Characteristics of cytokines in regeneration of injured peripheral nerves

Background Cytokines are essential cellular modulators of a variety of physiological and pathological activities, including peripheral nerve repair and regeneration. However, the molecular changes of these cellular mediators during peripheral nerve regeneration are not well clarified. The study is aimed to discover critical cytokines for the regenerative process of injured peripheral nerves. Methods The sequencing data of the injured nerve stumps and the dorsal root ganglia (DRGs) of Sprague-Dawley (SD) rats subjected to sciatic nerve (SN) crush injury was analyzed to determine expression patterns of genes coding for cytokines. PCR experiments were used to validate the accuracy of sequencing data. Results A total of 46, 52, and 54 upstream cytokines were differentially expressed in SNs at 1 day, 4 days, and 7 days after nerve injury. And a total of 25, 28, and 34 upstream cytokines were differentially expressed in DRGs at these time points. The expression patterns of some essential upstream cytokines were displayed in a heatmap and validated by PCR experiment. Bioinformatic analysis of these differentially expressed upstream cytokines after nerve injury demonstrated that inflammatory and immune responses were significantly involved.


Introduction
Peripheral nerves are vulnerable tissues that are generally defenseless to traumatic injuries caused by bump, stretch, crush, and penetrating wounds and non-traumatic injuries caused by genetic, metabolic, infectious, and medically induced factors (1,2). Fortunately, unlike nerves in the central nervous system, peripheral nerves can regenerate and achieve certain functional recovery after injury, although fully functional recovery is generally unexpected (3). After peripheral nerve injury, distal nerve stumps undergo Wallerian degeneration, activated Schwann cells and macrophages phagocytosis debris of axon and myelin sheaths, axons of survived neurons regrow toward target tissues for reinnervation (3,4).
Cytokines are a wide category of immunomodulating proteins or peptides including chemokines, interferons, interleukins, lymphokines, and tumour necrosis factors. Cytokines play essential roles in inflammation and immune responses and participate in the regulation of the maturation, growth, and responsiveness of a variety of cell populations (5,6). Cytokines have been identified to be constitutively involved in the nervous system in health and disease (7)(8)(9)(10). Cytokines are also extremely critical for peripheral nerve injury and repair as fine-tuned expressions of cytokines modulate the cellular behaviors of Schwann cells, macrophages, and neurons and regulate debris clearance, axon growth, and peripheral nerve regeneration (11).
Understanding the molecular changes of these cellular mediators during peripheral nerve regeneration opens new possibilities to improve the repair of injured nerves and to minimize the induction of neuropathic pain (11). On this purpose, previously obtained sequencing data of the injured nerve stumps of Sprague-Dawley (SD) rats subjected to sciatic nerve (SN) crush injury was analyzed to determine expression patterns of genes coding for cytokines (12). Moreover, considering that cytokines retrograde transport to the neuronal bodies and affects neuronal activities, sequencing data of the dorsal root ganglia (DRGs) after rat SN crush injury was also jointly investigated (13).
Differentially expressed genes in SNs and DRGs after nerve crush injury were identified and upstream cytokines of these differentially expressed genes were recognized by Ingenuity Pathway Analysis (IPA) bioinformatic tool. Differentially expressed upstream cytokines at 1 day, 4 days, and 7 days after nerve crush injury were subjected to functional enrichment of Gene Ontology (GO) categories and  data of rat DRGs at 0 hour, 3 hours, 9 hours, 1 day, 4 days, and 7 days after SN crush injury were conserved in NCBI database with the accession number PRJNA547681 (SRP200823). Differentially expressed genes in SNs and DRGs at certain time points after nerve crush injury were selected by comparing their expression levels under the injured status with the expression levels under the uninjured status (0 hour control). Genes with a fold changes < 2 or > -2 and a experimental false discovery rate (FDR) < 0.05 were defined as differentially expressed genes.

Bioinformatic analysis
Upstream cytokines of differentially expressed genes in SNs and DRGs were identified by IPA bioinformatic tool (Ingenuity Systems Inc., Redwood City, CA, USA) for Ingenuity pathway knowledge base (IPKB)-based upstream regulator analysis. Genes coding for cytokines with a fold changes < 2 or > -2 at 1 day, 4 days, or 7 days as compared with 0 hour were defined as differentially expressed cytokines and were subjected to subsequent bioinformatic analyses.
Commonly differentially expressed cytokines in SNs and DRGs at 1 day, 4 days, or 7 days after SN crush injury were identified by the Venny 2.1.0 online bioinformatic tool (http://bioinfogp.cnb.csic.es/tools/venny/index.html) (14). The expression profiles of these commonly differentially expressed cytokines were demonstrated by a heatmap. Signaling pathways and biological processes involved in differentially expressed upstream cytokines were discovered by DAVID bioinformatic enrichment tools (15, 16).

Animal surgery and collection of the dorsal root ganglia and SN stumps
The conduction of rat SN crush injury and the collection of SNs and DRGs of uninjured and injured rats were performed as previously described (12,13). Adult male SD rats weighting 180-220 g were obtained from the Experimental Animal Center of Nantong University (Animal licenses No. SCXK [Su] 2014-0001 and SYXK [Su] 2012-0031) and subjected to animal surgery. Rats were anaesthetizated intraperitoneally with a mixture of 85 mg/kg trichloroacetaldehyde monohydrate, 42 mg/kg magnesium sulfate, and 17 mg/kg sodium pentobarbital. SNs at 10 mm above the bifurcation into the tibial and common fibular nerves were exposed by a skin incision in the left outer mid-thigh. Exposed SNs were crushed with a forceps at a force of 54 N for 3 times with 10 seconds for each time. Rats underwent SN crush injury were sacrificed by decapitation at 1 day, 4 days, and 7 days after animal surgery. Rat underwent sham surgery were sacrificed and designated as 0 hour controls. Rat SNs and lumbar 4 to lumbar 6 DRGs were removed for RNA isolation. Relative mRNA abundances of Cxcl10 and Il1rn were determined using the comparative 2 −ΔΔCt method, in which ΔCt = Ct (injured) -Ct (uninjured) and ΔΔCt = Ct (target gene) -Ct (reference gene) (17).

Statistical analysis
Summarized PCR results were reported as means ± SEM with n = 3. Statistical analysis and graphs were generated using GraphPad Prism 6.0 (GraphPad Software, Inc., San Diego, CA, USA). The p-value was determined from one-way analysis of variance (ANOVA) and a p-value < 0.05 was considered as statistically significant.
3. Results 3.1. Identification of differentially expressed upstream cytokines in SNs and DRGs following peripheral nerve injury Previously, the expression patterns of genes in SNs (12) and DRGs (13) at multiple time points after rat SN crush injury were determined and a global view of genetic changes following peripheral nerve injury was obtained. Considering the essential roles of cytokines in tissue remodeling and organ regeneration, IPA bioinformatic analysis was applied to screen upstream cytokines of differentially expressed genes in SNs and DRGs after nerve crush injury. The expression levels of genes coding for these upstream cytokines were further examined and differentially expressed upstream cytokines in SNs and DRGs at 1 day, 4 days, and 7 days after nerve injury were recognized (Table S1). Venn diagrams were generated to compare differentially expressed upstream cytokines in SNs and DRGs at certain time points after nerve injury and to obtain a comprehensive view of altered cytokines during nerve regeneration (Fig. 1A-1C). A total of 46 upstream cytokines were differentially expressed in SNs at 1 day after nerve injury. At later time points, a relatively larger number of upstream cytokines were differentially expressed in SNs (Fig. 1D). In DRGs, a smaller group of upstream cytokines were differentially expressed as compared with in SNs. And the numbers of differentially expressed upstream cytokines also increased at later time points after nerve injury (Fig. 1D). Detailed investigation of these differentially expressed upstream cytokines showed that the majority of cytokines were up-regulated and only a few cytokines were down-regulated in SNs.
However, in DRGs, the percentage of down-regulated cytokines was much higher (Table S1). The intersection set of Venn diagrams discovered cytokines that were differentially expressed in both SNs and DRGs at the same time points. And the expression changes of these intersected cytokines were listed besides the Venn diagrams ( Fig. 1A-1C). Some cytokines, such as interleukin-6 (Il6), were kept up-regulated in SNs and DRGs after nerve injury while some cytokines, such as Cxcl10, were upregulated in SNs but down-regulated in DRGs (Fig. 1A-1C).

Demonstration of the expression patterns of upstream cytokines in SNs and DRGs following peripheral nerve injury
To identify the dynamic changes of critical cytokines during nerve regeneration, intersected cytokines in SNs and DRGs were further studied. A total of 27 cytokines were differentially expressed in both SNs and DRGs at 1 day, 4 days, or 7 days after nerve injury. The expression levels of these cytokines were investigated and displayed in a heatmap (Fig. 2). Some cytokines showed similar expression treated in both SNs and DRGs. For example, tumor necrosis factor ligand superfamily member 10 (Tnfsf10) was down-regulated in both SNs and DRGs after nerve injury, CD40 ligand (Cd40lg) was upregulated in both SNs and DRGs at 4 days after nerve injury, and interleukin-9 (Il9) was up-regulated in both SNs and DRGs at 7 days after nerve injury. Some cytokines, such as Il1rn and C-C motif chemokine ligand 2 (Ccl2), exhibited higher expression changes in SNs as compared with DRGs.
The expression patterns of representative cytokines revealed by sequencing assay were further validated by PCR experiments. Different batch of SD rats used for sequencing were collected for sciatic nerve crush injury surgery and subsequent PCR validation. Outcomes from PCR experiments demonstrated that the relative abundances of gene coding for cytokine Cxc10 were increased in SNs ( Fig. 3A) but decreased in DRGs (Fig. 3B) following nerve injury. And the relative abundances of gene coding for Il1rn were up-regulated in both SNs (Fig. 3C) and DRGs (Fig. 3D). These outcomes were consistent with the expression trends determined by sequencing data (shown in red lines), indicating that sequencing data were of high accuracy.

Identification of significantly involved signaling pathways of differentially expressed upstream cytokines following peripheral nerve injury
Bioinformatic analyses were performed to evaluate significantly involved signaling pathways of differentially expressed upstream cytokines in SNs and DRGs after nerve injury. Activated signaling pathways that were related to nerve regeneration in up-regulated cytokines and down-regulated cytokines in SNs and DRGs were separately explored (Fig. 4). We focus on upstream cytokines in the current study, so it stands to reason that cytokine-cytokine receptor interaction and chemokine signaling pathway were most strongly enriched signaling pathways. Other significantly enriched signaling pathways included Toll-like receptor signaling pathway, TNF signaling pathway, NOD-like receptor signaling pathway, NF-κB signaling pathway, and JAK-STAT signaling pathway. And these signaling pathways were most robustly involved in up-regulated upstream cytokines in SNs.

Identification of significantly involved GO biological process categories of differentially expressed upstream cytokines following peripheral nerve injury
Critical nerve regeneration-related biological processes occurred after sciatic nerve crush injury were further discovered by categorizing differentially expressed upstream cytokines to GO terms.
Inflammatory response and immune response were the most significantly involved biological processes and were also most strongly involved in up-regulated upstream cytokines in SNs (Fig. 5).
Some other inflammatory response and immune response-related biological processes, such as neutrophil chemotaxis, monocyte chemotaxis, cellular response to interleukin-1, also exhibited low pvalues, indicating the significance of inflammation and immune response.
To further reveal the importance of inflammation in nerve regeneration, inflammation-related biological processes were interconnected to a network (Fig. 6A). The inflammation-centered network showed that both acute and chronic inflammatory responses were activated after nerve repair. The chemotaxis, migration, and extravasation of various types of cells, including lymphocytes, macrophages, and monocytes, contribute to activated inflammatory response.
Similarly, a network of immune response-related biological processes was also generated (Fig. 6B).
Many Differentially expressed cytokines in the injured SNs might essentially benefit the infiltration and polarization of monocytes, macrophages, and Schwann cells, encourage the phagocytosis and clearance of axon and myelin debris, and promote axon regrowth and regeneration. Actually, a large range of cytokines were found to be up-regulated in the injured nerve stumps. These cytokines might be secreted and released by Schwann cells and macrophages after peripheral nerve injury (18,19).
Moreover, it was worth noting that many cytokines might carry out opposing effects at multiple time points during peripheral nerve regeneration and represent a "double-edged sword" (11). Our current study suggested that differentially expressed upstream cytokines in the injured SNs after peripheral nerve injury were highly related with inflammation and immune responses. Therefore, the controversial biological roles of cytokines might be due to the degree and timing of inflammation and immune responses induced by different expression levels of cytokines (11). These results were also consistent with our previous finding that robust immune and inflammatory responses were sustained significantly involved during nerve degeneration and regeneration (24). These outcomes implied that, to achieve orchestrated regulation of cytokines, it was of great importance to obtain an overview of the expression patterns of cytokines in the injured nerve stumps at different time points after peripheral nerve injury.
Besides affecting the injured nerve stumps and reconstructing the regenerative microenvironment, cytokines could influence the expressions of neurotrophins and their receptors and thus could affect the neurite outgrowth of neurons (11). For instance, the addition of interleukin 4 (IL-4) or interferon-γ (IFN-γ) to neurotrophin-4 (NT-4)-treated DRG neurons would increase NT-4-induced neurite outgrowth and the addition of TNF-α to neurotrophin-treated DRG neurons would decrease neurotrophin-induced neurite outgrowth (25). In addition, cytokine induced inflammation and immune response would activate retrograde signaling and might induce the death or survival of DRG neurons (11,26).
Consequently, in our current study, we also jointly determine the dynamic expression levels of cytokines in DRGs and discovered some significantly changed cytokines, such as interferon alpha 4 (Ifna4), Il6, and interleukin 24 (Il24).
Interestingly, some cytokines, such as Cxcl10, were discovered to be up-regulated in nerve stumps but down-regulated in DRGs after nerve injury. It was shown that Cxcl10 could promote the invasion of lymphocytes and macrophages and affect myelination in a viral model of multiple sclerosis (27) and could induce neuropathic pain in DRGs after chronic constriction injury (28). Therefore, it was possible that up-regulated Cxcl10 in SNs after nerve injury could contribute to debris clearance in the injured nerve stumps while down-regulated Cxcl10 in DRGs could reduce neuropathic pain. Further functional studies would reveal the specific roles of these cytokines during peripheral nerve repair and regeneration and would provide new targets of the treatment of peripheral nerve injuries.

Conclusions
In summary, the findings provided an overview of the dynamic changes of cytokines in SNs and DRGs

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.