Intercellular transmission of chronic ER stress: a new mechanism of metabolic diseases

Cell–cell communication plays a critical role in cell proliferation, differentiation, morphogenesis, functiogenesis, and tissue homoeostasis in multicellular organisms, while abnormal communication among cells may disrupt biological processes [1]. Intercellular communication is largely mediated by gap junctions (GJs), consisting of arrays of intercellular channels that enable cells to communicate by allowing ions and small molecules to directly transfer between neighboring cells [2]. Three families of integral membrane proteins forming GJ have been identified as innexins, connexins, and pannexins [2]. In vertebrates, six connexins (Cxs) oligomerized to connexons function as hemichannels and assemble into GJs, providing a channel for direct exchange of small signaling molecules between adjacent cells [2]. Cx43, a 43-kDa protein, as the most widely expressed Cx, plays essential roles in regulating many physiological and pathological processes in various tissues [3–7]. Recently, accumulating studies implicated that increased intercellular GJ activity and Cx43 expression play a critical role in the pathogenesis of atherosclerosis [3], chronic kidney disease [4], rheumatoid arthritis [5], osteoarthritis [6], and neurodegenerative diseases [7], which are associated with chronic inflammation and cellular stress. These findings suggested that increased intercellular communication might be a common feature of conditions characterized by various cellular stresses. Both obesity and type 2 diabetes are common metabolic diseases, accompanied by low-grade, chronic inflammation or metaflammation and cellular stress, including oxidative stress and endoplasmic reticulum (ER) stress [8]. Various insults disrupt protein homeostasis (proteostasis) and result in an accumulation of misfolded/unfolded proteins in the ER lumen, thus provoking ER stress. The unfolded protein response (UPR) is triggered to cope with the stress and restore ER homeostasis. Nevertheless, if the UPR fails to restore the proteostasis, the UPR signals will trigger cell death [8]. For a long time, ER stress response has been considered to be a cell-autonomous process; however, recent studies demonstrated that through certain soluble factors, ER stress signaling can be transmitted between cell models, e.g. from various cancer cells to tumor-infiltrating myeloid cells [9] or dendritic cells [10], from myocardiocytes to macrophages [11], from neuronal cells and astrocytes to unstressed central nervous system cells [12], and so on. Furthermore, Mahadevan et al. [9] tested transmissible ER stress in vivo, and they demonstrated that intraperitoneal injection of transmissible ER stress factors into normal mice elicited a generalized ER stress in the liver. As known, ER stress not only can occur in cells as part of normal physiology but also can occur frequently in many diseases, especially metabolic diseases involving inflammation [8]. Remarkably, Zhang et al. [13] reported that hepatic ER dysfunction and/or ER stress involved in the network of stresses, impairing systemic glucose homeostasis, thus, promoting the progression of insulin resistance and diabetes. However, whether ER stress in liver can propagate between communicated cells and how to coordinate the tissue response to ER stress remain unclear. A recent report by Tirosh et al. [14] showed that Cx43 expression was increased in response to ER stress in hepatocyte cell lines (Hepa1-6, AML12) and isolated primary mouse hepatocytes, thus increasing GJ-mediated cell–cell communication. Additionally, C57BL/6 mice fed with high-fat diet (HFD) for 16 weeks led to increased hepatic ER stress and also elevated Cx43 expression, indicating that Cx43 expression is associated with ER stress in vivo. By detecting UPR activity, it appeared that increased cell–cell coupling mediated by Cx43 allowed ER stress to disseminate from stressed donor cells to unstressed recipient cells. Furthermore, silencing of Cx43 expression by its small interfering RNA in donors or in recipient cells prevented cell–cell transfer of ER stress, which indicated that the transfer is in a Cx43-dependent manner. Intercellular communication may be beneficial or may be harmful (the ‘bystander’ effect) by opening a GJ channel through which ions, toxins, and inflammatory signals can rapidly spread to adjacent cells. Indeed, most studies so far indicated that intercellular communication is beneficial and considered as a universal way for tissue adaptation to various acute stresses. Kasper et al. [15] identified a mechanism of cell–cell communication that rapidly diffuses pro-inflammatory signals to uninfected cells through GJs, thereby amplifying innate immunity during bacterial infection. However, Maes et al. [16] found that hepatic GJs deteriorated upon acetaminophen-induced liver toxicity, which was associated with Cx43 production. In addition, the deficiency of Cx43 exacerbated the prognosis for liver injury, resulting in increased hepatocytic necrosis [17]. These findings highlighted the adaptive response of Cx43 and intercellular communication to acute injury in the liver. In agreement with these findings, Tirosh et al. [14] demonstrated that acute ER stress induced by tunicamycin injection led to an increase in hepatic Cx43-mediated intercellular communication, also identified as an adaptive response, as liver-specific Cx43 deletion in Cx43L-KO mice exhibited exacerbated hepatocellular damage and enhanced liver ER stress, while adenoviral delivery of Cx43 was hepatoprotective. Nevertheless, under chronic, prolonged stress conditions, Cx43-mediated intercellular communication becomes deleterious, facilitating the spread of ER stress, promoting ER dysfunction and metabolic disorders. HFD-induced obese mice initiates a chronic ER stress, which also enhances cell–cell coupling, but significantly decreases insulin sensitivity, glucose tolerance, and non-alcoholic fatty liver disease activity score. Meanwhile, hepatocyte Cx43 deletion markedly prevents this transmission and restores the glucose, lipid metabolic homeostasis induced by HFD [14]. Taken together, the increase in Cx43-mediated intercellular communication is an adaptive response to acute stress, which may turn into a maladaptive response to chronic, non-resolving stress. Remarkably, the expression of hepatic Cx43 in wild-type mouse is rather low but can be highly regulated under stressful conditions [18]. Furthermore, Cx43 was found to be upregulated very early during HFD-induced obesity [14]. Additionally, oxidative stress, inflammatory cytokines, glucotoxicity, and lipotoxicity have all been shown to promote Cx43 expression and increase intercellular communication [19]. Thus, it would be beneficial to reduce Cx43 expression for the prevention tissue spread of multiple harmful signals besides ER stress. In summary, Tirosh et al. [14] first demonstrated that hepatocytic Cx43-mediated intercellular transfer of obesity-induced ER stress led to fatty liver disease and metabolic alterations (Fig. 1). In the future, targeting hepatocytic Cx43 by various drugs would be an attractive therapeutic strategy for the treatment of metabolic disorders. Cx43-mediated intercellular transmission of hepatic ER stress Cx43-mediated intercellular communication is an adaptive response under conditions of acute ER stress. However, Cx43-mediated cell–cell communication allows chronic ER stress to spread, as a maladaptive response, resulting in metabolic disorders. Inhibition of Cx43 prevents intercellular transfer of chronic ER stress and restores metabolic homeostasis. ERS: ER stress, NAFLD: non-alcoholic fatty liver disease. Cx43-mediated intercellular transmission of hepatic ER stress Cx43-mediated intercellular communication is an adaptive response under conditions of acute ER stress. However, Cx43-mediated cell–cell communication allows chronic ER stress to spread, as a maladaptive response, resulting in metabolic disorders. Inhibition of Cx43 prevents intercellular transfer of chronic ER stress and restores metabolic homeostasis. ERS: ER stress, NAFLD: non-alcoholic fatty liver disease. This work was supported by the grants from the National Natural Science Foundation of China (No. 41877390), the Nature Science Fund of Hunan Province (Nos. 2019JJ40240 and 2016JJ2113), the Education and Innovation Fund of University of South China (No. 2019JG029), and the Graduate Research and Innovation Project of University of South China (No. 203YXC019). The authors declare that they have no conflict of interest.