Thrombolysis by PLAT/tPA increases serum free IGF1 leading to a decrease of deleterious autophagy following brain ischemia

Abstract

Cerebral ischemia is a pathology involving a cascade of cellular mechanisms, leading to the deregulation of proteostasis, including macroautophagy/autophagy, and finally to neuronal death. If it is now accepted that cerebral ischemia induces autophagy, the effect of thrombolysis/energy recovery on proteostasis remains unknown. Here, we investigated the effect of thrombolysis by PLAT/tPA (plasminogen activator, tissue) on autophagy and neuronal death. In two in vitro models of hypoxia reperfusion and an in vivo model of thromboembolic stroke with thrombolysis by PLAT/tPA, we found that ischemia enhances neuronal deleterious autophagy. Interestingly, PLAT/tPA decreases autophagy to mediate neuroprotection by modulating the PI3K-AKT-MTOR pathways both in vitro and in vivo. We identified IGF1R (insulin-like growth factor I receptor; a tyrosine kinase receptor) as the effective receptor and showed in vitro, in vivo and in human stroke patients and that PLAT/tPA is able to degrade IGFBP3 (insulin-like growth factor binding protein 3) to increase IGF1 (insulin-like growth factor 1) bioavailability and thus IGF1R activation.Abbreviations: AKT/protein kinase B: thymoma viral proto-oncogene 1; EGFR: epidermal growth factor receptor; Hx: hypoxia; IGF1: insulin-like growth factor 1; IGF1R: insulin-like growth factor I receptor; IGFBP3: insulin-like growth factor binding protein 3; Ka: Kainate; MAP1LC3/LC3: microtubule-associated protein 1 light chain 3; MAPK/ERK: mitogen-activated protein kinase; MTOR: mechanistic target of rapamycin kinase; MTORC1: MTOR complex 1; OGD: oxygen and glucose deprivation; OGD_reox: oxygen and glucose deprivation + reoxygentation; PepA: pepstatin A1; PI3K: phosphoinositide 3-kinase; PLAT/tPA: plasminogen activator, tissue; PPP: picropodophyllin; SCH77: SCH772984; ULK1: unc-51 like kinase 1; Wort: wortmannin.

Keywords: IGF1R; IGFBP3; LC3; MTORC1; SQSTM1/p62; stroke.

Figure 1.

Autophagy pathways are increased by oxygen and glucose deprivation in cortical neurons. (A) Pie charts representing significantly (p < 0.05) up- and downregulated transcripts in Ctrl Vs OGD_reox, Ctrl Vs OGD_reox + PLAT/tPA and OGD_reox Vs OGD_reox + PLAT/tPA conditions. Heatmaps of genes significantly affected (p < 0.05) in both OGD_reox and OGD_reox + PLAT/tPA conditions Vs Ctrl condition for (B) MAPK signaling pathway, (C) PI3K-AKT signaling pathway, (D) MTOR signaling pathway, and (E) autophagy pathway. Heatmaps of genes significantly affected (p < 0.05) between Ctrl and OGD_reox conditions for (F) MAPK signaling pathway, (G) PI3K-AKT signaling pathway, (H) MTOR signaling pathway, and (I) autophagy pathway. Heatmaps of genes significantly affected (p < 0.05) between Ctrl and OGD_reox + PLAT/tPA conditions for (J) MAPK signaling pathway, (K) PI3K-AKT signaling pathway, (L) MTOR signaling pathway, and (M) autophagy pathway.

Figure 2.

PLAT/tPA decreases OGD-induced autophagy and cell death in cortical neurons. (A) Schematic representation of in vitro ischemic model consisting in 1 h of OGD followed by 3 h of reoxygenation (OGD_reox), on 12 days old murine primary cortical neurons in the presence or not of treatments. (B) Representative western blots of LC3-II, SQSTM1/p62 and ACTB in neurons subjected or not to OGD_reox with and without PLAT/tPA (300 nM). Densitometric quantification of LC3-II (C) and SQSTM1/p62 (D) normalized to ACTB (mean±S.E.M. n = 5 and 6 independent experiments for LC3 and SQSTM1/p62 respectively; ##: p < 0.01 and #: p < 0.05 compared to Ctrl; **: p < 0.01 and *: p < 0.05 compared between OGD_reox conditions; Mann–Whitney test). (E) Neuronal death of pure cortical neurons assessed by LDH release after OGD_reox in the presence or not of 10 µg/ml of E64d and 10 µg/ml pepstatin A (mean±S.E.M. n = 12 for Ctrl, n = 16 for Ctrl+E64d-PepA, n = 31 for OGD_reox, n = 33 for OGD_reox+E64d-PepA, n = 28 for OGD_reox+ PLAT/tPA from 4 independent experiments; ###: p < 0.001 compared to Ctrl; ***: p < 0.001 compared between OGD_reox; Mann–Whitney test). (F) Representative western blots of LC3-II, SQSTM1/p62 and ACTB in neurons subjected or not to OGD_reox in the presence or not of 10 µg/ml of E64d and 10 µg/ml Pepstadin A. Densitometric quantification of LC3-II (G) and SQSTM1/p62 (H) normalized to ACTB (mean±S.E.M. n = 8 independent experiments; ##: p < 0.01 and ###: p < 0.001 compared to Ctrl; ***: p < 0.001 compared between OGD_reox; Mann–Whitney test.

Figure 3.

PLAT/tPA prevents hypoxic/excitotoxic-induced autophagy and neuronal death. (A-C) PLAT/tPA (300 nM) affords strong neuroprotection against cell death induced by the combination of kainate (Ka, 30 µM) and hypoxia (Hx, 6% oxygen) for 30 min in primary cortical neuron cultures as shown by a reduction in both (A) the release of lactate dehydrogenase (LDH) and (B-C) propidium iodide (PI)-staining 6 h after KaHx treatment. (A) Quantification of released LDH expressed as a percentage of the KaHx value (data are represented as mean±S.E.M. n = 8 from 3 independent experiments; ###: p < 0.001 compared to Ctrl; ***: p < 0.001 compared between KaHx; Tukey’s multiple comparisons tests) (B) Quantification of the PI-positive nuclei expressed as a percentage of the number of KaHx PI-positive nuclei (data are represented as mean±S.E.M. n = 24 from 3 independent experiments; ###: p < 0.001 compared to Ctrl; ***: p < 0.001 and **: p < 0.01 compared between KaHx; Tukey’s multiple comparisons tests) (C) Representative images of PI- (magenta) and Hoechst- (cyan) stained nuclei. Bars = 100 µm. (D-F) PLAT/tPA treatment prevents KaHx-induced autophagy as indicated by a suppression of both KaHx-induced (E) increase in LC3-II level and (F) degradation of SQSTM1/p62. (D) Representative western blots of LC3-II and SQSTM1/p62 in neurons subjected or not to KaHx in the presence of PLAT/tPA 0 to 300 nM. ACTB was used as loading control. (E) Densitometric quantification of LC3-II normalized to ACTB. Levels expressed as a percentage of KaHx value (data are represented as mean±S.E.M. n = 8 independent experiments; ##: p < 0.01 compared to Ctrl; **: p < 0.01 compared between KaHx; Tukey’s multiple comparisons tests). (F) Densitometric quantification of SQSTM1/p62 normalized to ACTB. Levels expressed as a percentage of KaHx value (data are represented as mean±S.E.M. n = 8 independent experiments; ###: p < 0.001 and ##: p < 0.01 compared to Ctrl; ***: p < 0.001 compared between KaHx; Tukey’s multiple comparisons tests). (G) Representative confocal images of transfected cortical neurons with a tandem mRFP-GFP-LC3-expressing plasmid in control (Ct) or kainate/hypoxia (KaHx) condition with or without PLAT/tPA (300 nM) treatment. (H-I) Corresponding quantifications of the number of dots per µm² of (H) GFP⁺ (yellow) RFP⁺ (magenta) (early autophagosomes) (I) GFP⁻ RFP⁺ (mature autophagosomes/autolysosomes) and (J) the total of the two types (total dots). PLAT/tPA treatment prevents the increase in the autophagy flux (both autophagosome formation and autophagosome/lysosome fusion) induced after 3 h of KaHx stimulation. Data are represented as mean±S.E.M. n = 20; ###: p < 0.001 compared to Ctrl; ***: p < 0.001 compared with KaHx; Tukey’s multiple comparisons tests).

Figure 4.

OGD-enhanced autophagy is dependent on MTOR pathway. (A) Representative western blots of p-MTORC1 (Ser2448), p-MTORC2 (Ser2481), MTOR total, p-ULK1-ULK2 (Ser757) and ACTB in neurons subjected or not to OGD_reox with and without PLAT/tPA (300 nM). Densitometric quantification of p-MTORC1 (Ser2448) normalized to MTOR total (B), of p-MTORC2 (Ser2481) normalized to MTOR total (C), of p-ULK1-ULK2 (Ser757) normalized to ACTB (D) (mean±S.E.M. n = 4 to 6 independent experiments; ##: p < 0.01 and #: p < 0.05 compared to Ctrl; **: p < 0.01 and *: p < 0.05 compared to OGD_reox; Mann–Whitney test).

Figure 5.

PLAT/tPA decreases OGD-induced autophagy through the PI3K-AKT pathway. (A) Representative western blots of p-MAPK/ERK1/2 (Thr202, Tyr204), MAPK/ERK1/2 total, p-AKT (Ser473) and AKT total in neurons subjected or not to OGD_reox with and without PLAT/tPA (300 nM). Densitometric quantification of p-MAPK/ERK1/2 (Thr202, Tyr204) normalized to MAPK/ERK1/2 total (B) and of p-AKT (Ser473) normalized to AKT total (C) (mean±S.E.M. n = 5 independent experiments; ##: p < 0.01 compared to Ctrl; **: p < 0.01 and p < 0.05 compared between OGD_reox conditions; Mann–Whitney test). (D) Representative western blots of LC3-II, SQSTM1/p62 and ACTB in neurons subjected or not to OGD_reox with and without PLAT/tPA (300 nM) in the presence or not of 5 μM SCH772984 (SCH77; MAPK/ERK1/2 inhibitor). Densitometric quantification of LC3-II (E) and SQSTM1/p62 (F) normalized to ACTB (mean±S.E.M. n = 6 independent experiments; #: p < 0.05 compared to Ctrl; **: p < 0.01 and *: p < 0.05 compared between OGD_reox conditions; Mann–Whitney test). (G) Representative western blots of LC3-II, SQSTM1/p62 and ACTB in neurons subjected or not to OGD_reox with and without PLAT/tPA (300 nM) in the presence or not of 10 μM A6730 (AKT inhibitor). Densitometric quantification of LC3-II (H) and SQSTM1/p62 (I) normalized to ACTB (mean±S.E.M. n = 5 and 6 independent experiments for LC3 and SQSTM1/p62 respectively; ##: p < 0.01 and #: p < 0.05 compared to Ctrl; **: p < 0.01 and *: p < 0.05 compared between OGD_reox conditions; Mann–Whitney test). (J) Representative western blots of LC3-II, SQSTM1/p62 and ACTB in neurons subjected or not to OGD_reox with and without PLAT/tPA (300 nM) in the presence or not of 250 nM wortmannin (Wort; PI3K inhibitor). Densitometric quantification of LC3-II (K) and SQSTM1/p62 (L) normalized to ACTB (as mean±S.E.M. n = 5 independent experiments; ##: p < 0.01 and #: p < 0.05 compared to Ctrl; *: p < 0.05 compared between OGD_reox conditions; Mann–Whitney test).

Figure 6.

PLAT/tPA decreases OGD-induced PI3K-AKT signaling through IGF1R receptor. (A) Representative western blots of LC3-II, SQSTM1/p62 and ACTB in neurons subjected or not to OGD_reox with and without PLAT/tPA (300 nM) in the presence or not of 5 µM AG1478 (EGFR inhibitor). Densitometric quantification of LC3-II (B) and SQSTM1/p62 (C) normalized to ACTB (mean±S.E.M. n = 4 and 5 independent experiments for LC3 and SQSTM1/p62 respectively; ##: p < 0.01 and #: p < 0.05 compared to Ctrl; **: p < 0.01 and*: p < 0.05 compared between OGD_reox conditions; Mann–Whitney test). (D) Representative western blots of p-AKT (Ser473), AKT total, p-MTORC1 (Ser2448) and MTOR total in neurons subjected or not to OGD_reox with and without PLAT/tPA (300 nM) in the presence or not of 100 nM Picropodophyllotoxin. Densitometric quantification of p-AKT (Ser473) normalized to AKT total (E) and of MTORC1 (Ser2448) normalized to MTOR total (F) (mean±S.E.M. n = 4 independent experiments; #: p < 0.05 compared to Ctrl; *: p < 0.05 compared between OGD_reox conditions; Mann–Whitney test). (G) Representative western blots of LC3-II, SQSTM1/p62 and ACTB in neurons subjected or not to OGD_reox with and without PLAT/tPA (300 nM) in the presence or not of 100 nM Picropodophyllotoxin (PPP; IGF1R inhibitor). Densitometric quantification of LC3-II (H) and of SQSTM1/p62 (I) normalized to ACTB (mean±S.E.M. n = 5 independent experiments; ##: p < 0.01 compared to Ctrl; **: p < 0.01 compared between OGD_reox; Mann–Whitney test). (J) Representative western blots of LC3-II, SQSTM1/p62 and ACTB in neurons subjected or not to OGD_reox with and without IGF1 (250 nM). Densitometric quantification of LC3-II (K) and SQSTM1/p62 (L) normalized to ACTB (mean±S.E.M. n = 6 and 7 independent experiments for LC3 and SQSTM1/p62 respectively; ###: p < 0.005 and ##: p < 0.01 compared to Ctrl; **: p < 0.01 compared between OGD_reox conditions; Mann–Whitney test).

Figure 7.

Active PLAT/tPA protects neurons from cell death by decreasing autophagy through IGF1R/PI3k-AKT/MTORC1. Neuronal death of pure cortical neurons assessed by LDH release after OGD_reox in the presence or not of PLAT/tPA and 5 μM SCH772984 (SCH77; MAPK/ERK1/2 inhibitor) (A), 250 nM Wortmannin (Wort; PI3K inhibitor) (B), 10 μM A6730 (AKT inhibitor) (C), 100 nM Picropodophyllotoxin (PPP; IGF1R inhibitor) (D), 300 nM d-GGACK- PLAT/tPA (inactive PLAT/tPA) (E) or 1 μM Aprotinin (plasmin inhibitor) (F) (mean±S.E.M. n = 14 to 30 from 6 independent experiments; ###: p < 0.001 compared to Ctrl; ***: p < 0.001, **: p < 0.01 and *: p < 0.05 compared between OGD_reox; Mann–Whitney test).

Figure 8.

PLAT/tPA decreases autophagy and ischemic lesion in a mouse model of cerebral ischemia. (A) Schematic representation of Middle Cerebral Artery occlusion (MCAo) based on thrombin injection in the MCA. (B) Representative MRI images and infarct volumes (mm³) 24 h after MCAo in mice treated with vehicle or PLAT/tPA. (mean±S.E.M; n = 10 for Vehicle; n = 9 for PLAT/tPA; *: p < 0.05; t test). Grip strength of both forepaws (C), right forepaw (D), or left forepaw (E) in mice treated with vehicle or PLAT/tPA after MCAo. Measurements were picked up the day before MCAo surgery (baseline acquisition) and 24 h after MCAo surgery, ratio between the two values is calculated in order to normalize each mouse with its own baseline values (mean±S.E.M; n = 10 for Vehicle; n = 9 for PLAT/tPA; ***: p < 0.001, *: p < 0.05; t test) (F) Representative western blots of LC3-II, SQSTM1/p62 and ACTB in ipsilateral cortex of mice after MCAo. Densitometric quantification of LC3-II (G) or SQSTM1/p62 (H) normalized to ACTB (mean±S.E.M. n = 4 independent experiments; *: p < 0.05; Mann–Whitney test). (I) Representative western blots of p-MTOR (Ser2448), MTOR tot, LC3-II, SQSTM1/p62, ATG5, ATG7 and ACTB in cortex treated with vehicle or PLAT/tPA after MCAo. Densitometric quantification of p-MTORC1 (Ser2448) normalized to MTOR total (J), of LC3-II normalized to ACTB (K), of SQSTM1/p62 normalized to ACTB (L), of ATG5 normalized to ACTB (M), of ATG7 normalized to ACTB (N) (mean±S.E.M. n = 8 to 10; *: p < 0.05; Mann–Whitney test).

Figure 9.

PLAT/tPA cleaves IGFBP3 and increases the bioavailability of IGF1 in vitro, in vivo in mice and in humans. (A) Human recombinant IGFBP3 was incubated or not with PLAT/tPA during 15 min, 1 h, 2 h or 4 h. Samples were then analyzed to SDS/PAGE and western blotting probed with anti-IGFBP3 antibody. (B) Concentration of IGFBP3 (ng/ml) measured by ELISA in cell culture medium of cortical neurons subjected or not to OGD_reox with or without PLAT/tPA (300 nM) (mean±S.E.M. n = 8 independent experiments; ***: p < 0.001; t test). (C) Concentration of IGFBP3 (ng/ml) measured by ELISA in serum of mice treated with vehicle or PLAT/tPA after MCAo. Serum was collected 1 h and 24 h after treatment (mean±S.E.M; n = 8 for vehicle 1 h; n = 9 for PLAT/tPA 1 h; n = 8 for vehicle 24 h; n = 7 for PLAT/tPA 1 h; ##: p < 0.01 vehicle 1 h compared to PLAT/tPA 1 h; $: p < 0.05 vehicle 24 h compared to PLAT/tPA 24 h; Mann-Whitney test). Total IGF1 (ng/ml) (D), Free IGF1 (ng/ml) (E), and ratio between free and total IGF1 (F) measured by ELISA in serum of mice treated with vehicle or PLAT/tPA after MCAo. Serum was collected 1 h and 24 h after treatment (mean±S.E.M; n = 8 for vehicle 1 h; n = 9 for PLAT/tPA 1 h; n = 8 for vehicle 24 h; n = 7 for PLAT/tPA 1 h; ###: p < 0.001, ##: p < 0.01 vehicle 1 h compared to PLAT/tPA 1 h; $$$: p < 0.001, $$: p < 0.01 vehicle 24 h compared to PLAT/tPA 24 h; ***: p < 0.001, *: p < 0.05 PLAT/tPA 1 h compared to PLAT/tPA 24 h; Mann-Whitney test). (G) Concentration of IGFBP3 (ng/ml) was measured by ELISA in serum of stroke patients before PLAT/tPA injection (baseline) and 1 h, 2 h and 24 h after t PLAT/tPA injection (mean±S.E.M; n = 5 patients; *: p < 0.05; Paired t test).