Modulation of gene expression on a transcriptome-wide level following human neural stem cell transplantation in aged mouse stroke brains

Abstract

Introduction: Neural stem cell (NSC) transplantation offers great potential for treating ischemic stroke. Clinically, ischemia followed by reperfusion results in robust cerebrovascular injury that upregulates proinflammatory factors, disrupts neurovascular units, and causes brain cell death. NSCs possess multiple actions that can be exploited for reducing the severity of neurovascular injury. Our previous studies in young adult mice showed that human NSC transplantation during the subacute stage diminishes stroke pathophysiology and improves behavioral outcome.

Methods: We employed a well-established and commonly used stroke model, middle cerebral artery occlusion with subsequent reperfusion (MCAO/R). Here, we assessed the outcomes of hNSC transplantation 48 h post-MCAO (24 h post-transplant) in aged mouse brains in response to stroke because aging is a crucial risk factor for cerebral ischemia. Next, we tested whether administration of the integrin α5β1 inhibitor, ATN-161, prior to hNSC transplantation further affects stoke outcome as compared with NSCs alone. RNA sequencing (RNA-seq) was used to assess the impact of hNSC transplantation on differentially expressed genes (DEGs) on a transcriptome-wide level.

Results: Here, we report that hNSC-engrafted brains with or without ATN-161 showed significantly reduced infarct size, and attenuated the induction of proinflammatory factors and matrix metalloproteases. RNA-seq analysis revealed DEGs and molecular pathways by which hNSCs induce a beneficial post-stroke outcome in aged stroke brains. 811 genes were differentially expressed (651 downregulated and 160 upregulated) in hNSC-engrafted stroke brains. Functional pathway analysis identified enriched and depleted pathways in hNSC-engrafted aged mouse stroke brains. Depletion of pathways following hNSC-engraftment included signaling involving neuroinflammation, acute phase response, leukocyte extravasation, and phagosome formation. On the other hand, enrichment of pathways in hNSC-engrafted brains was associated with PPAR signaling, LXR/RXR activation, and inhibition of matrix metalloproteases. Hierarchical cluster analysis of DEGs in hNSC-engrafted brains indicate decreased expression of genes encoding TNF receptors, proinflammatory factors, apoptosis factors, adhesion and leukocyte extravasation, and Toll-like receptors.

Conclusions: Our study is the first to show global transcripts differentially expressed following hNSC transplantation in the subacute phase of stroke in aged mice. The outcome of our transcriptome study would be useful to develop new therapies ameliorating early-stage stroke injury.

Keywords: Blood-brain barrier; Inflammation; Ingenuity pathway analysis; Neural stem cells; RNA sequencing; Stem cell transplantation; Stroke; Transcriptome.

Fig. 1.. Human NSC engraftment decreases infarct size of aged mouse stroke brains.

Infarct sizes in hNSC-transplanted stroke brains (MCAO/R + Tx and MCAO/R + ATN-161 + Tx groups) are significantly reduced compared with non-transplanted control stroke brains (MCAO/R). ***P < .001 vs. sham group; ^#P < .05, ^##P < .01 vs. MCAO/R group (n = 9, sham; n = 7, MCAO/R; n = 9, MCAO/R + Tx; n = 8, MCAO/R + ATN-161 + Tx). Data are presented as mean ± SEM. MCAO/R, middle cerebral artery occlusion with reperfusion; Tx, hNSC transplantation.

Fig. 2.. Global transcriptional changes induced by human NSC transplantation in aged mouse brains post-stroke.

RNA-seq analysis was performed for gene expression profiling 48 h post-stroke. (A) Principal component analysis (PCA) of post-stroke mouse brain transcriptomes [MCAO+hNSCs (n = 3) and MCAO (n = 4)] showing reproducibility of biological replicates within each group and separation of groups along main principal component (PC1). The amounts of variance in data represented by each principal component (PC1 and PC2) are highlighted. (B) Volcano plot comparing the transcriptomes of MCAO+hNSCs versus MCAO groups. Significantly downregulated (651) and upregulated (160) genes in MCAO+hNSCs versus MCAO groups are indicated in blue and red colors, respectively. Differentially expressed genes (DEGs) were identified using DESeq2 Wald test, BH adjusted p-value < .05, fold-change threshold of 1.5. Several representative downregulated genes in hNSC-transplanted brains are highlighted. (C) Ingenuity Pathway Analysis (IPA) of DEGs highlights the reduction of inflammation-mediated signaling in the hNSC-engrafted group (MCAO+hNSCs) compared with non-transplanted stroke controls (MCAO). The activation or inhibition states (activation z-scores from IPA) of the pathways are indicated along the x-axis. Sizes of the circles correspond to the number of genes involved in different signaling pathways. Circle colors represent pathway enrichment p-values. (D) Heatmap indicates downregulated expression of genes encoding TNF receptors, proinflammatory factors, apoptosis factors, adhesion and leukocyte extravasation, and Toll-like receptors in hNSC-engrafted brains (MCAO+hNSCs vs. MCAO mice). Row-wise z-scores were computed using transcripts per million (TPM). MCAO, middle cerebral artery occlusion with reperfusion.

Fig. 3.. Human NSC engraftment attenuates the induction of proinflammatory factors in aged mouse stroke brains.

Reduced expression of proinflammatory cytokines in hNSC-transplanted stroke brains from aged mice. Transcript levels of TNF-α, IL-6, and IL-1β are upregulated in MCAO/R only brains. However, the induction of proinflammatory factors is attenuated in hNSC-engrafted groups (MCAO/R + Tx and MCAO/R + ATN-161 + Tx brains) in comparison to the non-transplanted MCAO/R control group. (A) Reduced expression of TNF-α in hNSC-transplanted groups. ****P < .0001 vs. sham; ^#P < .05, ^##P < .01 vs. MCAO/R (n = 9, sham; n = 5, MCAO/R; n = 6, MCAO/R + Tx; n = 6, MCAO/R + ATN-161 + Tx). (B) Reduced expression of IL-6 in hNSC-transplanted groups. *P < .05, **P < .01, ****P < .0001 vs. sham; ^##P < .01, ^###P < .001 vs. MCAO/R group (n = 9, sham; n = 7, MCAO/R; n = 6, MCAO/R + Tx; n = 9, MCAO/R + ATN-161 + Tx). (C) Reduced expression of IL-1β in hNSC-transplanted groups. ***P < .001, ****P < .0001 vs. sham; ^##P < .01 vs. MCAO/R group (n = 6, sham; n = 6, MCAO/R; n = 4, MCAO/R + Tx; n = 4, MCAO/R + ATN-161 + Tx). RT-PCR was employed to assess the levels of proinflammatory cytokines, normalized to GAPDH. Data are presented as mean ± SEM. MCAO/R, middle cerebral artery occlusion with reperfusion; Tx, hNSC transplantation.

Fig. 4.. Human NSC engraftment attenuates the induction of ICAM-1, MMP-9 and MMP-3 in aged mouse stroke brains.

(A) ICAM-1 is elevated in non-transplanted MCAO/R brains while induction of ICAM-1 is attenuated in hNSC-transplanted stroke brains (MCAO/R + Tx and MCAO/R + ATN-161 + Tx groups). ****P < .0001 vs. sham; ^#P < .05 vs. MCAO/R group (n = 8, sham; n = 10, MCAO/R; n = 7, MCAO/R + Tx; n = 10, MCAO/R + ATN-161 + Tx). (B) Transcript level of MMP-9 is elevated in MCAO/R only stroke brains. However, induction of MMP-9 is mitigated in hNSC-transplanted groups (MCAO/R + Tx and MCAO/R + ATN-161 + Tx brains), compared with the non-transplanted MCAO/R control group. ***P < .001, *P < .05 vs. sham; ^#P < .05 vs. MCAO/R (n = 7, sham; n = 7, MCAO/R; n = 5, MCAO/R + Tx; n = 7, MCAO/R + ATN-161 + Tx). (C) MMP-3 is upregulated in MCAO/R stroke brains. However, induction of MMP-3 is attenuated in hNSC-engrafted transplanted brains (MCAO/R + Tx and MCAO/R + ATN-161 + Tx groups). ****P < .0001, **P < .01 vs. sham; ^#P < .05 vs. MCAO/R group (n = 9, sham; n = 9, MCAO/R; n = 7, MCAO/R + Tx; n = 7, MCAO/R + ATN-161 + Tx). RT-PCR was utilized to assess MMP-9 and MMP-3 gene expression, normalized to GAPDH. Data are presented as mean ± SEM. MCAO/R, middle cerebral artery occlusion with reperfusion; Tx, hNSC transplantation.