Erythropoietin is a paracrine mediator of ischemic tolerance in the brain: evidence from an in vitro model

Abstract

In an in vitro model of cerebral ischemia (oxygen glucose deprivation, OGD) we investigated whether erythropoietin (EPO) plays a critical role in ischemic preconditioning. We found that EPO time and dose-dependently induced protection against OGD in rat primary cortical neurons. Protection was significant at 5 min and reached a maximum at 48 hr after EPO application. Protection was blocked by the coapplication of a soluble Epo receptor (sEpoR) or an antibody against EpoR (anti-EpoR). Medium transfer from OGD-treated astrocytes to untreated neurons induced protection against OGD in neurons, which was attenuated strongly by the application of sEpoR and anti-EpoR. In contrast, medium transfer from OGD-treated neurons to untreated neurons induced protection against OGD that did not involve EPO. In astrocytes the OGD enhanced the nuclear translocation of hypoxia-inducible factor 1 (HIF-1), the major transcription factor regulating EPO expression. Consequently, transcription of EPO-mRNA was increased in astrocytes after OGD. Cultured neurons express EpoR, and the Janus kinase-2 (JAK-2) inhibitor AG490 abolished EPO-induced tolerance against OGD. Furthermore, EPO-induced neuroprotection as well as phosphorylation of the proapoptotic Bcl family member Bad was reduced by the phosphoinositide-3 kinase (PI3K) inhibitor LY294002. The results suggest that astrocytes challenged with OGD provide paracrine protective signals to neurons. We provide evidence for the following signaling cascade: HIF-1 is activated rapidly by hypoxia in astrocytes. After HIF-1 activation the astrocytes express and release EPO. EPO activates the neuronal EPO receptor and, subsequently, JAK-2 and thereby PI3K. PI3K deactivates BAD via Akt-mediated phosphorylation and thus may inhibit hypoxia-induced apoptosis in neurons. Our results establish EPO as an important paracrine neuroprotective mediator of ischemic preconditioning.

Fig. 1.

Primary cortical neurons express the erythropoietin receptor (EpoR), and human recombinant EPO (rhEPO) dose- and time-dependently induces tolerance against OGD in primary cortical neurons. A, Primary cortical neurons express EpoR; expression of EpoR was analyzed by immunocytochemistry in primary cultures of cortical neurons. a, Unstained control (omission of primary antibody). b, Expression of EpoR on neurons after staining with a polyclonal antibody against EpoR (magnification, 400×). Scale bar, 50 μm. B, rhEPO dose-dependently induces tolerance against OGD in primary cortical neurons. Before 120 min of OGD primary cortical neurons were pretreated for 48 hr with 1, 10, or 100 U/l rhEPO, respectively. rhEPO (100 U/l) significantly protected primary cortical neurons from cell death.C, rhEPO time-dependently induces tolerance against OGD in primary cortical neurons. Pretreatment with 100 U/l rhEPO for the indicated intervals induced a fast and prolonged neuroprotection. After 5 min of EPO exposure the neurons already were protected from OGD-induced cell death by 50%. The maximum of protection was observed after a pretreatment period of 24 and 48 hr. D, Compared with EPO-untreated OGD control neurons (Co), preincubation of neurons with 100 U/l EPO for 48 hr resulted in a statistically significant reduction of cell death 24 hr after lethal OGD (Pre), whereas post-treatment with 100 U/l EPO immediately after lethal OGD was not neuroprotective (Post). Cell death in B–D was measured as LDH release into the medium for 24 hr after OGD. Data were obtained from three independent experiments with 16 cell cultures each; data were normalized and presented as means ± SD. Multiple comparisons (Dunn's method) were performed after Kruskal–Wallis one-way ANOVA on ranks (*p > 0.05 vs Co). Normoxic (BSS) stimulation was set to 0% cell death and OGD stimulation (Co) to 100% cell death, respectively.

Fig. 2.

EPO-induced neuroprotection is mediated by interaction with the cognate receptor of EPO. Neuroprotection induced by pretreatment with 100 U/l rhEPO 48 hr before lethal OGD (120 min) was inhibited by the coapplication of either a soluble erythropoietin receptor (sEpoR) or an antibody against the erythropoietin receptor (anti-EpoR, aEpoR). Cell death was measured as LDH release into the medium for 24 hr after OGD. Data were obtained from three independent experiments with 16 cell cultures each; data were normalized and presented as means ± SD. Multiple comparisons (Dunn's method) were performed after Kruskal–Wallis one-way ANOVA on ranks (*p > 0.05 vs all other groups). Normoxic (BSS) stimulation was set to 0% cell death and OGD stimulation (Co) to 100% cell death, respectively.

Fig. 3.

Fluorescence microscopy (A), TUNEL assay (B), and DNA laddering (C) demonstrate the neuroprotective effect of EPO in OGD-induced neuronal apoptosis. Shown are micrographs of ethidium bromide/acridine orange-stained (A; magnification 400×) and TUNEL-stained (B; magnification 200×) primary cortical neurons 24 hr after OGD with or without EPO pretreatment.a, Normoxic control; b, OGD for 120 min without EPO pretreatment; c, OGD for 120 min with EPO pretreatment. Cultured neurons were stimulated by 100 U/l rhEPO 48 hr before 120 min of OGD. d, OGD for 120 min with EPO pretreatment and block by anti-EpoR. Stimulation of cultured neurons by 100 U/l rhEPO 48 hr before 120 min of OGD was blocked by the coapplication of 2.5 μg/ml anti-EpoR. The insets(magnification 400×) in micrographs Bb andBd demonstrate apoptotic bodies. Scale bars: (ind) a–d, 50 μm; inset, 20 μm. C, OGD-induced DNA fragmentation was abolished by pretreatment with EPO. DNA fragmentation was analyzed 24 hr after lethal OGD (120 min) of primary cortical neurons with or without EPO pretreatment. Lane M, 1 kb ladder; lane 1, BSS-treated cells; lane 2, OGD-treated cells;lane 3, OGD-treated cells preincubated with 100 U/l rhEPO 48 hr before OGD; lane 4, OGD-treated cells preincubated with 100 U/l rhEPO and 2.5 μg/ml sEpoR 48 hr before OGD.

Fig. 4.

Astrocytic HIF-1 activation and erythropoietin expression are sequential events that follow oxygen glucose deprivation. A, Astrocytes produce erythropoietin in response to OGD. Astrocytes were stimulated by OGD or control conditions for 180 min. Shown are different time points after OGD cells were harvested and total RNA was isolated and reverse transcribed. Using a quantitative competitive RT-PCR approach and the housekeeping gene β-actin as an internal standard, we determined EPO mRNA expression. Data were obtained from three independent experiments, presented as the means ± SD of arbitrary units (AU). B, In accordance with the mRNA expression of EPO, the protein was detectable 60 min after OGD. Shown are Western blot analysis from control astroglial cultures (lane 1), OGD-stimulated cultures (lane 2), and positive control (lane 3). The molecular size of the protein standard (lane M) is indicated. C, Induction of astroglial EPO expression is associated with the preceding induction of HIF-1 binding activity. Astrocytes were stimulated by OGD or control conditions for 180 min. Nuclear extracts were prepared immediately, and 30 μg of each was tested for HIF-1 DNA binding activities, using an fEMSA approach. Shown are probes without nuclear extract (lane 1), probes and nuclear extracts from BSS-stimulated astrocytes (lane 2), probes and nuclear extracts from OGD-stimulated astrocytes (lane 3), probes and nuclear extract from OGD-stimulated astrocytes and the addition of an unspecific competitor (lane 4; 50-fold), probes and nuclear extract from OGD-stimulated astrocytes and the addition of a specific competitor (lane 5; 50-fold), and probes and nuclear extract from OGD-stimulated astrocytes and the addition of a specific antibody against HIF-1α (lane 6).C, Constitutive DNA binding activity; H, specific HIF-1 DNA complex; S, supershifted HIF-1 DNA complex.

Fig. 5.

Preconditioned medium from astrocytes induces ischemic tolerance in cortical neurons. A, Experimental paradigm. Astrocytes were stimulated for 180 min either by OGD (preconditioning) or with BSS medium under normoxia (no preconditioning), and media were harvested 24 hr later.B, Neuronal cultures, pretreated with medium either from OGD-stimulated (preconditioning, P, bottom row) or from BSS-stimulated (no preconditioning, N, bottom row) astrocytes for 48 hr, were exposed to OGD (H, top row) or BSS (N, top row) for 120 min, respectively. Cell death, measured as LDH release into the medium for 24 hr after OGD, was reduced significantly by OGD-conditioned medium from astrocytes. This neuroprotection was diminished by 2.5 μg/ml sEpoR and 2.5 μg/ml anti-EpoR, indicating that EPO is the neuroprotective factor. Data were obtained from three independent experiments with 16 cell cultures each; data were normalized and presented as means ± SD. Multiple comparisons (Dunn's method) were performed after Kruskal–Wallis one-way ANOVA on ranks (*p > 0.05 vs N/H, P/H + sEpoR, or P/H + aEpoR). N represents normoxic (BSS) and H represents hypoxic (OGD) treatment.N/N treatment was set to 0% cell death and N/H treatment to 100% cell death, respectively.

Fig. 6.

Neuro-neuronal preconditioning is not mediated by erythropoietin. A, Experimental paradigm. Cortical neurons were stimulated either by OGD for 60 min (preconditioning) or with BSS medium under normoxia (no preconditioning), and both media were harvested 24 hr later. B, Neuronal cultures, pretreated with medium either from OGD-stimulated (preconditioning,P, bottom row) or from BSS-stimulated (no preconditioning, N, bottom row) cortical neurons for 48 hr, were exposed to OGD (H, top row) or BSS (N, top row) for 120 min. Cell death, measured as LDH release into the medium for 24 hr after OGD, was reduced significantly in neuronal cultures treated with OGD-conditioned medium from neurons. However, this neuroprotection was diminished neither by 2.5 μg/ml sEpoR nor by 2.5 μg/ml anti-EpoR, excluding EPO as the neuroprotective factor. Data were obtained from three independent experiments with 16 cell cultures each; data were normalized and presented as means ± SD. Multiple comparisons (Dunn's method) were performed after Kruskal–Wallis one-way ANOVA on ranks (*p > 0.05 vs N/H treatment). N represents normoxic (BSS) andH represents hypoxic (OGD) treatment.N/N treatment was set to 0% cell death and N/H treatment to 100% cell death, respectively.

Fig. 7.

Erythropoietin-induced neuroprotection is mediated by Janus kinase-2 (JAK-2). Cortical neurons, pretreated with 100 U/l rhEPO and/or AG490 (5 μm) for 48 hr, were exposed to 120 min of OGD or BSS. Coapplication of the JAK-2 inhibitor AG490 abolished the neuroprotective effect of 100 U/l rhEPO (100 U/l vs100 U/l + AG). Neuronal viability under hypoxic (AG) or normoxic (data not shown) conditions is not altered by JAK-2 inhibition alone. Cell death was measured as LDH release into the medium for 24 hr after OGD. Data were obtained from three independent experiments with eight cell cultures each; data were normalized and presented as means ± SD (*p > 0.05 vs 100 U/l EPO; Mann–Whitney rank sum test). Normoxic (BSS) stimulation was set to 0% cell death and OGD stimulation (Co) to 100% cell death, respectively.

Fig. 8.

JAK-2/STAT pathway is not involved in EPO-mediated neuroprotection. A, EPO stimulation of neurons does not activate the transcription factors STAT1, STAT3, and STAT5, as assessed by fEMSA. Shown are probes without nuclear extract (lane 1); probes and positive controls for STAT1 from phorbol ester-treated HeLa cells and for STAT3 and STAT5 from phorbol ester-treated K562 cells (lane 2); probes, positive controls, and a 50-fold excess of specific competitor against STAT1, STAT3, or STAT5, respectively (lane 3); probes and extract from untreated neurons (lane 4); and probes and extract from neurons treated by 100 U/l rhEPO for 30 min (lane 5). F, Free probe;S, specific protein DNA complex. B, Transcriptional pattern of anti- and pro-apoptotic Bcl-2 family genes is not altered in neurons by erythropoietin stimulation. Cortical neurons were stimulated by 100 U/l rhEPO for 0, 12, or 48 hr, respectively. EPO-stimulated neurons (black columns) were compared with untreated control neurons (white columns) and with neurons treated with the combination of 100 U/l rhEPO and 2.5 μg/ml sEpoR (gray columns). Using a quantitative real-time RT-PCR approach and the housekeeping gene β-actin as an internal standard, we found no induction or repression of mRNA expression of Bcl-2, Bcl-X_L, Bag-1, Bax, and Bad. Data were obtained from three independent experiments, presented as means ± SD of arbitrary units (AU).

Fig. 9.

Erythropoietin-induced neuroprotection is mediated by phosphoinositol-3 kinase (PI3K). Pretreatment of neurons for a short (1 hr, A) and a prolonged (24 hr, B) incubation period with the specific PI3K inhibitor LY294002 (10 μm) immediately before the application of 100 U/l EPO partially abolished the neuroprotective effect of EPO against 120 min of OGD (100 U/l vs 100 U/l + LY). Pretreatment with LY294002 alone did not influence neuronal viability under hypoxic (LY) or normoxic (data not shown) conditions. Cell death was measured as LDH release into the medium for 24 hr after OGD. Data were obtained from three independent experiments with eight cell cultures each; data were normalized and presented as means ± SD (*p > 0.05 vs 100 U/l EPO; Mann–Whitney rank sum test). Normoxic (BSS) stimulation was set to 0% cell death and OGD stimulation (Co) to 100% cell death, respectively.

Fig. 10.

Erythropoietin-induced neuroprotection is associated with the activation of Akt kinase and the phosphorylation of BAD. rhEPO (100 U/l) for 30 min activated Akt kinase and phosphorylated BAD, but not p38 and p44/42 MAPK, in cortical neurons. EPO-induced activation of Akt kinase and BAD phosphorylation is blocked by 2.5 μg/ml sEpoR as well as by the specific PI3K inhibitor LY294002 (10 μm). Shown are Western blot analysis from control cultures (lane 1), EPO-stimulated cultures (lane 2), sEpoR- and EPO-treated cultures (lane 3), and LY294002- and EPO-treated cultures (lane 4). The molecular sizes of the protein standards (lane M) are indicated (left). GSK3 indicates the substrate protein in the Akt kinase assay.