**5.2 Angiogenesis**

220 Breast Cancer – Focusing Tumor Microenvironment, Stem Cells and Metastasis

To begin to define the role played by TAFs in tumour progression, Micke *et al.* (2007) conducted cDNA microarray analyses comparing the transcriptome of TAFs from basal cell carcinoma with normal dermal fibroblasts (Micke *et al*., 2007). This study showed that TAFs overexpress multiple growth factors such as PDGF, EGF, and VEGF, chemokines such as SDF1 and CXCL12 and matrix proteins such as MMP11, LAMA2 and COL5A2. In fact, these TAFs are known to secrete IGF-2, FGF-7, TGF-β, leptin, and NGF, which bind to their cognate receptors on BCA cells to stimulate HA production (Szabo *et al*., 2011). This then promotes expression of cytokines such as TGF-β that attract and stimulate TAFs to proliferate. This paracrine effect is a positive feedback mechanism, because proliferating TAFs secrete additional growth factors, cytokines, chemokines, and MMPs that sustain BCA transformation and promote BCA progression. Additionally, VEGF, produced by TAFs, and HA oligosaccharides induce angiogenesis. HA itself also impairs immune surveillance, and/or activates TAMs and neutrophils that have tumour enhancing potential. Overexpression of HAS in a non-transformed rat fibroblast, 3Y1, increases high MW HA production and the resultant pericellular HA coat provides cells with a proliferation advantage that is accompanied by loss of contact inhibition of growth. This is achieved through HA-mediated activation of PI3 kinase. Lower MW HA also increases proliferation in these cells but has no effect on the HA matrix (Itano *et al*., 2002). TAFs affect not only BCA cells but also normal cells in which the tumour is embedded. For example, TAFs induce stem cell-like behaviour and aberrant differentiation in normal fibroblasts, which can affect BCA progression. TAFs promote the expression of stem-cell markers such as Oct4 and Sox2 in 3T3 cells (Szabo *et al*., 2011) and stimulate trans-differentiation of normal fibroblasts into

Adipose tissue in mammary glands is important for its secretory and endocrinal functions as well as metabolism, energy homeostasis and stem cell compartment. Adipocytes contribute to the mammary tissue ECM and this effect is at least partly regulated by HA. There are not many studies that focus on HA and its relationship to adipocytes, however, the importance of this polysaccharide on adipose-stromal interactions in the breast tissue is becoming apparent. For example, HA increases the crosslinking of collagen-HA matrices, supports proliferation and differentiation of pre-adipocytes and induces a higher proportion

Chen *et al.* (2007) also showed that HA extends the lifespan, reduces cellular senescence and enhances differentiation potential of murine adipose-derived stromal cells (mADSCs) *in culture*. Collectively, these results provide preliminary evidence for a key role of HA in controlling the adipose component of the breast tissue and allude to a potential role of this

Considerable evidence indicates that HA fragmentation is required for immune cell trafficking, fibroblast migration, stem cell migration from niches to the wound site and endothelial cell migration during angiogenesis. For example, acellular hydrogel matrix composed of fibronectin and HA, which simulates a wound microenvironment, supports

**5. HA regulates mammary cell functions that promote BCA progression** 

**4.2.2 HA/stromal fibroblast/epithelial cell interaction and tumour progression** 

myofibroblasts when they are confronted with primary BCA cells.

**4.2.3 HA, adipocytes and adipose tissue** 

of cycling cells (Davidenko *et al*., 2010).

regulation in BCA (Chen *et al*., 2007).

**5.1 Cell migration** 

Hypoxic conditions within tumours require neovascularisation of the microenvironment for the tumour to continue to grow and metastasize. Hypoxia, a condition often found within the TME, induces the activation, as seen by nuclear translocation, of either or both of NFκB and HIF-1α. This effect has been shown both *in vitro* in MCF-7 BCA cells, and *in vivo* (Tafani *et al.,* 2010). Invasion, migration, and proliferation of endothelial cells, as well as tissue remodelling, are essential processes during angiogenesis, which directly and indirectly help to promote tumour growth and metastasis. Necrotic cells, which have died as a result of hypoxia, also release chemokines that recruit macrophages and a pro-inflammatory response conducive to tissue remodelling. Hypoxia may produce ROS which in turn cause HA fragmentation and Noble *et al.* (1996) showed that NFκB transcription in macrophages is activated by HA fragments (Noble *et al*., 1996). Later, Rockey *et al.* (1998) were the first to show in hepatocytes that HA activation of NFκB induces NOS2 production, which can be synergistically increased in the presence of cytokines such as IFN-γ (Rockey *et al*., 1998). It has since been shown that HA fragments activate the NFκB pathway through TLR4 in both DC and macrophages (Termeer *et al*., 2002). Hypoxia induced activation of HIF-1α and NFκB induces pro-inflammatory gene expression and both mRNA and protein levels of inflammatory mediators such as RAGE, PTX3, NOS2, COX2, and CXCR4 are increased. Increased expression of CXCR4, which is the receptor for SDF-1, is seen on MCF-7 cells subjected to hypoxic conditions (Tafani *et al*., 2010). This increases the migratory and invasive capacity of these cells, which are usually non-invasive. In these same studies it was found that nuclear translocation of NFκB is at least partly dependent on HIF-1α, indicating that it may be under hypoxic regulation, as inhibition of HIF-1α decreases nuclear localisation of NFκB, and in turn RAGE and P2X7R expression, inhibiting cell invasion (Tafani *et al*., 2010).

In general, high MW HA inhibits angiogenesis while fragments promote angiogenesis. Overexpression of HA and HYALs has been linked to an increase in angiogenesis in several types of cancers including breast (Tan *et al*., 2010), bladder (Lokeshwar *et al.*, 2000, Golshani

Hyaluronan Associated Inflammation and Microenvironment

**7. HA and receptor antagonists in clinical trials** 

Remodelling Influences Breast Cancer Progression 223

Since it is evident that HA and its receptors play an important role in BCA and other tumours, it is unsurprising that reagents blocking HA metabolism are being assessed as therapeutic agents in certain types of cancer. In pre-clinical models, Kultti *et al. (*2009) demonstrated that the HAS inhibitor 4-MU (4-Methylumbelliferone) specifically depletes intracellular levels of UDP-Glc-UA (Kakizaki *et al*., 2004) by serving as a glucoronidation substrate in A2058 melanoma cells, MCF-7, MDA-MB-361 BCA cells, SKOV-3 ovarian, and UT-SCC118 squamous carcinoma cells. Additionally, Lokeswar *et al.* (2010) used 4-MU to block growth of human prostate cancer cell line xenografts in immunocompromised mice. 4-MU induces apoptosis in these tumours and also strongly inhibits cell proliferation, motility and invasion. These effects can be reversed by addition of HA, which demonstrates that, although 4-MU does not specifically block HAS and has other off target effects, its

HA has also proven to be a good adjunct therapeutic option *in vivo* in human cancers since it promotes targeting of active anti-cancer compounds. For example, when patients with Calmette-Guérin refractory bladder cancer were included in a Phase I clinical trial using Paclitaxel-HA (ONCOFID-P-BTM) for treatment of their cancers, 60% of the patients treated exhibited a clinical response with minimal toxicity reported (Bassi *et al*., 2010). HA has been successfully used to carry/target other chemotherapeutics, thus reducing cytotoxic side effects of the active drug. Hyung *et al.* (2008) demonstrated the efficacy of HA-coated drug carriers by delivering doxorubicin to MDA-MB-231 and ZR-75-1 human BCA cell lines (Hyung *et al*., 2008). Similarly, after coating nanoparticles containing paclitaxel with HA, cytotoxicity is reduced while cellular uptake of the drug by S-180 sarcoma cell line is

In light of fairly recent evidence for the display of CD44 on BCA tumour initiator cells, interest in developing CD44 targeted therapies has increased. Riechelmann *et al.* (2008) exploited the potential of CD44 in a Phase I clinical trial using an antimicrotubule agent (mertansine) and a monoclonal antibody to CD44v6 (bivatuzumab), (BIWI 1), to treat patients with recurrent or metastatic head and neck squamous cell carcinoma (Riechelmann *et al*., 2008). The response to the treatment was unexpectedly variable and the trials using these agents were stopped after one patient died of toxic epidermal necrolysis (Tijink *et al*., 2006). Targeting the HA binding ability of activated CD44 may result in decreased toxicity. RHAMM peptide vaccination (e.g. R3, which is HLA-A2-restricted) has recently been assessed in PhaseI/II clinical trials for treatment of MM, AML, and CLL (Giannopoulos *et al*., 2010, Greiner *et al*., 2008, 2010, Schmitt *et al*., 2008). Additionally, vaccination with DC pre-stimulated against the same peptide has also undergone Phase I and II clinical trials for treatment of CLL (Hus *et al*., 2008). Vaccination with RHAMM peptide has the attractive advantage of very low toxicity because it is not expressed in healthy bone marrow tissue. RHAMM vaccination resulted in leukemic blast lysis, blast reduction in the bone marrow and avoided the need for blood transfusions for one patient. Furthermore, an immunological response, marked by an increase in T cell frequency, was observed in 70% of AML, MM, and MDS patients in an initial study (Schmitt *et al*., 2008). Subsequently, RHAMM peptide was shown to be non-toxic at high dosage (1000 µg/vaccination), however, there was no dose-dependent effect, indicating that RHAMM is an effective therapeutic target even at low levels (Greiner *et al*., 2010). A similar response was seen in CLL patients vaccinated with RHAMM peptide, as well as RHAMM peptide-stimulated DC.

effects on tumour cell growth result from inhibition of HAS (Ekici *et al*., 2004).

enhanced 9.5 fold *in vitro* and in a mouse model (He *et al*., 2009).

*et al.*, 2008), prostate (Ekici *et al*, 2004, Bharadwaj *et al.*, 2007), and endometrial (Paiva *et al.*, 2005). Koyama *et al.* (2007) demonstrated that an increase in HAS2 expression by genetic modifications in a mouse model of BCA causes a higher incidence of adenocarcinoma accompanied by an increase in angiogenesis (Koyama *et al*., 2007). An increase in HA by overexpression of HAS2 in transgenic mice induces a more aggressive BCA phenotype and an increase in blood and lymphatic vessels (Kobayashi *et al*., 2010). In these tumours, the stromal cells also secrete a variety of pro-angiogenic factors. Furthermore, HA concentration in stroma and blood vessels is increased, as well as the amount of small HA fragments.

The pro-angiogenic effects of HA fragments result from the display of CD44 and RHAMM (Wang *et* al., 2011, Slevin *et al.*, 2007) on the surfaces of endothelial, BCA or leukocyte cells. and Interaction of HA fragments with these cells produces the factors required for stimulating endothelial cells to form new blood vessels. HA fragments stimulate endothelial cell proliferation, migration and tube formation. Increased expression of HYALs in conjunction with MMPs and Cathepsin-D induce a more invasive phenotype in the endothelial cell line ECV-304 as detected by matrigel invasion assay (Wang *et al*., 2009). Additionally, pro-inflammatory cytokines, secreted by leukocytes activated by CD44-HA mediated interactions, stimulate endothelial cells to produce HA. When HUVEC cells are stimulated with IL-1B, TNF-α and β1, they secrete HA. CD44-HA interaction stimulates early morphogenic events, such as tube formation and proliferation in HUVECs (Wang *et al*., 2011). Furthermore, HA works synergistically with macrophage recruitment to promote vascular formation and HA in the stroma promotes lymphangiogenesis at the invasive tumour front in BCA through the activation of endothelial LYVE-1 (Itano *et al*., 2002).

### **6. HA and multi-drug resistance in BCA**

Most tumours initially respond to chemotherapy treatment but later acquire resistance, resulting in treatment failure and tumour recurrence. Some mechanisms by which tumour cells acquire resistance include inhibition of apoptosis, stimulation of cell proliferation and enhanced expression and activity of drug export pumps, particularly ATP driven pumps (ABC transporters), which reduce the intracellular, and therefore active, concentration of several chemotherapeutic agents. HA fragments augment expression and activity of MDR1, a member of the ABC drug transporter family, in primary BCA cells (Toole and Slomiany, 2008). This HA induced upregulation involves the Akt/PI3 kinase signalling pathway and is CD44 dependent. CD44-HA interactions stimulate MDR1 expression via multiple signalling mechanisms including epigenetic gene expression regulation. CD44-HA binding results in activation of PKC as well as increased phosphorylation and nuclear translocation of Nanog, a stem cell specific transcription factor. Moreover, interaction of Nanog with Stat-3 in the nucleus increases Stat-3 regulated gene expression, resulting in increased expression of MDR1. Activation of Nanog also results in production of the micro RNA miR-21 and down-regulation of PDCD4, a tumour suppressor protein (Bourguignon *et al*., 2008, 2009). CD44-HA interaction increases an association between MDR1 and the cytoskeletal protein ankyrin, resulting in enhanced drug export (Bourguignon *et al*., 2008). Additionally, CD44-HA interactions upregulate the expression of the histone acetyl-transferase, p300, inducing the acetylation of catenin and NFB. This stimulates expression of MDR1 and the anti-apoptotic protein, Bcl-xL (Bourguignon *et al*., 2009). It is very likely that BCA tumours with high HA metabolisms are also highly resistant to treatment with drugs that can be exported by MDR1.

*et al.*, 2008), prostate (Ekici *et al*, 2004, Bharadwaj *et al.*, 2007), and endometrial (Paiva *et al.*, 2005). Koyama *et al.* (2007) demonstrated that an increase in HAS2 expression by genetic modifications in a mouse model of BCA causes a higher incidence of adenocarcinoma accompanied by an increase in angiogenesis (Koyama *et al*., 2007). An increase in HA by overexpression of HAS2 in transgenic mice induces a more aggressive BCA phenotype and an increase in blood and lymphatic vessels (Kobayashi *et al*., 2010). In these tumours, the stromal cells also secrete a variety of pro-angiogenic factors. Furthermore, HA concentration in stroma and blood vessels is increased, as well as the amount of small HA fragments. The pro-angiogenic effects of HA fragments result from the display of CD44 and RHAMM (Wang *et* al., 2011, Slevin *et al.*, 2007) on the surfaces of endothelial, BCA or leukocyte cells. and Interaction of HA fragments with these cells produces the factors required for stimulating endothelial cells to form new blood vessels. HA fragments stimulate endothelial cell proliferation, migration and tube formation. Increased expression of HYALs in conjunction with MMPs and Cathepsin-D induce a more invasive phenotype in the endothelial cell line ECV-304 as detected by matrigel invasion assay (Wang *et al*., 2009). Additionally, pro-inflammatory cytokines, secreted by leukocytes activated by CD44-HA mediated interactions, stimulate endothelial cells to produce HA. When HUVEC cells are stimulated with IL-1B, TNF-α and β1, they secrete HA. CD44-HA interaction stimulates early morphogenic events, such as tube formation and proliferation in HUVECs (Wang *et al*., 2011). Furthermore, HA works synergistically with macrophage recruitment to promote vascular formation and HA in the stroma promotes lymphangiogenesis at the invasive

tumour front in BCA through the activation of endothelial LYVE-1 (Itano *et al*., 2002).

also highly resistant to treatment with drugs that can be exported by MDR1.

Most tumours initially respond to chemotherapy treatment but later acquire resistance, resulting in treatment failure and tumour recurrence. Some mechanisms by which tumour cells acquire resistance include inhibition of apoptosis, stimulation of cell proliferation and enhanced expression and activity of drug export pumps, particularly ATP driven pumps (ABC transporters), which reduce the intracellular, and therefore active, concentration of several chemotherapeutic agents. HA fragments augment expression and activity of MDR1, a member of the ABC drug transporter family, in primary BCA cells (Toole and Slomiany, 2008). This HA induced upregulation involves the Akt/PI3 kinase signalling pathway and is CD44 dependent. CD44-HA interactions stimulate MDR1 expression via multiple signalling mechanisms including epigenetic gene expression regulation. CD44-HA binding results in activation of PKC as well as increased phosphorylation and nuclear translocation of Nanog, a stem cell specific transcription factor. Moreover, interaction of Nanog with Stat-3 in the nucleus increases Stat-3 regulated gene expression, resulting in increased expression of MDR1. Activation of Nanog also results in production of the micro RNA miR-21 and down-regulation of PDCD4, a tumour suppressor protein (Bourguignon *et al*., 2008, 2009). CD44-HA interaction increases an association between MDR1 and the cytoskeletal protein ankyrin, resulting in enhanced drug export (Bourguignon *et al*., 2008). Additionally, CD44-HA interactions upregulate the expression of the histone acetyl-transferase, p300, inducing the acetylation of catenin and NFB. This stimulates expression of MDR1 and the anti-apoptotic protein, Bcl-xL (Bourguignon *et al*., 2009). It is very likely that BCA tumours with high HA metabolisms are

**6. HA and multi-drug resistance in BCA** 
