**3. Tumor models**

### **3.1 Tumor models in general**

Mouse tumor-angiogenesis models include xenograft/transplant models and spontaneous models in genetically engineered mice. Xenograft/transplant models are the most widely used, due to the homogeneous and fast onset and progression of the disease (Y Cao, 2005). In these models, human tumor cells are transplanted into athymic nude mice or severely compromised immunodeficient (SCID) mice (xenograft model), or murine tumor cells are transplanted into mice of the same genetic background (homologous transplantation model). Tumor cells can be injected under the skin (heterotypic transplantation, except skin

samples. Thin sections of 15 µm should be adhered on Superfrost Plus microscope slides, subsequently fixed with cold acetone and stained with specific primary antibodies. However, it is important to note, when staining for lymphatic vessels in the adipose tissue that LYVE-1 is not as widespread on the lymphatic endothelium as podoplanin in this particular tissue (Fig. 9). Here, unlike other tissues, LYVE-1 staining mainly detects

Fig. 9. Double cryosection immunostaining of the WAT of a mouse with podoplanin (red) and LYVE-1 (green). Podoplanin-positive lymphatic staining is indicated with arrowheads

Since hypoxia is one of the driving forces of angiogenesis, detection of the presence of hypoxia in tissues could also be performed. Hypoxyprobe-1 (pimonidazole hydrochloride) staining can be used to detect cell and tissue hypoxia. Pimonidazole hydrochloride has a molecular weight of 290.8 kD, ultraviolet absorbance at 324 nm and a plasma half-life of approximately 25 minutes in mice. The detection of hypoxia in tissues can be performed by intravenous or intraperitoneal injection or oral ingestion of pimonidazole hydrochloride at a dosage of 60 mg/kg, 15-90 minutes before sacrificing mice. Dissected tissues should be fixed in 4% PFA followed by paraffin embedding. Paraffin-embedded tissue sections of 3-5 µm is further stained using a peroxidase conjugated anti-pimonidazole antibodies and

Mouse tumor-angiogenesis models include xenograft/transplant models and spontaneous models in genetically engineered mice. Xenograft/transplant models are the most widely used, due to the homogeneous and fast onset and progression of the disease (Y Cao, 2005). In these models, human tumor cells are transplanted into athymic nude mice or severely compromised immunodeficient (SCID) mice (xenograft model), or murine tumor cells are transplanted into mice of the same genetic background (homologous transplantation model). Tumor cells can be injected under the skin (heterotypic transplantation, except skin

and LYVE-1-positive non- lymphatic staining is indicated with arrows.

inflammatory cells.

Scale bar = 100 μm.

**2.8.5 Hypoxia staining** 

**3. Tumor models** 

counterstained with nuclei staining.

**3.1 Tumor models in general** 

cancer) or into the same organ from which the tumors originate (orthotropic transplantation). It is important to remember when studying tumor angiogenesis that the tumor microvasculature might differ depending on the implantation sites. Orthotropic transplantation is considered to show a similar tumor vessel phenotype as the primary lesion, and thus more closely recapitulate the clinical situation. However, it is often easier and more accurate to follow the tumor growth in heterotypic models such as when the tumor grows under the skin, where it is easily visualized. Furthermore, depending on the tumor growth rate and size, the vessel structure in tumors can be different. Fast growing and big tumors (≥ 1.0 cm3) usually have extensive necrosis in the center, which affects the vascular network in these areas. The appropriate tumor size for evaluation of tumor angiogenesis is considered to be 0.5-0.8 cm3. Tumor cells can be modified to express high levels of angiogenic factors, such as VEGF and FGF, or reporter genes, such as green fluorescent protein (GFP), red fluorescent protein (RFP) or luciferase. These modifications allow us to study the function of specific angiogenic factors on tumor microenvironment in association with invasion and metastasis (R Cao et al, 2004). Recent technological developments have permitted us to use genetically manipulated animal models for tumor studies, in which the mice are over-expressing or deficient in angiogenesis related genes. The majority of these mice are however immunocompetent, requiring the use of murine tumor cells in such animals.

Another type of tumor-angiogenesis models are the spontaneous models. Such models are based on genetically engineered mice, where a tumor suppressor has been deleted, an oncogene is being over-expressed or both, often in a particular cell type such as the pancreatic beta-cell of the islets of Langerhans (Hanahan, 1985). These models provide a disease history that is more closely recapitulating the one in human patients, as tumors are generated from one single hyperplastic cell, and progress through steps which are highly similar to pathological progression of pre-malignant to malignant lesions in humans. In later stages of tumor development in these models, angiogenesis also becomes important and contributes to growth and metastasis of these primary lesions (Koh et al, 2010).

To study tumor angiogenesis using mouse models, control animals or tumor conditions must be incorporated in the study. If tumor cells are genetically altered, the proper control would be tumor cells which have an empty expression vector inserted in their genome instead of one coding for a particular angiogenic factor. For evaluation of treatment efficacy of anti-angiogenic drugs, the control group must be given the vehicle (solvent) in which the drug was prepared, because some vehicles can themselves affect the vascular structure in tumors. Age and genetic background of the mice must also be standardized for the different groups. Standard age for tumor experiments is 6-10 weeks. Sex is usually not important, unless the tumor cells under investigation are gender-specific such as in the case of breast or prostate cancer, but should nonetheless be the same in for example treatment and control groups.

### **3.2 Tumor models: Assessment of angiogenesis**

Dr. Judah Folkman proposed in 1971 that all tumor growth is angiogenesis-dependent (Folkman, 1971). In general, because a limiting factor of tumor growth is the supply of sufficient levels of oxygen and nutrition, fast growing tumors are characterized by a more aggressive angiogenic phenotype which arises from the ability of the tumor cells to secrete angiogenic growth factors to support and change their microenvironments. For instance, tumor cells transduced with the gene coding for VEGF grow much faster than the emptyvector transduced control cells, due to much higher intra-tumoral blood vessel density (Eriksson et al, 2002). In these tumors however, the vasculature consists of highly irregular and immature vessels and many vascular plexuses, which does not support efficient perfusion of blood compared to a less chaotic vasculature in VEGF non-transduced tumor tissue (Fig. 10). To assess angiogenesis in mouse models, there are thus three essential points which should be addressed.

Fig. 10. Histological examinations of tumor tissue. (A) H&E staining of T241 fibrosarcoma. (B) Immunohistchemical staining of T241 fibrosarcoma visualized by fluorescence microscopy. Anti-endomucin labeling of endothelial cells is shown in red and anti-α-smooth muscle actin (SMA) labeling of pericytes/smooth muscle cells is shown in green. (C) Confocal imaging of T241 fibrosarcoma. GFP transfected tumors are shown in green, anti-CD31 labeling of endothelial cells is shown in red. (D) Confocal imaging of VEGF overexpresing T241 fibrosarcoma. GFP transfected tumors are shown in green, anti-CD31 labeling of endothelial cells is shown in red.

1. Tumor size (volume). Tumor growth depends on angiogenesis. Thus, tumor size is usually proportional to the degree of which tumor-angiogenesis is induced by the tumor. If the xenograft/transplant models are used and especially if tumors are transplanted dorsally under the skin, the size is easily measured and an accurate tumor growth curve can be generated by daily measurements of tumor width (W) and length (L) using calipers and calculating the volume (V) as V=0.52\*W2\*L. If the tumor is not

tumor cells transduced with the gene coding for VEGF grow much faster than the emptyvector transduced control cells, due to much higher intra-tumoral blood vessel density (Eriksson et al, 2002). In these tumors however, the vasculature consists of highly irregular and immature vessels and many vascular plexuses, which does not support efficient perfusion of blood compared to a less chaotic vasculature in VEGF non-transduced tumor tissue (Fig. 10). To assess angiogenesis in mouse models, there are thus three essential points

Fig. 10. Histological examinations of tumor tissue. (A) H&E staining of T241 fibrosarcoma.

microscopy. Anti-endomucin labeling of endothelial cells is shown in red and anti-α-smooth

1. Tumor size (volume). Tumor growth depends on angiogenesis. Thus, tumor size is usually proportional to the degree of which tumor-angiogenesis is induced by the tumor. If the xenograft/transplant models are used and especially if tumors are transplanted dorsally under the skin, the size is easily measured and an accurate tumor growth curve can be generated by daily measurements of tumor width (W) and length (L) using calipers and calculating the volume (V) as V=0.52\*W2\*L. If the tumor is not

(C) Confocal imaging of T241 fibrosarcoma. GFP transfected tumors are shown in green, anti-CD31 labeling of endothelial cells is shown in red. (D) Confocal imaging of VEGF overexpresing T241 fibrosarcoma. GFP transfected tumors are shown in green, anti-CD31

(B) Immunohistchemical staining of T241 fibrosarcoma visualized by fluorescence

muscle actin (SMA) labeling of pericytes/smooth muscle cells is shown in green.

labeling of endothelial cells is shown in red.

which should be addressed.

visible because it is growing in a location inside the animal, special imaging systems, such as CT, magnetic resonance imaging (MRI) or bioluminesence imaging analysis (if tumors express luciferase) are needed to estimate the tumor size. The effect of antiangiogenic drugs can be inferred from the reduction in tumor growth relative to vehicle treated controls.


### **3.3 Lymphangiogenesis and lymphatic metastasis**

It is well known that blood vessels can support tumor growth by providing oxygen and nutrients, and removing waste products, but the function of tumor lymphatic vessels remains poorly understood (Y Cao, 2008).

Lymphatic networks consist of lymphatic capillaries, collecting lymphatic vessels, and lymph nodes. Unlike blood vessels, lymphatic capillaries consist of one layer of lymphatic endothelial cells (LECs), discontinuous basement membrane and few vascular smooth muscle cells (VSMCs). Lymphatic vessels lack tight junctions between endothelial cells, but are instead equipped with one-way lymphatic valves which give these vessels the ability to collect fluids and macromolecules from the tissue and transport it back to the circulation. On the other hand, these features of the lymphatic endothelium also mean that malignant cells can easily enter into and disseminate via the lymphatic system, leading to lymphatic metastasis. Tumor lymphangiogenesis is therefore associated with cancer metastasis. In some common cancers, such as lung and breast cancer, lymphatic metastasis is the dominant route for tumor metastasis. Invasion of intra- or peri-tumoral lymphatics may result in dissemination of malignant cells to the lymphatic system, leading to lymphatic metastasis in regional lymph nodes (Y Cao Y, 2005; R Cao et al, 2004).

Similar to blood vessels, lymphatic vessels are quiescent in healthy individuals. The formation of lymphatic vessels in tumors is a multistep process that involves LEC proliferation, migration, tube formation and remodeling, which require up-regulation of lymphangiogenic stimulators and down-regulation of lymphangiogenic inhibitors (Y Cao, 2005). Lymphatic vessel growth may represent the imbalanced consequence between positive and negative regulators tipping toward positive regulation. Understanding the molecular mechanisms that control lymphangiogenesis is therefore an important step in the development of therapeutic agents in the prevention and treatment of cancer metastasis.

Among the list of lymphangiogenic factors, members of the VEGF family are the best characterized. VEGF-A, which binds to VEGFR-2 and VEGFR-1, PlGF and VEGF-B which bind to VEGFR-1 and especially VEGF-C and VEGF-D which bind to VEGFR-3 are lymphangiogenic factors, the -C and -D isoforms being the most potent, which regulate both physiological and pathological lymphangiogenesis. VEGF-C/VEGF-D-VEGFR-3-mediated signals are also critical for the sprouting of the first lymphatic vessel from the developing veins in the embryo. This signaling pathway is essential for differentiation of endothelial progenitor cells into the lymphatic lineage (Kukk et al, 1996; Alitalo et al, 2005).

Primary tumors produce several lymphangiogenic factors, including VEGF-A, VEGF-C, VEGF-D, Insulin-like growth factor (IGF), hepatocyte growth factor (HGF) and PDGF-B. These factors induce angiogenesis and lymphangiogenesis both in the local environment and in the regional lymph nodes. Furthermore, they play an important role in establishing the pre-metastatic niche. The term pre-metastatic niche describes the adaptations of for example the lymph nodes which are needed to allow disseminating tumor cells, arriving at a later stage, to meet optimal conditions for growth in that particular site. In addition to signals produced by the tumor cells themselves, inflammatory cells such as macrophages are recruited to tumors by a wide range of tumor cell-derived cytokines and growth factors. At the tumor site, inflammatory cells play a critical role in mediating lymphangiogenesis most likely through the secretion of several lymphangiogenic cytokines.

To detect lymphatic vessels in or around tumors, immunohistochemical staining with lymphatic specific markers, such as VEGFR-3, LYVE-1 and podoplanin is recommended (see Fig. 13). Lymphangiogenesis can be evaluated using all of the models described in this chapter, by using antibodies against one or a combination of these factors either in whole mount staining or on thin sections of frozen or paraffin embedded tissue (Fig. 13).
