The effect of Jun dimerization on neurite outgrowth and motif binding

Abstract

Axon regeneration is a necessary step toward functional recovery after spinal cord injury. The AP-1 transcription factor c-Jun has long been known to play an important role in directing the transcriptional response of Dorsal Root Ganglion (DRG) neurons to peripheral axotomy that results in successful axon regeneration. Here we performed ChIPseq for Jun in mouse DRG neurons after a sciatic nerve crush or sham surgery in order to measure the changes in Jun's DNA binding in response to peripheral axotomy. We found that the majority of Jun's injury-responsive changes in DNA binding occur at putative enhancer elements, rather than proximal to transcription start sites. We also used a series of single polypeptide chain tandem transcription factors to test the effects of different Jun-containing dimers on neurite outgrowth in DRG, cortical and hippocampal neurons. These experiments demonstrated that dimers composed of Jun and Atf3 promoted neurite outgrowth in rat CNS neurons as well as mouse DRG neurons. Our work provides new insight into the mechanisms underlying Jun's role in axon regeneration.

Keywords: AP-1; Chromatin immunoprecipitation; Dorsal Root Ganglion; Enhancer.

Figure 1:

Peripheral axotomy results in a change to Jun’s binding localization. A) Venn Diagram of sites bound by Jun. 701 sites are significantly more bound after injury, 185 sites are significantly less bound after injury, and 160 sites are bound to a similar extent both before and after injury. B-D) Distribution of peaks more bound (B), less bound (C), and commonly bound (D) after injury by genomic regulatory annotation demonstrates that the majority of Jun’s injury-responsive binding site changes occur at putative enhancer elements while the majority of Jun’s injury-invariant binding sites occur proximal to TSSs. Non-TSS and non-enhancer proximal intronic, exonic, intergenic, and transcription termination sites (TTS) are also shown. E) Bar plot of odds ratios (error bars correspond to 95% confidence intervals) obtained by Fisher’s exact test on the overlap of each peak set with each regulatory element set. All bars plotted are significant relative to random genomic binding (Odds Ratio of 1, shown by horizontal line). All pairwise comparisons of bars are significant except for the two pairs specifically marked as non-significant (n.s.). This means that the distribution of differentially bound peaks across enhancer and TSS-proximal regions is different from that of the commonly bound peaks, but the distribution of more bound peaks is not different from that of less bound peaks.

Figure 2:

Differential Jun binding near transcription start sites. A) Bar plot showing the proportion of differentially- and commonly-bound Jun sites within 2kb of a TSS whose associated genes are upregulated (left) or downregulated (right) 7 days after sciatic nerve axotomy. This plot indicates that differentially bound peaks near gene transcription start sites are not predictive of the differential expression of those genes. B) Bar plot of enriched Biological Process GO terms from the set of significantly upregulated genes with Jun peaks within 2kb of their TSS which are significantly more bound after injury. This list is a sampling of the complete list found in Supplemental Table S2 and shows a strong enrichment for neuronal and axon growth-related terms. C) Bar plot of enriched Biological Process GO terms from the set of significantly downregulated genes with Jun peaks within 2kb of their TSS which are significantly less bound after injury. This list is a sampling of the complete list found in Supplemental Table S2 and does not show any enrichment for neuronal terms at all.

Figure 3:

The DNA motif transactivation potential of eight Jun-containing dimers. A-C) Bar plot of the minimal promoter-normalized ratio of Firefly luciferase signal to Renilla luciferase signal for Oxr1 and each of the eight tethered dimers when using a Firefly luciferase plasmid containing a CRE (A), TRE (B), or Degenerated CRE (C) motif upstream of a minimal promoter. In each panel, * indicates p-value <0.05, ** indicates p-value <0.01, *** indicates p-value <0.001 in a one-way ANOVA with a post-hoc Dunnett’s test. N=6 in panels A and C. N=3 in panel B. Horizontal dashed line in each panel indicates average signal produced by Oxr1 in that condition. Error bars represent standard error of the mean among biological replicates. Jun~Atf3 significantly promotes luciferase signal when exposed to the CRE or Degenerated CRE motifs. Jun~Fosl1 significantly promotes luciferase signal when exposed to the CRE or TRE motifs. D) EMSA dose-response experiment showing the binding of Jun~Atf3 to 50fmol of CRE oligo in each reaction and increasing quantities of TRE oligo, with quantities shown at the top. E) Quantification of the CRE oligo signal expressed as a percentage of the signal obtained when no TRE oligo is added, plotted as a function of increasing concentration of TRE oligo. This sigmoidal plot was fit using the Hill equation to predict the concentration of TRE oligo necessary to decrease the CRE signal to half of its initial value. This IC₅₀ value was found to be 82 fmol of TRE.

Figure 4:

Jun~Atf3 promotes neurite outgrowth in DRG and cortical neurons. In each panel, * indicates p-value <0.05, ** indicates p-value <0.01, *** indicates p-value <0.001 in a one-way ANOVA with a post-hoc Dunnett’s test. N=3 for each condition. Error bars represent standard error of the mean among biological replicates. A-B) DRG neuron results. A) Effect of monomers, co-transduced monomers and tethered dimers on Neurite Total Length (NTL). B) Effect of monomers, co-transduced monomers and tethered dimers on the Length of the Longest Neurite (LOLN). C-J) Cortical neuron results. C) Effect of monomers and co-transfected monomers on NTL. D-F) Effect of tethered dimers consisting of Jun and an Atf (D), Jun (E), or Fos (F) family member on NTL. G) Effect of monomers and co-transfected monomers on LOLN. H-J) Effect of tethered dimers consisting of Jun and an Atf (H), Jun (I), or Fos (J) family member on LOLN. K) Representative images of a cortical neuron transfected with Oxr1–2A-GFP (left) and another transfected with Jun~Atf3 (right). βIII-tubulin staining is shown in white and GFP (left) or Flag (right) staining is shown in green. Cellomics tracing of neurites is overlayed in white to improve visibility.