Transplantation of neural stem progenitor cells from different sources for severe spinal cord injury repair in rat

Abstract

Neural stem progenitor cell (NSPC) transplantation has been regarded as a promising therapeutic method for spinal cord injury (SCI) repair. However, different NSPCs may have different therapeutic effects, and it is therefore important to identify the optimal NSPC type. In our study, we compared the transcriptomes of human fetal brain-derived NSPCs (BNSPCs), spinal cord-derived NSPCs (SCNSPCs) and H9 embryonic stem-cell derived NSPCs (H9-NSPCs) in vitro and subsequently we transplanted each NSPC type on a collagen scaffold into a T8-9 complete SCI rat model in vivo. In vitro data showed that SCNSPCs had more highly expressed genes involved in nerve-related functions than the other two cell types. In vivo, compared with BNSPCs and H9-NSPCs, SCNSPCs exhibited the best therapeutic effects; in fact, SCNSPCs facilitated electrophysiological and hindlimb functional recovery. This study demonstrates that SCNSPCs may be an appropriate candidate cell type for SCI repair, which is of great clinical significance.

Keywords: Brain-derived NSPCs; Collagen scaffolds; H9 embryonic stem cell-derived NSPCs; Spinal cord injury; Spinal cord-derived NSPCs.

Graphical abstract

Fig. 1

Characterization of human fetal BNSPCs, H9–NSPCs, and SCNSPCs, and morphology of collagen scaffolds loaded with BNSPCs, H9–NSPCs, and SCNSPCs. (A) Schematic diagram of the workflow for isolation and culturing of human fetal BNSPCs, H9–NSPCs, and SCNSPCs. Each cell type was immunofluorescence stained and pictures were captured under an inverted microscope. Immunofluorescence staining with (B) the NSPCs marker Nestin, (D) the NSPCs marker SOX2, (F) the immature neuron marker TUJ1, (H) the astrocyte marker GFAP. Scale bar = 200 μm. Quantification of percentages of (C) Nestin⁺, (E) SOX2⁺, (G) TUJ1⁺, (I) GFAP⁺ cells. *p < 0.05, **p < 0.01. (J) Macroscopic image showing the collagen scaffold structure with a diameter of 4 mm. Scale bar = 2 mm. Scanning electron microscope images showing (K) cross-sectional morphology of a collagen scaffold, Scale bar = 25 μm, (L) longitudinal sectional morphology of a collagen scaffold, Scale bar = 100 μm, and culturing (M) BNSPCs, (N) H9–NSPCs, (O) SCNSPCs on collagen scaffolds for 2 weeks. Scale bar = 20 μm. Confocal images of AO/PI staining of (P) BNSPCs, (Q) H9–NSPCs, and (R) SCNSPCs cultured on collagen scaffolds for 2 weeks. Scale bar = 100 μm.

Fig. 2

SCNSPCs had more highly expressed genes related to nerve functions than human fetal BNSPCs or H9–NSPCs. (A) Schematic diagram of the workflow for RNA sequencing of human fetal BNSPCs, H9–NSPCs, and SCNSPCs. (B) Venn diagram showing overall gene expression profiles of BNSPCs, H9–NSPCs, and SCNSPCs. Genes with counts <30 were not included in the analysis. (C) Heatmap showing expression values and (D) GO term enrichment analysis of the 50 most highly expressed genes unique to BNSPCs, H9–NSPCs, and SCNSPCs. (E) Venn diagram showing overlapping and uniquely expressed genes between SCNSPCs and BNSPCs and between SCNSPCs and H9–NSPCs. The threshold value for inclusion in this analysis was log2 (fold change) > 1. (F) Heatmap and (G) GO term enrichment analysis of the 2461 genes differentially expressed in both BNSPCs and H9–NSPCs compared to SCNSPCs.

Fig. 3

Heatmap showing gene expression of human fetal BNSPCs, H9–NSPCs, and SCNSPCs cultured in vitro. Data are shown for genes in 11 neuronal signaling pathways associated with neuron differentiation, myelination, synapse organization, neurotransmitter secretion and axon extension.

Fig. 4

GFP-SCNSPCs survived better and differentiated into more neurons than the other cell types in SCI rats at 1mpi. (A) Schematic diagram of the in vitro and in vivo experimental processes. We seeded GFP-BNSPCs, GFP-H9-NSPCs, or GFP-SCNSPCs on collagen scaffolds in vitro. Collagen scaffolds loaded with GFP-BNSPCs, GFP-H9-NSPCs, or GFP-SCNSPCs were transplanted into T8-9 severe spinal cord transection model rats. Rats in each group were sacrificed at one mpi, then grafted cells survival and differentiation were assessed. Immunofluorescence staining of (B) NeuN⁺ mature neurons and (E) NF⁺ mature nerve fibers in the lesion site. Scale bar = 50 μm. Quantification of percentage of GFP⁺ cells (C) and (F) in the lesion site. *p < 0.05, **p < 0.01. Quantification of percentage of (D) GFP⁺/NeuN⁺ mature neurons in GFP⁺ cells and (G) NF⁺ mature nerve fibers in the lesion site. *p < 0.05, **p < 0.01. (H) Expression of genes related to positive regulation of neuron differentiation of BNSPCs, H9–NSPCs or SCNSPCs as determined by RNA sequencing.

Fig. 5

GFP-SCNSPCs could be myelinated in the SCI site at one mpi in T8-9 severe spinal cord transection model rats. Immunofluorescence staining of (A) S100β⁺ Schwann cell-wrapped myelinated nerve fibers, scale bar = 50 μm, and (D) MBP⁺ myelinated nerve fibers in the lesion site, scale bar = 50 μm. Quantification of percentage of GFP⁺ cells (B) and (E) in the lesion site. **p < 0.01. Quantification of percentage of (C) GFP⁺/S100β⁺ myelinated nerve fibers in GFP⁺ cells, *p < 0.05, **p < 0.01. and (F) GFP⁺/MBP⁺ myelinated nerve fibers in GFP⁺ cells in the lesion site. **p < 0.01, ***p < 0.001. Expression of genes related to (G) axon extension, (H) myelination, and (I) oligodendrocyte differentiation in BNSPCs, H9–NSPCs, and SCNSPCs as determined by RNA sequencing.

Fig. 6

GFP-SCNSPCs formed synapses and secreted neurotransmitters in the SCI region at 1mpi in T8-9 severe spinal cord transection model rats. Immunofluorescence staining of (A) SYP⁺ presynaptic cells, (D) PSD95⁺ postsynaptic cells, and (H) GABA⁺ inhibitory neurotransmitters, GAD65/67⁺ inhibitory neurotransmitters, VGLUT1⁺ excitatory neurotransmitters in the lesion site. Scale bar = 50 μm. Quantification of percentage of GFP⁺ cells (B) and (E) in the lesion site. *p < 0.05, **p < 0.01, ***p < 0.001. Quantification of percentage of (C) GFP⁺/SYP⁺ cells in GFP⁺ cells, *p < 0.05, **p < 0.01, ***p < 0.001, (F) GFP⁺/PSD95⁺ cells in GFP⁺ cells, ***p < 0.001, and (I) GABA⁺, GAD65/67⁺ and VGLUT1⁺ cells in the lesion site. *p < 0.05, **p < 0.01, ***p < 0.001. Expression of genes related to (G) synapse organization, and (J) neurotransmitter secretion of BNSPCs, H9–NSPCs, and SCNSPCs as determined by RNA sequencing.

Fig. 7

Transplantation of GFP-SCNSPCs promoted electrophysiological and motor function recovery at one mpi. (A) BBB score of rats were evaluated in a transparent glass jar with a rubber pad at the bottom to increase friction. (B) Schematic diagram of rat CMEPs evaluation. (C) BBB scores of rats in each group. *p < 0.05 vs. scaffold group, **p < 0.01 vs. scaffold group, ^Op < 0.05 vs. scaffold + BNSPCs group, ^#p < 0.05 vs. scaffold + H9–NSPCs group. (D) CMEPs waves of rats in each group. Quantification of (E) latent periods and (F) amplitude for rats in each group. **p < 0.01, ***p < 0.001.