SUMOylation promotes survival and integration of neural stem cell grafts in ischemic stroke

Abstract

Background: Neural stem cell (NSC)-based therapies hold great promise for treating diseases of the central nervous system (CNS). However, several fundamental problems still need to be overcome to fully exploit the clinical potential of NSC therapeutics. Chief among them is the limited survival of NSC grafts within hostile microenvironments.

Methods: Herein, we sought to engineer NSCs in an effort to increase graft survival within ischemic brain lesions via upregulation of global SUMOylation, a post-translational modification critically involved in mediating tolerance to ischemia/reperfusion.

Findings: NSCs overexpressing the SUMO E2-conjugase Ubc9 displayed resistance to oxygen-glucose-deprivation/restoration of oxygen/glucose (OGD/ROG) and enhanced neuronal differentiation in vitro, as well as increased survival and neuronal differentiation when transplanted in mice with transient middle cerebral artery occlusion in vivo.

Interpretation: Our work highlights a critical role for SUMOylation in NSC biology and identifies a biological pathway that can be targeted to increase the effectiveness of exogenous stem cell medicines in ischemic stroke. FUND: Intramural Research Program of the NINDS/NIH, the Italian Multiple Sclerosis Foundation (FISM), the Bascule Charitable Trust, NIH-IRTA-OxCam and Wellcome Trust Research Training Fellowships.

Keywords: Cell therapy; Cellular engineering; Ischemia/reperfusion; Neural stem cells (NSCs); Regenerative medicine; SUMOylation; Stroke; Ubc9.

Graphical abstract

Fig. 1

Overexpression of Ubc9 induces genes and proteins related to cell signaling, cell metabolism and cell cycle. (A) Representative confocal images of WT NSCs and Ubc9 NSCs growing in vitro as SOX2⁺/Vimentin (Vim)⁺ neurospheres. Nuclei are stained with DAPI. See also Fig. S1. (B) Genomic PCR for CAG-UBE2I transgene in Ubc9 NSCs (+: positive control). (C) Representative immunoblots and densitometries normalized to corresponding actin levels and expressed as a fold induction (FI) relative to WT NSCs. Data are means ± SEM. N ≥ 3 per group, * p-value<0.05 (2-tailed Student's t-test). (D) Global SUMO-1 pulldown in WT NSCs and Ubc9 NSCs. Input = precleared lysate, IP = immunoprecipitated lysate. See also Data S1. (E) Venn diagram of protein mass spectrometry data showing that 125 proteins co-immunoprecipitated with SUMO-1 in WT NSCs only, 815 proteins in Ubc9 NSCs only, and 759 were common to both. Pathways and functions significantly upregulated in Ubc9 NSCs compared to WT NSCs based on protein mass spectrometry results are reported. N = 2 per group (Fisher's Exact Test). See also Data S1 and Fig. S1. (F) Heatmap of genes differentially expressed in Ubc9 NSCs vs WT NSCs. N = 4 per group, adjusted p-value<0.05. See also Data S2. (G) Bar chart of the fold change (log2) in Ubc9 NSCs vs WT NSCs of the 10 most upregulated and downregulated genes. Adjusted p-value<0.05. See also Data S2. (H) Plot of the GO enrichment results for genes differentially expressed in Ubc9 NSCs vs WT NSCs. The x-axis shows the enrichment ratio, the color indicates the enrichment p-value (Kolmogorov–Smirnov statistic) and the size of each point indicates the number of significant genes in the corresponding GO category. See also Data S2.

Fig. 2

Ubc9 NSCs show a quiescent metabolic profile upon differentiation and are predisposed to neuronal differentiation in vitro. (A) Heatmap of genes differentially expressed in Ubc9 Diff vs WT Diff cells. N = 4 per group, adjusted p-value<0.05. See also Data S2. (B) Bar chart of the fold change (log2) in Ubc9 Diff vs WT Diff cells of the 10 most upregulated and downregulated genes. All genes have an adjusted p-value<0.05. See also Data S2. (C) Plot of the GO enrichment results for genes differentially expressed in Ubc9 Diff vs WT Diff cells. The x-axis shows the enrichment ratio, the color indicates the enrichment p-value (Kolmogorov–Smirnov statistic), and the size of each point indicates the number of significant genes in the corresponding GO category. The table at the bottom shows the 6 KEGG pathways showing the strongest enrichment (GAGE analysis p-value<0.05 for all pathways, see methods) in the same comparison. See also Data S2. (D) OCR vs ECAR graph of undifferentiated and differentiated WT and Ubc9 NSCs. Data are expressed as mean pmoles/min (OCR) and mean milli-pH units (mpH)/min (ECAR) ± SEM. N = 7 per group from 2 independent experiments. (E) Quantification and representative immunofluorescence-staining of WT Diff and Ubc9 Diff NSCs. Data are means ± SEM. N ≥ 6 per group from 2 independent experiments, **p-value<0.01 (2-tailed Student's t-test). Nuclei are stained with DAPI, scale bars: 50 μm.

Fig. 3

Ubc9 gain of function NSCs are protected against OGD/ROG in vitro. (A) Heatmap showing expression profile (z-score normalized intensities) of genes with a statistically-significant interaction (adjusted p-value<0.05) between the effect of OGD/ROG vs normoxia (control) and the effect of Ubc9 overexpression vs WT NSCs. The colored dendrogram shows the three expression clusters described in the text. See also Fig. S2, S3, S4 and Data S3. (B) Representative density plot and quantification of WT and Ubc9 undifferentiated NSCs post-OGD/ROG for live/dead analysis. Data are means ± SEM. N = 11 per group from 3 independent experiments, ***p-value<0.001 (2-tailed Student's t-test). (C) Representative density plot and quantification of WT and Ubc9 differentiated NSCs post-OGD/ROG for live/dead analysis. X axis (Hoechst blue), Y axis (PI). Data are means ± SEM. N ≥ 8 per group from 3 independent experiments, *p-value<0.05 (2-tailed Student's t-test). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 4

Transplanted Ubc9 NSCs display increased survival and form more neurons in the stroke brain. (A) Stereological quantification and representative IHC images of transplanted GFP⁺ WT and Ubc9 NSCs at 5 dpt revealing greater survival of Ubc9 NSCs. N ≥ 3 per group. (B) Quantification and representative images of Ki67⁺ and TUNEL⁺ transplanted GFP⁺ cells at 5 dpt. N ≥ 3 per group. Nuclei are stained with DAPI. (C) Representative 3D reconstructions of ischemic brains at 30 days post-transplantation (dpt) of mice transplanted with WT or Ubc9 NSCs (blue: contralateral hemisphere, green: healthy ischemic ipsilateral hemisphere, red: ischemic lesion). Data at 5 and 30 dpt are expressed as percentage of contralateral hemisphere. N ≥ 3 per group. (D) Representative images and quantification of GFP⁺ WT and Ubc9 NSCs stained for GFAP, OLIG2, and NeuN retrieved at 30 dpt within the ischemic hemisphere. Significantly more numerous GFP⁺ Ubc9 NSCs were found to be NeuN⁺ compared to GFP⁺ WT NSCs. (E) Representative images and quantification of GFP⁺ WT and Ubc9 NSCs stained for MAP2, and PSD95 retrieved at 30 dpt within the ischemic hemisphere. Significantly higher MAP2 and PSD95 expression is found in GFP⁺ areas, suggesting an increase of dendrite formation and integration of synapses, respectively. Nuclei are stained with DAPI. Data in A-E are means ± SEM. *p-value<0.05 (2-tailed Student's t-test). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)