Inhibition of prolyl hydroxylases by dimethyloxaloylglycine after stroke reduces ischemic brain injury and requires hypoxia inducible factor-1α

Abstract

Pathological oxygen deprivation inhibits prolyl hydroxylase (PHD) activity and stimulates a protective cellular oxygen-sensing response in part through the stabilization and activation of the Hypoxia Inducible Factor (HIF) 1α transcription factor. The present investigation tested the therapeutic potential of enhanced activation of oxygen-sensing pathways by competitive pharmacologic PHD inhibition after stroke, hypothesizing that post-ischemic PHD inhibition would reduce neuronal cell death and require the activation of HIF-1α. The PHD inhibitor dimethyloxaloylglycine (DMOG, 100 μM) reduced cell death by oxygen glucose deprivation (OGD), an in vitro model of ischemia, and the protection required HIF-1α. In vivo, DMOG (50 mg/kg, i.p.) administered 30 or 60 min after distal occlusion of the middle cerebral artery (MCA) in mice enhanced the activation of HIF-1α protein, enhanced transcription of the HIF-regulated genes vascular endothelial growth factor, erythropoietin, endothelial nitric oxide synthase, and pyruvate dehydrogenase kinase-1, reduced ischemic infarct volume and activation of the pro-apoptotic caspase-3 protein, reduced behavioral deficits after stroke, and reduced the loss of local blood flow in the MCA territory after stroke. Inhibition of HIF-1α in vivo by Digoxin or Acriflavine abrogated the infarct sparing properties of DMOG. These data suggest that supplemental activation of oxygen-sensing pathways after stroke may provide a clinically applicable intervention for the promotion of neurovascular cell survival after ischemia.

Fig 1. DMOG induces normoxic HIF-1α expression in cortical neurons

A. Primary cortical neurons were treated with indicated concentration of DMOG for 24h. Western blot demonstrates an increase in HIF-1α (115 kDa) protein expression with DMOG treatment. β-actin is a loading control. B. Neurons were treated for 24h with vehicle or 250μM DMOG. Quantitative RT-PCR was used to determine the expression of HIF-1-dependent gene VEGF. (n=3, Mean ± SEM, * p<0.05).

Fig 2. PHD inhibitor attenuates OGD-induced cell death and requires HIF-1α

Cortical neurons were subjected to OGD for 2h with 24h reperfusion. Cell death was assessed by trypan blue exclusion 24h after OGD. Cortical neurons were pre-treated 24h prior to OGD or post-treated after OGD with 100μM DMOG. (A) DMOG reduced cell death in both pre- and post- treatment. (*p<0.05 compared to control; #p<0.05 compared to OGD; n=4). (B) Western blot of protein collected from cortical neurons 3h after OGD or OGD + DMOG demonstrate that HIF-1α is enhanced after OGD in the DMOG treatment group (n=3). (C) Inset demonstrates successful HIF-1α knockdown. Neuronal cultures were infected with lentiviral shRNA directed against HIF-1α for 48h. Parallel control and HIF-1α-shRNA cells were treated with 250μM DMOG for 24h to induce HIF-1α protein stabilization. Control cultures but not shRNA treated cultures show HIF-1α protein expression. DMOG did not reduce cell death from OGD in cells with HIF-1α knockdown. Data are the mean ± SEM from at least three independent experiments. (*compared to parallel control cultures, p<0.05, n=3).

Fig 3. PHD inhibitor DMOG attenuates apoptotic cell death in cortical neurons

(A) Pure neuronal cultures were subjected to an apoptotic insult of B-27 supplement withdrawal for 24h and dead cells were counted by trypan blue staining. (B) Neuronal lysate was collected after 24h and immunoblotted for HIF-1α and β-actin loading control. (C) Western blot for the caspase-3 active fragment 3 and 6h after initiation of B-27 withdrawal. Data are the mean ± SEM from at least three independent experiments. (*p<0.05; n=3).

Fig 4. DMOG intraperitoneal injection stabilizes HIF-1α protein in the brain

(A) Adult male mice were injected with 50mg/kg DMOG and sacrificed at the indicated time intervals (hours). Cortical tissue was harvested and immunoblotted for HIF-1α or β actin. Representative western is shown. (B) Densitometry summary of western blots. (n=3–4 animals per time-point; mean ± SEM, *compared to vehicle injected animal, p<0.05).

Fig 5. PHD inhibitor post-ischemic treatment reduces ischemic infarct formation and attenuates peri-infarct loss of cerebral perfusion

Representative TTC staining of ischemic infarct 72h after stroke in (A) control and (B) 60 min DMOG (50mg/kg) post-treatment. (C) Quantification of TTC infarct volume (mm³) 72h after stroke demonstrates the protective effect of PHD inhibitor post-ischemic treatment. (n=10–12 animals per group, *p<0.05). (D) Local cerebral blood flow (CBF) was examined over the territory of the right middle cerebral artery covering the ischemic border region in our focal ischemic stroke model. (E) Pseudo-colored representation of intensity of CBF before, during, and 72h after stroke in control or DMOG post-treated animals. (F) Quantification of CBF relative to initial flow. (n=8, *p<0.05).

Fig 6. PHD inhibitor treatment reduced activation of caspase-3 in the ischemic cortex

(A) Immuno-staining for active caspase-3 and neuronal nuclei (NeuN) 24h after ischemia. (B) Quantification of the number of active caspase-3 positive cells in the peri-infarct region. DMOG post-treatment reduced the number of apoptotic neurons. (n=4, *p<0.05).

Fig 7. PHD inhibitor post-ischemic treatment reduces sensorimotor behavioral deficits after stroke

Adhesive removal test for sensorimotor forelimb function before, 1–3, 7 and 14 days after stroke. Deficits are measured as increased amount of time to remove the adhesive dot in the affected left forepaw (A) and the unaffected right forepaw (B). (*p<0.05, n=10–12 animals per group)

Fig 8. Post-ischemic DMOG therapy enhances HIF-1α expression

(A) Expression of HIF-1α 3h after stroke with vehicle or DMOG 30 min post-treatment. (* p<0.05, n=4) (B) mRNA expression of HIF-1 responsive genes in control stroke animals or DMOG-treated animals 12h after stroke. (* p<0.05, n=3–4)

Fig 9. Inhibition of HIF-1α abrogates the post-ischemic DMOG-mediated neuroprotection

(A) Adult male mice were pre-treated for 6h with saline or HIF-1α inhibitors 2mg/kg DIG, or 2mg/kg ACF. Mice in each group were then treated with DMOG (50mg/kg) or vehicle. After 24h, cortical tissues were immunoblotted for HIF-1α. Densitometry is represented as fold change to vehicle for each condition. DMOG enhances HIF-1α expression compared to vehicle in control conditions, but not under DIG treatment. ACF shows a trend for enhanced expression of HIF-1α with DMOG, but it is not statistically significant. (n=3 animals per group). (B) RNA was extracted from cortical tissues treated with DIG, ACF, and DMOG. Quantitative RT-PCR for the HIF-1-dependent gene VEGF illustrates that DIG and ACF inhibit the DMOG-induced transcription of HIF-1 dependent VEGF expression. (C) Mice were pre-treated for 24h with saline, DIG, or ACF and then underwent surgical dMCAO. Mice in each group were treated 30m after stroke with either DMOG or vehicle. After 12h, peri-infarct cortical tissue was immunoblotted for HIF-1α and actin loading control. DMOG enhances HIF-1α expression compared to sham in control conditions; DIG and ACF reduce the DMOG-mediated increase in HIF-1a. Quantification of densitometry is fold change compared to sham control. (n=3–4, p<0.05) (D) Quantitative RT-PCR analysis of VEGF and EPO from peri-infarct cortical tissue 12h after stroke shows that DIG and ACF reduce the DMOG-mediated increase in expression. All data are fold change relative to sham control. (n=3–4 animals per group, p<0.05) (E) Mice were pre-treated for 24h with HIF inhibitors DIG or ACF. TTC staining of ischemic infarct 72h after stroke with or without DMOG post-treatment shows no protection in the HIF-1α inhibitor groups. (F) Quantification of infarct volume. (n=8–10 in each group, * p<0.05).