Gap junctions are membrane structures made of intercellular channels which permit the diffusion from cytoplasm to cytoplasm of small hydrophilic molecules. Nearly 40 years ago, the loss of functional gap junctions has been described in cancer cells and led to the hypothesis that such type of intercellular communication is involved in the carcinogenesis process. From this time, a lot of data has been accumulated confirming that gap junctions are frequently decreased or absent in cancer cells whatever their tissue and species origins. Here, we review such data by insisting on the possible links existing between altered gap-junctional intercellular communication capacity (or the altered expression of their constitutive proteins, the connexins) and the stages of cancer progression in various cancer models. Then, we analyse particular aspects of the disturbance of connexin-mediated communication in cancer such as the cytoplasmic localization of connexins, the lack of heterologous communication between cancer cells and normal cells, the role of connexin gene mutations in cancer. In a separate part of the review, we also analyse the disturbance of gap-junctional intercellular communication during the late stages of cancer (invasion and metastasis processes).

Cancer was one of the first pathologies to be associated with gap-junction defect. Despite the evidence accumulated over the last 40-year period, the molecular involvement of gap junctions and their structural proteins (connexins) in cancer has not been elucidated. The lack of a satisfying explanation may come from the complexity of the disease, evolving through various stages during tumor progression, with cancer cells exhibiting different phenotypes. Here, the question of the involvement of gap junctions has been readdressed by considering the connexin expression/function level at different fundamental stages of carcinogenesis (cell proliferation, cell invasion, and cancer cell dissemination). By performing this analysis, it becomes clear that gap junctions are probably differently involved, depending on the stage of the cancer progression considered. In particular, the most recent data suggest that connexins may act on cell growth by controlling gene expression through a variety of processes (independent of or dependent on the gap-junctional communication capacity). During invasion, connexins have been demonstrated to enhance adherence of cancer cells to the stroma, migration, and probably their dissemination by establishing communication with the endothelial barrier. All these data present a complex picture of connexins in various functions, depending on the cell phenotype.

Despite the tremendous amount of data over the last 40years, lack of gap junctional intercellular communication (GJIC) or altered expression of gap junction proteins is still a lesser known 'hallmark' of cancer. Expression of astrocytic gap junction protein, connexin43 (Cx43), is often reduced in astrocytomas, the most common neoplasia of the central nervous system (CNS) in adults. Supported by a number of evidences, the global decrease of Cx43 expression appears to be advantageous for the growth of glioma cells. Although the mechanisms by which Cx43 regulates the expression levels of proteins involved in cell growth is unclear, there are evidences to suggest that it might be independent of their channel forming properties. In this regard, the carboxyl tail of Cx43 may have the ability to control the translocation of transcription factor regulators into the nucleus. However, this putative tumor suppressor effect of Cx43 is counterbalanced by its capacity to enhance the migration of glioma cells out of the tumor core through mechanisms that seems to implicate its carboxyl tail, possibly by interacting with the actin cytoskeleton. This ambivalence between the tumor suppressor effect and promotion of cell migration may partly be explained by the heterogeneous expression of Cx43 in the glioma core especially at the malignant glioblastoma stage; some tumor cells would be expected to migrate (Cx43 expressing cells) and others to proliferate (non-expressing Cx43 cells). Moreover, the involvement of Cx43 in glioma progression seems to be more complex since, in addition, GJIC may increase their resistance to apoptosis and Cx43 may also affect cell homeostasis in a paracrine fashion via hemichannel action. In conclusion, Cx43 appears to be involved at different levels of the glioma progression by acting on cell growth regulation, promotion of cell migration and resistance to apoptosis. This article is part of a Special Issue entitled: The Communicating junctions, composition, structure and characteristics.

Inflammation is a highly regulated process with common but also specific characteristics in each tissue affected. Recruitment of leukocytes from the blood to the injured tissue is an important early step in the inflammatory cascade. This review highlights the role of connexins (Cxs) in the regulation of both acute and chronic inflammatory processes. Cxs form gap junction channels that provide a cytoplasmic continuity between adjacent cells allowing the intercellular exchange of ions and metabolites. Their structural halves form connexons or hemichannels. Each of them consists of 6 Cx proteins and hemichannels not taking part in gap junction formation but facilitating the release of small molecules such as ATP. Based on the differential distribution of various Cxs in different tissues such as the brain, lung capillaries and large blood vessels, our aim was to analyze the specific roles of Cxs in the inflammatory process in these tissues. Three typical sites of inflammation were chosen to shed light on similarities and differences in several types of responses: (1) atherosclerosis as a model for chronic inflammation, (2) the lung as an example of acute inflammation and (3) the ‘immune-privileged’ environment of the brain to highlight specific reactions of the vasculature to ischemic damage and inflammation at this site.

Connexin43 (Cx43) is a ubiquitously expressed member of the gap junction protein family that mediates gap junction intercellular communication (GJIC) by allowing exchange of cytosolic materials. Previous studies have used Cx43 truncated at the cytoplasmic tail (C-tail) to demonstrate that the C-tail is essential to regulate cell growth and motility. Therefore, the aim of our study was to delineate the respective role of the truncated Cx43 and the C-tail in mediating Cx43-dependent signaling. A truncated Cx43 expressing the channel part of the protein (TrCx43, amino acid 1-242) and a construct encompassing only the C-tail from amino acid 243 (243Cx43) were transduced into LN18 human glioma cells. Our results showed that the ability of Cx43 to suppress growth was independent of GJIC as assessed by dye transfer, but was dependent on the presence of a rigid extracellular matrix. We further demonstrated that the C-tail alone is sufficient to promote motility. Surprisingly, Cx43 is also able to increase migration in the absence of the C-tail, suggesting the presence of at least two distinct signaling mechanisms utilized by Cx43 to affect motility. Finally, we used time-lapse imaging to examine the behavior of migrating cells and it was apparent that the C-tail was associated with a lamellipodia-based migration not observed in either mock or TrCx43 expressing LN18 cells. Our study shows for the first time that a free C-tail is sufficient to induce Cx43-dependent changes in cell morphology and that Cx43 signaling is linked to the regulation of the actin cytoskeleton.

For decades, cancer was associated with gap-junction defects. However, more recently it appeared that the gap junction proteins (connexins) could be re-expressed and participate to cancer cell dissemination during the late stages of tumor progression. Since primary tumors of prostate cancer (PCa) are known to be connexin deficient, it was interesting to verify whether their bone-targeted metastatic behaviour could be influenced by the re-expression of the connexin type (connexin43) which is originally present in prostate tissue and highly expressed in bone where it participates to the differentiation of osteoblastic cells. Thus, we investigated the effect of the increased Cx43 expression, by retroviral infection, on the metastatic behaviour of two well-characterized cell lines (PC-3 and LNCaP) representing different stages of PCa progression. It appeared that Cx43 differently behaved in those cell lines and induced different phenotypes. In LNCaP, Cx43 was functional, localized at the plasma membrane and its high expression was correlated with a more aggressive phenotype both in vitro and in vivo. In particular, those Cx43-expressing LNCaP cells exhibited a high incidence of osteolytic metastases generated by bone xenografts in mice. Interestingly, LNCaP cells were also able to decrease the proliferation of cocultured osteoblastic cells. In contrast, the increased expression of Cx43 in PC-3 cells led to an unfunctional, cytoplasmic localization of the protein and was correlated with a reduction of proliferation, adhesion and invasion of the cells. In conclusion, the localization and the functionality of Cx43 may govern the ability of PCa cells to metastasize in bones.

Precise diagnosis of low and high grades of brain tumors permits determining therapeutical strategies. So far, diagnosis and prognosis of gliomas were based on histological and genetic criteria which need being completed by a panel of molecular markers. Highly distributed in brain, gap junction proteins, connexins, could be considered as markers of glioma progression as previous studies indicated that expression of a connexin type, connexin43 (Cx43), is inversely correlated to tumor grading. However, this assumption was weakened by the low number of glioma samples used. Taking advantage of tissue microarray technique, we pursued this analysis by studying in situ expression of Cx43 on 85 samples (37 grade IV, 18 grade III, 24 grade II, and 6 grades II to III). Our analysis confirmed the global diminution of Cx43 expression in glioblastomas that was observed in previous studies. However, this analysis brought new insights such as the following ones. First, the high number of samples permitted to show that more than 60% of glioblastomas still express Cx43. Second, no gradual decrease in Cx43 expression was observed between grades II and III, but Cx43 appeared to be a marker distinguishing oligodendrocytic and astrocytic grade III tumors. Third, independently from tumor grade, a Cx43 nuclear staining was detected in areas where leukocytes are present. In conclusion, our study emphasizes the importance of in situ immunohistochemical approaches by giving more precise insights in the subcellular localization of Cx43. It also emphasizes the necessity to carry out such analysis on a wide range of samples to circumvent the high glioma heterogeneity.

The recovery of an intact epithelium following injury is critical for restoration of lung homeostasis, a process that may be altered in cystic fibrosis (CF). In response to injury, progenitor cells in the undamaged areas migrate, proliferate and re-differentiate to regenerate an intact airway epithelium. The mechanisms regulating this regenerative response are, however, not well understood. In a model of circular wound injury of well-differentiated human airway epithelial cell (HAEC) cultures, we identified the gap junction protein Cx26 as an important regulator of cell proliferation. We report that induction of Cx26 in repairing HAECs is associated with cell proliferation. We also show that Cx26 is expressed in a population of CK14-positive basal-like cells. Cx26 silencing in immortalized cell lines using siRNA and in primary HAECs using lentiviral-transduced shRNA enhanced Ki67-labeling index and Ki67 mRNA, indicating that Cx26 acts a negative regulator of HAEC proliferation.Cx26 silencing also markedly decreased the transcription of KLF4 in immortalized HAECs.We further show that CF HAECs exhibited deregulated expression of KLF4, Ki67 and Cx26 as well enhanced rate of wound closure in the early response to injury. These results point to an altered repair process of CF HAECs characterized by rapid but desynchronized initiation of HAEC activation and proliferation.