷ did not show a neuroprotective effect at the same test concentration (Figure 1D). We previously found that the EP2 receptor activation regulated the delayed death of microglia induced by LPS/IL-13, which was largely prevented by one of our EP2 antagonists—TG4-155.²³ Considering that both neurons and microglia express EP2,¹⁶ the receptor likely plays different roles in these two types of hippocampal cells. The specific effects of these EP2 antagonists on microglia and neurons remain to be resolved for future studies. Nevertheless, these results together suggest that EP2 inhibition in a brain-mimicking neuron–glia mixed culture setting overall is protective against NMDA/glycine-induced excitotoxicity.

Figure 1.

EP2 receptor inhibition is protective against excitotoxicity in neuron–glia cultures. (A) Chemical structures of selective antagonists for EP2 receptor. (B) Immunostaining for NeuN, Iba1, and GFAP was performed to visualize neurons, microglia, and astrocytes in hippocampal primary cultures, respectively. Scale bar = 50 μm. (C) Cell counts in eight random fields (20×) showed that the hippocampal mixed cultures typically contain about 55% neurons, 22% microglia, and 23% astrocytes. (D) Lactate dehydrogenase release into the culture medium was measured to indicate the neuronal excitotoxicity induced by NMDA (30 μM)/glycine (10 μM) treatment in hippocampal primary cultures (**P < 0.01; ***P < 0.001, one-way ANOVA and posthoc Dunnett’s multiple comparisons test). Data are shown as mean ± SEM (N = 8).

EP2 Inhibition by TG6-10-1 Decreases the Infarct Volumes in a Model of Transient Focal Cerebral Ischemia

Several earlier studies revealed that the global congenital ablation of the EP2 receptor aggravated the cerebral ischemic injuries in mice.¹³⁻¹⁵ However, more recent evidence suggested that the postnatal deletion of the EP2 receptor or conditional ablation of its neuronal form reduced infarct volumes and neurological deficits in a similar ischemic model.¹⁶ We next wanted to test our EP2 antagonists in a mouse model of ischemic stroke to determine the effects of pharmacological inhibition of EP2 on ischemia-induced brain injury and inflammation (Figure 2A). We elected to use compound TG6-10-1 for this animal study because, among the second-generation EP2 antagonists that we recently developed, it has adequate in vivo half-life (∼1.8 h) and favorable brain penetration (brain/plasma = ∼1.6) when systemically administered in mice,^18,24 although it was not the most potent analogue for protecting hippocampal neurons against the glutamate receptor-mediated excitotoxicity (Figure 1D). The occlusion of MCA in mice lasted for 45 min, followed by a reperfusion (Figure 2A). The mice then were intraperitoneally injected with vehicle or TG6-10-1 (5 or 10 mg/kg) at 4.5, 12, and 24 h after MCAO. We use a modified Bederson scale to evaluate the neurological deficits of mice after experimental ischemia (Figure 2B). There was no difference in Bederson’s neuroscore between the vehicle group and the two EP2 antagonist groups at day 1, indicating mice from these three experimental groups had been effectively randomized. However, the EP2 antagonist-treated mice began to manifest a trend in the reduction of neuroscores 2 days after ischemic stroke when compared to the vehicle-treated mice (Figure 2B). At day 3 after MCAO, all vehicle-treated mice (5 out of 5) had the highest neuroscore—5, as they all completely lost their spontaneous locomotion and body movement. In contrast, nearly 40% (5 out of 13) of mice treated with TG6-10-1 (5 or 10 mg/kg) showed milder functional deficits (neuroscores 1 to 3) (Figure 2C).

Figure 2.

Effects of TG6-10-1 treatment on neurological and histological outcomes in mice subjected to transient MCAO. (A) Adult male C57BL/6 mice (12 weeks old) were subjected to transient MCAO for 45 min, followed by reperfusion. The animals were treated by vehicle or EP2 antagonist TG6-10-1 (5 or 10 mg/kg, i.p.) at 4.5, 12, and 24 h after MCAO (N = 5–7). Three days after MCAO, all mice were sacrificed for histological analyses. (B) Modified Bederson exam was performed to assess the neurological deficits of the animals after ischemic stroke. Post-MCAO treatment with TG6-10-1 showed a trend in reducing the neuroscores of mice following ischemic stroke (P = 0.4993 and 0.1480 for 5 and 10 mg/kg treatment compared to vehicle treatment, respectively, two-way ANOVA and posthoc Dunnett’s multiple comparisons test). Data are shown as mean ± SEM. (C) Treatment with TG6-10-1 (5 or 10 mg/kg) overall decreased the neuroscores of mice 3 days after MCAO (P = 0.0236 compared to the Bederson score 5—complete loss of locomotion and movement, column statistics analysis). (D) Triphenyltetrazolium chloride (TTC) staining was performed to visualize the infarct areas in the brain. Note that only those viable parenchymal tissues became pinkish and reddish upon TTC staining. (E) Treatment with TG6-10-1 reduced the infarct size, measured 3 days after MCAO (P = 0.0063 and 0.5608 for 5 and 10 mg/kg treatment compared to vehicle treatment, respectively, one-way ANOVA and posthoc Dunnett’s multiple comparisons test). Data are shown as mean + SEM.

All mice were subsequently sacrificed at the end points, and brain tissues were carefully harvested. Coronal slices were prepared and subjected to triphenyltetrazolium chloride (TTC) staining to differentiate between metabolically active and inactive brain tissues (Figure 2D). The 45 min episode of MCAO on average led to >65% of brain areas damaged in vehicle-treated control mice, quantified by calculating the TTC staining-negative regions on the sections (Figure 2E). However, post-MCAO treatment with 5 or 10 mg/kg TG6-10-1 in mice decreased the metabolically inactive brain areas to about 45% (P = 0.0063) and 60% (P = 0.5608), respectively, after experimental ischemic stroke (Figure 2E). These results together suggest that the post-MCAO treatment with EP2 antagonist TG6-10-1 improved the functional recovery of mice from the ischemic stroke and protected brain cells and tissues from ischemic injury. Our data overall validated the beneficial effects of the postnatal ablation of the EP2 receptor in mice after a transient ischemia.

Post-ischemia Inhibition of EP2 Receptor Downregulates the Proinflammatory Cytokines

The initial cerebral ischemic events are typically followed by widespread long-lasting aseptic inflammatory processes in the affected brain areas.¹⁶ The strong innate immune responses are consistently implicated in the progression of brain tissue secondary injury, resolution, and long-term remodeling.³ As the major inflammatory markers and mediators, proinflammatory cytokines are mainly released from injured neurons and activated glial cells.⁵ We next investigated the effects of post-ischemia inhibition of the EP2 receptor with TG6-10-1 treatment on the levels of the three prototypical proinflammatory cytokines including interleukin 1β (IL-1β), IL-6, and tumor necrosis factor α (TNF-α) in the injured brain areas. We found that the transient occlusion of MCA for 45 min in mice powerfully induced the mRNA expression of IL-1β by 24-fold (Figure 3A), IL-6 by 50-fold (Figure 3B), and TNF-α by 12-fold (Figure 3C). Systemic treatment with TG6-10-1 after MCAO on the whole showed a trend in decreasing the mRNA expression of both IL-1β and IL-6 in the ischemia-injured hemisphere (Figure 3A,B). However, the ischemic injury-induced mRNA expression of cytokine TNF-α in the same brain areas was substantially reduced by about 62% (P = 0.0092) and 56% (P = 0.0227) by treatments with 5 and 10 mg/kg TG6-10-1, respectively, when compared to the vehicle treatment. These quantitative polymerase chain reaction (qPCR) results together revealed that the post-ischemic inhibition of EP2 receptor overall downregulated the expression of the proinflammatory cytokines in the injured brain areas.

Figure 3.

EP2 inhibition alters the brain transcription of proinflammatory cytokines after focal cerebral ischemia. Quantitative PCR was performed to measure the mRNA expression of the three prototypical proinflammatory cytokines including IL-1β (A), IL-6 (B), and TNF-α (C) in the ipsilateral hemisphere, measured 3 days after MCAO (N = 5–7). TG6-10-1 treatment after MCAO in mice did not significantly alter the expression of IL-1β (P = 0.9636 and 0.3668 for 5 and 10 mg/kg treatment compared to vehicle treatment, respectively) and IL-6 (P = 0.8519 and 0.6700 for 5 and 10 mg/kg TG6-10-1, respectively) in the ipsilateral hemisphere. However, the mRNA induction of TNF-α in the brain by ischemic injury was largely prevented by post-MCAO treatment with TG6-10-1 (P = 0.0092 and 0.0227 for 5 and 10 mg/kg TG6-10-1, respectively). The statistical analysis was performed using one-way ANOVA and posthoc Dunnett’s multiple comparisons test. Data are shown as mean + SEM.

Discussion

The inducible global deletion of the EP2 receptor has recently been shown to prevent cerebral injury in mice after transient MCAO, and these benefits were further confirmed using pharmacological inhibition of the receptor by a selective antagonist benzoxazepine 52.¹⁶ The present study was designated to further validate the feasibility of pharmacologically targeting the EP2 receptor for excitotoxicity and cerebral ischemia utilizing other EP2 antagonists that we recently developed for CNS conditions. We found that the EP2 inhibition by several novel EP2 antagonists decreased the NMDA-induced cellular damage in rat primary hippocampal mix cultures containing both neurons and glial cells. Systemic treatment with a highly bioavailable and brain-permeable EP2 antagonist TG6-10-1 after a 45 min episode of MCAO in mice improved functional recovery, reduced cerebral injury, and decreased the robust induction of proinflammatory cytokines in the ischemia-injured hemisphere. These preclinical findings provide reinforcing evidence supporting the feasibility of pharmacologically targeting the EP2 receptor for transient ischemic stroke. Using a different small-molecule EP2 antagonist in this study also suggests that our findings are very unlikely dependent on a specific chemical scaffold, thereby lowering the likelihood that the benefits are derived from off-target activities of the antagonists.

As the rate-limiting enzyme for the biosynthesis of prostanoids, COX-2 has been implicated in the pathophysiological mechanisms of inflammation-associated CNS disorders, including stroke, epilepsy, and several neurodegenerative diseases.^18,25,26 COX-2 expression is rapidly upregulated in brain cells including neurons and glia by cerebral ischemia in both human patients and experimental rodents. Overexpression of neuronal COX-2 increases whereas COX-2 inhibition decreases the cerebral infarction.^9,25 However, the notion of COX-2 as a therapeutic target has been deeply dampened over the past two decades, as the long-term use of selective COX-2 inhibitors can break the balance between thromboxanes and prostacyclins, thereby leading to severe cardiovascular and cerebrovascular complications.²⁷ New therapeutic strategies targeting the downstream molecules that mediate the deleterious effects of COX-2 have been widely proposed and explored.^12,21,28 As such, the EP2 receptor is emerging as a promising alternative target for cerebral ischemia and other brain conditions. Results from the present study suggest that EP2 receptor inhibition by several different compounds decreased the neuronal injury induced by NMDA treatment in hippocampal primary mixed cultures containing both neurons and glia, consistent with the proposal that glial EP2 signaling might exacerbate an excitotoxic effect on neurons.²¹ The evident benefits of TG6-10-1 treatment after transient MCAO in this study further support the feasibility of targeting the EP2 receptor for ischemic stroke, reinforcing the emerging notion that EP2 antagonism represents an alternative strategy to COX-2 inhibition and provides higher therapeutic specificity.

The functions of EP2 receptor in ischemic stroke have been controversial over the past decade. Two early studies reported that the global congenital deletion of the EP2 receptor exacerbated the cerebral ischemic injuries in mice after both transient MCAO¹³ and permanent MCAO.¹⁴ These results were confirmed a few years later by another study showing that the genetic deletion of EP2 with the same strategy enlarged, but EP2-selective agonist ONO-AE1-259-01 decreased, the infarct volumes and neurological deficits in both transient and permanent models of MCAO.¹⁵ However, a more recent study by the former research group showed that the induced postnatal deletion of the EP2 receptor or conditional ablation of its neuronal form in mice reduced infarct volumes and neurological deficits after the transient focal ischemia.¹⁶ Markedly, these benefits from specific ablation of the EP2 receptor have been recapitulated in wild-type mice treated by a selective EP2 antagonist benzoxazepine 52 in the same mouse model of transient ischemia. The seemingly discrepant results from these earlier and recent studies might suggest a potential confound of global congenital deletion of EP2, as the receptor has some important basic functions under normal physiological conditions in both the periphery and CNS. Intriguingly, the global EP2 knockout mice are well-known for their complications by developmental and other homeostatic adjustments that result in hypertension and reduced litter size, which had been independently reported by multiple research groups.¹²

The NCBI mouse transcript database shows that the expression of EP2 gene is highly dynamic (https://www.ncbi.nlm.nih.gov/gene/19217). For instance, the EP2 transcript expression in the CNS continually increases by nearly 8 times from E11.5 to E18. In contrast, its transcript level in the adult cortex is only about one-fourth of that in E18, suggesting that the EP2 receptor is highly expressed in the developing brain but low in the adult forebrain. Therefore, the EP2 receptor may play some essential roles in the early CNS development when its expression remains high. Indeed, global prenatal deletion of the EP2 receptor can impair many CNS functions, such as sensorimotor gating, synaptic plasticity, hippocampal long-term potentiation, and depression, leading to significant cognitive deficits and other CNS complications.^29,30 Therefore, the global congenital ablation may not be an ideal strategy to study the roles of EP2 receptor in ischemic stroke, as these deficits and complications would occur before, during, and after the ischemic injury and make the data interpretation very difficult if not impossible. In contrast, mice with either the postnatal deletion of the EP2 receptor or the conditional ablation of its neuronal form preserve the normal behavioral phenotypes, as their EP2 functions are mostly retained during the CNS development, justifying their use as proper tools to study the pathophysiological roles of the receptor in animal models of CNS conditions.

In summary, this proof-of-concept study provides reinforcing evidence for the feasibility that pharmacologically targeting a specific downstream prostanoid receptor can offer an alternative to the blockade of the entire COX cascade. Our findings also support the notion that the EP2 antagonism represents a novel therapeutic strategy to treat cerebral ischemia presumably with higher therapeutic specificity than the COX-2 inhibition. Our current lead EP2 antagonist TG6-10-1 is well positioned as a prime candidate for progression to more intense preclinical and clinical studies. Future studies should be directed to assess the therapeutic effects of the delayed administration of TG6-10-1 or other analogues several hours after cerebral ischemia possibly for a longer therapeutic time window than the 4.5 h that is generally required for the current thrombolytic therapy. The activation of G_s-coupled EP2 receptor can stimulate three signaling effectors, namely, G-protein-dependent PKA and cAMP/exchange protein directly activated by cAMP (Epac) and G-protein-independent β-arrestin.¹² Studies in the future are also necessitated to determine the downstream pathway(s) directly responsible for the EP2 antagonism-mediated benefits in this mouse MCAO model, which might provide new molecular targets for cerebral ischemia.

Methods

Chemicals and Drugs

Test compounds were synthesized as previously described (Figure 1A).¹⁷ Compound TG6-10-1 was also obtained from Advanced ChemBlocks for the animal experiments, and the purity was confirmed by LC/MS and NMR in the Medicinal Chemistry Core at the University of Tennessee Health Science Center. The selectivity and potency of compounds from different sources and batches were evaluated blindly and compared for consistency in potency and selectivity as we previously described.²⁶

Primary Hippocampal Neuron–Glia Mixed Culture

Hippocampal cells were isolated from embryos (E18) of timed-pregnant Sprague–Dawley rats as we previously described³¹ and were plated onto poly-d-lysine (Sigma-Aldrich) coated 24-well plates at a density of 150,000 cells/well in Neurobasal medium supplemented with B27, 5% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 μg/mL streptomycin (Invitrogen). Cells were incubated at 37 °C in a humidified atmosphere consisting of 5% CO₂ and 95% air, and half of the culture medium was replaced twice a week with Neurobasal medium without FBS. The neurons and glial cells in the cultures were visualized by immunostaining. In brief, cultured cells were fixed for 20 min with 4% paraformaldehyde in phosphate-buffered saline (PBS) (pH 7.4), permeabilized by 0.2% Triton X-100 for 15 min, and blocked with 10% goat serum for 1 h. The cells were then incubated in mouse anti-NeuN monoclonal antibody (1:250, EMD Millipore, #MAB377), rabbit anti-Iba1 polyclonal antibody (1:250, Wako Chemicals, #019-19741), rabbit anti-GFAP polyclonal antibody (1:250, Abcam, # PA1-10019), or mouse anti-GFAP monoclonal antibody (1:250, SCBT, #SC-33673) at 4 °C overnight. Cells were then washed and incubated with anti-rabbit or mouse secondary antibodies conjugated with Alexa Fluor 488 or 546 (1:500, Invitrogen) for 2 h and DAPI (1 μg/mL in PBS, Invitrogen) for 10 min. Slides were mounted using DPX mountant (Electron Microscopy Sciences). Images were obtained using a fluorescence microscope BZ-X800 (Keyence) (Figure 1B). Cells were counted in eight random fields (20×) over four wells, and the percentages of NeuN-positive neurons, Iba1-positive microglia, and GFAP-positive astrocytes were calculated. There was no palpable colocalization between neuronal and either microglial or astrocytic nuclei (Figure 1B), indicating the cell type specificity of immunostaining for these biomarkers.

Neuronal Excitotoxicity Assay

Primary hippocampal cultures (DIV14) were pretreated with test compounds for 15 min. Excitotoxicity was then induced by incubating cells with NMDA (30 μM) plus glycine (10 μM) in the medium overnight in the continual presence of test compounds. Neurotoxic damage was assessed by measuring the fraction of lactate dehydrogenase released into the culture medium using a cytotoxicity detection kit (Roche Applied Science).^31,32 Released LDH % was calculated as 100 × LDH released/(LDH released + LDH in the cells), where the LDH in the cells was determined in the cell lysate.

Animal Procedures

All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Tennessee Health Science Center and performed in accordance with the Guide for the Care and Use of Laboratory Animals (the Guide) from the NIH. Adult male C57BL/6 mice (∼12 weeks old) from Charles River Laboratories were housed in standard humidity (∼45–50%) at room temperature (∼21–25 °C) and under a 12 h light/dark cycle with food and water ad libitum.

Transient focal ischemia was induced by middle cerebral artery occlusion for 45 min as we previously described.³³ The anesthetic status of animals was induced and maintained via the continuous inhalation of vaporized isoflurane (Henry Schein) at 3 and 1.5–2%, respectively. The rectal temperature was monitored via a digital thermometer and maintained at 37 °C using a heating pad throughout the procedure. With the anesthetic status being confirmed, a midline skin incision was made on the neck, and the subcutaneous soft tissues were subjected to blunt separations. The right common carotid artery (CCA) was carefully separated from its adjacent vagal nerve, which was carefully protected from surgical damage. The superior thyroid artery was ablated before the right external carotid artery (ECA) was ligated and cut apart with the cauterizer (low temp cautery kit, Bovie Medical Corporation). The occipital artery was ablated before the bifurcation of internal carotid artery (ICA) distal part and the pterygopalatine artery (PPA) became visible. Both the CCA and ICA were clamped using microvascular clips. A knot with a 5–0 suture (LOOK Suture Silk) was loosely prepared close to the origin of ECA, and a tiny nick was prepared along the ECA stump for the insertion of filament into the ECA lumen. MCAO was achieved via the gentle delivery of 2 cm long 6–0 silicon-coated filament (Doccol Corporation) into 10 mm to block the origin of MCA. The loose knot around the origin of ECA was then tightened to prevent bleeding from the nick. After the surgery, the animals were caged above a heating pad to maintain their body temperature before the restoration of consciousness. The occlusion was maintained for 45 min, followed by the withdrawal of the filament for reperfusion. For the analgesia, the mice received a dose of Buprenorphine SR-LAB (1.0 mg/kg, s.c.) 1 h before the surgery. Buprenorphine SR-LAB releases over 72 h and provides blood levels greater than 1.0 ng/mL in mice for postoperative analgesia.

Drug Treatment

With recovery for MCAO surgery, animals were randomized and administered either vehicle (10% DMSO, 50% PEG 400, 40% ddH₂O) or EP2 antagonist TG6-10-1 (5 or 10 mg/kg, i.p.,) at 4.5, 12, and 14 h after MCAO. The mice were fed moistened rodent chow, monitored daily, and injected with lactated Ringer’s and 5% dextrose injection (Baxter) (0.5–1 mL, s.c. or i.p.) when necessary. The end point of this study was 72 h after the MCAO surgery.

Neurological Assessment

Experimental mice were assessed for neurological deficits on each day after MCAO using a modified Bederson scoring system in a blinded manner (Table 1).³⁴

Table 1. Modified Bederson Scale for Global Neurological Assessment.

Score 0	no deficit: when suspended by tail, the animal fully extends bilateral forelimbs forward
Score 1	forelimb flexion: the animal flexes the forelimb on its contralateral side; forelimb flexion involves the wrist, elbow, and shoulder
Score 2	decreased resistance to lateral push: a normal or mildly injured animal shows equivalent resistance to unilateral push-induced passive movement; mouse with more severe ischemic damage typically shows weaker resistance on the contralateral side than that on the ipsilateral side
Score 3	unidirectional circling–circling and tail-chasing usually suggest that large ischemic damage is obtained
Score 4	longitudinal spinning (barrel rolling) or seizure activity
Score 5	no movement: the animal completely loses its spontaneous locomotion and body movement

Quantification of Infarct Volumes

Mice were euthanized under deep isoflurane anesthesia, and then brains were collected and rinsed in cold PBS. The mouse coronal brain matrix (Alto) was used to prepare the 1 mm thick coronal sections. All sections were collected and stained with 0.2% TTC solution (2,3,5-triphenyltetrazolium chloride, Santa Cruz Biotechnology) for 20 min and imaged. The infarct and the entire hemisphere volumes were measured using ImageJ software (NIH) in a blinded manner. The calculation of global infarct size follows Cavalieri’s principle in three-dimensional case (or “trapezoidal rule”). The infarct size was adjusted for edema formation in compliance with Swanson’s correction.³⁵ Given that either artifactual extension or shrinkage of tissue may exist, a relative value of infarct size was reported in this study. All absolute values were therefore normalized to its contralateral tissue volume constituted by continuous eight slices approximately ranging from +3.00 to −5.00 mm from the bregma.

Quantitative PCR

The total RNA from mouse hippocampus was isolated using TRIzol (Invitrogen) with the PureLink RNA Mini Kit (Invitrogen). RNA purity and concentration were measured by the A260/A280 ratio and A260 value, respectively, using a NanoDrop One microvolume spectrophotometer (Thermo Fisher). The first-strand complementary DNA (cDNA) was synthesized using the SuperScript III First-Strand Synthesis SuperMix (Invitrogen) following the product manual. The qPCR was performed using 8 μL of 10× diluted cDNA, 0.4 μM of primers (Table 2), and 2× B-R SYBR Green SuperMix (Quanta BioSciences) with a final volume of 20 μL in a CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories). Cycling conditions were as follows: 95 °C for 2 min followed by 40 cycles of 95 °C for 15 s and then 60 °C for 1 min. Melting curve analysis was used to verify the single-species PCR product. Fluorescent data were acquired at the 60 °C step. The cycle of quantification for β-actin was subtracted from the cycle of quantification measured for each gene of interest to yield ΔCq.^31,36 Samples without the cDNA template served as negative controls.

Table 2. Primers for qPCR.

gene	forward primer (5′ → 3′)	reverse primer (5′ → 3′)	amplicon size (bp)	NCBI reference sequence
IL-1β	TGAGCACCTTCTTTTCCTTCA	TTGTCTAATGGGAACGTCACAC	101	NM_008361.3
IL-6	TCTAATTCATATCTTCAACCAAGAGG	TGGTCCTTAGCCACTCCTTC	119	NM_031168.2
TNF-α	TCTTCTGTCTACTGAACTTCGG	AAGATGATCTGAGTGTGAGGG	108	NM_013693.3
β-actin	AAGGCCAACCGTGAAAAGAT	GTGGTACGACCAGAGGCATAC	109	NM_007393.5

Statistical Analysis

Statistical analyses were performed using Prism (GraphPad Software) by analysis of variance (ANOVA) with posthoc Dunnett’s test or column statistics analysis as fully described in the figures. P < 0.05 was considered statistically significant. Data are presented as mean ± or + SEM.

Author Contributions

L.L., Y.Y., J.H., and J.J. designed the research study; L.L., Y.Y., R.H., and J.J. conducted the experiments; L.L., Y.Y., J.H., and J.J. analyzed the data; L.L. and J.J. wrote the manuscript; all authors reviewed and edited the manuscript.

The authors declare no competing financial interest.