Keywords

Introduction

Heat shock proteins are highly conserved molecular chaperones involved in proper folding of newly synthesized proteins. They are also able to interact with naïve and denatured proteins to avoid inappropriate interactions and the formation of protein aggregates of aberrantly folded proteins (Liberek et al. 2008). Furthermore, HSP facilitate protein translocation, exhibit cytoprotective (Sharp et al. 1999; Giffard et al. 2004) and antiapoptosis functions (Martin et al. 1992; Aquino et al. 1993; Gao et al. 1995; Aquino et al. 1997), directly interacting with various components of the tightly regulated programmed cell death machinery, upstream and downstream of the mitochondrial events (Garrido et al. 2001; Beere 2004; Madden et al. 2008). More recently, it has been demonstrated that HSP are also involved in controlling cell signaling (Calderwood et al. 2007), in modulation of both immune response (Johnson and Fleshner 2006) and chronic disease conditions (Kampinga et al. 2007). Furthermore, HSP has cytostimolatory functions (Asea 2008). Almost all the HSP families consist of constitutively expressed members playing housekeeping roles and stress-induced members, which display a crucial role in recovery after different cellular stresses (e.g. heat shock, ultraviolet radiation, viral or bacterial infections, oxidative stress, ischemia, exercise, metabolic stress, heavy metals and so on) (Collins and Hightower 1982; Lindquist 1986; Patel et al. 1995; Richard et al. 1996; Yang et al. 1996; Feder and Hofmann 1999; Said Ali et al. 2010; Pierzchalski et al. 2014). For many years heat shock proteins were thought to be exclusive cytoplasmic proteins, acting only in intracellular compartments. The biology of both intracellular constitutive and inducible HSP has been extensively summarized by a large number of reviews and it will be not treated anymore (Lindquist and Craig 1988; Morimoto 1991; Bukau et al. 2006; Hartl and Hayer-Hartl 2009).

A new twist in the stress field was given by the finding that HSP could be also extracellular protein. In fact, it is now widely accepted that almost all the HSP are released in the extracellular environment, exerting multiple effects on different cell types (Tytell 2005). For example, in 1986 Tytell and coworkers identified an HSP-like protein as a glia-axon transfer protein in squid giant axon (Tytell et al. 1986). In addition, Hightower and Guidon found that Hsp70 is released extracellularly by a mechanism different from the classical secretion endoplasmic reticulum (ER)-Golgi pathway (Hightower and Guidon Jr 1989), as later confirmed by Hunter-Lavin et al. (Hunter-Lavin et al. 2004). Therefore, HSP could have a dual function depending on their intracellular or extracellular location. In the past years heat shock proteins were classified into different families depending on their molecular mass (i.e., Hsp110, Hsp90, Hsp70, Hsp60, Hsp40 and the small HSP). They comprise members with different inducibility and intracellular localization (Feder and Hofmann 1999). In 2009, Kampinga and coworkers proposed new guidelines for human HSP nomenclature (Kampinga et al. 2009). However, as most literature refers to HSP with their old name we decided to maintain the old nomenclature.

Hsp70 Family

One of the most conserved and abundant subsets of HSP is the Hsp70 family (Muchowski and Wacker 2005; Noble et al. 2008), consisting of at least four distinct proteins, two are cytosolic/nuclear, whereas the other two are localized within mitochondria and endoplasmic reticulum (ER) (Kregel 2002; Voos 2013). Furthermore, cytosolic Hsp70 includes the stress inducible form Hsp70 (Voellmy 2004) and the constitutively expressed Hsc70. Although Hsp70 synthesis is induced by several types of stressors (Wu et al. 1985; Milarski and Morimoto 1989), it is now well accepted that a basal concentration of Hsp70 is present in many tissues without any stress. Indeed, in addition to their cytoprotective properties member of the Hsp70 families are involved in physiological processes, such as cell differentiation, cell maturation and proliferation (Luft and Dix 1999; Lui and Kong 2007).

Extracellular Hsp70 and Its Release Mechanisms

As for other HSP (Chabas et al. 2001; Stadelmann et al. 2005), it has been demonstrated that Hsp70, in both basal and stress-induced conditions, is not only an intracellular (iHsp70) but also an extracellular protein (eHsp70), functioning as intercellular-signaling ligand (Asea et al. 2000; Asea 2003; Calderwood et al. 2007; Turturici et al. 2011). Indeed, it has been demonstrated that a variety of cell types, including neural cells (Guzhova et al. 2001; Taylor et al. 2007), epithelial cells (Broquet et al. 2003), embryo cells (Hightower and Guidon Jr 1989), B lymphocytes and dendritic cells (Théry et al. 1999, Clayton et al. 2005), maturing erythrocytes (Mathew et al. 1995) and tumor cells (Gastpar et al. 2005) is able to release Hsp70 in the culture medium.

Early studies on eHsp70 hypothesized that this protein was released by necrotic or dead cells, due to its aminoacidic sequence, and in particular to its lack of any exocytosis signals, (Gallucci et al. 1999, Basu et al. 2000; Saito et al. 2005). On the contrary, it is now well known that eHsp70 is also actively released by mammalian cultured cells through non classical secretory pathways, excluding the ER/Golgi compartment (Hightower and Guidon Jr 1989; Broquet et al. 2003), as already demonstrated for other proteins that lack secretion leader sequence. eHsp70 was also detected in the peripheral circulation of either healthy subjects or in several pathological states (Pockley et al. 1998, 2002; 2003; Asea 2007; Giraldo et al. 2008). All these data demonstrated that there are at least two different methods of Hsp70 release: one active due to an unconventional secretory pathway (Nickel and Seedorf 2008), and the other one passive, due to cell death and subsequent lysis. One of the possible active Hsp70 release pathway is similar to that identified for IL-1β (Eder et al. 2008). Specifically, Mambula and Calderwood demonstrated the involvement of a lysosome-endosome pathway, requiring the ABC-1 transporter (Mambula and Calderwood 2006). Hsp70 has also been proposed to be released by secretory-like granules (Evdonin et al. 2006), typical of specialized endocrine and exocrine cells, but already identified in non-secretory cells (Beuret et al. 2004). Another interesting idea images that Hsp70 release was dependent on its initial insertion into plasma membrane, depending on phosphatidylserine presence, membrane fluidity and, for certain cell types, lipid raft integrity (Triantafilou et al. 2002; Broquet et al. 2003; Arispe et al. 2004; Hunter-Lavin et al. 2004; Chen et al. 2005; Wang et al. 2006a; Horvath et al. 2008; Vega et al. 2008; Evdokimovskaya et al. 2010). The importance of Hsp70 insertion in cellular membrane is confirmed by the finding of Hsp70 containing extracellular vesicles (EVs) in cell culture medium, even in the absence of detectable cell death (Hunter-Lavin et al. 2004; Gastpar et al. 2005; Vega et al. 2008). Two principal EV types, i.e. exosomes and membrane vesicles, with different cellular origins have been demonstrated to be involved in Hsp70 release in the extracellular environment. Exosomes are nanovesicles derived from the multivesicular endosomal cell compartment (Heijnen et al. 1999; Fevrier and Raposo 2004; Théry 2011). Several authors have demonstrated that in vitro different cell types are able to release Hsp70 in the extracellular environment by an unconventional pathway involving exosomes (Bausero et al. 2005; Gastpar et al. 2005; Lancaster and Febbraio 2005; Zhan et al. 2009). Indeed, for many years Hsp70/Hsc70 have been considered as exosome markers (Mathivanan and Simpson 2009). Contrary to this idea, for the first time Barreca et al. in 2017 demonstrated that mesoangioblasts, mouse vessel associated progenitor stem cells, are able to release Hsp70 in the extracellular milieu, even under basal growth conditions, by EVs originating directly from the plasmatic membrane (Barreca et al. 2017). Schematic model of different Hsp70 release mechanisms is represented in Fig. 1.

Fig. 1
figure 1

Mechanism of Hsp70 release from cells. Hsp70 is able to leave cells using different way: I through endolysosome pathway; II by using secretory-like pathways; IIIa it can penetrate through the lipid bilayer via phospatidylserine IIIb or through of lipid raft interaction; IV it can be released inside exosomes; V or within membrane vesicles. In the extracellular milieu Hsp70 can interact with several cell types (target cells)

Full size image

Extracellular Hsp70 Roles

As several mechanisms for extracellular Hsp70 release have been identified, there should be also several physiological roles for this protein in the external environment. In particular, its passive release from dying cells is considered as a danger signal. On the contrary, Hsp70 active release from living cells could indicate a successful stress response, and eHsp70 could have an active signaling role. An intriguing aspect of Hsp70 roles is its ability to induce antagonistic events, depending on its localization (Rodrigues-Krause et al. 2012). For example, iHsp70 exerts a strong anti-inflammatory effects through the interaction with the nuclear factor kB (NF-kB), blocking its activation (Jones et al. 2011), whereas eHsp70 has the opposite role, inducing the activation of several proinflammatory pathways. To date, there is evidence that eHsp70 can be internalized by several cell types, localizing both in the cytoplasm and in the nucleus, often promoting cell survival (Guzhova et al. 2001; Novoselova et al. 2005; Tytell 2005).

eHsp70 and Immune Responses

eHsp70 possesses powerful immunological properties. Two are the possible origins of eHsp70: pathogen derived Hsp70, which signals a local infection, and mammalian Hsp70, usually indicating an intracellular trauma. Hsp70, as other HSP associates itself with a broad variety of peptides, generated within the cells (Zhu et al. 1996). These peptides consist of normal self-peptides as well as antigenic peptides (i.e., tumor, bacterial and viral antigens) (Nieland et al. 1996; Castelli et al. 2001; Zugel et al. 2001). Several studies have demonstrated that eHsp70 is able to initiate both innate and adaptive immunity response as illustrated in Fig. 2 (Johnson and Fleshner 2006; Tsan and Gao 2009). For example, it is involved in the activation of cells of the innate immune pathway, such as macrophages, monocytes, which are directly induced in cytokine synthesis, neutrophils, dendritic cells (DCs), natural killer (NK) cells (Asea et al. 2000; Basu et al. 2001; Vabulas et al. 2002a; Gastpar et al. 2004; Aneja et al. 2006; Kovalchin et al. 2006; Wang et al. 2006a; Vega et al. 2008). In addition, eHsp70 also increases microbicidal capacity and chemiotaxis of neutrophils (Ortega et al. 2006, 2009) and phagocytosis (Wang et al. 2006a).

Fig. 2
figure 2

Schematic model of eHsp70 immune response activation. eHsp70 is able to initiate both innate and adaptive immunity. On the one end, eHsp70 is involved in the activation of cells belonging to the innate immune pathway, such as macrophages, DCs and NK cells, directly induced in cytokine synthesizing (chaperokine role). On the other end, eHsp70 can acts as chaperone of antigens, which are transferred to APCs, processed, and cross-presented to T cells, inducing a strong T cell activation (chaperone role)

Full size image

Phagocytosis characterizes the innate immune response and is initiated and executed by the antigen presenting cells (APCs). Some of these cells, including macrophages, are activated either by the presence of pathogens, or by physical, chemical stress, trauma and so on. All these conditions are responsible for HSP, and especially Hsp70, release in the extracellular environment. Immunologically, eHsp70 binds different receptors on macrophages membrane regulating several functions, such as cytokine release, phagocytosis, tumor-rejection and upregulation of co-stimulatory molecules (Basu et al. 2001; Delneste et al. 2002; Srivastava et al. 2002; Vabulas et al. 2002b; Asea 2005). Wang et al. demonstrated that eHsp70 is able to enhance murine macrophage-mediated antigen uptake. This process is concentration and time dependent. They also showed that eHsp70 mediated macrophage activation depends on actin cytoskeleton reorganization, and is not influenced by de novo protein synthesis. All these observations have suggested to the authors that the process of eHsp70 mediated phagocytosis occurs through the activation of a short signaling pathway, not involving gene upregulation. Moreover, they proved that eHsp70 and macrophages interaction occurs on lipid raft microdomains of macrophage plasmamembrane. Indeed, lipid raft disruption by cholesterol removal partially abrogates eHsp70 effects on phagocytosis. Finally, eHsp70 mediated phagocytosis enhances antigenic processing and presentation to CD4+ T lymphocytes in a MHC II limited manner (Wang et al. 2006a). The involvement of eHsp70 in macrophage activation was also demonstrated by Vega and collaborators, confirming the paracrine role in cytokine expression and secretion by macrophages, already observed by Svensson and coworkers (Svensson et al. 2006). In particular they demonstrated that extracellular membrane bound Hsp70 activated macrophages that produced higher level of TNF-α in comparison with unstimulated cells. Furthermore, TNF-α production after stimulation with membrane positive eHsp70 was much higher than that observed with recombinant or purified Hsp70. According to these results the authors hypothesized that the membrane environment makes eHsp70 a better target to be recognized by macrophages that can engulf it via phagocytosis. On the other end, it is possible that eHsp70 insertion within extracellular membrane is responsible for its multimerization, making it more active in triggering a response by macrophages (Vega et al. 2008).

In vivo wound healing experiments carried out by injecting into naïve mice mimicking a wounding model, macrophages pretreated with Hsp70 showed an enhanced macrophage wound closure ability, compared to buffer treated cells (Kovalchin et al. 2006). It is well known that macrophage-mediated phagocytosis plays a fundamental role in wound healing by clearing debris and neutrophils. Kovalchin’s group results indeed demonstrated that eHsp70 mediates some of its action on wound closure through the regulation of macrophage phagocytic activity (Kovalchin et al. 2006). According to HSP capability to interact with other cell types, such as neutrophils and platelets (Hilf et al. 2002; Radsak et al. 2003), it is also possible that eHsp70 stimulates wound closure not only by stimulating macrophage-mediated phagocytosis, but also altering the wound milieu via cytokine release.

The extracellular chaperone is also responsible for DC activation through their surface binding. In their immature form these cells are prevalently involved in antigen capture and processing, whereas mature DCs are potent APCs. Indeed, when their maturation is induced, they become potent stimulator of naïve T cells (Banchereau and Steinman 1998). According to this dual role, DCs represent a connection between the adaptive and the innate immune system. Therefore, in its function of DC activation, eHsp70 may stimulate a cross-talk between the two immune responses (Srivastava 2002). In particular, Kuppner et al. have observed that eHsp70 bound and stimulated maturation of immature differentiated DCs, but reduced their maturation from monocyte precursor cells, probably due to its incapacity to interact with these cells. In addition, DCs matured by eHsp70 increased cell proliferation level of PBL and showed an enlarged capability to stimulate IFN-γ synthesis from specifically stimulated CTL (Kuppner et al. 2001). DCs are also able to bind tumor eHsp70-peptide complex, which is endocyted through a receptor mediated process and cross-presented to MHC I (Noessner et al. 2002). Milani et al. demonstrated that eHsp70-peptide complex is not only involved in providing to DCs the peptide for presentation for T cell stimulation, but also stimulates DCs to release TNF-α inducing by an autocrine loop their maturation (Milani et al. 2002).

All the presented data indicate that eHsp70, acting as a danger signal to the innate immune system, on one hand induces: a cascade of proinflammatory cytokines (e.g. TNF-α, IL-1-IL-6, Il-12); NO and C-C chemokines release by monocytes, macrophages and DCs (Asea et al. 2000, 2002; Chen et al. 2002; Panjwani et al. 2002; Asea 2006); the stimulation of DCs and the activation of NK cells. Qiao et al. demonstrated a positive switch of the DC-NK cells cross activation. In particular, they found that soluble Hsp70, which in vivo is released as results of pathological cell stress and death (Basu et al. 2000), could amplify DC-NK cell interaction, leading to a higher IFN-γ response (Qiao et al. 2008). On the other hand, eHsp70 acts as chaperone of bound peptides, which are transferred to DCs, also induced to mature, and cross-presented to T cells, inducing stronger cell activation compared with antigenic peptide alone (Li et al. 2002; Bendz et al. 2007).

In the last years various papers have demonstrated that HSP-reactive T cells have an immunoregulatory phenotype, suggesting that these proteins, and in particular Hsp60 and Hsp70, could activate immunoregulatory pathways, which can suppress the immune response occurring in human inflammatory diseases. In fact, several data highlighted that in various inflammatory diseases, eHsp70 exerts under certain circumstances also immunoregulatory and suppressive functions. (Wieten et al. 2007). In various autoimmune models, such as rheumatoid arthritis and diabetes, eHsp70 can also downregulate the immune response through its ability to stimulate T cells (Th2 and Tregs) to produce IL-10, the main anti-inflammatory and immunosuppressive cytokine (Moore et al. 2001; Borges et al. 2012; Stocki and Dickinson 2012). This process, depending on the form of eHsp70 (i.e., associated with peptide, with membranes, with nucleotides, peptide free) and the receptor activated (Li et al. 2012) shifts cell phenotype towards a tolerogenic one exhibiting anti-inflammatory properties. It has been demonstrated that several adaptive mechanisms could be involved in eHsp70 specific induced Tregs. In particular, peptides derived from eHsp70, either pathogen associated or secreted endogenous Hsp70, can be presented to MHC II on IFN stimulated APCs or non APCs. Most Tregs subset induce IL-10, responsible for inflammatory response suppression (Wendling et al. 2000; Prakken et al. 2001; Bluestone 2005). Another example of the anti-inflammatory eHsp70 role was observed in a mouse model of allergic bronchial asthma inflammatory process. eHsp70 treated mice showed a significant decrease in Th2 associated cytokines IL-4, IL-5 and IL-13 secretion compared to untreated mice. In addition, it has been observed that eHsp70 administration promoted survival of neutrophils at the site of inflammation, and reduced the influx of the eosinophils to the lungs (Shevchenko et al. 2016).

Under certain circumstances also crossreaction with self-Hsp70 induce T cells to synthesize of IL-10 (Tanaka et al. 1999; Wendling et al. 2000). Indeed, in animal models of experimental arthritis T cell reactivity to self Hsp70 downregulates inflammation developing Th2 CD4+ T cells producing the regulatory cytokines IL-4 and IL-10 (Tanaka et al. 1999; Wendling et al. 2000; Wieten et al. 2007). Another anti-inflammatory role for eHsp70 was observed in rheumatoid arthritis. In particular, eHsp70 exerts an anti-inflammatory role on fibroblast-like synoviocytes inhibiting TNF-α induced IL-6, IL-8 and MCP-1 secretion. The silencing of this inflammatory response is due to the downregulation of NF-kB nuclear translocation (Luo et al. 2008). Immunization with the recognized bacterial Hsp70 peptide protects rats against adjuvant-induced arthritis (Tanaka et al. 1999). A protective effect from experimentally induced arthritis was also observed in rats pre-immunized with mycobacterial Hsp70 (Wendling et al. 2000; Prakken et al. 2001).

eHsp70 Role in Cancer Immunity

The eHsp70 or the one expressed on the tumor cell surface have been shown to elicit a strong anti-tumor immune response mediated by T cells, APCs and NK cells (Multhoff et al. 1997; Asea et al. 2000; Dressel et al. 2000; Clark and Ménoret 2001). It is by now demonstrated that many cancer cells contain a high level of chaperones, especially Hsp70 (Ciocca and Calderwood 2005). Indeed, contrary to normal cells, cancer cells overexpress Hsp70 in the cytoplasm, and also display the chaperone on their plasma membrane (mHsp70) (Multhoff et al. 1997; Gehrmann et al. 2003). This overexpressed Hsp70 has been found to stimulate an immune response specific to cancer cells (Suto and Srivastava 1995). Literature data highlighted that eHsp70 exerts a dual role in cancer, promoting survival of tumor cells, as well as contributing to anti-tumor immunity. In fact, in tumor environment eHsp70 can inhibit early-stage tumor growth by activating the immune response, whereas it stimulates late-stage tumor growth by inhibiting the immune system (Wu et al. 2012). Thus some eHsp70 features help cancer cells to escape from cytotoxic cells, while others facilitate an immune attack. According to the mechanism of Hsp70 release in the extracellular environment, i.e. as a free soluble molecule or in association with extracellular vesicles, it mediates distinct functions through the interaction with different target cells (e.g., immune cells, endothelial cells), and the activation of different intracellular pathways. Three different mechanisms could be involved in Hsp70 anti-tumor immunity:

  1. (a)

    Hsp70, due to its chaperone role can act as carrier for tumor derived peptides which are re-presented by APCs, such as macrophages and DCs, the most powerful APC. This complex is recognized by specific receptors, internalized within the APCs, and cross-presented to CD4+ T cells on MHC II (Tamura et al. 1997; Li et al. 2002), initiating an adaptive tumor-specific immune response (Arnold-Schild et al. 1999; Schild et al. 1999; Noessner et al. 2002; Enomoto et al. 2006). DCs are also able to efficiently transfer tumor and self-antigens on to MHC I (cross-presentation) (Heath and Carbone 2001). eHsp70 has also immunogenic functions related to its C-terminal substrate-binding domain, which is involved in interaction with APCs and consequently in antigen cross-presentation (MacAry et al. 2004).

  2. (b)

    The interaction between eHsp70 and its receptors on APCs stimulates these cells to secrete immunostimulatory cytokines, playing a proinflammatory role, responsible for enhanced programming of cytotoxic lymphocytes (Calderwood et al. 2012). For this “chaperokine” effect immunogenic peptides are not required (Asea et al. 2000, 2002; Asea 2006).

  3. (c)

    The last possible mechanism of action against tumor cells is NK dependent, and is correlated to plasmamembrane localization of Hsp70 (Multhoff et al. 1999). In fact, resting NK cell activation either by eHsp70 or by mHsp70 derived from tumor, even in the absence of immunogenic peptides (Multhoff et al. 1998), leads to an increased proliferative, migratory and cytotoxic activity against transmembrane Hsp70 positive tumor cells, whereas no killing activity was observed for Hsp70 negative cells (Multhoff et al. 1997; Gastpar et al. 2004, 2005).

Schematic model of different eHsp70 roles in cancer immunity is represented in Fig. 3.

Fig. 3
figure 3

Summary of Hsp70 activities in inducing anti-tumor immune responses. eHsp70 either alone or in combination with immunogenic peptides is able to induce an immune response. mHsp70 acts as a tumor-specific antigen, which is recognized by NK cells inducing their cytolytic, proliferative, and migratory capacity. eHsp70 also induces macrophages (M), NK and DCs to release proinflammatory cytokines. Moreover, eHsp70 could act as a carrier for tumor antigens and could support antigen uptake, processing, and presentation on MHC I to CD8+ cytotoxic T lymphocytes, or on MHC II s to CD4+ helper T cells

Full size image

Due to the observation that preactivated CD94+ NK cells (Gross et al. 2003a, b), through an Hsp70 peptide (TDK) plus low doses of interleukin 2, were able to induce regression of Hsp70 membrane positive tumors in immunodeficient mice (Moser et al. 2002; Stangl et al. 2006), Hsp70 was used in clinical trials of patients with colorectal and non small lung cell cancer who failed standard therapies such as chemotherapy, radiotherapy or laser induced thermotherapy (Krause et al. 2004; Specht et al. 2015). The recognition by NK cells of fragment crystallizable region of antibodies that have bound to tumor specific antigens activate antibody-dependent cellular cytotoxicity, establishing a link between B cell and NK cell mediated immune responses.

It is now well demonstrated that increased membrane bound Hsp70 expression and its extracellular release can stimulate an anti-tumor immune response, rendering cancer cells more susceptible to immune cells. For this reason two possible methods have been tested to increase eHsp70 amount: pharmacological agents (Arispe et al. 2002), and physical factors (e.g., hyperthermia, photodynamic therapy and ionizing radiations) (Stangl et al. 2011). Some in vivo data indicate the prospect of therapeutic vaccines based on Hsp70 administration (Ito et al. 2004; Geng et al. 2006; Kumar et al. 2009; Abkin et al. 2013). eHsp70 released by tumor cells can serve as autocrine and paracrine cytokine and through the interaction with receptors on tumor cells or APCs it stimulates the secretion of chemokines, and proinflammatory cytokines and nitric oxide respectively (Asea et al. 2000; Panjwani et al. 2002; Vega et al. 2008), providing an inflammatory microenvironment. This inflammation exerts anti-tumor activity at early tumor stage, whereas it support tumor growth and metastasis formation at a chronic stage. Equally, eHsp70 has an effect also on dendritic cells, which were activated, and chemoattracted from tumor tissues to secondary lymphoid organs, where they showed an improved capacity in the induction of tumor antigen-specific cytotoxic T lymphocytes (Kuppner et al. 2001; Wang et al. 2005; Chen et al. 2009).

In confirmation of the protective role of the protein in tumors, it has been demonstrated that its inhibition is lethal, and its silencing induces in vitro cell death as well as in tumor xenografts in mice (Wei et al. 1995; Kaur et al. 2000; Nylandsted et al. 2000). For this reason, in several cancer patients a more elevated eHsp70 level was found in blood compared to patients with non-cancerous pathologies (Gehrmann et al. 2014) (e.g., chronic myeloid leukaemia, acute leukaemia, colorectal cancer, glioblastoma, pancreatic cancer) (Yeh et al. 2009, 2010; Kocsis et al. 2010; Elstner et al. 2011; Dutta et al. 2012). Wu and coworkers demonstrated that eHsp70 encourages hepatocarcinoma growth by promoting tumor cell proliferation and apoptosis resistance (Wu et al. 2012). Both the events were induced in a dose dependent manner through the activation of the NF-kB pathway. The involvement of eHsp70/Hsp70-peptide complex in hepatocellular carcinoma cell proliferation was also confirmed by Zhe et al. (2016). Protective effect of eHsp70 on induced apoptosis by H2O2 on lymphoma cells was also observed. Indeed, pretreatment of U937 cells with eHsp70 inhibited hydrogen peroxide apoptosis induction (Franco et al. 2016).

eHsp70 and the Nervous System

To date it has been extensively demonstrated that Hsp70 exerts a cellular protective role under stress conditions. However, some cell types are not able to express this protein, among them there are certain types of neurons (e.g., hippocampal neurons) (Sprang and Brown 1987; Guzhova et al. 2001; Robinson et al. 2008). These mature neurons are particularly susceptible to toxic conditions due to their inability to increase Hsp70 level after stress (Morimoto et al. 1997), even if they contain Hsc70 (Brown 1991). The inability to activate the stress response is responsible for a wide range of neurodegenerative diseases (Planas et al. 1997), such as amyotrophic lateral sclerosis, Parkinson’s disease, Alzheimer’s disease, and polyglutamine diseases (Adachi et al. 2009). In this context, it has been demonstrated that Hsp70 released by surrounding glial cells is able to be endocyted by them (Guzhova et al. 2001; Novoselova et al. 2005; Robinson et al. 2005). The internalization of eHsp70 makes the neuronal cells resistant to cytotoxic effects and thermal stresses (Guzhova et al. 2001). A protective effect of exogenous Hsp70, on moto or sensory neurons and glial cells has been demonstrated by several authors (Edbladh et al. 1994; Guzhova et al. 2001; Tidwell et al. 2004; Robinson et al. 2005; Tytell 2005). The eHsp70 neuroprotective effect was also observed both under stress growth conditions (e.g., oxidative stress common in several neurodegenerative disease) and for those neurons undergoing natural cell death (Robinson et al. 2005).

Luo and colleagues demonstrated that eHsp70 in a dose dependent fashion is responsible for Schwann cell protection against apoptosis induced by H2O2. In particular, this protective effect was exerted through failure of caspase 3 and 9 activation and Bcl2 upregulation (Luo et al. 2012). Oxidative stress is the cause of several neurodegenerative diseases, as ROS induces membrane lipid peroxidation, nitration of proteins, and degradation of DNA, all inducing cell apoptosis (Fiskum 2004; Park et al. 2005). Because Schwann cells can regulate neuron survival and nerve regeneration, their preservation by eHsp70 is fundamental for peripheral nerve protection. Oxidative stress also impairs motoneurons during exercise. In fact, exercise induces ROS production, depending on the intensity, the type and the duration of the exercise. Although motoneurons during exercise are submitted to stress they still continue to function without any apparent damage, even if they are not able to increase iHsp70 synthesis (Robinson et al. 2005). This resistance can be explained by the release of eHsp70 by the liver, or by in vitro external addiction, and its internalization in motoneurons. eHsp70 could also act as extracellular signaling molecule, and this ability, together with its chaperone role after endocytosis, helps the cells against stress and damage (Robinson et al. 2005; Krause and Rodrigues-Krause 2011).

On the one end, many papers have demonstrated the involvement of HSP in central nervous system (CNS) diseases, such as those characterized by the presence and accumulation of misfolded proteins. Because HSP are found in protein aggregates, along with disease proteins, ubiquitin or other cellular molecules, we can hypothesize that both iHSP and eHSP have a role in refolding the misfolded proteins. In the other end, diseases in which CNS immune activation is a prominent feature, such as ischemia, immune-mediated disorders, infections, and trauma, may involve eHSP, because they are able to induce the innate immune response and to enhance the adaptive immunity. The most important HSP implicated in the immune response is eHsp70 (for a review see Turturici et al. 2011). eHsp70, through the interaction with specific cell surface receptors, are responsible for the expression of proinflammatory cytokines, chemokines and for DC activation, which are involved in antigen presentation to B and T lymphocytes.

High levels of anti-Hsp70 autoantibodies were found in the cerebrospinal fluid (CSF) of patients affected by multiple sclerosis (MS), than in that of healthy controls. In particular, the highest levels were detected in patients with progressive MS, in contrast to those with a stable disease (Chiba et al. 2006). This increase induces a higher production of IL-8 in THP-1 monocytes with consequent higher inflammatory levels (Yokota et al. 2010). eHsp70 was also found in and around MS lesions, and it may be involved in the induction or exacerbation of the immunologic response, because of its ability to act as a proinflammatory cytokine (for a review see Turturici et al. 2014). In addition, Lund and collaborators demonstrated that eHsp70 was associated with myelin basic protein (MBP)-derived peptides in normal appearing white matter of both MS ad normal human brain. They hypothesize that part of eHsp70-MBP peptide is secreted by stressed oligodendrocytes and it can act as an adjuvant molecule that stimulates an adaptive immune response against specific autoantigens. This event could be involved in the initiation of MS, and in the subsequent immune mediated destruction of myelin (Lund et al. 2006).

Galazka et al., obtained conflicting results which demonstrated that mouse immunization with eHsp70 fraction associated with peptide complexes isolated from animals with experimental autoimmune encephalomyelitis (EAE) reduced its subsequent induction (Galazka et al. 2006). On the contrary, the disease was not induced using eHsp70-complexes isolated from healthy donors. These results suggest substantial differences in the peptide that bind Hsp70 in normal versus pathological CNS. In contrast, in EAE, Hsp70 pharmacologically induced (e.g., with geldamycin) has a protective role because it suppresses the glial inflammatory response and ameliorates clinical signs (Dello Russo et al. 2006). All the evidences according to which Hsp70 is not only an intracellular chaperone but has also extracellular functions, such as neuroprotective effects in many brain diseases, could open a new scenario in CNS disease therapy.

Other Functions of eHsp70

In recent years new roles for eHsp70 have been found. For example, several studies have demonstrated that circulating eHsp70 is a marker for cardiovascular disease (Pockley et al. 2003). In addition to its marker role, eHsp70 actively participate in the inflammatory damage typical of the cardiovascular damage through its binding to TLR4 (Yang et al. 2005). Furthermore, in 2013 Gonzáles-Ramos et al. demonstrated that eHsp70 acts as a profibrotic regulator of the ECM protein synthesis (fibronectin and type I collagen) by the vascular smooth muscle cells. Also this stimulation depends on eHsp70 binding to the membrane receptor TLR4 and the subsequent activation of the ERK and JNK pathways. These two kinases determine the activation of the transcription factor AP-1, which increased the transcriptional capacity of the TGF-β1 promoter involved in increasing fibronectin and collagen I amount (Gonzáles-Ramos et al. 2013).

In vitro and in vivo assays showed that eHsp70 could be also an angiogenic regulator. Indeed, experiments on HUVEC cells demonstrated that eHsp70 activates the ERK signaling pathway involved in cell migration, tube formation and microvessel formation (Kim et al. 2016). Furthermore, eHsp70 promotes osteogenic differentiation of mesenchymal stem cells. Chen et al. showed that the addition of Hsp70 to the culture medium of the stem cells stimulate the expression of the alkaline phosphatase, an early marker of osteoblastic differentiation. eHsp70 also enhanced calcium deposits in both dose and time dependent manners, via the activation of the ERK signaling pathway (Chen et al. 2015).

Additionally, eHsp70 is involved in stem cell migration as demonstrated by Barreca et al. (Barreca et al. 2017). One of the fundamental problems in stem cells based therapies is their homing to the site of injury. It is well known that MMPs play a fundamental role in mesenchymal stem cell migration (Ries et al. 2007). Barreca and coworkers demonstrated that mouse mesoangioblast release Hsp70 in membrane derived vesicles as a transmembrane protein. mHsp70 acts via an autocrine signaling pathway by increasing mesoangioblast MMP2 release and accentuating their migration and invasion capability. A predominant role in the activation of these processes was carried out by the eHsp70 receptors TLR4 and CD91 and the consequent NF-kB pathway activation (Barreca et al. 2017).

eHsp70 Receptors

As described above Hsp70 may be released in the extracellular environment by both active and passive mechanisms, it can travel through the circulation to reach cellular target cells, such as immune effector cells. HSP have two essential properties that allow them to activate the immune response. First they are chaperones of antigenic peptides produced by specific cell types and presented via MHC I and MHC II. Second HSP are capable of binding to membrane receptors on APCs. Indeed, for eHsp70 activating innate and adaptive immune responses, it must bind to specific cell receptors. In the last years many cell-surface receptors involved in eHsp70 mediated signaling have been identified, including TLR2 and TLR4 with their cofactor CD14 (Asea et al. 2000, 2002), the co-stimulatory molecule CD40 (Becker et al. 2002; Wang et al. 2001), the chemokine receptor CCR5, the scavenger receptors (SR) LOX-1 (Delneste et al. 2002; Thériault et al. 2006), SCREC-1 (Thériault et al. 2006), FEEL-1 (Thériault et al. 2006), CD94 (C-type lectin) (Gross et al. 2003a, b), CD91 (Basu et al. 2001), CD36 (Delneste et al. 2002). All these immune receptors are differentially expressed on various cell types and have different roles in eHsp70 signaling (Fig. 4). Indeed, the receptors can be divided in three groups: pattern recognition receptors (e.g., TLR2, TLR4), uptake of antigens by APCs (e.g., LOX-1, CD91), mediating co-stimulatory signals (e.g., CCR5, CD40).

Fig. 4
figure 4

Hsp70 release from the donor cells and its binding to cell surface receptors on the target cells. Several cell types release the Hsp70 into the extracellular microenvironment through different active secretion mechanism or through a passive release after necrosis. eHsp70 can be recognized by several receptors on the target cells. Among them there are TLR2, TLR4, LOX-1, CD91, CD94, CCR5 and CD40 expressed on APCs, NK cells, and T cells Their integrated function is discussed in the text

Full size image

The TLR family consists of several members recognizing specific ligands, e.g., TLR2 (triacyled lipoproteins) and TLR4 (LPS). Activation of TLR pathway by ligand binding is responsible for activation of the transcription factors NF-kB and AP1 (Takeda and Akira 2004). NF-kB is central to the inflammatory response by innate cells (neutrophils, macrophages and DCs). For the first time Asea et al. in 2002 demonstrated that eHsp70 induced proinflammatory cytokine by binding monocyte TLR2 and TLR4 with their cofactor CD14 and the adaptor molecule MyD88, and by activating the NF-kB pathway involved in IL-1β, IL-6 and TNF-α (Asea et al. 2000, 2002). Furthermore, LRs located on macrophages have been demonstrated to be involved in antigen uptake and phagocytosis induction (Blander and Medzhitov 2004 Science). In particular, another member of TLR family which is able to interact with eHsp70 is the lipid raft associated TLR7, generally involved in viral RNA recognition. Wang et al. demonstrated that exogenous Hsp70 binds TLR7 on the lipid raft, which is essential for this binding. The interaction between eHsp70 and this member of the TLR family stimulates phagocytosis and antigen presentation through PI3K and the p38 MAPK pathway (Wang et al. 2006b).

TLR2 and TLR4 are not only expressed on immune cells, but also on endothelial cells, and airway epithelial cells (Greene et al. 2005). In heart eHsp70 directly decreased cardiomyocyte contractility and increased cell death through the activation of the inflammatory response by the signaling pathway TLR2-MyD88-NF-kB (Mathur et al. 2011). On the other hand, eHsp70 induced inflammation in the lung is TLR4-NF-kB dependent (Chase et al. 2007).

Another receptor involved in eHsp70 binding is CD40, a member of the TNF family. This binding was competed by the cochaperone Bip, whereas it is enhanced by the antigenic peptide. As observed for other receptor, binding of eHsp70-peptide complex to CD40 increases peptide uptake and activate the p38 intracellular pathway (Becker et al. 2002). Moreover, CD40+ cells were stimulated to produce chemokines (Lehner et al. 2003). As TLRs, CD40 is not only expressed on immune cells, but also on other cell types, such as HUVEC endothelial cells, where it functions by competing CD40 ligand and decreasing tubular formation process (Futagami et al. 2008). By binding assays Multhoff’s group demonstrates the involvement of the C-type lectin receptor CD94 in the interaction of NK cells with eHsp70/mHsp70 (Gross et al. 2003b). This interaction induce both their proliferation and cytolytic activity (Multhoff et al. 1999, 2001).

Another receptor implicated in eHsp70 binding is CD91, a transmembrane scavenger receptor ubiquitously expressed in multiple tissues where it plays a key role in intracellular signaling and endocytosis. Basu et al. identified CD91 as a receptor for several HSP including eHsp70 (Basu et al. 2001). Its role in binding and internalization of HSP by APCs, necessary for HSP chaperoned peptides cross-presentation, was reported in several papers (Delneste et al. 2002; Martin et al. 2003; Tobian et al. 2004). CD91-HSP was recently also be shown to be a signaling complex regulating multiple pathways in APCs (Pawaria and Binder 2011). The signaling pathway started up by CD91 upon HSP stimulation includes the activation of NF-kB and p38 MAPK (Pawaria and Binder 2011). As a consequence of this intracellular signaling several cytokines are released by HSP stimulated APCs (such as TNF-α, IL-1β, IL-6, IL-12 and GM-CSF) (Pawaria and Binder 2011). APCs also upregulate the expression of costimulatory molecules, including CD80, CD86, CD40 and MHC II (Basu et al. 2000). The costimulatory molecules and cytokines expressed depends on both the APC stimulated and the HSP involved in stimulation. According to these data CD91 has a role in cross-presentation and in costimulation given by the APCs to T cells in response to eHSP. The receptor CD91 expressed on monocytes, after eHsp70-peptide binding stimulates antigen specific CD4+ T cell proliferation (Fischer et al. 2010). Although CD91 is highly expressed on macrophages, it is very low on immature DCs. Delneste and coworkers demonstrated that on these cells the scavenger receptor LOX-1 is the main eHsp70 binding receptor and is involved in antigen cross-presentation to MHC I in vitro and in vivo (Delneste et al. 2002).

One of the last eHsp70 receptor identified is the receptor for advanced glycation endproducts (RAGE), a receptor that interacts with several inflammation and stress mediators to activate the inflammatory pathway (Xie et al. 2013). The activation of RAGE by eHsp70 evokes strong proinflammatory changes in human lung carcinoma cell line. In fact, interaction between eHsp70 and RAGE trigger the typical cellular effects of this receptor activation (i.e., ERK phosphorylation and activation of NF-kB, increased expression of RAGE itself and secretion of proinflammatory cytokines) (Somensi et al. 2017).

Conclusions

Over the years several authors have detected Hsp70 in the extracellular environment. Early studies hypothesized that this protein was released consequently to cell death, as it lacks any exocytosis signals. Nowadays, it is well demonstrated that Hsp70 is released through active mechanism, involving non classical secretory pathways (e.g., extracellular vesicles, lysosome-endosome pathway and so on). Several roles have been demonstrated for eHsp70, especially at cell signaling and cell communication level, depending on its state (i.e. membrane-bound or membrane-free). An intriguing aspect of eHsp70 biology is its ability to induce antagonist responses, such as inflammatory and anti-inflammatory events. A particularly abundant field on eHsp70 functions regards its ability to modulate the immune system. Indeed, it has been displayed that eHsp70 possesses powerful immunological properties, acting both on innate and adaptive immune system. eHsp70 activates innate immune system cells by inducing cytokine release, or stimulate the adaptive immune response via antigen cross presentation. In addition to its immunological role, eHsp70 has protective functions in nervous cells, such as motoneurons, sensory neurons, glial cells. Finally, eHsp70 has been demonstrated to be involved in cell migration and angiogenesis, and it is a marker for several diseases. A better comprehension of eHsp70 multiple roles will be useful for therapeutic applications in inflammatory diseases, cancer and autoimmunity.