**3. Role of chemokines and their receptors in breast cancer progression and metastasis**

While most evidence presented above suggests that proinflammatory cytokines and enzymes play an important role in mediating tumorigenesis, and tumor progression, the molecular mechanisms of metastasis and its relationship with the organotropism of cancer cell remain unclear. However, recent studies focused on the chemokines and their receptors, and the different interactions with inflammatory cytokines in the tumor microenvironment have provided additional information that might better explain the non-random patterns of organotropism during metastasis, including atypical metastasis to rare organs (Franco-Barraza, et al., 2010, Valdivia-Silva, et al., 2009).

Chemokine activities in different malignancy including breast cancer are mediated primarily by their ability to induce chemotaxis of leukocytes, endothelial cells, and/or the tumor cells. Chemokines induce migration of leukocyte subpopulations to tumor sites that may promote antitumor activities (such as Th1 cells or natural killer cells), while other chemokines are responsible for large quantities of deleterious tumor-associated macrophages (TAM) at tumor sites (Allavena, et al., 2008, Ben-Baruch, 2008, Soria & Ben-Baruch, 2008) as discussed above. Moreover, specific chemokines upregulate endothelial cell migration and proliferation, and promoting angiogenesis, whereas other chemokines have powerful angiostatic properties (Strieter, et al., 2006, Struyf, et al., 2011). Another very important activity of chemokines is induction of tumor cell invasion and migration, thereby playing key roles in dictating site-directed metastasis formation (Ben-Baruch, 2008, Zlotnik, 2006). Chemokines and their receptors can execute such multifaceted roles in malignancy because cells of the tumor microenvironment, and in many cases also by the tumor cells themselves express them. As such, they can affect through autocrine pathways the ability of

Novel Insights Into the Role of Inflammation in Promoting Breast Cancer Development 135

receptors involved in regulation of metastasis. Most of the previously reported studies had focused in analysing chemokine receptors expressed in different neoplasias without evaluating their phenotypic changes and functionality during the progress of the disease (Ben-Baruch, 2008). However, it has not been clearly demonstrated any type of regulation of the microenvironment in these changes. Finally, the chemokine receptors expressed under non-stimulated conditions by cancer cells were considered biomarkers to specific homing to organs, but it does not explained atypical metastasis of cancer to rare organs

(Charalabopoulos, et al., 2004, Johnson, 2010, Kilgore, et al., 2007, Saisho, et al., 2005).

ones (Franco-Barraza, et al., 2010).

Within the tumor microenvironment, chemokines and their receptors play different roles in modulating several functions as described above, and through these processes, help to define the progression of the cancer. Stromal, and immune cells, including leukocytes differentiating into tumor-associated macrophages (TAM) at the tumor site, express numerous promoting factors, such as growth factors, angiogenic mediators, extracellular matrix-degrading enzymes, inflammatory cytokines, and more chemokines (Polyak & Kalluri, 2010). Interestingly, pro-inflammatory cytokines like IL-1, IL-6, IFN-g and TNF-a, which are important modulators of chemokine receptors expression in different tissues, have demonstrated to regulate their expression in cancer cells in a non–random manner (Valdivia-Silva, et al., 2009). Similar to cytokines regulate for CXCR4 and CCR5 in astrocytes (Croitoru-Lamoury, et al., 2003), CXCR2 in human mesangial cells(Schwarz, et al., 2002), and CX3CR1 in smooth muscle cells (Chandrasekar, et al., 2003), synovium (Nanki, et al., 2002), and different epithelial cells (Fujimoto, et al., 2001, Matsumiya, et al., 2001); different doses and times of exposition allowed the expression of specific type of chemokine receptor in several breast cancer cell lines and the change of their phenotypes into more invasiveness

We have analysed the human breast carcinoma MCF-7 cell line as a model of pre-invasive stage to demonstrate the regulation by an inflammatory microenvironment on chemokine receptor expression and functionality (Valdivia-Silva, et al., 2009). The comparison of the expression of CXCR4, CX3CR1, CXCR2, CCR9 and CCR5 at the transcriptional, protein, and functional levels under two different *in vitro* conditions (basal versus cytokine- stimulation) showed clearly the regulation of the specific cytokine over specific chemokine receptor, independently of the genetic background of MCF-7, which presents very low levels of these receptors under basal conditions. This was also observed in the highly metastatic MDA-MB-231, MDA-MB-361 and in the poorly metastatic T47D breast cancer cell lines; although the levels of expression observed after cytokine stimulation were different than those obtained in the MCF-7 cell line. A direct suggestion of these results, affirms that basal expression of a given chemokine receptor is not by itself a good marker of homing or aggressiveness and is subject to change by the microenvironment. Another important outcome in that work was the absence of correlation between the functionality of the receptor and their expression (gen or protein). For example, an increase in CXCR2 expression in MCF-7 cell line does not correlates with an increase in the migration index. In contrast, CX3CR, induced by TNF-a, had a small but significant increase at the protein level, which had an impact on their chemotactic activity. A considerable increase of chemokine receptors was found in nonmigratory cancer cells indicating that that chemokine receptor expression does not necessarily result in migration response to a chemoattractant ligand. It also suggests that only a fraction of the cells have the potential to form metastases and capable to invade different organs. In fact, genetic analysis of the MDA-MB-231 breast cancer cell line subpopulations, obtained from *in vivo* experiments, identified a gene set whose expression

the cancer cells to express tumor-promoting functions, and can also act in paracrine manners on host cells, thereby influencing their roles in malignancy.

Breast cancer metastasis is the result of several sequential steps and represents a highly organized, non-random and organ selective process dependent on intricate stroma-stroma interactions at the target organ (Ben-Baruch, 2006, Lu & Kang, 2007), causing high mortality by invasion of vital organs, such as bone, lung, brain and liver. Important evidence suggests that chemokines have an important role in regulating trafficking and metastasis (Bagley, et al., 2010). Indeed, breast cancer cells express chemokine receptors in a non-random manner, and these observations pointed to several chemokine/ receptor pairs that control cell–cell migration (Zlotnik, 2008). Association of chemokine receptors with various cancers including breast carcinoma has been widely documented (Ali & Lazennec, 2007, Karnoub & Weinberg, 2006, Koizumi, et al., 2007, Ruffini, et al., 2007). Accumulative evidence, in particular from clinical retrospective studies, presents a compelling picture indicating that the experimental evidence derived from *in vitro* experiments and animal models pointing to a pivotal role of chemokine receptors in cancer metastasis. CXCR4 and CCR7 are the most widely expressed in many different cancers, and the expression of CXCL12 and CCL21, their specific ligands, respectively, are highest in lung, liver bone marrow for the first one and lymph nodes for both (Nevo, et al., 2004, Schimanski, et al., 2008). Additionally, the expression of CCR7 in patients with several types of cancer has an excellent correlation with the ability of the tumor to spread to the lymph nodes (Takanami, 2003, Wang, et al., 2005). Other chemokine receptors may participate in the regulation of metastasis of specific cancers and in tumor progression. CX3CR1 is involved in homing metastasis to brain for glioblastoma and breast cancer (Andre, et al., 2010, Lavergne, et al., 2003) and to bone and bone marrow endothelial cell for prostate cancer (Shulby, et al., 2004). CCR9/CCL25 axis was found in melanoma (Letsch, et al., 2004), ovarian cancer (Johnson-Holiday, et al., 2007), prostate cancer (Singh, et al., 2004), nasopharyngeal carcinoma (Ou, et al., 2006), acute lymphoblastic leukaemia (Annels, et al., 2004) and probably breast cancer (Johnson-Holiday, et al., 2011); most of the cases are related to metastatic lesions in the gastrointestinal tract included the liver. Additionally, elevated expression levels of CXCR2 and CCR5 and their ligands, CXCL8 and CCL5, respectively, in breast carcinoma and other neoplasias were significant associated with increased malignancy, advance disease, early relapse and poor prognosis (Ben-Baruch, 2006, Yaal-Hahoshen, et al., 2006). Moreover, it has been demonstrated that tumor cells can generate autocrine gradients of ligands of chemokine receptors (i.e., CCR7) that guide their migration in direction of a physiological level of interstitial flow towards functional lymphatics, even if lymphatic endothelial cells are absent; although the effect is greatly amplified when both flow and cells are present (Shields, et al., 2007). This data suggests that the chemokine–chemokine receptor interaction is of particular importance in the metastatic destination of many cancers.

However, a couple of questions are very important to make in this point: Is the chemokine receptor expression in cancer cells constant? Or might the tumor microenvironment or inflammation regulate the chemokine receptor expression in cancer cells? Interestingly, these questions, which are product of logic thinking on the tumor microenvironment, were not made until recently by our group (Valdivia-Silva, et al., 2009). Indeed, the chemokine receptor expression has not been thoroughly studied under inflammatory conditions.

Although there are reports demonstrating that tumor and leukocytes increase expression of chemokines and cytokines during disease progression, it is not clear what are the chemokine

the cancer cells to express tumor-promoting functions, and can also act in paracrine manners

Breast cancer metastasis is the result of several sequential steps and represents a highly organized, non-random and organ selective process dependent on intricate stroma-stroma interactions at the target organ (Ben-Baruch, 2006, Lu & Kang, 2007), causing high mortality by invasion of vital organs, such as bone, lung, brain and liver. Important evidence suggests that chemokines have an important role in regulating trafficking and metastasis (Bagley, et al., 2010). Indeed, breast cancer cells express chemokine receptors in a non-random manner, and these observations pointed to several chemokine/ receptor pairs that control cell–cell migration (Zlotnik, 2008). Association of chemokine receptors with various cancers including breast carcinoma has been widely documented (Ali & Lazennec, 2007, Karnoub & Weinberg, 2006, Koizumi, et al., 2007, Ruffini, et al., 2007). Accumulative evidence, in particular from clinical retrospective studies, presents a compelling picture indicating that the experimental evidence derived from *in vitro* experiments and animal models pointing to a pivotal role of chemokine receptors in cancer metastasis. CXCR4 and CCR7 are the most widely expressed in many different cancers, and the expression of CXCL12 and CCL21, their specific ligands, respectively, are highest in lung, liver bone marrow for the first one and lymph nodes for both (Nevo, et al., 2004, Schimanski, et al., 2008). Additionally, the expression of CCR7 in patients with several types of cancer has an excellent correlation with the ability of the tumor to spread to the lymph nodes (Takanami, 2003, Wang, et al., 2005). Other chemokine receptors may participate in the regulation of metastasis of specific cancers and in tumor progression. CX3CR1 is involved in homing metastasis to brain for glioblastoma and breast cancer (Andre, et al., 2010, Lavergne, et al., 2003) and to bone and bone marrow endothelial cell for prostate cancer (Shulby, et al., 2004). CCR9/CCL25 axis was found in melanoma (Letsch, et al., 2004), ovarian cancer (Johnson-Holiday, et al., 2007), prostate cancer (Singh, et al., 2004), nasopharyngeal carcinoma (Ou, et al., 2006), acute lymphoblastic leukaemia (Annels, et al., 2004) and probably breast cancer (Johnson-Holiday, et al., 2011); most of the cases are related to metastatic lesions in the gastrointestinal tract included the liver. Additionally, elevated expression levels of CXCR2 and CCR5 and their ligands, CXCL8 and CCL5, respectively, in breast carcinoma and other neoplasias were significant associated with increased malignancy, advance disease, early relapse and poor prognosis (Ben-Baruch, 2006, Yaal-Hahoshen, et al., 2006). Moreover, it has been demonstrated that tumor cells can generate autocrine gradients of ligands of chemokine receptors (i.e., CCR7) that guide their migration in direction of a physiological level of interstitial flow towards functional lymphatics, even if lymphatic endothelial cells are absent; although the effect is greatly amplified when both flow and cells are present (Shields, et al., 2007). This data suggests that the chemokine–chemokine receptor interaction

on host cells, thereby influencing their roles in malignancy.

is of particular importance in the metastatic destination of many cancers.

However, a couple of questions are very important to make in this point: Is the chemokine receptor expression in cancer cells constant? Or might the tumor microenvironment or inflammation regulate the chemokine receptor expression in cancer cells? Interestingly, these questions, which are product of logic thinking on the tumor microenvironment, were not made until recently by our group (Valdivia-Silva, et al., 2009). Indeed, the chemokine receptor expression has not been thoroughly studied under inflammatory conditions. Although there are reports demonstrating that tumor and leukocytes increase expression of chemokines and cytokines during disease progression, it is not clear what are the chemokine receptors involved in regulation of metastasis. Most of the previously reported studies had focused in analysing chemokine receptors expressed in different neoplasias without evaluating their phenotypic changes and functionality during the progress of the disease (Ben-Baruch, 2008). However, it has not been clearly demonstrated any type of regulation of the microenvironment in these changes. Finally, the chemokine receptors expressed under non-stimulated conditions by cancer cells were considered biomarkers to specific homing to organs, but it does not explained atypical metastasis of cancer to rare organs (Charalabopoulos, et al., 2004, Johnson, 2010, Kilgore, et al., 2007, Saisho, et al., 2005).

Within the tumor microenvironment, chemokines and their receptors play different roles in modulating several functions as described above, and through these processes, help to define the progression of the cancer. Stromal, and immune cells, including leukocytes differentiating into tumor-associated macrophages (TAM) at the tumor site, express numerous promoting factors, such as growth factors, angiogenic mediators, extracellular matrix-degrading enzymes, inflammatory cytokines, and more chemokines (Polyak & Kalluri, 2010). Interestingly, pro-inflammatory cytokines like IL-1, IL-6, IFN-g and TNF-a, which are important modulators of chemokine receptors expression in different tissues, have demonstrated to regulate their expression in cancer cells in a non–random manner (Valdivia-Silva, et al., 2009). Similar to cytokines regulate for CXCR4 and CCR5 in astrocytes (Croitoru-Lamoury, et al., 2003), CXCR2 in human mesangial cells(Schwarz, et al., 2002), and CX3CR1 in smooth muscle cells (Chandrasekar, et al., 2003), synovium (Nanki, et al., 2002), and different epithelial cells (Fujimoto, et al., 2001, Matsumiya, et al., 2001); different doses and times of exposition allowed the expression of specific type of chemokine receptor in several breast cancer cell lines and the change of their phenotypes into more invasiveness ones (Franco-Barraza, et al., 2010).

We have analysed the human breast carcinoma MCF-7 cell line as a model of pre-invasive stage to demonstrate the regulation by an inflammatory microenvironment on chemokine receptor expression and functionality (Valdivia-Silva, et al., 2009). The comparison of the expression of CXCR4, CX3CR1, CXCR2, CCR9 and CCR5 at the transcriptional, protein, and functional levels under two different *in vitro* conditions (basal versus cytokine- stimulation) showed clearly the regulation of the specific cytokine over specific chemokine receptor, independently of the genetic background of MCF-7, which presents very low levels of these receptors under basal conditions. This was also observed in the highly metastatic MDA-MB-231, MDA-MB-361 and in the poorly metastatic T47D breast cancer cell lines; although the levels of expression observed after cytokine stimulation were different than those obtained in the MCF-7 cell line. A direct suggestion of these results, affirms that basal expression of a given chemokine receptor is not by itself a good marker of homing or aggressiveness and is subject to change by the microenvironment. Another important outcome in that work was the absence of correlation between the functionality of the receptor and their expression (gen or protein). For example, an increase in CXCR2 expression in MCF-7 cell line does not correlates with an increase in the migration index. In contrast, CX3CR, induced by TNF-a, had a small but significant increase at the protein level, which had an impact on their chemotactic activity. A considerable increase of chemokine receptors was found in nonmigratory cancer cells indicating that that chemokine receptor expression does not necessarily result in migration response to a chemoattractant ligand. It also suggests that only a fraction of the cells have the potential to form metastases and capable to invade different organs. In fact, genetic analysis of the MDA-MB-231 breast cancer cell line subpopulations, obtained from *in vivo* experiments, identified a gene set whose expression

Novel Insights Into the Role of Inflammation in Promoting Breast Cancer Development 137

microcirculation, direction of lymph or circulation drainage, etc.). B) A second view, based on the participation of an inflammatory microenvironment, takes into account constant interactions between tumor cells and surrounding cells during the different stages of cancer development. Therefore, the final response is the result of positive and negative effects and not only dependent on internal genetic changes in cancer cells but on interactions and epigenetic control of multiple inflammatory into molecules released into the tumor microenvironment. Therefore, the final metastatic homing, which is mediated by expression of chemokines and chemokine receptors, will be dependent on the deregulation of the host

As a consequence of studies focusing almost exclusively on cancer cells, nearly all of the currently used cancer therapeutic agents target the cancer cells that, due to their inherent genomic instability, frequently acquire therapeutic resistance (Rajagopalan, et al., 2003). In part due to frequent therapeutic failures during the course of treatment of advanced stage tumors, increasing emphasis has been placed on targeting various stromal cells, particularly endothelial cells, via therapeutic interventions. Since these cells are thought to be normal and genetically stable, they are less likely to develop acquired resistance to cancer therapy. Thus, isolating, and characterizing each cell type (epithelial, myoephitelial, and various stromal cells) comprising non-malignant and cancerous breast tissue would not only help us to understand the role these cells play in breast tumorigenesis, but would likely give us new

There is now an abundant literature documenting the associations of chemokine receptors with various types of cancer (Zlotnik, 2006) and their importance to mediate the establishment or development of metastatic foci. In fact, some anticancer drugs currently in use -like Herceptin- may involve the downregulation of chemokine receptors as part of their mechanism of action (Li, et al., 2004). This would provide the ultimate validation of the hypothesis, and would also point to future opportunities for therapeutic intervention as we discussed below. Current therapies such as surgery, radiotherapy and chemotherapy are primarily concerned with destruction of cancer. Targeting chemokines and chemokine receptors will allow limiting angiogenesis or metastasis and may enable such therapies to act as chemotherapeutic agents alone or in synergism with conventional agents. The upregulation of certain chemokine molecules in tumor as compared with normal cells offers a potential avenue—where cancer cells and their metastases can be specifically targeted. This

selective destruction of cells is also pre-requisite of non-toxic treatment regimens.

Manipulation of the tumor microenvironment by treatment with chemokines can be used to recruit either immature dendritic cells for the initiation of anti-tumor responses or effector cells for cytotoxic responses. Intratumoral delivery of CCL21 using pox virus vaccine into established tumors derived from murine colon cancer line, CT26 results in enhanced infiltration of CD4 T cells which correlated with inhibition of tumor growth (Flanagan, et al., 2004). Non-immunogenic murine breast carcinoma is rejected after transducing cells with CCL19. The rejection of tumor was mediated by activated NK and CD4+ cells (Braun, et al., 2000). Adenoviral delivery of the CCL16 is able to inhibit growth of mammary tumors and prevent metastatic growth (Okada, et al., 2004). Importantly, in treatment involving delivery of chemokines to the tumor environment, there is a major problem of heterogeneity of the tumor cells. Chemokines may have dual effects, can be beneficial to one patient might be

**4. Targeting chemokines for breast cancer metastasis** 

molecular targets for cancer intervention and treatment.

immune response

pattern is associated to metastasis to bone but not adrenal medulla (Kirschmann, et al., 1999, Xu, et al., 2010). Interestingly this signature is retained through repeated passage of the metastatic cell population both in vitro and in vivo. Therefore, breast cancer cells with a defined tissue-specific metastatic ability pre-exist in the parental tumor cell population and may have a distinctive metastasis gene expression signature. Thus, these data suggested that inflammatory stimulation in the tumor microenvironment might affect cancer cells migration by different mechanisms. Importantly, not all cancer cell population, including cell lines, had the same behaviour under the same cytokine stimulation. Finally, other important finding in this study suggested that cancer cells require constant inflammatory stimuli by the microenvironment to trigger their invasive and metastatic activity, because of after a short time without stimuli (hours to days), the cells diminished their specific-stimuli chemokine receptor expression.

Altogether, these data allowed us to propose that exist sub-populations expressing different levels of chemokine receptor expression, which under a particular stimuli in the host microenvironment, change their expression levels and thus their aggressiveness. Then, atypical metastasis of breast cancer to others organs, which are relatively rare, could fall under this scheme. The biological inflammatory global response in the tumor microenvironment might be triggering the expression of different chemokine receptors and determining a new homing for these cancer cells. More broadly, these observations strongly support the overall model where chemokines determine the metastatic destinations of cancer cells (Fig 1.)

Fig. 1. Microenvironment and cancer progression.

Two theories have been proposed to explain this process, a conventional theory based on genetic alterations and a second view that involves participation of an inflammatory microenvironment. A) Initially, susceptible cells to different carcinogenic factors (e.g., genetic susceptibility obtained by inheritance) suffer specific DNA mutations that trigger tumorigenesis. The conventional theory is focused on the view that cancer progression is initially dependent on a sequence of genetic alterations and, finally, purely mechanical factors regulate the fate of blood-borne metastasis tumour cells (e.g. proximity,

pattern is associated to metastasis to bone but not adrenal medulla (Kirschmann, et al., 1999, Xu, et al., 2010). Interestingly this signature is retained through repeated passage of the metastatic cell population both in vitro and in vivo. Therefore, breast cancer cells with a defined tissue-specific metastatic ability pre-exist in the parental tumor cell population and may have a distinctive metastasis gene expression signature. Thus, these data suggested that inflammatory stimulation in the tumor microenvironment might affect cancer cells migration by different mechanisms. Importantly, not all cancer cell population, including cell lines, had the same behaviour under the same cytokine stimulation. Finally, other important finding in this study suggested that cancer cells require constant inflammatory stimuli by the microenvironment to trigger their invasive and metastatic activity, because of after a short time without stimuli (hours to days), the cells diminished their specific-stimuli

Altogether, these data allowed us to propose that exist sub-populations expressing different levels of chemokine receptor expression, which under a particular stimuli in the host microenvironment, change their expression levels and thus their aggressiveness. Then, atypical metastasis of breast cancer to others organs, which are relatively rare, could fall under this scheme. The biological inflammatory global response in the tumor microenvironment might be triggering the expression of different chemokine receptors and determining a new homing for these cancer cells. More broadly, these observations strongly support the overall model where chemokines determine the metastatic destinations of

Two theories have been proposed to explain this process, a conventional theory based on genetic alterations and a second view that involves participation of an inflammatory microenvironment. A) Initially, susceptible cells to different carcinogenic factors (e.g., genetic susceptibility obtained by inheritance) suffer specific DNA mutations that trigger tumorigenesis. The conventional theory is focused on the view that cancer progression is initially dependent on a sequence of genetic alterations and, finally, purely mechanical factors regulate the fate of blood-borne metastasis tumour cells (e.g. proximity,

chemokine receptor expression.

Fig. 1. Microenvironment and cancer progression.

cancer cells (Fig 1.)

microcirculation, direction of lymph or circulation drainage, etc.). B) A second view, based on the participation of an inflammatory microenvironment, takes into account constant interactions between tumor cells and surrounding cells during the different stages of cancer development. Therefore, the final response is the result of positive and negative effects and not only dependent on internal genetic changes in cancer cells but on interactions and epigenetic control of multiple inflammatory into molecules released into the tumor microenvironment. Therefore, the final metastatic homing, which is mediated by expression of chemokines and chemokine receptors, will be dependent on the deregulation of the host immune response
