Unfolding anti-tumor immunity: ER stress responses sculpt tolerogenic myeloid cells in cancer
Role of the UPR in malignant cells
The key interaction between the UPR and tumorigenesis has been comprehensively discussed in previous reviews [1, 4, 5, 34]. Malignant cells thrive under ER stress-inducing conditions such as hypoxia, nutrient deprivation, and low pH. In addition, cancer cells generate reactive metabolic byproducts that avidly modify ER-resident proteins and chaperones. Notably, the induction of various UPR-related factors has been commonly reported in patients with various cancer types and their overexpression usually correlates with poor prognosis and resistance to therapy [21, 44–46]. Interestingly, treatment of tumor-bearing mice with the ER stress inducer thapsigargin increased tumor growth, whereas global UPR inhibition using chemical chaperones, such as 4-Phenylbutyric acid (4-PBA) or tauroursodeoxycholic acid (TUDCA), delayed tumor progression and metastasis [9, 47].
Seminal studies have determined the cancer cell-intrinsic protumoral role of the IRE1?- XBP1 and the PERK-eIF2? pathways in vivo. Implantation of malignant cells or transformed fibroblasts lacking IRE1?/XBP1 or PERK/eIF2? in mice resulted in reduced tumor growth, which was attributed to low angiogenesis and increased sensitivity of the cancer cells to ER stress inducers, including hypoxia and high levels of ROS [35]. Accordingly, targeting IRE1? or PERK signaling in vivo with specific small-molecule inhibitors has shown significant therapeutic effects in various preclinical models of disease [48–52]. More recently, XBP1 was demonstrated to foster triple negative breast cancer progression by cooperating with HIF1? to support tumor-initiating cell function and metastatic capacity under hypoxia [21]. XBP1 contributes to the pathogenesis of multiple myeloma [53], and has been implicated in cancer cell de-differentiation, susceptibility to oncovirus infection and the epithelial-to-mesenchymal transition [54]. Andrew Hu and colleagues have elegantly demonstrated constitutive IRE1?-XBP1 activation in chronic lymphocytic leukemia cells, which promoted their pathogenesis in vivo [48]. In addition, inhibiting IRE1? function by overexpressing a dominant negative IRE1? variant significantly increased overall host survival by decreasing tumor growth rate and angiogenesis in a model of glioma [55]. Recent studies have also indicated that IRE1?-XBP1 signaling supports the aggressiveness of pancreatic cancer cells in xenograft models [56].
Similar to the effect induced by IRE1?-XBP1 signaling, the activation of PERK-eIF2? has also been implicated in the development of several malignancies, including breast, lung, and liver carcinoma [36, 47]. In those models, deletion of Perk rendered malignant cells highly susceptible to the cell death induced after exposure to hypoxia, DNA damage, low levels of nutrients, and high levels of reactive oxygen species [57]. Furthermore, the absence of PERK-eIF2? signaling impaired the ability of breast cancer cells to migrate and invade, thereby decreasing their ability to metastasize in vivo [49, 58, 59]. Therefore, the inhibition of PERK resulted in cancer cell apoptosis and significant anti-tumor effects [43]. As such, silencing of Perk increased the therapeutic efficacy of treatments based on the depletion of amino acids in T cell leukemia [60], and sensitized chronic myeloid leukemia (CML) cells to the apoptosis induced by the BCR/ABL inhibitor, imatinib mesylate [61]. Thus, the intrinsic effects of a controlled UPR in cancer cells appear to favor tumor growth and metastasis through the promotion of malignant cell survival, angiogenesis and chemoresistance, thus justifying the use of specific UPR inhibitors for the treatment cancer.
Although activation of the UPR has been primarily associated with cancer cell survival and tumor progression, some studies suggest that molecular factors in this pathway could also suppress tumor development in certain contexts. For instance, increased oncogenic transformation has been evidenced in fibroblasts after inhibiting the PERK target eIF2? [62], and increased proliferation and mammary tumor formation has been reported upon expression of a dominant-negative form of PERK in mammary epithelial cells [63]. Furthermore, in the context of acute myeloid leukemia, increased expression of ER stress response markers correlates with better prognosis in patients with this disease [64]. Taken together, these studies indicate that the effects of the UPR in cancer cells is context-dependent and that variables such as the stage of cancer progression and the cellular source of malignancy are critical determinants of whether this pathway plays either a pro-tumorigenic or anti-tumoral role.