**3.1 Receptor tyrosine kinase inhibition**

Given the critical role that RTKs play in signal transduction profoundly affecting tumor progression, several molecules to block the aberrant activation of these pathways have been developed (Zwick et al., 2002). Due to the high level of complexity of these pathways, investigators have attempted to interfere at different levels of the signaling cascade, from the extracellular ligands, to the membrane receptors, to the intracellular non-receptor tyrosine kinases (Fig.2). Several modalities are currently employed to inhibit kinase activity such as small inhibitors, molecular antibodies, but also peptide mimetics and antisense oligonucleotides (Dancey & Sausville, 2003; Imai & Takaoka, 2006).

*Small molecule tyrosine kinase inhibitors* - Tyrosin kinases are present in an active or inactive form. Active kinases are characterized by high structural homology whereas the structures of the inactive forms are dissimilar. This characteristic suggests that molecules able to recognize and bind to the inactive form of the enzyme, could present higher efficacy because of lower cross-reactivity and increased specificity. Despite the fact that more specific inhibitors allow to precisely control cellular activity, several studies have demonstrated the importance of inhibiting multiple critical nodes in order to increase treatment response. Most cancers in fact are characterized by multiple aberrant signaling pathways and thus may require inhibition of a number of signaling endpoints for an optimal response. In addition, the ability of multi-targeted drugs to inhibit different kinases can represent an advantage in case of the development of drug resistance. In some occasions cells treated with a selective kinase inhibitor activate alternative pathways to maintain their activity. Hence, a chemical compound able to simultaneously inhibit several kinases bares the potential to be less prone to drug resistance. On the other hand, the use of multitargeted inhibitors is usually accompanied by worse side effects and limits on the maximum dose

Therapeutical Cues from the Tumor Microenvironment 243

Fig. 2. Schematic representation of the therapeutic strategies targeting receptor tyrosine kinases; PDGFR: Platelet-derived growth factor Receptor; EGFR: Epidermal Growth Factor Receptor; VEGFR: Vascular Endothelial Growth Factor Receptor; FGFR: Fibroblast Growth

used in clinical trials for the treatment of metastatic GISTs and glioblastoma.

Sorafenib (Nexavar®) is a multi-kinase small molecule inhibitor able to block CRAF, BRAF, VEGFR2, PDGFRβ, KIT and FLT3; is used in clinical trials as monotherapy or in combination with standard therapy (Wilhelm et al., 2006; Liu et al., 2006). Brivanib, a dual tyrosine kinase inhibitor of VEGFR and FGFR signaling, was active that in phase I clinical trials against metastatic solid tumors refractory to standard therapy (Park et al., 2011). Imatinib mesilate (Gleevec®) binds to the inactive form of the BCR-ABL fusion protein inhibiting its activity and is used for the treatment of chronic myeloid leukemia. It also functions as a multi-target inhibitor of KIT and PDGFRβ. Due to this specificity it is also

Other anti-VEGF therapeutic approaches are under pre-clinical and clinical trials (Roskoski,

*Anti-EGFR therapy* - EGFR receptor was one of the first tyrosine kinase to be linked to tumor development and the dysregulation of one or more components of its pathway are a common trait of several tumor types. For this reason many molecules have been developed to try to inhibit the EGFR aberrant activity at different levels. Many of the mAbs and small molecules directly target the EGF receptor and some are currently used in clinical practice. Gefitinib (Iressa®) was the first EGFR small molecule inhibitor to enter clinical trials (Schiller, 2003). In combination with standard therapy its activity has been evaluated in several tumor types such as non-small cell lung cancer and squamous cell carcinoma of the head and neck

although no improvement in overall survival and response rate has been reported yet.

Factor Receptor; KIT: c-Kit Receptor aka CD117.

Jr., 2007; Kowanetz & Ferrara, 2006).

that can be used for treatment, thus limiting the real benefits. Most of the small molecule kinase inhibitors currently in preclinical development and in clinical use are ATPcompetitive inhibitors. These molecules are characterized by a high affinity for the kinase ATP-binding sites and their interaction with these sites prevents ATP binding and therefore kinase activation. One of the limitations of this kind of inhibitors is represented by the susceptibility to mutations that enhance target kinase activity, which is common in several types of cancer. In addition to ATP-competitive inhibitors, several non-competitive inhibitors have also been developed. These molecules bind to an allosteric site distant from the ATP-binding site and induce a conformational change of the receptor thus preventing the ATP binding. These kinds of inhibitors are active independently on the ATP binding site sequence, thus maintaining their efficacy also in the presence of mutant proteins.

*Monoclonal antibodies* - An alternative approach to impair the RTKs' activation is represented by the use of blocking monoclonal antibodies (mAbs) (Hinoda et al., 2004). These antibodies act directly preventing the ligand-receptor interaction and therefore inhibiting ligandmediated intracellular signaling and increasing RTK down-regulation and internalization. In addition, mAbs act also indirectly by stimulating the immune response. Following their binding to the cell surface receptors the subsequent activation of the immune system causes antibody-dependent cellular cytotoxicity through the activation of macrophages or natural killer cells. Finally, recent studies demonstrated that some monoclonal antibodies are able to influence tumor growth by modulating the receptor signaling activity and therefore altering the intracellular signaling inducing apoptosis and/or growth inhibition.

Compared to the small molecule inhibitors, mAbs are more expensive to produce and, due their large molecular weight, the tissue penetration, tumor retention and blood clearance is less efficient especially in some tissues such as the brain. Moreover, because of their inability to penetrate through the cellular membrane, they can only be used to target cell surface molecules and not intracellular proteins. Despite these limitations, in clinical trials the mAbs have encountered higher approval success rates compared to new drugs, including smallmolecule agents (Reichert et al., 2005). It must be pointed out that the activation of the immune system is likely crucial to efficiently eliminate tumor cells and for this reason, the use of small molecule inhibitors in combination with mAbs should maximize the therapeutic effects.

*RTKs as Antiangiogenic targets* - The prevention of angiogenesis is a potentially successful mean to impair tumor growth and progression. In contrast to standard chemotherapy, inhibition of angiogenesis generally does not lead to tumor regression but induces longterm stable disease creating a favorable window of opportunity for radiotherapy or chemotherapy to exert stronger tumor-killing effects. Growth factors such as VEGF-A, FGF2, IL8 and PDGF play an important role in promoting blood vessels formation and several antiangiogenic molecules have been developed to target either growth factors or their receptors.

The use of Bevacizumab (Avastin®), a humanized mAb able to specifically bind to VEGF-A and blocking its interaction with VEGFR2, in combination with 5-fluorouracil-based chemotherapy was approved for the first-line treatment of metastatic colorectal cancer (CRC) by FDA in February 2004 (Ferrara et al., 2005). Sunitinib (Sutent®) is an orally active small molecule inhibitor of VEGFR2, PDGFRβ, KIT and FLT3 (Mendel et al., 2003; Abrams et al., 2003) and is used in clinical trials for patients with advanced solid malignancies (Motzer et al., 2006; Motzer & Bukowski, 2006).

that can be used for treatment, thus limiting the real benefits. Most of the small molecule kinase inhibitors currently in preclinical development and in clinical use are ATPcompetitive inhibitors. These molecules are characterized by a high affinity for the kinase ATP-binding sites and their interaction with these sites prevents ATP binding and therefore kinase activation. One of the limitations of this kind of inhibitors is represented by the susceptibility to mutations that enhance target kinase activity, which is common in several types of cancer. In addition to ATP-competitive inhibitors, several non-competitive inhibitors have also been developed. These molecules bind to an allosteric site distant from the ATP-binding site and induce a conformational change of the receptor thus preventing the ATP binding. These kinds of inhibitors are active independently on the ATP binding site

*Monoclonal antibodies* - An alternative approach to impair the RTKs' activation is represented by the use of blocking monoclonal antibodies (mAbs) (Hinoda et al., 2004). These antibodies act directly preventing the ligand-receptor interaction and therefore inhibiting ligandmediated intracellular signaling and increasing RTK down-regulation and internalization. In addition, mAbs act also indirectly by stimulating the immune response. Following their binding to the cell surface receptors the subsequent activation of the immune system causes antibody-dependent cellular cytotoxicity through the activation of macrophages or natural killer cells. Finally, recent studies demonstrated that some monoclonal antibodies are able to influence tumor growth by modulating the receptor signaling activity and therefore altering

Compared to the small molecule inhibitors, mAbs are more expensive to produce and, due their large molecular weight, the tissue penetration, tumor retention and blood clearance is less efficient especially in some tissues such as the brain. Moreover, because of their inability to penetrate through the cellular membrane, they can only be used to target cell surface molecules and not intracellular proteins. Despite these limitations, in clinical trials the mAbs have encountered higher approval success rates compared to new drugs, including smallmolecule agents (Reichert et al., 2005). It must be pointed out that the activation of the immune system is likely crucial to efficiently eliminate tumor cells and for this reason, the use of small molecule inhibitors in combination with mAbs should maximize the

*RTKs as Antiangiogenic targets* - The prevention of angiogenesis is a potentially successful mean to impair tumor growth and progression. In contrast to standard chemotherapy, inhibition of angiogenesis generally does not lead to tumor regression but induces longterm stable disease creating a favorable window of opportunity for radiotherapy or chemotherapy to exert stronger tumor-killing effects. Growth factors such as VEGF-A, FGF2, IL8 and PDGF play an important role in promoting blood vessels formation and several antiangiogenic molecules have been developed to target either growth factors or their

The use of Bevacizumab (Avastin®), a humanized mAb able to specifically bind to VEGF-A and blocking its interaction with VEGFR2, in combination with 5-fluorouracil-based chemotherapy was approved for the first-line treatment of metastatic colorectal cancer (CRC) by FDA in February 2004 (Ferrara et al., 2005). Sunitinib (Sutent®) is an orally active small molecule inhibitor of VEGFR2, PDGFRβ, KIT and FLT3 (Mendel et al., 2003; Abrams et al., 2003) and is used in clinical trials for patients with advanced solid malignancies

sequence, thus maintaining their efficacy also in the presence of mutant proteins.

the intracellular signaling inducing apoptosis and/or growth inhibition.

therapeutic effects.

receptors.

(Motzer et al., 2006; Motzer & Bukowski, 2006).

Fig. 2. Schematic representation of the therapeutic strategies targeting receptor tyrosine kinases; PDGFR: Platelet-derived growth factor Receptor; EGFR: Epidermal Growth Factor Receptor; VEGFR: Vascular Endothelial Growth Factor Receptor; FGFR: Fibroblast Growth Factor Receptor; KIT: c-Kit Receptor aka CD117.

Sorafenib (Nexavar®) is a multi-kinase small molecule inhibitor able to block CRAF, BRAF, VEGFR2, PDGFRβ, KIT and FLT3; is used in clinical trials as monotherapy or in combination with standard therapy (Wilhelm et al., 2006; Liu et al., 2006). Brivanib, a dual tyrosine kinase inhibitor of VEGFR and FGFR signaling, was active that in phase I clinical trials against metastatic solid tumors refractory to standard therapy (Park et al., 2011).

Imatinib mesilate (Gleevec®) binds to the inactive form of the BCR-ABL fusion protein inhibiting its activity and is used for the treatment of chronic myeloid leukemia. It also functions as a multi-target inhibitor of KIT and PDGFRβ. Due to this specificity it is also used in clinical trials for the treatment of metastatic GISTs and glioblastoma.

Other anti-VEGF therapeutic approaches are under pre-clinical and clinical trials (Roskoski, Jr., 2007; Kowanetz & Ferrara, 2006).

*Anti-EGFR therapy* - EGFR receptor was one of the first tyrosine kinase to be linked to tumor development and the dysregulation of one or more components of its pathway are a common trait of several tumor types. For this reason many molecules have been developed to try to inhibit the EGFR aberrant activity at different levels. Many of the mAbs and small molecules directly target the EGF receptor and some are currently used in clinical practice.

Gefitinib (Iressa®) was the first EGFR small molecule inhibitor to enter clinical trials (Schiller, 2003). In combination with standard therapy its activity has been evaluated in several tumor types such as non-small cell lung cancer and squamous cell carcinoma of the head and neck although no improvement in overall survival and response rate has been reported yet.

Therapeutical Cues from the Tumor Microenvironment 245

administered SPARC can inhibit the growth of neuroblastoma xenografts in mice (Chlenski et al., 2002). Moreover, exogenous SPARC can reverse the acquired resistance to chemotherapeutic agents (Tai et al., 2005). As previously explained, integrins are involved in anchorage independent growth, anoikis resistance and metastasis formation, which are all characteristics of malignant tumor cells. For this reason small molecule antagonists able to block the integrin binding to the ECM molecules are currently under evaluation. As an example RGD-mimetic peptides have been extensively used in pre-clinical studies (Eble & Haier, 2006). Cilengitide, a cyclic peptidic antagonist of αvβ3 and αvβ5, to block also RGDindependent effects and to achieve a more efficacious anti-tumoral effect, for example (Paolillo et al., 2009). Furthermore, thanks to the capacity of the HS side chains of HSPGs to bind a multitude of growth factors, chemochines, cytochines and enzymes, neutralizing antibodies or antisense oligonucleotides have been demonstrated to enhance apoptosis in a number of tumor cells and may represent promising therapeutic approaches (Fukumoto et al., 2000). The abnormal PG expression in cancer and stromal cells may serve as a biomarker for tumor progression and patient survival. In this context, the better understanding of role of PGs in cancer progression may represent a novel approach in directly targeting the tumor

Fig. 3. Schematic representation of the therapeutic strategies targeting apoptosis. PGs: Proteoglycans; α4β1 and αvβ5: integrin receptors; CCN1, SPARC and EMILIN2: extracellular matrix molecules; TNF α: Tumor Necrosis Factor α; TNF-R1: Tumor Necrosis Factor Receptor 1; Fas and TRAIL-R (TNF-Related Apoptosis Inducing Ligand Receptor) death receptors; FADD (Fas-Associated protein with Death Domain): adapter molecule of the extrinsic apoptotic pathway; ROS: Reactive Oxygen Species; FLIP: FLICE-like inhibitory protein.

Finally, also the ECM molecule EMILIN2 may represent a good target in this context and increasing its levels at the tumor site may suggest a therapeutic use since EMILIN2 exerts pro-apoptotic effects for tumor cells enhancing the extrinsic apoptotic pathway but leaves normal cells unharmed. This hypothesis is well supported by pre-clinical studies (Mongiat

microenvironment (Theocharis et al., 2010).

et al., 2007; Mongiat et al., 2010).

Cetuximab (Erbitux®), a human-mouse chimeric antibody able to bind and inhibit EGF receptor is used for the treatment of metastatic colorectal cancer (Goldberg, 2005). A higher response rate to the therapy was observed in combination with irinotecan, oxaliplatin or radiotherapy in different tumor types, highlighting the importance of acting on different cellular activities to inhibit tumor growth.

Panitumumab (Vectibix®) is another anti-EGFR antibody (Giusti et al., 2007) that has shown efficacy in chemotherapy refractory colorectal cancers and is being used in clinical trials for the treatment of esophageal, head and neck and lung tumors. Trastuzumab (Herceptin®) is a mAb approved in 1998 for the treatment of HER2-positive breast cancer patients in combination with paclitaxel and doxorubicin. This molecule is able to bind to the extracellular domain of ERBB2 preventing receptor dimerization and the subsequent signaling cascade.

Most of the failures with the use of some of these drugs may be due to a wrong selection of the patients. Indeed the utility of a personalized therapy derived from a patient selection based on tumor molecular profile has been demonstrated for several of these targeted therapeutics, including cetuximab and trastuzumab. The presence of a specific genomic and/or proteomic profile in fact correlates with the response to the therapy. Finally, other molecules that target multiple members of the EGF receptor family are currently under investigation (Reid et al., 2007) and some of them such as XL647 and AEE788 have shown to be active *in vitro* and are in different phases of clinical trials.

Considering the complexity of the mechanisms regulating tumor progression, the most effective approach in impairing tumor development is to simultaneously target different pathways. The increasing knowledge about the crosstalk between cancer cells and tumor microenvironment will allow the development of new drugs in order to block different players of the signaling cascades. The goal is that with the use of therapies involving complementary and synergistic combinations will increase drug efficacy and reduce toxicity as well as the risk of development of resistance to the treatment.

#### **3.2 Apoptotic molecules**

The ability of tumor cells to evade apoptosis is one of the hallmarks of cancer progression and can play a significant role in their resistance to conventional therapeutic regimens. To overcome resistance to apoptosis through the development of therapeutic drugs tackling the single components of the pathway is one of the efforts that is being attempted to potentiate the anti-tumoral therapy (Reed & Pellecchia, 2005). Researchers have also focused the attention on different agents able to indirectly modulate apoptosis (Fig. 3). Histone deacetylase (HDAC) inhibitors, for example, can reduce transcription of anti-apoptotic molecules of the Bcl2 family and are currently under investigation in clinical trials. Smallmolecule drugs and mAbs directed against various RTKs analyzed in the previous section also affect this pathway and induce tumor cells to become more sensitive to apoptosis. Personalized treatments can be achieved given that the tumor-specific genetic lesions that affect sensitivity or resistance to apoptosis are well known (Zivny et al., 2010). In addition, different therapeutic strategies have been hypothesized or are currently under evaluation taking into account all the microenvironment molecules found to affect apoptosis. For instance, the *in vitro* anti-proliferative effect of exogenous SPARC could be exploited therapeutically by administering recombinant SPARC or one of its derivate peptides to patients. Pre-clinical studies are encouraging and demonstrate that subcutaneously

Cetuximab (Erbitux®), a human-mouse chimeric antibody able to bind and inhibit EGF receptor is used for the treatment of metastatic colorectal cancer (Goldberg, 2005). A higher response rate to the therapy was observed in combination with irinotecan, oxaliplatin or radiotherapy in different tumor types, highlighting the importance of acting on different

Panitumumab (Vectibix®) is another anti-EGFR antibody (Giusti et al., 2007) that has shown efficacy in chemotherapy refractory colorectal cancers and is being used in clinical trials for the treatment of esophageal, head and neck and lung tumors. Trastuzumab (Herceptin®) is a mAb approved in 1998 for the treatment of HER2-positive breast cancer patients in combination with paclitaxel and doxorubicin. This molecule is able to bind to the extracellular domain of ERBB2 preventing receptor dimerization and the subsequent

Most of the failures with the use of some of these drugs may be due to a wrong selection of the patients. Indeed the utility of a personalized therapy derived from a patient selection based on tumor molecular profile has been demonstrated for several of these targeted therapeutics, including cetuximab and trastuzumab. The presence of a specific genomic and/or proteomic profile in fact correlates with the response to the therapy. Finally, other molecules that target multiple members of the EGF receptor family are currently under investigation (Reid et al., 2007) and some of them such as XL647 and AEE788 have shown to

Considering the complexity of the mechanisms regulating tumor progression, the most effective approach in impairing tumor development is to simultaneously target different pathways. The increasing knowledge about the crosstalk between cancer cells and tumor microenvironment will allow the development of new drugs in order to block different players of the signaling cascades. The goal is that with the use of therapies involving complementary and synergistic combinations will increase drug efficacy and reduce toxicity

The ability of tumor cells to evade apoptosis is one of the hallmarks of cancer progression and can play a significant role in their resistance to conventional therapeutic regimens. To overcome resistance to apoptosis through the development of therapeutic drugs tackling the single components of the pathway is one of the efforts that is being attempted to potentiate the anti-tumoral therapy (Reed & Pellecchia, 2005). Researchers have also focused the attention on different agents able to indirectly modulate apoptosis (Fig. 3). Histone deacetylase (HDAC) inhibitors, for example, can reduce transcription of anti-apoptotic molecules of the Bcl2 family and are currently under investigation in clinical trials. Smallmolecule drugs and mAbs directed against various RTKs analyzed in the previous section also affect this pathway and induce tumor cells to become more sensitive to apoptosis. Personalized treatments can be achieved given that the tumor-specific genetic lesions that affect sensitivity or resistance to apoptosis are well known (Zivny et al., 2010). In addition, different therapeutic strategies have been hypothesized or are currently under evaluation taking into account all the microenvironment molecules found to affect apoptosis. For instance, the *in vitro* anti-proliferative effect of exogenous SPARC could be exploited therapeutically by administering recombinant SPARC or one of its derivate peptides to patients. Pre-clinical studies are encouraging and demonstrate that subcutaneously

cellular activities to inhibit tumor growth.

be active *in vitro* and are in different phases of clinical trials.

as well as the risk of development of resistance to the treatment.

signaling cascade.

**3.2 Apoptotic molecules** 

administered SPARC can inhibit the growth of neuroblastoma xenografts in mice (Chlenski et al., 2002). Moreover, exogenous SPARC can reverse the acquired resistance to chemotherapeutic agents (Tai et al., 2005). As previously explained, integrins are involved in anchorage independent growth, anoikis resistance and metastasis formation, which are all characteristics of malignant tumor cells. For this reason small molecule antagonists able to block the integrin binding to the ECM molecules are currently under evaluation. As an example RGD-mimetic peptides have been extensively used in pre-clinical studies (Eble & Haier, 2006). Cilengitide, a cyclic peptidic antagonist of αvβ3 and αvβ5, to block also RGDindependent effects and to achieve a more efficacious anti-tumoral effect, for example (Paolillo et al., 2009). Furthermore, thanks to the capacity of the HS side chains of HSPGs to bind a multitude of growth factors, chemochines, cytochines and enzymes, neutralizing antibodies or antisense oligonucleotides have been demonstrated to enhance apoptosis in a number of tumor cells and may represent promising therapeutic approaches (Fukumoto et al., 2000). The abnormal PG expression in cancer and stromal cells may serve as a biomarker for tumor progression and patient survival. In this context, the better understanding of role of PGs in cancer progression may represent a novel approach in directly targeting the tumor microenvironment (Theocharis et al., 2010).

Fig. 3. Schematic representation of the therapeutic strategies targeting apoptosis. PGs: Proteoglycans; α4β1 and αvβ5: integrin receptors; CCN1, SPARC and EMILIN2: extracellular matrix molecules; TNF α: Tumor Necrosis Factor α; TNF-R1: Tumor Necrosis Factor Receptor 1; Fas and TRAIL-R (TNF-Related Apoptosis Inducing Ligand Receptor) death receptors; FADD (Fas-Associated protein with Death Domain): adapter molecule of the extrinsic apoptotic pathway; ROS: Reactive Oxygen Species; FLIP: FLICE-like inhibitory protein.

Finally, also the ECM molecule EMILIN2 may represent a good target in this context and increasing its levels at the tumor site may suggest a therapeutic use since EMILIN2 exerts pro-apoptotic effects for tumor cells enhancing the extrinsic apoptotic pathway but leaves normal cells unharmed. This hypothesis is well supported by pre-clinical studies (Mongiat et al., 2007; Mongiat et al., 2010).

Therapeutical Cues from the Tumor Microenvironment 247

under evaluation. Through a specific CD8+ T cell-mediated killing of CAFs, this vaccine successfully affected both primary tumor growth and metastases of multidrug-resistant murine colon and breast carcinomas. Accordingly, FAP vaccinated mice display a reduction of intra-tumoral collagen I and thus an increased doxorubicin uptake leading in many cases

*Targeting TAMs* - Possible therapeutic strategies targeting TAMs are aimed at reducing the number and the function of these cells and/or increasing their antitumor activity. Different groups reported that TAM are committed to produce high levels of the inhibitory cytokine IL10, which mediates defective IL12 production and NF-kB activation in TAMs, an event which is very important in determining TAMs activity and that is frequently correlated with poor prognosis. It has been demonstrated that the employment of anti-IL10 receptor antibodies and Toll-like receptor 9 ligand CpG can restore NF-kB activity and revert the macrophages phenotype from M2 to M1 status, thus activating an innate response against

Another way to reduce the number of TAMs is to target molecules that are involved in the recruitment of macrophages. In this perspective, the employment of antibodies against chemokines or chemokines receptors such as CFS1/CSF-1R could have important

Given the important role of TAMs in supporting angiogenesis, antiangiogenic drugs also represent a therapeutic tool against TAMs. Tumor infiltrating macrophages are a significant source of VEGF that act as a loop by recruiting other macrophages. Thus targeting this growth factor could reduce the number of TAMs in the tumor stroma. Linomide is an interesting antiangiogenic drug that is effective in the reduction of tumor growth in a murine prostate cancer model. Its activity includes the inhibition of macrophages recruitment and the increase of antiangiogenic molecules in TAMs such as IL12, IL18, CXCL10 and CXCL9. Hypoxia induces HIF, and hence VEGF expression in TAMs and HIF could represent another interesting target. Hypoxia signaling also leads to an increase in CXCR4 expression both on TAMs and ECs and CXCR4 antibodies such as AMD3100 inhibit

Anti-tumor agents with selective cytotoxic activity on monocyte-macrophages would be ideal therapeutic tools for their combined action on tumor cells and TAMs. Yondelis (Trabectedin), a natural product derived from the marine organism *Ecteinascidia turbinate* displays a potent anti-tumor activity and is specifically cytotoxic to macrophages and TAMs, while sparing the lymphocyte sub-set (Sessa et al., 2005). In addition, Yondelis inhibits the production of CCL2 and IL6 by TAMs which in turn contribute to growth

A legumain-based vaccine represents another strategy aimed at inducing the host to contrast the infiltration of TAMs in the tumor. In fact, legumain, a member of the endopeptidase family, is highly over-expressed by TAMs in murine and human breast tumor tissues, and hence provides an ideal strategy for breast tumor targeting (Lewen et al., 2008; Xiang et al., 2008). The vaccination of mice against legumain provides a robust CD8+ T cell response against TAMs, significantly reducing tumor growth. Moreover, a dramatic reduction in TGFβ, TNFα, MMP9 and VEGF expression is induced, with a consequent

*Targeting ECs and pericytes -* Besides targeting pro-angiogenetic growth factors and cytokines as mentioned above it is also possible to specifically target ECs and pericytes. The selective disruption of tumor vessels through the employment of the so called vascular disrupting

tumor angiogenesis *in vivo* (Mantovani et al., 2004; Mantovani et al., 2006).

suppression of inflammation-associated human tumors (Allavena et al., 2005).

reduction of both angiogenesis and metastatic rate (Luo et al., 2006).

to tumor rejection and increased lifespan (Loeffler et al., 2006).

tumor (Guiducci et al., 2005).

therapeutic effects.

#### **3.3 Cells of the microenvironment**

*Targeting CAFs* - CAFs are the primary cells within the tumor stroma that determine the tumor dynamics. Despite their frequent genomic instability, they are more sensitive to chemotherapeutic drugs than tumor cells and thus are considered potential innovative therapeutic targets (Fig. 4). Many epithelial tumor cells secrete PDGF but lack PDGFR that is expressed by CAFs and its activation is frequently associated with metastases in colon carcinomas (Pietras et al., 2001; Bouzin & Feron, 2007; Haubeiss et al., 2010). In a recent study, the activity of four different FDA approved PDGF inhibitors such as Dasatinib, Imatinib, Nilotinib and Sorafenib has been compared and all were found to reduce CAFs' viability. Dasatinib resulted the most efficacious and inhibited not only CAFs' viability, but also induced a quiescent state and partially reverting their phenotype to normal. In addition, the incubation of cancer cells with conditioned media from CAFs pre-incubated with Dasatinib resulted in a significant reduction of tumor cell proliferation (Haubeiss et al., 2010). PDGF inhibitors also reduced the contractility of myofibroblasts in the tumor, thereby reducing interstitial fluid pressure and increasing chemotherapeutic drugs uptake. Encouraging results in inhibiting PDGFβ-R were also obtained using CDP860, a Fab' fragment-polyethylene glycol conjugate (Bouzin & Feron, 2007).

Fig. 4. Schematic representation of the therapeutic strategies targeting tumor associated cells FAP: Fibroblast Activating Protein; CAF: Cancer Associated Fibroblasts; PDGFR: Platelet-Derived Growth Factor Receptor; α-CSF1 and α-CSF1R: antibodies against the Colony Stimulating Factor 1 and its receptor; α-CXCR4: antibodies against the C-X-C chemockine receptor type 4; α-IL10R: antibodies against the Interleukin 10 Receptor ; VEGFR2: Vascular Endothelial Growth Factor 2; VDAs: Vascular Disrupting Agents.

Another potential therapeutic target is the Fibroblast Activating Protein (FAP), a serine protease actively implicated in ECM remodeling and over-expressed in different tumors. Several studies underline its involvement in immunosuppression, promotion of tumor growth and increased metastatic potential (Bouzin & Feron, 2007; Loeffler et al., 2006; Kraman et al., 2010). Apart from a FAP-antibody an oral DNA vaccine against FAP is also

*Targeting CAFs* - CAFs are the primary cells within the tumor stroma that determine the tumor dynamics. Despite their frequent genomic instability, they are more sensitive to chemotherapeutic drugs than tumor cells and thus are considered potential innovative therapeutic targets (Fig. 4). Many epithelial tumor cells secrete PDGF but lack PDGFR that is expressed by CAFs and its activation is frequently associated with metastases in colon carcinomas (Pietras et al., 2001; Bouzin & Feron, 2007; Haubeiss et al., 2010). In a recent study, the activity of four different FDA approved PDGF inhibitors such as Dasatinib, Imatinib, Nilotinib and Sorafenib has been compared and all were found to reduce CAFs' viability. Dasatinib resulted the most efficacious and inhibited not only CAFs' viability, but also induced a quiescent state and partially reverting their phenotype to normal. In addition, the incubation of cancer cells with conditioned media from CAFs pre-incubated with Dasatinib resulted in a significant reduction of tumor cell proliferation (Haubeiss et al., 2010). PDGF inhibitors also reduced the contractility of myofibroblasts in the tumor, thereby reducing interstitial fluid pressure and increasing chemotherapeutic drugs uptake. Encouraging results in inhibiting PDGFβ-R were also obtained using CDP860, a Fab'

Fig. 4. Schematic representation of the therapeutic strategies targeting tumor associated cells FAP: Fibroblast Activating Protein; CAF: Cancer Associated Fibroblasts; PDGFR: Platelet-Derived Growth Factor Receptor; α-CSF1 and α-CSF1R: antibodies against the Colony Stimulating Factor 1 and its receptor; α-CXCR4: antibodies against the C-X-C chemockine receptor type 4; α-IL10R: antibodies against the Interleukin 10 Receptor ; VEGFR2: Vascular

Another potential therapeutic target is the Fibroblast Activating Protein (FAP), a serine protease actively implicated in ECM remodeling and over-expressed in different tumors. Several studies underline its involvement in immunosuppression, promotion of tumor growth and increased metastatic potential (Bouzin & Feron, 2007; Loeffler et al., 2006; Kraman et al., 2010). Apart from a FAP-antibody an oral DNA vaccine against FAP is also

**3.3 Cells of the microenvironment** 

fragment-polyethylene glycol conjugate (Bouzin & Feron, 2007).

Endothelial Growth Factor 2; VDAs: Vascular Disrupting Agents.

under evaluation. Through a specific CD8+ T cell-mediated killing of CAFs, this vaccine successfully affected both primary tumor growth and metastases of multidrug-resistant murine colon and breast carcinomas. Accordingly, FAP vaccinated mice display a reduction of intra-tumoral collagen I and thus an increased doxorubicin uptake leading in many cases to tumor rejection and increased lifespan (Loeffler et al., 2006).

*Targeting TAMs* - Possible therapeutic strategies targeting TAMs are aimed at reducing the number and the function of these cells and/or increasing their antitumor activity. Different groups reported that TAM are committed to produce high levels of the inhibitory cytokine IL10, which mediates defective IL12 production and NF-kB activation in TAMs, an event which is very important in determining TAMs activity and that is frequently correlated with poor prognosis. It has been demonstrated that the employment of anti-IL10 receptor antibodies and Toll-like receptor 9 ligand CpG can restore NF-kB activity and revert the macrophages phenotype from M2 to M1 status, thus activating an innate response against tumor (Guiducci et al., 2005).

Another way to reduce the number of TAMs is to target molecules that are involved in the recruitment of macrophages. In this perspective, the employment of antibodies against chemokines or chemokines receptors such as CFS1/CSF-1R could have important therapeutic effects.

Given the important role of TAMs in supporting angiogenesis, antiangiogenic drugs also represent a therapeutic tool against TAMs. Tumor infiltrating macrophages are a significant source of VEGF that act as a loop by recruiting other macrophages. Thus targeting this growth factor could reduce the number of TAMs in the tumor stroma. Linomide is an interesting antiangiogenic drug that is effective in the reduction of tumor growth in a murine prostate cancer model. Its activity includes the inhibition of macrophages recruitment and the increase of antiangiogenic molecules in TAMs such as IL12, IL18, CXCL10 and CXCL9. Hypoxia induces HIF, and hence VEGF expression in TAMs and HIF could represent another interesting target. Hypoxia signaling also leads to an increase in CXCR4 expression both on TAMs and ECs and CXCR4 antibodies such as AMD3100 inhibit tumor angiogenesis *in vivo* (Mantovani et al., 2004; Mantovani et al., 2006).

Anti-tumor agents with selective cytotoxic activity on monocyte-macrophages would be ideal therapeutic tools for their combined action on tumor cells and TAMs. Yondelis (Trabectedin), a natural product derived from the marine organism *Ecteinascidia turbinate* displays a potent anti-tumor activity and is specifically cytotoxic to macrophages and TAMs, while sparing the lymphocyte sub-set (Sessa et al., 2005). In addition, Yondelis inhibits the production of CCL2 and IL6 by TAMs which in turn contribute to growth suppression of inflammation-associated human tumors (Allavena et al., 2005).

A legumain-based vaccine represents another strategy aimed at inducing the host to contrast the infiltration of TAMs in the tumor. In fact, legumain, a member of the endopeptidase family, is highly over-expressed by TAMs in murine and human breast tumor tissues, and hence provides an ideal strategy for breast tumor targeting (Lewen et al., 2008; Xiang et al., 2008). The vaccination of mice against legumain provides a robust CD8+ T cell response against TAMs, significantly reducing tumor growth. Moreover, a dramatic reduction in TGFβ, TNFα, MMP9 and VEGF expression is induced, with a consequent reduction of both angiogenesis and metastatic rate (Luo et al., 2006).

*Targeting ECs and pericytes -* Besides targeting pro-angiogenetic growth factors and cytokines as mentioned above it is also possible to specifically target ECs and pericytes. The selective disruption of tumor vessels through the employment of the so called vascular disrupting

Therapeutical Cues from the Tumor Microenvironment 249

*Pro-angiogenic ECM proteins -* Many ECM proteins surrounding the vasculature including collagens, laminins, and fibronectins are pro-angiogenic promoting EC survival, proliferation, migration and tube formation. Pro-angiogenic factors such as VEGF, bFGF and TGF-β are bound and sequestered by the ECM via the heparan-like glycosaminoglycans. Fibronectin, a fairly ubiquitous and abundant ECM protein that becomes assembled into fibrils at the cell surface, is necessary for vasculogenesis, in fact if this gene is knocked out mice die before birth due among others, to vascular bed defects. An alternatively-spliced form of fibronectin that contains an extra B domain has been found in fetal and neoplastic tissues but not in normal adult tissues; this isoform was found to be synthesized by the vascular cells in malignant astrocytoma (Castellani et al., 2002) and its expression appears to be a precise diagnostic marker of the highest grade of glioma or glioblastoma. Also other studies demonstrate that this isoform may be a good marker of angiogenesis (Santimaria et

Osteopontin, a secreted cell attachment protein, seems to be over-expressed in association with increased tumor angiogenesis (Hirama et al., 2003). The proangiogenic function of osteopontin can be attributed to its ability to promote VEGF-directed dermal microvascular EC migration (Senger et al., 1996) and to increase MMP-2 levels in an RGD-dependent manner (Teti et al., 1998). Additionally, osteopontin-bound integrin αvβ3 inhibits NF-kB-

Tenascin-C is a glycoprotein composed of six subunits covalently associated by disulfide bonds. Different human Tenascin-C isoforms are generated by alternative splicing with aberrantly regulation in neoplastic tissues (Jones & Jones, 2000). It is expressed around angiogenic vessels in many tumors and there is evidence that it promotes and regulates angiogenesis *in vitro* and *in vivo*. Indeed the antibodies directed against Tenascin-C in

HSPGs are necessary for stable binding of the proangiogenic growth factor bFGF to its receptor (Ornitz et al., 1992) and alterations in HS glycosamminoglycan in breast carcinoma have been shown to result in an increase in bFGF binding and receptor complex assembly. Perlecan deposited along blood vessels basement membrane is thought to mediate structural and functional interactions with different molecules (Iozzo, 2005; Mongiat et al., 2003a). Intact perlecan is thought to be a pro-angiogenic HSPG as: *i)* its expression is altered during embryonic vasculogenesis and in neoplasia (Tapanadechopone et al., 2001; Zhou et al., 2004) perlecan-null mice show severe and sometimes fatal vascular and chondrogenic defects (Costell et al., 1999); *iii)* antisense targeting of perlecan blocks tumor growth and angiogenesis *in vivo* (Sharma et al., 1998); and *iv)* increased perlecan expression stimulates angiogenesis (Jiang & Couchman, 2003)*.* Tumor cell secreted perlecan is thought to promote EC sprouting and proliferation, thereby promoting angiogenesis (Jiang et al., 2004). Interestingly a degradation product of perlecan (endorepellin) exerts an opposite effect

Transmembrane chondroitin sulphate proteoglycan NG2 is another protein involved in tumor angiogenesis and has been shown to promote EC spreading perycite function (Ozerdem & Stallcup, 2003). An additional proangiogenic mechanism by which NG2 promotes angiogenesis includes its ability to bind and sequester angiostatin, thus blocking its antiangiogenic function (Chekenya et al., 2002). All these molecules may represent important tools for the development of drugs able to counteract blood vessel formation. *Anti-angiogenic ECM proteins -* Among the molecules that counteract blood vessel formation thrombospondin-1 (TSP-1) and thrombospondin-2 (TSP-2) are the best studied and their

glioma patients induced a significant inhibition of tumor angiogenesis.

al., 2003) and could also be a therapeutic target.

dependent EC apoptosis (Cooper et al., 2002).

(Mongiat et al., 2003b).

agents (VDAs) is a potentially useful alternative strategy. These compounds that predominantly affect the tumor periphery and small tumor masses have a different mechanism of action and tolerability profile compared with conventional antiangiogenic drugs and thus may provide an additional clinical benefit for a wide patient population. VDAs act to disrupt the established tumor vasculature in order to create an extensive necrosis in the tumor core. On the basis of their mechanism of action, VDAs are divided into 2 groups. The first includes the tubulin-binding agents that selectively target tumor ECs by disrupting their cytoskeleton. The second VDAs group includes compounds related to flavone acetic acid and they act in two ways; they induce ECs apoptosis and indirectly increase the intra-tumoral concentrations of TNFα and other cytokines, as well as nitric oxide, leading to an inhibition of tumor flow (Boehm et al., 2010; McKeage & Baguley, 2010).
