Connexin46 in the nucleus of cancer cells: a possible role as transcription modulator

Abstract

Background: Oncogenes drive cancer progression, but few are active exclusively in tumor cells. Connexins (Cxs), traditionally recognized as ion channel proteins, can localize to the nucleus and regulate gene expression, playing key roles in both physiological and pathological processes. Cx46, once thought to be restricted to the eye lens, has been implicated in tumor growth, though its underlying mechanisms remain unclear. This study investigates the nuclear presence of Cx46 in cancer cells and its potential role as a transcriptional modulator.

Methods: We employed ChIP-Seq, confocal immunofluorescence, and nuclear protein purification to assess Cx46 localization and DNA interactions. Functional assays were conducted to evaluate its effects on invasion, division, spheroid formation, and mesenchymal marker expression. Single-point mutations and molecular dynamics simulations were used to explore potential Cx46-DNA interactions.

Results: Cx46 mRNA upregulation was found in a variety of tumors compared to adjacent healthy tissue. In HeLa cells, which do not express Cx46, its transfection promoted proliferation, invasion and self-renewal capacity, cancer stem cell traits and mesenchymal features. Consistently, in Sk-Mel-2, which naturally express Cx46, reduced Cx46 expression led to a decrease in the similar parameters. In HeLa cells, nuclear Cx46 was detected in two forms, full length 46 kDa and a 30 kDa fragment (GJA3-30 k), ChIP-Seq experiments revealed that Cx46 binds to the DNA at intergenic and promoter regions, leading to the activation of oncogenic pathways. Molecular dynamics simulations suggest that GJA3-30 k dimerizes in a RAD50-like structure, forming stable DNA complexes. Cx46 and in some cases GJA3-30 k were detected in the nuclei of multiple cancer cell lines, including prostate, breast and skin cancers.

Conclusions: Our findings reveal a novel nuclear role for Cx46 in cancer, demonstrating its function as a transcriptional regulator and its potential as a therapeutic target.

Keywords: Cancer; Connexin46; IRES; Transcription factor.

Fig. 1

Gene expression levels of Cx46 in various cancer types. Gene expression data was obtained from the publicly available TCGA repository. This analysis compares the average expression of the GJA3 gene (encoding Cx46) in different tumor types (yellow bars) with the expression in corresponding healthy tissues (blue bars). Each bar represents the mean ± SD. Statistical significance was assessed using the Mann–Whitney U test. These findings provide valuable insights into the dysregulation of Cx46 gene expression across various cancer types, suggesting a potential for Cx46 in the pathogenesis of these malignancies

Fig. 2

Effects of Cx46 expression on the proliferation, invasion, and stemness of HeLa cells. a Representative Western blot of wild-type (WT) and Cx46-expressing HeLa cells. A Cx46-specific antibody targeting the C-terminal region of the protein was used, with β-actin as a loading control. Cx46 and actin bands were visualized using an HRP-conjugated secondary antibody. b Phase-contrast microscopy of HeLa WT and HeLa-Cx46 cells, revealing morphological differences. c Proliferation assay was performed by incorporating EdU (a thymidine analog) into dividing cells for 4 h, followed by detection using an epifluorescence microscope (red). Hoechst staining (blue) was used to determine the total number of cells after fixation. The graph presents the mean ± SE of EdU-positive cells relative to Hoechst-stained cells. Statistical analysis was performed using the Mann–Whitney U test. d Invasion assay was conducted using a transwell system, assessing the ability of cells to traverse an extracellular matrix over 72 h. The upper panel shows representative images of invaded cells with DAPI-stained nuclei. The lower panel quantifies the number of invaded cells (mean ± SE, from three independent experiments). Statistical analysis was conducted using the Mann–Whitney U test. e Mesenchymal marker expression was analyzed by qPCR. Graphs display the quantification of three independent experiments. F-G Stemness was assessed using a colony-forming unit (CFU) assay and a spheroid formation assay. For CFU, 35 mm dishes were seeded with 50 cells and after 14 days, the colonies were stained with crystal violet. (mean ± SE, from three independent experiments). Statistical analysis was conducted using the Mann–Whitney U test. g For the spheroid formation assay, cells were seeded cells in a non-adherent culture system for 14 days and images were taken using a Nikon Eclipse-Ti (n = 3)

Fig. 3

Cx46 exhibits nuclear localization in HeLa cells. a Representative confocal images captured using a 60 × oil objective. HeLa-Cx46 cells were fixed and stained with an anti-Cx46 antibody (red) and an anti-H3acK27 antibody (green). The zoomed-in inset in the lower left corner highlights a specific region of the field, marked by a white segmented square. Co-localization of the red and green appears as yellow, indicating the presence of Cx46 in the nucleus. Scale bar = 10 mm (b) Orthogonal view: Orthogonal projections generated from different planes provide a three-dimensional perspective of Cx46’s subcellular localization. These views confirm its presence in the nucleus. c Western blot analysis of cellular fractions: Cytoplasmic (C) and nuclear (N) protein samples were analyzed by Western blot. Cx46 was detected in both cytoplasmic and nuclear samples, appearing as two major bands (~30 and 46 kDa). An antibody against Laminin B (nuclear protein) was used as a control, which recognizes nuclear samples and not cytoplasmic samples (~68 kDa)

Fig. 4

Analysis of the DNA binding properties of Cx46. a Distribution of Cx46 Enrichment Peaks. This panel shows the distribution of Cx46 enrichment peaks throughout the genome, indicating the percentage of enrichment peaks in across genomic regions. b Heat Map of Cx46 enrichment: A heat map representation shows Cx46 binding around the transcription start site (TSS) within a ± 5 kbp region. The left side represents the unenriched chromatin IgG control, while the right-side displays where Cx46 is bound to DNA, highlighting its preferential binding sites. c Sequence analysis: Analysis of the DNA sequences that most commonly bind to Cx46. The left side displays sequences specific to Cx46 binding, whereas the right side shows sequences that interact with both Cx46 and other transcription factors. d The gene ontology of the genes associated with the enrichment cluster derived from the heat map is presented. The ontology of cluster 1 is shown above, and the ontology of genes derived from cluster 2 is shown below. This analysis provides insights into the functional categories and biological processes associated with those genes that show Cx46 binding. e Fold changes in gene expression: A comparison of gene expression levels between HeLa-Cx46 and HeLa wild-type cells is shown, indicating the fold changes in mRNA levels. This analysis reveals the impact of Cx46 on gene expression and its potential role in modulating transcriptional activity in HeLa cells

Fig. 5

Association of Cx46 with differential gene expression in HeLa cells. a Cx46 association with differentially expressed genes: This panel displays a representation of 23 genes that exhibit altered expression in HeLa-Cx46 cells. Genes with Cx46 enrichment peaks in their promoter regions are indicated by arrows. Up-regulated genes are represented in orange boxes, while down-regulated genes are represented by light blue boxes. Additionally, some genes are differentially expressed but lack direct Cx46 binding at their transcription start site (TSS), indicated by separate boxes with arrows. b Cx46-regulated gene network: A network diagram illustrates the regulatory network influenced by Cx46. This network was constructed by integrating data from Chip-seq, RNA-seq, and databases RegNetwork and DoRothEA. Up-regulated genes are depicted in orange boxes, while down-regulated genes are shown by light blue boxes

Fig. 6

GJA3-30 k has the potential to interact with DNA. a A representative model of the Cx46 hexamer and DNA interactions, as predicted by AlphaFold3, suggesting potential DNA-binding capacity for Cx46. b HeLa cells were transfected with wild-type Cx46 or mutants in which the second (M2), third (M3), fourth (M4), and fifth (M5) methionines were replaced with alanine. A zoomed-in Western blot (right panel) reveals that the 30 kDa fragment is absent exclusively in the M3 mutant, suggesting its critical role in the generation of the GJA3-30 k fragment. c A structural model of the DNA-bound GJA3-30 k complex, as predicted by AlphaFold3. d Superimposed structures of GJA3-30 k (green) and RAD50 (red), illustrating structural similarities between GJA3-30 k and RAD50, a protein essential for DNA repair and cell cycle regulation

Fig. 7

Western blot analysis of nuclear and cytoplasmic protein fractions from prostate cancer (DU145, LNCaP, and PC3), melanoma (SkMel2, 397Mel, and Mel1), and breast cancer (MDA-MB-231 and ZR75) cell lines. Lamin B1 was used as a nuclear purity control (N) and GAPDH and actin as cytoplasmic markers (C). The western blots demonstrate Cx46 expression across all analyzed cell lines, presented at both high and low exposure times. Cx46 is observed at molecular weights of approximately 68, 63, 46 kDa. Additionally, the melanoma cell lines show the presence of GJA3-30 K. Two exposure times for Cx46 are shown: one at 2 s and the other at 10 s, in order to highlight bands with lower abundance