Analysis of neuronal injury transcriptional response identifies CTCF and YY1 as co-operating factors regulating axon regeneration

Abstract

Injured sensory neurons activate a transcriptional program necessary for robust axon regeneration and eventual target reinnervation. Understanding the transcriptional regulators that govern this axon regenerative response may guide therapeutic strategies to promote axon regeneration in the injured nervous system. Here, we used cultured dorsal root ganglia neurons to identify pro-regenerative transcription factors. Using RNA sequencing, we first characterized this neuronal culture and determined that embryonic day 13.5 DRG (eDRG) neurons cultured for 7 days are similar to e15.5 DRG neurons in vivo and that all neuronal subtypes are represented. This eDRG neuronal culture does not contain other non-neuronal cell types. Next, we performed RNA sequencing at different time points after in vitro axotomy. Analysis of differentially expressed genes revealed upregulation of known regeneration associated transcription factors, including Jun, Atf3 and Rest, paralleling the axon injury response in vivo. Analysis of transcription factor binding sites in differentially expressed genes revealed other known transcription factors promoting axon regeneration, such as Myc, Hif1α, Pparγ, Ascl1a, Srf, and Ctcf, as well as other transcription factors not yet characterized in axon regeneration. We next tested if overexpression of novel candidate transcription factors alone or in combination promotes axon regeneration in vitro. Our results demonstrate that expression of Ctcf with Yy1 or E2f2 enhances in vitro axon regeneration. Our analysis highlights that transcription factor interaction and chromatin architecture play important roles as a regulator of axon regeneration.

Keywords: CTCF; E2F2; YY1; axon regeneration; bioinformatics analyses; dorsal root ganglia; sensory neurons; transcription factors.

FIGURE 1

eDRG spot culture contains mostly sensory neurons after treatment with FDU. (A) Representative images of eDRG culture stained with the neuronal marker TUJ1 (red) and the Satellite glia marker FABP7 (green) at DIV1 and at DIV7 with and without treatment with FDU. Scale Bar: 250 μm. (B) Quantification of the fluorescence intensity of FABP7 normalized to TUJ1 at DIV1 and DIV7 ± FDU. n = 4 biologically independent animals. P values are determined by One way ANOVA. Data are presented as mean values ± SEM. (C) Representative images of eDRG culture in high magnification, with and without FDU, stained with TUJ1 (red) and DAPI (blue). Scale Bar: 20 μm. (D) Quantification of selected neuronal and non-neuronal markers (counts per million) from RNAseq analysis of eDRG spot culture at DIV7 treated with FDU.

FIGURE 2

Neurons in the eDRG at DIV7 are similar to DRG neurons at the E15.5 developmental stage. (A) Fraction of 15 neuronal subtypes across four developmental stages of DRG neurons. (B) Heatmap shows the scaled z-score of the average CPM of 50 subtype-specific genes from eDRG controls and 4 developmental stages. Both row and column clustering were applied. High and low expressions are indicated in red and blue, respectively. (C) Plots of the counts of 50 subtype-specific genes expressed in each neuronal subtype during development using RPKM with cutoffs of 1 and 5.

FIGURE 3

Time course analysis (1, 3, 8, 16, and 24 h) of the transcriptional response to injury in eDRG spot culture. (A) Heatmap of correlation of samples from time-course RNAseq analysis after axotomy. (B) PCA analysis of time-course RNAseq analysis after axotomy illustrates the relative similarity between sample groups at control, 1, 3, 8, 16, and 24 hpi (3 replicates per time point). (C) Volcano plots of differentially regulated genes after injury (p adj < 0.05, FC > 2). (D) K-means clustering identified five unique gene profile clusters with similar gene expression dynamics according to the time-course gene expression data. (E) Heatmap shows the scale z-score of the average CPM of 365 differentially expressed genes from time-course data. Red and blue cells indicate relative gene upregulation and downregulation, respectively. Green cells indicate cluster types identified in D. the cells next to cluster type are genes that are significantly expressed at different time points between treated and control groups. (F) Venn diagram comparing the differentially expressed genes after injury in the eDRG spot culture and 2 datasets of in vivo adult injury using enriched neuronal populations. (G) Out of the 29 up-regulated TFs in the eDRG spot, 11 were also identified in adult DRG neurons following sciatic nerve crush.

FIGURE 4

Overexpression of TF combinations in eDRG axotomy model. (A) Schematic timeline of the experimental procedure for TF overexpression in injured cultured eDRG neurons. (B) Axon growth of infected cells with TF combinations relative to control. (C) Significant TF were tested alone and repeated in combinations. (D) Represented images of axotomized cultures infected with the pro-regenerative TFs alone and in combinations. Scale Bar: 250 μm. n = 3–9 biologically independent animals examined over 2 independent experiments P-values determined by one-way ANOVA followed by Dunnett’s multiple comparisons test. Data are presented as mean values ± SEM.