Keywords

1 Introduction

Malignant transformation is initiated by oncogenic mutations and loss of tumor suppressor genes, causing loss of growth control. However, the resulting proliferative imbalance is usually controlled by the normal surrounding tissue, thereby suppressing tumorigenesis (Bissell and Hines 2011). Tumors form when cancer cells succeed to distort local tissue homeostasis and recruit normal cells to support their sustained proliferation and evasion of immune surveillance. These non-transformed cells protect the cancer cells and support them by creating a pro-tumorigenic tumor microenvironment (TME) or tumor stroma (Fig. 7.1a). The TME comprises a variety of cell types, including lymphocytes, macrophages, neutrophils, fibroblasts, and endothelial cells as well as extracellular matrix (ECM) components (Hanahan and Weinberg 2011; Hance et al. 2014; Place et al. 2011). Endothelial cells give rise to the neoangiogenic vasculature, recruit immune cells, and modulate cancer cell dissemination and metastasis. Neutrophils and lymphocytes can mount cancer cell-killing responses or release pro-inflammatory factors that stimulate tumor progression (Hanahan and Coussens 2012; Sagiv et al. 2015). Cancer-associated fibroblasts (CAF), which are perhaps the most abundant cells in the TME of carcinomas (Hanahan and Weinberg 2011) support cancer cells through secretion of ECM, chemokines and cytokines (Alspach et al. 2014; Coppe et al. 2008; Erez et al. 2010; Finak et al. 2008). CAFs also promote the recruitment of tumor-associated macrophages (TAMs), with which they engage in a reciprocal relationship that promotes malignancy (Cirri and Chiarugi 2011; Comito et al. 2014; Erez et al. 2010). TAMs further stimulate angiogenesis, enhance cancer cell migration and invasion, and suppress anti-tumor immunity (Qian and Pollard 2010).

Fig. 7.1
figure 1

Tumors are stressful microenvironments. (a) Cancer cells are depicted in shades of grey, different cell types interacting with cancer cells in the tumor microenvironment are highlighted in colors as detailed. (b) Stress conditions inherent to the tumor microenvironment are shown

Full size image

Throughout evolution, stress responses have helped cells and organisms survive in harsh conditions and overcome population bottlenecks. Rapidly evolving tumors are exposed to oxidative stress, DNA damage, metabolic stress, hypoxia and low pH (Fig. 7.1b; (Leprivier et al. 2015)). Moreover, they are exposed to population bottlenecks imposed by natural steps of tumor progression such as invasion and metastasis, as well as by external forces such as surgery and drug treatments. Under these extreme conditions cells of the TME and cancer cells constantly communicate with one another via secreted factors such as cytokines, chemokines, growth factors, and proteases (Witz 2008). This ongoing cell-to-cell communication helps cancer cells adapt to the stressful environment by modifying signalling pathways and reprogramming normal neighbouring cells. Cancer cells need a means of adapting their own protein machinery to function under stressful conditions and to signal to the TME to enable survival and metastasis. One way of surviving under extreme conditions is via activation of molecular chaperones.

Molecular chaperones or heat-shock proteins (HSPs) are a large family of highly conserved proteins traditionally known to be involved in protein folding, protein trafficking and assembly/disassembly of oligomeric structures. Under stress conditions they facilitate correct folding and prevent toxic protein aggregation (Hartl et al. 2011). Chaperones can be found in various subcellular and extra-cellular localizations, such as the nucleus, cytosol, endoplasmic reticulum (ER), mitochondria, lysosomes, cell-surface and blood (Henderson 2010). Depending on their localization chaperones can play important roles in inter and intracellular signalling. They interact with transcription factors, hormone receptors and kinases (Whitesell and Lindquist 2005). They integrate a wide range of cellular signalling pathways which help cells adapt to extreme environments (Maguire et al. 2002).

In tumors, chaperones promote the survival of cancer cells in the harsh tumor microenvironment. Master regulators of cytosolic chaperones such as Heat-shock factor 1 (HSF1), and ER chaperones such as XBP1 were shown to be associated with poor patient outcome and to promote cancer in mice (Romero-Ramirez et al. 2004; Santagata et al. 2011). The pro-tumorigenic activity of master regulators of stress responses such as HSF1 and XBP1 is mediated in part by activation of chaperones. Importantly however, these transcription factors activate a large repertoire of cancer-specific targets, and their activity in cancer is distinct from the typical stress-induced activity (Chen et al. 2014; Mendillo et al. 2012; Scherz-Shouval et al. 2014). Cancer cell-autonomous activities are discussed in other chapters of this book. Here we focus on the non-cancer-cell-autonomous activity of stress-activated transcription factors and chaperones and on the various ways by which they modulate the TME (Fig. 7.2; Table 7.1).

Fig. 7.2
figure 2

Chaperones mediate cell-cell interactions in the tumor microenvironment. Signaling pathways mediated by stress-activated transcription factors and chaperones are presented. (See Table 7.1 for details and references)

Full size image
Table 7.1 Non-cell-autonomous signaling mediated by stress-activated transcription factors and chaperones in the tumor microenvironment
Full size table

2 Chaperones in Human Cancer

Increased expression of chaperones is a common feature of various human cancers. Heat-shock protein 70 (Hsp70) is overexpressed in most human cancers, including breast, colon, liver, prostate and acute myeloid leukaemia (AML) (Sherman and Gabai 2015; Steiner et al. 2006). Importantly however, the correlation between its expression and patient outcome is mixed and is cancer type specific: in gastric cancer Hsp70 expression has no prognostic value, and in renal and oesophageal cancer increased levels of Hsp70 are associated with better disease outcome (Maehara et al. 2000; Sherman and Gabai 2015; Shiozaki et al. 2000). Heat-shock protein 90 (Hsp90) is overexpressed in various human cancers as well (Ciocca et al. 2013; Pick et al. 2007), and has been a long sought-after therapeutic target in cancer. To date, 18 Hsp90 inhibitors have entered the clinic, of which 5 are still in active clinical trials. Unfortunately, none of these inhibitors have been FDA approved yet (Yuno et al. 2018). The main hindrance in developing cancer-selective Hsp90 inhibitors is the dependence of normal cells on Hsp90 for survival, and the fact that inhibition of Hsp90 leads to activation of HSF1 (Whitesell et al. 2014; Yuno et al. 2018). Both could potentially be overcome by applying low-level inhibition of Hsp90 in combination with chemotherapy or hormonal therapy (Whitesell et al. 2014). The ER Hsp70 chaperone glucose regulated protein 78 (GRP78), also known as the immunoglobulin heavy chain binding protein (BiP), is induced under stress conditions such as glucose depletion, anoxia, acidosis and ER stress. In patients GRP78 is thought to be involved in the development of castration-resistant prostate cancer and increased levels of it are associated with disease recurrence in prostate cancer (Daneshmand et al. 2007; Pootrakul et al. 2006). Not only chaperones, but also the factors driving their expression are activated in human cancer. Activation of HSF1 both in cancer cells and in CAFs is associated with poor patient outcome in a variety of human carcinomas (Liao et al. 2015; Mendillo et al. 2012; Scherz-Shouval et al. 2014). Overexpression of activated XBP1 correlates with poor prognosis in glioblastoma (Avril et al. 2017; Obacz et al. 2017; Pluquet et al. 2013; Rubio-Patino et al. 2018), and in luminal/ER+ as well as triple-negative breast cancer (TNBC) (Cancer Genome Atlas Network 2012; Chen et al. 2014). Another arm of the unfolded protein response (UPR), that of PKR-like ER kinase (PERK) is upregulated in human breast ductal carcinomas in situ (DCIS), where PERK phosphorylation is higher in DCIS tissue than normal breast (Avivar-Valderas et al. 2011).

3 Mechanisms of Chaperone Mediated Tumor-Stroma Interactions

Chaperones play an important role in linking stress at the cellular level to stress in the organism. Though classically considered cell autonomous survival pathways, evidence suggesting non-cell autonomous activation of the heat shock response (HSR) and the unfolded protein response (UPR) has been accumulating. This systemic activation of proteostasis mechanisms allows a coordinated response to stress and coordinated ageing of cells in different tissues (Taylor et al. 2014; van Oosten-Hawle et al. 2013). Hijacked by cancer, it also supports the growth of tumors at the expense of the host. Chaperones and stress-activated transcription factors in tumors exert their non-cell autonomous effects via two main routes: (1) Activated in cancer cells, they mediate intercellular communication with cells of the TME through direct secretion from cancer cells or by chaperoning of cell-surface proteins, ECM components, and secreted molecules (2) Activated in cells of the TME, they facilitate cancer-promoting activities such as immune modulation, ECM remodeling and angiogenesis.

4 Chaperones Are Secreted to the Extracellular Space

In addition to their canonical cytosolic activity, both Hsp70 and Hsp90 can act as cell surface and extracellular chaperones, and this activity could be tumor promoting but also anti-tumorigenic. The term ‘chaperokine’ was coined to describe extracellular Hsp70’s dual role as a chaperone and a cytokine (Asea 2005). Extracellular Hsp70 (eHsp70) can bind to cell surface receptors such as CD91 and LOX-1 on antigen presenting cells (APC) and elicit immune responses through presentation of peptides from its chaperoned clients (Theriault et al. 2005; Zhou and Binder 2014). This antigen presenting activity is shared with other Hsps including Hsp60 and Hsp90 (Asea et al. 2000; Quintana and Cohen 2005). Such immune modulating activities could explain why, in some cancers, Hsp70 expression is correlated with better prognosis.

Extracellular Hsp90 (eHsp90) can play a tumor-repressive role through similar antigen presentation activities to those of Hsp70. However, it has many tumor-promoting effects, since it supports wound healing, tissue regeneration and cell migration (Hance et al. 2014). Selective inhibition of eHsp90 activity by cell impermeant molecules results in reduced cancer cell motility and invasion (Tsutsumi et al. 2008). But how is it secreted and what is the mechanism by which it promotes migration and invasion? eHsp90 release is triggered by necrosis, disruption of membrane integrity, growth factors and a variety of stress factors including DNA damage, oxidative stress, heat stress, hypoxia and exposure to chemotherapy (Hance et al. 2014). In the extracellular space eHsp90 serves as a major regulator of metalloprotease (MMP) activity (Correia et al. 2013; Eustace et al. 2004). MMP-2 is a client of the Hsp70/Hsp90 organizing protein (HOP) complex and the stabilizing interaction of these chaperones with MMP-2 promotes tumor invasion (Eustace et al. 2004; Sims et al. 2011). In advanced stage gastric cancer patients Hsp90 expression is significantly correlated with MMP-9 expression, and associated with poor prognosis (Wang et al. 2013). Secreted eHsp90 also promotes mammary epithelial invasion through its activation of MMP-3 (Correia et al. 2013). By controlling the activity of MMPs, eHsp90 indirectly controls the structure and composition of the ECM. eHsp90 also regulates signalling through its interaction with cell surface receptors such as low-density lipoprotein receptor-related protein 1 (LRP1) and HER2 (Hance et al. 2014; Sidera et al. 2008).

5 Chaperones Mediate Cell-Cell Communication via Exosomes

Chaperones were spotted on the surface of cancer cells and in extracellular domains more than a decade ago (Shin et al. 2003; Tsutsumi et al. 2008), but the mechanism by which they are exported from cells was not clear. A growing body of evidence suggests that at least part of the extracellular activity of chaperones is mediated through exosomes. Exosomes are nano-extracellular vesicles released by different types of mammalian cells. They contain protein, RNA and DNA. In cancer, exosomes are critical mediators of the communication between cancer cells and cells of the TME (Becker et al. 2016). Under stress conditions and in response to drug treatment, the secretion of Hsp60, Hsp70 and Hsp90 from cancer cells via exosomes is increased (Lv et al. 2012), stimulating migratory and cytolytic activity of natural killer (NK) cells and thereby serving as anti-tumor agents (Gastpar et al. 2005). Beneficial effects were also reported for small Hsps (Hsp27 and Hsp20) secreted through exosomes in various pathological conditions (Reddy et al. 2018), and Hsp20 has been implicated in exosome biogenesis (Wang et al. 2016). Notably, intercellular transmission of chaperones mediated by exosomes was shown to maintain and improve proteostasis in recipient cells (Takeuchi et al. 2015). Though this hasn’t been reported in cancer, one could envision how such non-cell-autonomous maintenance of organismal proteostasis is subverted to promote the survival of tumors.

6 Chaperones Are Activated in Cells of the TME

With the growing understanding that cells of the TME play important roles in cancer, an increasing number of studies have recently shown activation of stress-induced transcription factors such as HSF1 and XBP1 and chaperones including Hsp70, Hsp90 and GRP78 in cells of the TME. HSF1 is activated in cancer cells as well as in fibroblasts and endothelial cells, and in each cell type it drives a distinct transcriptional program. In cancer cells HSF1 promotes proliferation, invasion and migration (Mendillo et al. 2012). In fibroblasts HSF1 drives a wound-healing CAF signature, promoting the growth and malignant properties of adjacent cancer cells (Ferrari et al. 2019; Scherz-Shouval et al. 2014). In tumor-associated endothelial cells exposed to thermal treatment, HSF1 drives classic thermotolerance, leading to decreased transport of therapeutic agents in mouse tumors (Bagley et al. 2015). How does the same transcription factor drive different transcriptional programs in different cell types and different contexts? Possibly epigenetic alterations alter promoter accessibility. Alternatively the combination of stresses in the tumor – hypoxia, nutrient deprivation, genotoxic and proteotoxic stress all together leads to activation of multiple stress responses, resulting in different transcriptional outputs in different cell types in the TME.

Nevertheless, evidence for activation of canonical chaperones in the TME is accumulating as well. In particular, Hsp70 activation in macrophages promotes their migration and infiltration into tumors, and inhibition of this activity by a pharmacological inhibitor (JG-98) or by knockout of Hsp70 in the stroma profoundly affects tumor growth in mice (Gabai et al. 2016). The ER Hsp70 GRP78 is also activated in macrophages, as well as on the surface of T-cells, in response to stress, and regulates the activity of several cancer-associated cytokines, such as interleukin-6 (IL-6), macrophage migration inhibitory factor (MIF), transformed growth factor β (TGF-β) and interleukin-10 (IL-10) (Mendillo et al. 2012). IL-6 independent proinflammatory conditioning of macrophages by cancer cells was shown in a pathway termed transmissible ER-stress (TERS), where macrophages cultured in conditioned medium from ER-stressed cancer cells become activated, and themselves undergo ER stress with the up-regulation of Grp78, Gadd34, Chop, and Xbp-1splicing (Mahadevan et al. 2011). GRP78 is also expressed on the surface of endothelial cells, where it promotes tumor vascularization and angiogenesis (Virrey et al. 2008). Its overexpression drives resistance to anti-angiogenic therapy in models of glioblastoma (Virrey et al. 2008), whereas loss of GRP78 in endothelial cells supresses tumor growth and angiogenesis in a mouse model of melanoma (Dong et al. 2011).

GRP78 controls all 3 arms of the UPR through its binding to activating transcription factor 6 (ATF6), PERK, and inositol-requiring enzyme 1 (IRE1) (Obacz et al. 2017). Upon ER-stress GRP78 releases these proteins, and the UPR is activated. All three arms of the UPR have been linked to cancer, mostly through activation in cancer cells. The IRE1/XBP1 arm has been known to regulate innate immune responses in macrophages in response to pathogen-induced TLR activation (Martinon et al. 2010) and was suspected to do so in tumors as well. Recently an immunomodulatory role was shown for IRE1/XBP1 in dendritic cells (DC) in ovarian cancer (Cubillos-Ruiz et al. 2015). Activation of XBP1 in tumor-associated DCs disrupts their homeostasis and prevents T cell mediated anti-tumor immunity. Silencing of XBP1 in DCs reversed this process and prolonged survival of ovarian cancer-bearing mice (Cubillos-Ruiz et al. 2015). The cellular stress sensor CHOP is also activated in the TME, where it modulates anti-tumor immunity. Deletion of CHOP in Myeloid-derived suppressor cells (MDSCs) leads to induction of IL-6, resulting in T cell proliferation and anti-tumor responses in tumor-bearing mice (Thevenot et al. 2014). Thus, activation of chaperones in cells of the TME does not merely promote survival of the cells in which they are activated, but actually drives specific cancer phenotypes, thereby supporting survival of the tumor in a non-cell-autonomous manner.

7 Chaperones Play a Role in Angiogenesis

Starvation for oxygen leading to hypoxia triggers angiogenesis. Cancer cells sense hypoxia through HIF1-alpha, and trigger angiogenesis by upregulation of vascular endothelial growth factor (VEGF). The UPR plays a critical role in this process, triggered by exposure of endothelial cells to different types of stress such as hypoxia, low pH and glucose deprivation. The PERK/ATF4 axis promotes pro-angiogenic factors including VEGF, FGF-2 and IL-6 while decreasing angiogenic inhibitors including THBS1, CXCL14 and CXCL10 mRNA (Cubillos-Ruiz et al. 2017; Wang et al. 2012). XBP1 and ATF4 can directly bind and transactivate the promoter of VEGFA (Pereira et al. 2010), and all three arms of the UPR (IRE1, ATF6 and PERK) can induce endothelial cells to secrete VEGF (Karali et al. 2014). Several reports imply that HSPs may also play a role in angiogenesis (Calderwood and Gong 2016). Hsp27 secreted from endothelial cells regulates angiogenesis via direct binding to VEGF (Lee et al. 2012). Hsp90 can also increase angiogenesis by regulating HIF1-alpha/VEGF signalling, and inhibition of Hsp90 induces HIF1-alpha and VEGF degradation (Okui et al. 2011). To continue positively regulating angiogenesis endothelial cells secrete Hsp90 to the TME where it acts through chaperoning of MMP2 (Song et al. 2010).

8 Concluding Remarks

Cancer cells co-evolve with cells of the TME. The balance between tumor promoting and tumor repressive activities of the TME dictates whether tumors will progress, invade, and metastasize, or whether they remain dormant. Normal cells will naturally repress cancer-promoting phenotypes, and must be massively rewired to do otherwise. How do chaperones play into this rewiring? As mechanisms of stromal rewiring are being discovered, the paradoxical role of chaperones in cancer unravels. Chaperones are activated in cancer-associated fibroblasts, macrophages, dendritic cells, NK cells, MDSCs, and endothelial cells. As crucial mediators of cell survival they are potentially ideal targets for anti-cancer therapy – they promote invasion, angiogenesis, ECM remodeling, immune evasion, and drug resistance. Yet they also activate anti-cancer immunity. Moreover, inhibition of chaperone activity could lead to protein aggregation and related pathological conditions. Deeper mechanistic study of the pathways activating chaperones in different types of cancer – the regulators, the clients, the targets and the balance between cytosolic and extracellular activities are crucial for our understanding of the interplay between cancer, stress and evolution, and for successful translation of this knowledge into useful therapies.