Prostate Cancer Model
Review Notes
Bone Met Models of Prostate Cancer
1. Natural history of prostate cancer
Prostate cancer is the most commonly diagnosed cancer in men at an estimated 24% of all new cancer cases. Even though the 5-year survival rate of localized disease is close to 100%, the 5-year survival rate of metastatic disease is still very low at 30%. In advanced prostate cancer, metastasis occurs in distant organs such as the brain, liver, lungs, lymph nodes, and bone. Particularly, 90% of patients with advanced prostate cancer present metastasis in the bone. Both localized and advanced prostate cancers do not respond to standard chemotherapeutic interventions and thus require alternative treatments. (is treatment of bone mets only palliative? If so, it can be said here) Metastasis of prostate cancer cells initially depends on the detachment of primary tumor cells and their migration to blood or lymphatic circulation. Not surprisingly, cell adhesion decreases as cancer cells begin to metastasize. …show more content…
Pre-metastatic cells undergo epithelial to mesenchymal transition (EMT) which is the biological process that defines these changes. This process includes the loss of epithelial markers such as E-cadherin, and the gain of mesenchymal markers such as N-cadherin. Patients with higher grade prostate cancers had higher N-cadherin expression than patients with lower grade prostate cancer. Similarly, changes in the binding between prostate cancer cells and the ECM are also implicated in increased metastatic potential. Reduced expression of α6β4 integrin results in weakened adhesion to the ECM and promote tumor cell migration. As prostate cancer becomes migratory, it must also acquire the necessary proteins to break down the ECM, chiefly proteins of the matrix metalloproteinase (MMP) family. Urokinase-type plasminogen activator is an important protein in the MMP cascade, and functions to initiate MMP action in prostate cancer cells. Migration is facilitated in part through increased signalling through focal adhesion kinases and Src family of kinases (SFKs). Intravasation of malignant prostate cancer cells is the end result of these processes. Tumour cells migrate through patient circulation until they attach to endothelial surfaces and extravasate into secondary tissue. This process is site specific, which is explained by the numerous local adhesion molecules and chemoattractants that help guide the tumour cells to a secondary site. Once at the secondary site, extravasation occurs through the generation of cellular structures called invadopodia, a structure that extends from the cells located within the vasculature into the secondary tissue. Invadopodia are comprised of an MMP, cortactin, Tks4 and Tks5 which are all necessary for extravasation to occur. When malignant prostate cancer cells colonize the bone, paracrine stimulation between tumour cells and bone marrow cells occur which stimulates both types of cells in a vicious cycle. Prostate cancer cells release numerous growth factors that stimulates the maturation of osteoblast and osteoclasts. These cells in turn, release growth factors that stimulate the growth of the prostate cancer cells. Metastasis of prostate cancer to the bone usually causes osteoblastic lesions to form as a result of an imbalance between osteoclast and osteoblast function. Proteins such as endothelin-1, bone morphogenic proteins and IGF all contribute to aberrant osteoblast activity. Progression of prostate cancer is also androgen-driven.
The androgen receptor on prostate cancer cells is a nuclear receptor that is essential for cell growth and survival. As well, AR signalling contributes to metastasis by facilitating EMT. Advanced prostate cancers are initially treated with hormone deprivation therapies to curb prostate cancer progression, however the efficacy is limited to 2-3 years before cancer progresses to castration resistant prostate cancer (CRPC). In CRPC, the androgen receptor signalling pathway has been reactivated and facilitates the continual expansion and metastasis of the tumour. This signalling axis can become dysregulated for a number of reasons including de novo androgen synthesis, overexpression of the AR, non-specific ligand recognition, or ligand-independent activation. As well, IL-6, Src and insulin-like growth factor pathways have been implicated in promoting androgen receptor-mediated gene expression in the absence of
androgens. Steps in pc progression leading to bone met
Pc invasion and mets – insights from genomic data
Molecular mech of met in pc
(WNT,SCR, NFKB, AR) – for mets?
(RANKL, AXII/AXIIR, TGF-β, IGFs, FGFs, BMPs, PDFGs, PSA, WNT, uPA, and endothelin I) - genes for bone mets?
(ZIP1/SLC39A1 loss) – loss of zinc to citrate function = more ATP (Zinc also inhibits NFKB) – happens in early PC
(P53 mutations happen closer to metastasis)
(PI3k/Akt, TGFB/SMAD, XIAP, MIC1, FAK)
2. Earliest models of mouse prostate cancer
Due to their ability to spontaneously develop disease, the first models of prostate cancer were developed using rats as the model organism. Initially, the Dunning rat model was the go-to model for prostate cancer, however early models were unable to present metastasis at all and the few refined models which could metastasize, failed to metastasize to bone. Canines also spontaneously develop prostate cancer, and they too develop bone metastases, however its incidence in dogs is rare and thus canines were inadequate to use as a model system. As well, canines lack androgen receptors and thus prostatic hyperplasia in dogs is always androgen insensitive which further discounts it as a good model system. Of the earliest prostate cancer models, the first model which could show tumour invasion into bone occurred when Pollard injected rat adenocarcinoma subcutaneously over a scratched cavalarium in late 1980s (This sentence is almost verbatim from a different paper since I cannot access this primary source). Though this method demonstrated prostate cancer cell growth within bone, it poorly represents most aspects related to the in vivo establishment of bone metastases. Around that time, a model demonstrating skeletal metastasis in an aythmice nude mice was developed by Shevrin et al. They injected PC3 prostate cancer cells into a mouse whose vena cava was temporarily occluded with a clamp at the time of the injection. Bone metastases were thought to have occurred as a result of the clamping which allowed for the newly injected cells to enter the veterbral venous plexus and divert flow to the lung capillary bed. This methodology was also used by Geldof and Rao who inoculated Copenhagen rats with R3327-Mat-LyLu prostate tumour cells. Until this point, rat prostate cancer models using this cell line were only able to demonstrate metastasis to the lymph nodes and lungs. Geldof and Rao showed that caval occlusion could also produce bone metastases in rats. Because these models success largely depended on surgical manipulation of the animal during the tumor injection, these models were not really that feasible to use in the study of prostate cancer bone metastasis. As well, the bone metastases found in the animals produced by this method did not follow the same patterns as human bone metastases (get primary source for more details). In 1991, Wang and Stearns took a different approach to develop a mouse model of bone metastasis. First, they selected for the most invasive progeny of a PC3 parent cell line by collecting the cells which could cross a Matrigel membrane. Then they inoculated SCID mice with these highly invasive cells through the mouse tail vein and harvested cells from any resulting metastatic lesions. These cells were then culture in vitro, and run through the Matrigel assay again to once more select for the most invasive cells. The results of this experiment were cell lines which could metastasize to SCID mouse bone tissue with greater than 80% efficacy. Creating bone metastases using this model did not require any surgical manipulation of these mice, however in order to use this model, the highly invasive cell line was required to be used within a very low passage number or else the cell line would lose its metastatic capacity. In an attempt to create a bone metastasis model that was less cumbersome than the vena caval occlusion models and reproducible without requiring maintenance of highly invasive cell lines, Haq et al. proposed an intracardiac injection model using Dunning rats and the R3327-Mat-LyLu cell line. They showed that intracardiac injections of this cell line produced spinal metastasis in 100% of the inoculated rats, and proved that serial inoculation and selection of invasive cells was not necessary in order to create a reproducible bone metastasis model. Intracardiac injections allow the tumour cells to by-pass the lung capillaries, just like the venal occlusion models, but in a much more manageable fashion. Though this technique is easier than the aforementioned models, it still requires a fair bit of technical skill to inject cells into the left ventricle of the animal heart. It is to-date, one of the most frequently used techniques performed to establish prostate cancer bone metastases. Another advantage of this inoculation approach is that the bone metastases formed after injection were more pathologically similar to the bone metastases that form naturally in humans (i.e. they initially colonize the bone marrow before invasion into the spinal canal).
(should I talk about LNCaP C4-2? – generated by culturing tumour cells of a mouse inoculated with LNCaP. 1994)
Concepts in prostatic cancer biology: Dunning R-3327 H, HI, and AT tumors.
3. Patient derived xenograft models
4. Direct models for forming bone mets
5. Transgenic models
In the late 1990s, some labs were working on developing transgenic mouse models for prostate cancer. Gingrich et al. established a mouse model whose epithelial cells expressed proteins from a probasin-SV40 T antigen transgene which allowed the mice to spontaneously develop prostate adenocarcinoma in as early as 18 weeks of age. This model was adequately named the TRAMP (transgenic adenocarcinoma mouse prostate) model. They had also showed that in this model, metastasis occurs frequently in the lungs and lymph nodes but only occasionally to the bone. Even though bone metastases are relatively infrequent, there are still advantages to a transgene model which cannot be ignored. Unlike models which required injection of foreign tumour cells, this model produces spontaneous autochtonous disease. Additionally, in all of the models created from injection, there is a requirement for immunodeficiency in order to ensure that the foreign cells are not immediately rejected. In a transgene model like TRAMP, the mice have a fully functional immune system and thus the immunobiology of tumorigenesis and metastasis can be better studied. One concern of many other transgenic models including TRAMP is that they primarily form a neuroendocrine differentiation of prostate cancer. This is a more severe form of disease however it is very uncommon in humans.
6. Summary
Bone Met Models of Prostate Cancer
1. Natural history of prostate cancer
Prostate cancer is the most commonly diagnosed cancer in men at an estimated 24% of all new cancer cases. Even though the 5-year survival rate of localized disease is close to 100%, the 5-year survival rate of metastatic disease is still very low at 30%. In advanced prostate cancer, metastasis occurs in distant organs such as the brain, liver, lungs, lymph nodes, and bone. Particularly, 90% of patients with advanced prostate cancer present metastasis in the bone. Both localized and advanced prostate cancers do not respond to standard chemotherapeutic interventions and thus require alternative treatments. (is treatment of bone mets only palliative? If so, it can be said here) Metastasis of prostate cancer cells initially depends on the detachment of primary tumor cells and their migration to blood or lymphatic circulation. Not surprisingly, cell adhesion decreases as cancer cells begin to metastasize. …show more content…
Pre-metastatic cells undergo epithelial to mesenchymal transition (EMT) which is the biological process that defines these changes. This process includes the loss of epithelial markers such as E-cadherin, and the gain of mesenchymal markers such as N-cadherin. Patients with higher grade prostate cancers had higher N-cadherin expression than patients with lower grade prostate cancer. Similarly, changes in the binding between prostate cancer cells and the ECM are also implicated in increased metastatic potential. Reduced expression of α6β4 integrin results in weakened adhesion to the ECM and promote tumor cell migration. As prostate cancer becomes migratory, it must also acquire the necessary proteins to break down the ECM, chiefly proteins of the matrix metalloproteinase (MMP) family. Urokinase-type plasminogen activator is an important protein in the MMP cascade, and functions to initiate MMP action in prostate cancer cells. Migration is facilitated in part through increased signalling through focal adhesion kinases and Src family of kinases (SFKs). Intravasation of malignant prostate cancer cells is the end result of these processes. Tumour cells migrate through patient circulation until they attach to endothelial surfaces and extravasate into secondary tissue. This process is site specific, which is explained by the numerous local adhesion molecules and chemoattractants that help guide the tumour cells to a secondary site. Once at the secondary site, extravasation occurs through the generation of cellular structures called invadopodia, a structure that extends from the cells located within the vasculature into the secondary tissue. Invadopodia are comprised of an MMP, cortactin, Tks4 and Tks5 which are all necessary for extravasation to occur. When malignant prostate cancer cells colonize the bone, paracrine stimulation between tumour cells and bone marrow cells occur which stimulates both types of cells in a vicious cycle. Prostate cancer cells release numerous growth factors that stimulates the maturation of osteoblast and osteoclasts. These cells in turn, release growth factors that stimulate the growth of the prostate cancer cells. Metastasis of prostate cancer to the bone usually causes osteoblastic lesions to form as a result of an imbalance between osteoclast and osteoblast function. Proteins such as endothelin-1, bone morphogenic proteins and IGF all contribute to aberrant osteoblast activity. Progression of prostate cancer is also androgen-driven.
The androgen receptor on prostate cancer cells is a nuclear receptor that is essential for cell growth and survival. As well, AR signalling contributes to metastasis by facilitating EMT. Advanced prostate cancers are initially treated with hormone deprivation therapies to curb prostate cancer progression, however the efficacy is limited to 2-3 years before cancer progresses to castration resistant prostate cancer (CRPC). In CRPC, the androgen receptor signalling pathway has been reactivated and facilitates the continual expansion and metastasis of the tumour. This signalling axis can become dysregulated for a number of reasons including de novo androgen synthesis, overexpression of the AR, non-specific ligand recognition, or ligand-independent activation. As well, IL-6, Src and insulin-like growth factor pathways have been implicated in promoting androgen receptor-mediated gene expression in the absence of
androgens. Steps in pc progression leading to bone met
Pc invasion and mets – insights from genomic data
Molecular mech of met in pc
(WNT,SCR, NFKB, AR) – for mets?
(RANKL, AXII/AXIIR, TGF-β, IGFs, FGFs, BMPs, PDFGs, PSA, WNT, uPA, and endothelin I) - genes for bone mets?
(ZIP1/SLC39A1 loss) – loss of zinc to citrate function = more ATP (Zinc also inhibits NFKB) – happens in early PC
(P53 mutations happen closer to metastasis)
(PI3k/Akt, TGFB/SMAD, XIAP, MIC1, FAK)
2. Earliest models of mouse prostate cancer
Due to their ability to spontaneously develop disease, the first models of prostate cancer were developed using rats as the model organism. Initially, the Dunning rat model was the go-to model for prostate cancer, however early models were unable to present metastasis at all and the few refined models which could metastasize, failed to metastasize to bone. Canines also spontaneously develop prostate cancer, and they too develop bone metastases, however its incidence in dogs is rare and thus canines were inadequate to use as a model system. As well, canines lack androgen receptors and thus prostatic hyperplasia in dogs is always androgen insensitive which further discounts it as a good model system. Of the earliest prostate cancer models, the first model which could show tumour invasion into bone occurred when Pollard injected rat adenocarcinoma subcutaneously over a scratched cavalarium in late 1980s (This sentence is almost verbatim from a different paper since I cannot access this primary source). Though this method demonstrated prostate cancer cell growth within bone, it poorly represents most aspects related to the in vivo establishment of bone metastases. Around that time, a model demonstrating skeletal metastasis in an aythmice nude mice was developed by Shevrin et al. They injected PC3 prostate cancer cells into a mouse whose vena cava was temporarily occluded with a clamp at the time of the injection. Bone metastases were thought to have occurred as a result of the clamping which allowed for the newly injected cells to enter the veterbral venous plexus and divert flow to the lung capillary bed. This methodology was also used by Geldof and Rao who inoculated Copenhagen rats with R3327-Mat-LyLu prostate tumour cells. Until this point, rat prostate cancer models using this cell line were only able to demonstrate metastasis to the lymph nodes and lungs. Geldof and Rao showed that caval occlusion could also produce bone metastases in rats. Because these models success largely depended on surgical manipulation of the animal during the tumor injection, these models were not really that feasible to use in the study of prostate cancer bone metastasis. As well, the bone metastases found in the animals produced by this method did not follow the same patterns as human bone metastases (get primary source for more details). In 1991, Wang and Stearns took a different approach to develop a mouse model of bone metastasis. First, they selected for the most invasive progeny of a PC3 parent cell line by collecting the cells which could cross a Matrigel membrane. Then they inoculated SCID mice with these highly invasive cells through the mouse tail vein and harvested cells from any resulting metastatic lesions. These cells were then culture in vitro, and run through the Matrigel assay again to once more select for the most invasive cells. The results of this experiment were cell lines which could metastasize to SCID mouse bone tissue with greater than 80% efficacy. Creating bone metastases using this model did not require any surgical manipulation of these mice, however in order to use this model, the highly invasive cell line was required to be used within a very low passage number or else the cell line would lose its metastatic capacity. In an attempt to create a bone metastasis model that was less cumbersome than the vena caval occlusion models and reproducible without requiring maintenance of highly invasive cell lines, Haq et al. proposed an intracardiac injection model using Dunning rats and the R3327-Mat-LyLu cell line. They showed that intracardiac injections of this cell line produced spinal metastasis in 100% of the inoculated rats, and proved that serial inoculation and selection of invasive cells was not necessary in order to create a reproducible bone metastasis model. Intracardiac injections allow the tumour cells to by-pass the lung capillaries, just like the venal occlusion models, but in a much more manageable fashion. Though this technique is easier than the aforementioned models, it still requires a fair bit of technical skill to inject cells into the left ventricle of the animal heart. It is to-date, one of the most frequently used techniques performed to establish prostate cancer bone metastases. Another advantage of this inoculation approach is that the bone metastases formed after injection were more pathologically similar to the bone metastases that form naturally in humans (i.e. they initially colonize the bone marrow before invasion into the spinal canal).
(should I talk about LNCaP C4-2? – generated by culturing tumour cells of a mouse inoculated with LNCaP. 1994)
Concepts in prostatic cancer biology: Dunning R-3327 H, HI, and AT tumors.
3. Patient derived xenograft models
4. Direct models for forming bone mets
5. Transgenic models
In the late 1990s, some labs were working on developing transgenic mouse models for prostate cancer. Gingrich et al. established a mouse model whose epithelial cells expressed proteins from a probasin-SV40 T antigen transgene which allowed the mice to spontaneously develop prostate adenocarcinoma in as early as 18 weeks of age. This model was adequately named the TRAMP (transgenic adenocarcinoma mouse prostate) model. They had also showed that in this model, metastasis occurs frequently in the lungs and lymph nodes but only occasionally to the bone. Even though bone metastases are relatively infrequent, there are still advantages to a transgene model which cannot be ignored. Unlike models which required injection of foreign tumour cells, this model produces spontaneous autochtonous disease. Additionally, in all of the models created from injection, there is a requirement for immunodeficiency in order to ensure that the foreign cells are not immediately rejected. In a transgene model like TRAMP, the mice have a fully functional immune system and thus the immunobiology of tumorigenesis and metastasis can be better studied. One concern of many other transgenic models including TRAMP is that they primarily form a neuroendocrine differentiation of prostate cancer. This is a more severe form of disease however it is very uncommon in humans.
6. Summary