In a series of experiments, we demonstrated that TERS impacts adversely upon myeloid DC cross-presentation and cross-priming (Mahadevan et al. 2012), two events associated with the induction of CD8 T cell-mediated immunity.

Initial lineage analysis of CD8⁺ T cells cross-primed by TERS-imprinted bone marrow derived DC showed transcriptional upregulation of the cytokines Il–10 and Tnf- α but not Il–17, upregulation of Foxp3, and downregulation of the costimulatory molecule CD28. LAG3, a negative costimulatory molecule (Huard et al. 1994) found on tumor-infiltrating T cells (Grosso et al. 2007), was slightly up-regulated. When we analyzed the 96-h TERS-imprinted myeloid DC:T cell co-culture supernatant, we observed increased secretion of IL-2 but no elevation of IL-10, IL-17, IFN-γ or TNF-α compared to control (Fig. 18.3). A provisional conclusion is that CD8 T cells cross-primed by TERS-imprinted bone marrow-derived DC display an uncommitted phenotype with potential suppressive characteristics (CD28 downregulation and Il–10 upregulation) (Filaci et al. 2007). Surprisingly, CD8⁺ T cells cross-primed by TERS-imprinted BMDC also demonstrated disproportionately high splicing of Xbp–1 compared to only modest upregulation of other UPR elements, the significance of which remains unknown.

In sum, the phenotype of CD8+ T cells cross-primed by TERS-imprinted myeloid DC appears similar to that of CD8⁺/CD28^– regulatory T cells secreting IL-10 and TNF-α, and expressing FOXP3, which have been found to infiltrate a variety of human tumors (Becker et al. 2000; Kruger et al. 2001; Filaci et al. 2007; Mahic et al. 2008). It still remains to be seen whether, like human CD8 suppressor T cells, TERS-directed CD8 T cells have suppressor functions effected by surface ecto-ATPases (e.g. CD39) and/or soluble mediators (e.g. IL-10). A comparison of the CD8+ T cell phenotype derived from TERS-imprinted APC with CD8+T cells infiltrating human tumors is presented in Table 18.2.

Several lines of evidence suggest that TERS is operational in vivo. First C57BL/6 mice injected intra-peritoneally with TERS develop an ER stress response in liver cells characterized by the up-regulation of Grp78, Chop and spliced Xbp–1. This suggests that a tissue that is sensitive to ER stress induction, the liver, readily becomes a target of TERS administered systemically.

In the previous sections we discussed the cell-intrinsic role of the UPR in tumor adaptation and survival, as well as its putative cell-extrinsic role in polarizing myeloid antigen presenting cells to a phenotype that facilitates tumor outgrowth via T cell-dependent and independent mechanisms. Considering this dual role, targeting the UPR in the tumor microenvironment will likely have a dual benefit: impairing tumor cell microenvironmental adaptation and survival, and disabling a mechanism of host immune subversion. Based on our current understanding, the cellular targets, of any such intervention would be the tumor cell, myeloid antigen presenting cells, and CD8+ T cells (Fig. 18.5a). It remains to be seen whether CD4 T cell immunity is also adversely affected by the cell-extrinsic effects of the UPR.

The UPR is tumor microenvironment-specific as demonstrated by studies showing that peritumoral areas do not express UPR genes and that a constitutive UPR takes place within spontaneously growing tumors, though heterogeneously within a tumor mass (Spiotto et al. 2010). In addition, several lines of evidence indicate that UPR inhibitors selectively target tumor cells, as discussed below.

As the UPR represents an adaptive mechanism to cope with ER stress, targeting the UPR will likely take the following forms: (1) inhibition of the UPR in tumor cells with high levels of basal ER stress (eg. microenvironment-driven: hypoxia, glucose deprivation; tumor-intrinsic: secretory tumors, like myeloma), or (2) exacerbation of ER stress and consequent induction of cytotoxic/apoptotic effects. While each of these approaches will individually exploit tumor microenvironmental ER stress, either by its induction or by targeting its adaptive response (the UPR), they alone may not be sufficient to control the UPR within the complex and heterogeneous tumor microenvironment. For instance, exacerbating ER stress alone may exhibit cytotoxicity, especially in hypoxic/nutrient deprived areas; however, tumor cells mounting a UPR that leads to survival will have a UPR-mediated adaptive advantage, including resistance to chemotherapy (Pyrko et al. 2007) and host immunity. On the other hand, only inhibiting the UPR will target tumor cells with increased basal ER stress due to heterogenous microenvironmental noxae, sparing cells in more vascularized areas. We propose that optimal targeting of the UPR should take the form of inducing ER stress in tumor cells (fueling the fire) while concomitantly inhibiting the UPR (locking up the extinguisher) (Fig. 18.5b). In sum, this combinatorial mechanism will simultaneously take advantage of the cytotoxic potential of ER stress while inhibiting the response mechanism needed to adapt.

These strategies have already gained some experimental support. Bortezomib, a proteasome inhibitor that induces accumulation of proteins thus exacerbating ER stress, causes significantly higher cytotoxicity in hypoxic HeLa and human colorectal cancer cells than in normoxic cells, an effect dependent on ER protein load and consequent ER stress (Fels et al. 2008). Similarly, the induction of ER stress with a targeted thapsigargin pro-drug, celecoxib, or bortezomib, induces glioblastoma cell death, especially in hypoxic areas of the tumor (Johnson et al. 2002; Denmeade et al. 2012; Schonthal 2013). Combination of the ER stress inducers bortezomib and celecoxib, or its non-coxib analogue, 2,5-dimethyl-celecoxib (DMC), causes severe ER stress and apoptosis in murine glioblastoma cells in vitro and in vivo (Kardosh et al. 2008).

Inhibition of Xbp–1 splicing in multiple myeloma with the IRE1α endoribonuclease small molecule inhibitor, STF-083010, results in tumor cell-specific death in vitro and in vivo (Papandreou et al. 2011). Similarly, irestatin, a small molecule inhibitor of IRE1α endoribonuclease activity, inhibits hypoxic human myeloma and colon cancer cell survival and colony formation in vitro, as well as in vivo tumorigenesis in a heterotopic xenograft model (Papandreou et al. 2011). Targeting cell surface GRP78 in colon and lung cancer in mice with a monoclonal antibody (mAb159) causes tumor regression in vivo (Liu et al. 2013). Lastly, inhibition of GRP78 activation with active compounds present within the herbal medicine Ponciri fructis or the pyrone-type polyketide, verrucosidin, exhibits selective cytotoxicity in human pancreatic cancer cells or colon cancer cells undergoing glucose deprivation-induced ER stress (Park et al. 2007; Cha et al. 2009).

Inducing ER stress while concomitantly inhibiting the adaptive UPR has also begun to find experimental support. For instance, it has been shown that the mechanism of bortezomib’s cytotoxic activity in myeloma cells is its ability to inhibit Xbp1 splicing via stabilization of unspliced XBP-1, which acts as a dominant negative suppressor of XBP1-s, while inducing ER stress via proteasome inhibition (Lee et al. 2003a). Congruently, the induction of ER stress with bortezomib or 17-AAG in myeloma cells was shown to synergize with the activity of transgenic or small molecule-mediated inhibition of Xbp–1 splicing resulting in the induction of greater and irreparable cytotoxicity than either agent alone in vitro and in vivo (Lee et al. 2003a; Mimura et al. 2012). In human pancreatic cancer cells, bortezomib reduces GRP78 and CHOP expression under ER stress conditions and sensitizes them to ER stress-inducing compounds, including thapsigargin, tunicamycin, and cisplatin, yielding synergistic cytotoxicity in vitro and in vivo (Nawrocki et al. 2005). GSK2606414, a small molecule inhibitor of PERK autophophorylation and downstream eIF2α phosphorylation, cooperates with ER stress induced by hypoxia or thapsigargin, causing greater inhibition of in vitro clonogenic survival of pancreatic and colon cancer cells than either PERK inhibition or ER stress induction alone (Axten et al. 2012; Cojocari et al. 2013). Epigallocatechin gallate, which inhibits GRP78 by targeting its ATP-binding domain, sensitizes human glioma cells to ER stress induced by the chemotherapeutic agent, temozolomide, resulting in synergistic cyotoxicity, greater than either agent alone (Pyrko et al. 2007). There are several chemical UPR inhibitors that have displayed efficacy against tumor growth in vitro and in vivo (reviewed in (Li et al. 2011) and (Schonthal 2013)) these are presented in Tables 18.3 and 18.4.

While there has been recent interest in developing UPR inhibitors active against tumor cells, there has been little or no investigation the effect of tumor UPR inhibition on the host anti-tumor immune response. A link between the tumor UPR and the immune response was originally suggested by the finding that silencing of Grp78 in mouse fibrosarcoma cells inhibited growth in an in vivo syngeneic transplantation model due, in part, to increased tumor cell-specific memory T cell generation (Jamora et al. 1996). In addition, overexpression of GRP78 in murine insulinoma cells leads to impaired CD8 T cell priming and inhibition of killing, when GRP78-overexpressing tumor cells were used to prime cytotoxic T cell lines, as targets, respectively (Wang et al. 2007). Discovery and characterization of the effect of TERS on host immunity has continued this line of inquiry (Zanetti 2013).

Based on our findings on transmissible ER stress, it appears that a fruitful avenue for therapeutic development will be to develop decoy systems (antibodies, aptamers, etc.) to intercept the TERS factor(s) in the extracellular space (Fig. 18.5a). In this scenario neutralization of TERS would also inhibit the polarization of myeloid cells to a pro-inflammatory/suppressive phenotype, and in turn prevent and unfetter the untoward effects on T cell-mediated immunity, perhaps permitting more effective autochthonous or vaccine-induced anti-tumor immune responses. In addition, TERS may induce tumor-infiltrating myeloid cells to produce tumorigenic cytokines and adversely affects antigen presentation (see Sects. 3.2 and 3.3 for discussion). Lastly, as the downstream effects of TERS on T cell priming are elucidated (i.e. polarization toward a suppressive phenotype), new targets for therapy will come to light (e.g., ecto-ATPases, immune checkpoint molecules, UPR signaling components). Targeting the tumor-infiltrating myeloid cell UPR, the tumor cell UPR, and ultimately suppressive T cells, will reset the multifaceted dysregulation of the tumor microenvironment that hinders anti-tumor immunity.