Animal Models – An Overview

Scott Winter and Lisette Maddison
Baylor College of Medicine

There are several different classes of animal models of prostate cancer including transgenic and knockout mice, canine and rat models and xenografts, . The general classes of models are summarized below and a few examples given for each class. The entire list of models is too extensive to list here and appropriate review articles are listed as sources for more in-depth information.

Transgenic models
A limited number of promoters including probasin and C3(1) are currently identified that direct expression of transgenes to prostate epithelial cells. The probasin promoter, including the minimal promoter, long promoter and composite promoter containing multiple androgen response elements, has been used most frequently to generate transgenic mice.

There are two general classes of transgenic models of prostate cancer. The first consists of models resulting from enforced expression of SV40 early genes. These models include the TRAMP model that utilizes the minimal rat probasin promoter to express the SV40 early genes (T and t antigens; Tag) as well as a number of transgenic lines using the long probasin promoter to express large T antigen, collectively termed the 'LADY' model. These two well characterized and widely distributed models display progressive disease from epithelial hyperplasia or PIN to adenocarcinoma and development of metastases. Also in this class of models are the C3(1)-Tag mice that also develop progressive prostate cancer but develop tumors in other tissues as well. Two additional models, Cryptdin-2-T and Gg-SV40 T, also develop progressive prostate cancer although the promoters used to drive SV40 large T antigen expression in these cases are not inherently prostate specific.

The second general class of transgenic mice utilizes the promoters mentioned above to express molecules of interest in the prostate. Generally, these molecules have previously been suggested to play a role in development of prostate cancer. The list is extensive but includes c-myc, Bcl-2 , Fgf8b and dominant negative TGFß. It is interesting to note that in the majority of these models, only a relatively mild phenotype, primarily epithelial hyperplasia or PIN, is observed and often these phenotypes do not arise until the mice are of advanced age.

Reviews
Matusik RJ, Masumori M, Thomas TZ, Case T, Paul M, Kasper S, Shappell SB. Transgenic mouse models of prostate cancer. In: Transgenics in Endocrinology, ed. by MM Matzuk, CW Brown, and TR Kumar. The Humana Press Inc (Totowa, NJ) Chapter 19, pp 401-425, 2001.

Shappell SB, Masumori M, Thomas TZ, Case T, Paul M, Kasper S, Matusik RJ. Transgenic mouse models of prostate carcinoma: Anatomic, Histopathologic, and molecular considerations. In: Prostate Cancer: Scientific and Clinical Aspects of Bridging the Gap, ed. by PD Abel and E-N Lalani. Imperial College Press (London) In press.

Knockout models
On the opposite pole of enforced transgene expression is the deletion of a gene of interest. A number of germline knockout mice have displayed prostate phenotypes. These include PTEN heterozygous knockout mice, Nkx3.1 null and p27 null mice. The loss of these genes has been suggested to play a role in the development of human prostate cancer , and this is supported by observations in the knockout mice. As with the transgenic models misexpressing single proteins of interest such as c-myc, Bcl-2, or TGFß, the resulting phenotypes in these mice consist primarily of hyperplasias that often develop in animals of advanced age. Data from both single transgenics and knockout mice suggests that expression or loss of a single gene is not sufficient to lead to carcinoma. An additional caveat to the use of germline knockout models is that all cells in the animal harbor the deletion and it is impossible to attribute the phenotype solely to prostate epithelial cells. Conditional genetic deletions will further clarify the contributions of these and other genes in development of prostate cancer.

Reviews
Sharma, P. & Schreiber, A. N.
Mouse models of prostate cancer.
Oncogene 18, 5349-5355. (1999)

Abate-Shen, C. & Shen, M. M.
Molecular genetics of prostate cancer.
Genes Dev 14, 2410-2434. (2000)

Huss, W. J., Maddison, L. A. & Greenberg, N. M.
Autochthonous mouse models for prostate cancer: past, present and future.
Semin Cancer Biol 11, 245-260. (2001)

Canine models
Dogs are the only species besides humans that develop spontaneous prostate cancer with high frequency. Interestingly, dogs also display a high frequency of metastasis, especially to bone. Compared to other models systems, the relatively large size of dogs facilitates their use in evaluating new imaging techniques and therapy regimens. The main limitations facing investigators are a long latency period related to the lifespan of canines and the high cost of maintaining dog colonies.

Review
Waters, D. J. et al.
Workgroup 4: spontaneous prostate carcinoma in dogs and nonhuman primates.
Prostate 36, 64-7. (1998)

Rat models
A number of rodent models have been used to study human prostate cancer. Several of these include the classic Lobund-Wistar, Dunning and Noble rat models that have been used extensively to study hormonal carcinogenesis. However, these models are limited by long tumor latencies, stochastic variability and lack of spontaneous metastases and have been largely supplanted by alternative mouse models.

Reviews
Bostwick, D., Ramnani, D. & Qian, J.
Prostatic intraepithelial neoplasia: animal models 2000.
Prostate 43, 286-294. (2000)

Lucia, M. S. et al.
Workgroup I: rodent models of prostate cancer.
Prostate 36, 49-55. (1998)

Xenografts and orthotopic models
Among the mouse models of prostate cancer are xenograft and orthotopic models. Human prostate cancer cells are injected either subcutaneously (xenograft) or intra-prostatically (orthotopic) in immune compromised mice. A number of prostate cancer cell lines have been used for this purpose including, but not limited to, CWR22, the LAPC series and LuCaP series of cell lines. These cell lines were derived from different sources ranging from primary tumors to distant site metastases. They represent a range of malignant potential and also display differential responses to androgens. These models give researchers the ability to follow tumor burden through monitoring PSA produced by human prostate epithelial cells. However, the cells used for these experiments are primarily derived from prostate cancers and thus are already transformed, hampering their use in the investigation of initiating events in prostate cancer.

Review
van Weerden, W. M. & Romijn, J. C.
Use of nude mouse xenograft models in prostate cancer research.
Prostate 43, 263-271. (2000)