TRAF family member-associated NF-kappa B activator (TANK) expression increases in injured sensory neurons and is transcriptionally regulated by Sox11

Abstract

Peripheral nerve injury evokes rapid and complex changes in gene transcription and cellular signaling pathways. Understanding how these changes are functionally related is essential for developing new approaches that accelerate and improve nerve regeneration. Toward this goal we found that nerve injury induces a rapid and significant up-regulation of the transcription factor Sox11 in dorsal root ganglia (DRG) neurons. Gain and loss of function studies have shown this increase is essential for normal axon regeneration. To determine how Sox11 impacts neuronal gene expression, DRG neurons were treated with Sox11 siRNA to identify potential transcriptional targets. One gene significantly reduced by Sox11 knockdown was TRAF (tumor necrosis factor (TNF) receptor-associated factor)-associated NF-κB activator (TANK). Here we show that TANK is expressed in DRG neurons, that TANK expression is increased in response to peripheral nerve injury and that Sox11 overexpression in vitro increases TANK expression. Injury and in vitro overexpression were also found to preferentially increase TANK transcript variant 3 and a larger TANK protein isoform. To determine if Sox11 regulates TANK transcription bioinformatic analysis was used to identify potential Sox-binding motifs within 5kbp of the TANK 5' untranslated region (UTR) across several mammalian genomes. Two sites in the mouse TANK gene were examined. Luciferase expression assays coupled with site-directed mutagenesis showed each site contributes to enhanced TANK promoter activity. In addition, chromatin immunoprecipitation assays showed direct Sox11 binding in regions containing the two identified Sox motifs in the mouse TANK 5'-UTR. These studies are the first to show that TANK is expressed in DRG neurons, that TANK is increased by peripheral nerve injury and that the regulation of TANK expression is, at least in part, controlled by the injury-associated transcription factor Sox11.

Figure 1. TANK is expressed in DRG neurons

A. TANK antibody staining of mouse DRG shows reactivity in virtually all neurons. B. Preincubation with a TANK blocking peptide significantly reduces TANK immunoreactivity in DRG. Images were taken at the same exposure and magnification, merged into one figure and then manipulated for brightness and contrast using Photoshop. C. Dissociated DRG neurons grown 3d in culture immunolabeled with the neuronal marker anti-beta-III-tubulin (green) exhibit TANK immunoreactivity (red). DAPI (blue) labeling shows TANK expression in DRG neurons and in a subpopulation of non-neuronal cells. D. Overlap of staining shown in (C). Arrows indicated triple-labeled neurons, arrowheads indicate TANK/DAPI labeled non-neuronal cells. Bar = 100µm.

Figure 2. Sox11 and TANK are upregulated in DRG following nerve injury

A. Sox11 mRNA is increased in lumbar DRG in response to sciatic nerve axotomy (n = 4 per group, p < 0.001 vs. naïve, ANOVA with Bonferroni post-hoc test). B. RT-PCR products run on an ethidium bromide labeled agarose gel show increase in TANK mRNA variants. The larger TANK mRNA variant 3 (at 685 bp) is preferentially increased in lumbar DRG at 1D and 4D post axotomy. T, total TANK mRNA; V3, variant 3 mRNA. C. Western blot shows upregulation of TANK band at ~53kDa (indicted by asterisk) in DRG at 1d and 3d post axotomy, relative to naïve control (C). Arrowhead indicates 48 kDa isoform. D. Densitometric analysis using larger groups (n=8 mice per group) show the 48kDa isoform also increases in response to injury at 3d post-injury (38%, p < 0.05, vs. naïve, Student’s t-test) in DRG in response to axotomy. Values were normalized to GAPDH expression. E. siRNA-mediated knockdown of Sox11 in cultured DRG neurons reduces intensity of the TANK immunoreactive band at ~53kDa (arrowhead) relative to cultures treated in parallel with nontargeting (NT) siRNA. GAPDH is shown as loading standard. Asterisks on charts indicate significance.

Figure 3. Sox11 increases in vitro TANK expression

A. Neuro2a cells transfected with pCMV or pCMV-Sox11 increase total TANK mRNA expression at 24h post transfection (n = 4 per group, p < 0.05, two-way ANOVA with Bonferroni post-hoc test). Fold changes are relative to 0h baseline. TANK expression also increases overall as a function of time (n = 8 per time point, main effect using two-way ANOVA, p < 0.05). B. Neuro2a cells transfected with Sox11 plasmid also upregulate expression of TANK variant 3 mRNA. Gel shows PCR products from RT-PCR reaction using primers to detect total (T) and variant 3 (V3) mRNAs, as seen in Fig. 2B. C. TANK protein increases by 36h in pCMV-Sox11 transfected neuro2a cells (n = 3 per group, p < 0.05, Student’s t-test). Plot is for TANK variant that runs at ~48 kDa with values normalized to GAPDH. D. Image of western blot whose values are plotted in panel C. E. TANK protein is increased by Sox11 transfection in HEK cells. F. TANK protein is increased in DRG neuron cultures treated with HSV-Sox11 viral vector. In contrast to neuro2a cells, the 53kDa TANK isoform (marked by asterisk) is prominent in HSV-Sox11 infected cultured DRG neurons. This band also increases in DRG from nerve cut animals (see Fig. 2D).

Figure 4. Sox11 regulates TANK promoter expression

A. (Top) A 2.3-kb fragment of the TANK promoter was PCR cloned into the pGL2-basic vector. Forward-direction Sox binding sites at nt-2267 (SoxCS#1) and nt-728 (SoxCS#2) are indicated relative to the TANK transcript (variant 3) translational start site. (Bottom) Diagram illustrating site-directed mutagenesis constructs. Plasmid pGL2-TANK was mutated to selectively inactivate one or both Sox consensus sequence sites. B. Phylogenetic analysis of TANK promoter/5’UTR region in vertebrates was done following ClustalW analysis using MacVector software (Cary, North Carolina). Nucleotide alignment is shown for a highly conserved region within 1-kbp of the translation start site; box indicates a conserved Sox-binding motif (AACAAAG). C. Cotransfection of pGL2-TANK with pCMV-Sox11 increases TANK promoter activity in a dose-dependent manner, ranging from 2.1-fold (with 25 ng pCMV-Sox11) to 5.5-fold (with 500 ng pCMV-Sox11). TANK promoter was not activated by pCMV (empty vector) or the internal control reporter vector, pRL-TK, the activity of which remained constant for all combinations of vector + pCMV-Sox11 dosage. D. Luciferase reporter activity in neuro2a cells containing wildtype (TANK-luc) or a mutagenized TANK promoter (n = 4 per group). In this experiment TANK-pGL2/pCMV-Sox11 cotransfection caused a 17.9-fold increase in TANK promoter activity (p < 1 ×10⁻⁸, ANOVA with Bonferroni post-hoc test). Mutagenesis of the first binding site resulted in a 50% decrease in Sox11-mediated TANK promoter activation (p < 1 ×10⁻⁴, ANOVA with Bonferroni post-hoc test). Mutation of the second binding site resulted in an 85% decrease in Sox11-mediated TANK promoter activation (p < 1 × 10⁻⁶, ANOVA with Bonferroni post-hoc test), while ablation of both binding sites resulted in a 75% decrease in Sox11-mediated TANK promoter activation (p < 1 × 10⁻⁶, ANOVA with Bonferroni post-hoc test).

Figure 5. Sox11 binds two sites on the mouse TANK promoter/5’UTR

Primers designed against the two most proximal Sox motifs (nt-2267, nt-728) in the mouse TANK promoter/5’UTR were used in ChIP assays performed on pFlag-Sox11 transfected neuro2a cells. Lanes from left to right: Input DNA only (no IP), PCR performed after IP with no precipitating antibody, PCR after IP with control IgG antibody, PCR performed after IP with anti-FLAG IgG. Appropriate sized bands were amplified in regions flanking both putative Sox binding sites on the TANK promoter.