An RNA-sequencing transcriptome of the rodent Schwann cell response to peripheral nerve injury

Abstract

Background: The important contribution of glia to mechanisms of injury and repair of the nervous system is increasingly recognized. In stark contrast to the central nervous system (CNS), the peripheral nervous system (PNS) has a remarkable capacity for regeneration after injury. Schwann cells are recognized as key contributors to PNS regeneration, but the molecular underpinnings of the Schwann cell response to injury and how they interact with the inflammatory response remain incompletely understood.

Methods: We completed bulk RNA-sequencing of Schwann cells purified acutely using immunopanning from the naïve and injured rodent sciatic nerve at 3, 5, and 7 days post-injury. We used qRT-PCR and in situ hybridization to assess cell purity and probe dataset integrity. Finally, we used bioinformatic analysis to probe Schwann cell-specific injury-induced modulation of cellular pathways.

Results: Our data confirm Schwann cell purity and validate RNAseq dataset integrity. Bioinformatic analysis identifies discrete modules of genes that follow distinct patterns of regulation in the 1st days after injury and their corresponding molecular pathways. These findings enable improved differentiation of myeloid and glial components of neuroinflammation after peripheral nerve injury and highlight novel molecular aspects of the Schwann cell injury response such as acute downregulation of the AGE/RAGE pathway and of secreted molecules Sparcl1 and Sema5a.

Conclusions: We provide a helpful resource for further deciphering the Schwann cell injury response and a depth of transcriptional data that can complement the findings of recent single cell sequencing approaches. As more data become available on the response of CNS glia to injury, we anticipate that this dataset will provide a valuable platform for understanding key differences in the PNS and CNS glial responses to injury and for designing approaches to ameliorate CNS regeneration.

Keywords: Injury response; Macrophage; Neuroinflammation; Peripheral nerve; Regeneration; Repair cell; Schwann cell; Transcriptome.

Fig. 1

Purification of Schwann cells after sciatic nerve crush. A Schematic of local response to sciatic nerve crush. Immediately after crush (day 0), debris and inflammatory cytokines are abundant. By 3 days, macrophages proliferate at the site of injury and Schwann cells adopt a reparative phenotype. 7 days after crush, most debris has been engulfed and Schwann cells align. B Purification of Schwann cells by immunopanning. Macrophages and perineural fibroblasts were depleted with anti-CD45 and anti-Thy1 antibodies, respectively. Schwann cells were positively selected by anti-O4 hybridoma. C Purified Schwann cells in culture after immunopanning. D Experimental timeline. Sciatic nerve crush occurred at day 0. At days 3, 7, and 7 post-crush, both whole nerve and purified Schwann cells were collected for downstream RNA-sequencing

Fig. 2

Purity of Schwann cell isolation. A Expression of typical Schwann cell, macrophage, fibroblast, and endothelial markers in whole nerve samples and purified Schwann cells at each day post-crush. B Expression of macrophage markers, CD45 and Cx3cr1, in purified Schwann cells as compared to whole nerve. Purified Schwann cells are almost completely devoid of myeloid markers. C Hierarchical clustering of all samples (top 2000 most variable genes; Linkage, average; ColumnPdistance, Spearman). D Schwann cell response to injury confirmed by expected upregulation of Sox10 and p75, and downregulation of myelin genes MBP and MPZ

Fig. 3

qRT-PCR validation. A Differentially expressed genes from the RNA-seq analysis were selected for validation by qRT-PCR. Schwann cells were purified from uncrushed or post-injury sciatic nerves and cDNA was generated for primer-based amplification using exon-spanning primers. UC: uncrushed. Spp1: secreted phosphoprotein 1, Runx2: runx family transcription factor 2, Tes: Testin LIM domain protein, Tnc: Tenascin C, PRX: periaxin, Gpc1: Glypican 1, Cited2: Cbp/P300 interacting transactivator with Glu/Asp rich carboxy-terminal domain 2, Sparcl1: SPARC Like 1, Btc, betacellulin, Mdk: midkine, Gfra1: GDNF family receptor alpha 1, Gpr83: G protein-coupled receptor 83, MaI: Mal, T cell differentiation protein, Edn3: endothelin 3, Cldn19: claudin 19

Fig. 4

In situ hybridization spatial validation. A–D Four genes (GFRA1, MDK, GPC1, and SPARCL1) were selected for validation based on their clear response to injury in Schwann cells and lack of change in myeloid populations after injury. Sciatic nerves were counterstained with S100 (red) to confirm Schwann cell localization of expression changes. All four genes exhibit expected temporal expression patterns within the nerve (GFRA1, MDK, GPC1 increasing and SPARCL1 decreasing). Right, 63× zoom. Arrowheads highlight examples of S100 colocalization with the in situ probe for each candidate gene. Scales = 50 μm (left images), and 20 μm (right images)

Fig. 5

Schwann cell gene changes post-crush. A–C Volcano plots of purified Schwann cells at day 3, 5, or 7 post-crush as compared to Schwann cells purified from uncrushed nerves. D–F Enriched pathways in Schwann cells from each timepoint based on Gene Ontogeny (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathway analysis of the top 2000 most upregulated genes at each post-crush timepoint (all genes with padj < 0.05). G Most enriched pathways (or GO categories) at day 3, 5 and 7. GO terms were binned across each time point, with the top five most frequently observed terms presented here. Direction of arrows highlight that some pathways, e.g., cell division and cell cycle arrest, decrease with time post-crush, while others (cytoskeletal rearrangement) peak at intermediate timepoints post-injury

Fig. 6

WGCNA of purified Schwann cells after sciatic nerve crush. A, B WGCNA on purified Schwann cells. The power value within the blockwise consensus module was determined by examining the soft-threshold-mean-connectivity curve (p = 21). The minimum WGCNA module size was 30 genes. Modules with < 0.25 similarity were merged to produce 10 final modules. C–E Enriched GO terms for 3 modules (1, 5, and 8) with distinct expression patterns