Nur77 target genes in neurons
We considered as Nur77 target genes, those genes that bind Nur77 in a window of - 2kb to + 500 bp from TSS in their promoters, according to our re-analysis of ChIP-Seq data of endogenous Nur77 of mouse neural stem cells (NSC) and NSC differentiated to neurons (NC) (17). To select genes in which the binding of Nur77 influences transcription, we used a microarray of mRNA from mouse hippocampal pyramidal neurons overexpressing Nur77 (12) and selected genes with a significant change of expression after Nur77 overexpression (Fig. 1A). We identified 113 Nur77 target genes in NSC (Table S.1) and 16 out these 113 genes changed their expression with a fold-change ≥ 2 after Nur77 overexpression (Fig. 1B). In NC, we identified 116 Nur77 target genes (Table S.2) and 17 out of these 116 genes changed their expression with a fold-change ≥ 2 after Nur77 overexpression (Fig. 1C). We found 53 Nur77 target genes common to NSC and NC, suggesting that these genes maintain Nur77 binding to their promoter during neuronal differentiation (Table S.3).
Binding of Nur77 on promoters of genes in the immune system and the central nervous system
To find out whether Nur77 exerts a similar function in the nervous and immune systems and its conservation in human and mouse, we compared the binding peaks from two high quality experimentally generated ChIP-Seq databases: 1.- ChIP-Seq from the ENCODE project of overexpressing EGFP-Nur77 in the human chronic myelogenous leukemia cell line K562 (16) and 2.- ChIP-Seq of mouse NC (17) (Fig. 2A). Despite the differences between these two ChIP-Seq protocols, we found 271 genes with Nur77 common binding on their promoter (-2kb to +500bp from the TSS) in immune and neuronal cells (Fig. 2A, Table S.4).
GO analysis of these 271 genes (Fig. 2B and Table S.5) showed a significant enrichment of Nur77 target genes in ribonucleoprotein complex binding (fold of enrichment 5.2), cadherin bindings (fold of enrichment 3.94), cell adhesion molecule binding (fold of enrichment 3.05) and protein domain specific binding (fold of enrichment 2.66) (Fig. 2B). In the cellular component classification, Nur77 target genes were mainly enriched in categories of adhesion and junction: focal adhesion (fold of enrichment 3.45), cell-substrate adherent junction (fold of enrichment 3.43), cell-substrate junction (fold of enrichment 3.39), adherens junction (fold of enrichment 3.02), and anchoring junction (fold of enrichment 2.93). Two categories of nuclear localization were enriched: nuclear speck (fold of enrichment 3.33) and nuclear body (fold of enrichment 2.86) (Fig. 2B). Many proteins encoded by Nur77 target genes, which are common to the nervous and immune system, are ribonucleoproteins and adhesion molecules. This fact was strengthened by the enrichment of Nur77 target genes in nuclear bodies and areas of adherents and anchoring junctions (Fig. 2C).
In the biological process classification, Nur77 target genes were enriched in regulation of Endoplasmic-Reticulum-Associated protein Degradation (ERAD) pathway (fold of enrichment 12.91) and regulation of response to endoplasmic reticulum stress (fold of enrichment 8.01). Three GO terms related to interleukin signaling were enriched: interleukin-12-mediated signaling pathway (fold of enrichment 10.11) cellular response to interleukin-12 (fold of enrichment 9.68) and response to interleukin-12 (fold of enrichment 9.49). Nur77 target genes were also enriched in the regulation of protein autophosphorylation (fold of enrichment 9.49) and cell aging (8.22) GO terms (Fig. 2B).
Interestingly, 9 out of the 271 genes were also found in the set of genes that significantly changed their expression after Nur77 overexpression in neurons: AGAP3, BIRC5, DYM, ITGB3, KIF21B, MORN5, RREB1, STRIP2 and WEE1, suggesting that these genes are also regulated in the immune system. Previous evidence supports that Nur77 controls the expression of BIRC5 gene (21), validating our results. Further studies are required to fully validate Nur77 control over these genes, both in the nervous and immune system.
Characterization of Nur77 binding sites throughout the genome.
To further characterize the binding profile of EGFP-Nur77, we compared the enrichment of ChIP-Seq peaks across 15 chromatin states defined by Ernest et al. for K562 cell line with the peaks of Nur77 (22). We calculated the overlap enrichment across the EGFP-Nur77 peaks and chromatin states for K562 cell line, obtaining a significant enrichment of Nur77 peaks across chromatin states associated with transcription (Fig. 3A). The analysis of the coordinates of EGFP-Nur77 peaks, reported by ENCODE, with respect to the nearest transcription start site (TSS), revealed a high frequency of Nur77 binding between +1000 and –1000 nucleotides from TSS (Fig. 3B).
An enrichment analysis using the human chromatin segmentation model generated by Ernst, (2011) showed significant enrichment of Nur77 peaks in chromatin states associated with transcriptional regulatory regions, particularly in strong enhancers and active promoters (Fig. 3C). Two of chromatin states are described as strong enhancers (states 4 and 5), which differentiate in the occurrence of specific chromatin marks and distance to the TSS. Strong enhancer state number 4 presented a higher occurrence of histone 3 lysine 4 trimethylation (H3K4me3) and histone 3 lysine 9 acetylation (H3K9ac) and was closer to TSS than Strong enhancer state number 5. Nur77 was enriched in both chromatin states described as strong enhancers, exhibiting a log2 enrichment greater than 4 in chromatin state 4 (Fig. 3C). High enrichment of Nur77 binding was also observed in transcriptional transition states of chromatin. These areas presented similar characteristics to transcriptional elongation areas, but with an increased presence of H4K20me1 and H3K4me1 and more sensitive to DNAse (22), suggesting an intermediate state between the promoter activation and effective elongation. In contrast, we found a negative enrichment for Nu77 binding in transcriptional elongation areas. Nur77 was poorly enriched in states numbers 6 and 7, both described as weak enhancers which differ in the occurrence of H3K4me2 and DNase sensitivity (22). Finally, negative enrichment was observed in heterochromatin regions, indicating the absence of Nur77 in inactive areas of the chromatin (Fig. 3C).
Altogether these data indicate that Nur77 is mostly associated with active sites of chromatin, concordant with the role of Nur77 as a transcriptional activator, also confirmed by the high presence of Nur77 in the TSS. Our data also shows that Nur77 binds transcriptional transition areas, suggesting that Nur77 is present in the promoters of its target genes independently of their state (active or inactive). On the other hand, the data suggests that the presence of Nur77 in enhancers would be limited to the active state.
In conclusion, our data analyses show that Nur77 is bound to the promoter of its target genes independently of their transcriptional state (weak, poised or active). In addition, our data suggests that the presence of Nur77 in enhancers is limited to the active state. We propose that Nur77 is always present in promoters but only binds to enhancers when it is upregulated, modulating thus transcription in response to stimuli.
Our analysis showed a strong participation of Nur77 target genes in anchoring and adhesion functions, which is consistent with the previously described roles of Nur77 in the modulation of neurite growth in neurons (13), and in the immune response (14,15), both processes that require interaction between cells and with the extracellular matrix.
Finally, genes found in this work as common targets of Nur77 in the nervous and immune systems are new and undescribed targets of this transcription factor. The work presented here is an approach pretending to guide the experimental focus regarding Nur77 investigation, solving in part the problem of lack of knowledge about Nur77 target genes and presenting new functions that can be attributed to this transcription factor in both the immune and nervous systems.
Limitations
The work presented here is a re-analysis of previously validated databases. However, differences in protocols or the overexpression of Nur77 could generate biases in the analyses. To be sure that genes described here are really modulated by Nur77, we were very restrictive in the selection process, this could lead to an underrepresentation of all genes regulated by Nur77 in neurons.