A. Human Prostate Cancer
After lung cancer, cancer of the prostate (CaP) is the second most common cause of cancer death in American males. A latent disease, many men have prostate cancer cells long before overt signs of the disease are apparent. The annual incidence of CaP is over 100,000 in the United States, of which over 40,000 will die of the disease. Nearly a third of patients present with locally advanced or metastatic disease, and androgen deprivation therapy forms the basis of conventional therapy for the majority of these patients. However, currently available approaches for advanced CaP are not curative , primarily because the cells lose their dependence on androgenic stimulation. The mechanisms of progression of CaP cells to hormone independence under androgen ablation therapy remain unclear. To investigate the factors and mechanisms that underlie the development of androgen resistance and metastasis, reliable in vivo models that mimic human CaP progression are essential. Moreover, it is critical that tumor models mirror the pathology and cellular and molecular characteristics of human CaP if it is to serve as a useful tool for basic research, drug screening, or the evaluation of new therapeutic strategies.
B. Spontaneous and Transgenic Models of Human Prostate Cancer
Currently, a single animal model cannot epitomize the multifaceted aspects of CaP pathogenesis and progression. Rodent models of prostate carcinoma have been developed by hormone treatment , spontaneous development , transgenic prostate-specific oncogene expression , and knockout of CaP-tumor suppressor genes . However, these models are largely inadequate in recapitulating the progression of human disease as bone metastasis,  the major cause of clinical morbidity attributable to CaP. Despite pitfalls, the mouse transgenic TRAMP model has been useful for studying the development and progression of prostatic adenocarcinoma. TRAMP mice, generated by expressing SV40-T antigen specifically in prostatic epithelium , develop prostatic intraepithelial neoplasia (PIN) by 10-12 weeks of age and eventually progress to adenocarcinoma with metastasis to lymph nodes and lungs . As in human disease, androgen ablation therapy in these mice contributes to the emergence of androgen-independent disease with a poorly differentiated phenotype .
As for lung cancer, investigators have chosen a number to utilize xenograft models of CaP. Unfortunately, CaP xenografts are far more fastidious than lung cancer xenografts, and the generation of models that are representative of typical human disease has only recently been accomplished. Until recently, the majority of research conducted for CaP relied on the cell lines PC-3, DU145, and LNCaP. Among these, only LNCaP cells exhibit androgen responsiveness and express the prostate-specific antigen (PSA) and androgen receptor (AR). Thus, the relevance of DU-145 and PC-3 cells to clinical CaP has been questioned. To overcome the shortage of representative models of human CaP, a number of investigators began establishing xenografts in immune-deficient scid/scid mice using samples obtained directly from patients [145-149]. These xenografts offered the following advantages: (1) the expansion of small amounts of starting clinical material, (2) the enrichment of relatively homogeneous cell populations from heterogeneous tumor cell populations, (3) the ability to investigate progression to metastasis and androgen independence [145, 146, 148], and (4) representative diversity that provided a more realistic picture of the heterogeneous nature of this disease. Investigators at UCLA established six distinct CaP xenografts from patients with locally advanced or metastatic diseases into scid/scid mice. Two of these xenografts, LAPC-4 and LAPC-9, have been maintained continuously for more than 2 years by serial passage in scid/scid mice [145, 146], and LAPC-4 has also been successfully established as a cell line in tissue culture to enable correlation with investigations performed in vitro. . LAPC-4 and LAPC-9 offer several advantages over previous models; both express the wild-type androgen receptor (AR), both xenografts have intact AR-signal transduction pathways, and both secrete high levels of the androgen-dependent protein PSA. Accordingly, they grow as androgen-dependent cancers in male scid mice and respond to androgen ablation treatment, but interestingly, they eventually progress to a hormone-refractory, androgen-independent state [145, 146]. LAPC-4 and LAPC-9 can be implanted subcutaneously, orthotopically into the mouse prostate, or intratibially. Orthotopic tumors metastasize reproducibly to regional lymph nodes and lung, providing an opportunity to study prostate cancer metastasis. Intratibial injection results in the formation of osteoblastic tumors typical of human CaP where bony metastasis is the major cause of morbidity.
From a research standpoint, the generation of these xenografts has provided significant dividends. Given the inability to culture CaP by other means, the xenografts have been used to identify chromosomal abnormalities and to pinpoint the genes important to the pathogenesis of CaP. For example, loss of chromosome 10q was a frequently observed genetic defect in prostate cancer. Recently, the PTEN/MMAC tumor suppressor gene was identified and mapped to chromosome 10<j23.3 [150, 151]. PTEN encodes a protein/lipid phosphatase which has been clearly established to function as a negative regulator of the PI3-kinase/Akt signaling pathway [152-158]. Loss of PTEN leads to constitutive activation of PI3-kinase, and in turn the Akt-signaling pathway . PI3-kinase is also a downstream target of several growth factors implicated in CaP pathogenesis including epidermal growth factor receptor (EGFR), insulin-like growth factor receptor (IGFR) and Her2/neu, and it is possible that deregulation of this pathway in PTEN-deficient cells may indeed be responsible for the cancer phenotype. Of note, knockout mice lacking PTEN as a consequence of targeted deletion develop multiple cancers, including prostatic hyperplasia and prostatic intraepithelial neoplasia [141, 159], Correspondingly, 50-60% of all prostate cancer xenografts established contain deletions, mutations, or absent expression of PTEN [160,161], making the xenografts a relevant and valuable source for biological and therapeutic discovery. Prostate cancer gene therapy approaches that specifically target this pathway are now underway in these models.
In addition to modeling the abnormalities of the PTEN/MMAC pathway, xenografts are important in delineating the role of androgens and androgen receptor (AR) signaling in CaP. Prostate epithelial cells utilize androgen as a growth and differentiation factor and are dependent on androgen for survival. Once transformed, androgen deprivation is associated with a transition of CaP cells through a range of diminishing androgen-dependence, and ultimately androgen independence. Although not well understood, this process likely involves perturbations in AR signaling of cellular growth control. Potential AR-related perturbations may involve (1) AR mutation or gene amplification, (2) cross-talk between AR and other signal pathways, and/or (3) alterations in transcriptional coregulators. Greater than 80% of clinical CaP specimens have confirmed AR expression, even in advanced androgen-independent diseases [162, 163]. Among these, AR-gene mutation or amplification has been documented in 20-40% of CaP cases [164-166], Both LNCaP and the CWR22 xenografts bear AR mutations that enable the receptor to be activated by nonandrogenic steroid hormones such as progesterone and estrogen. In addition, in a patient who had failed androgen ablation, it was recently demonstrated that his CaP-cells possessed a mutated AR with altered lig-and affinity. Essentially, the mutant AR functioned as a high-affinity Cortisol receptor, enabling the CaP cells to circumvent the androgen requirement for growth . Another emergent theme is that some hormone refractory cancers have activated the AR signaling pathway through a ligand-independent mechanism. For example, in LAPC-4 cells expressing wild-type AR, the overexpression of Her-2/neu has been shown to activate AR . Not surprisingly, the LAPC-4 xenograft progresses to androgen-independence after androgen ablation and differential gene expression studies reveal a consistent increase in Her-2/neu protein expression in androgen-independent tumors. Furthermore, forced overexpression of Her-2/neu in androgen-dependent CaP cells is sufficient to confer androgen-independent growth in vitro and to accelerate androgen-independent growth in castrated animals. Thus, Her-2/neu overexpression activates the AR signaling pathway in the absence of ligand and enhances the magnitude of AR response in the presence of low levels of androgen. Last, reconstitution experiments in a heterologous cell type expressing low levels of endogenous AR suggest that these effects of Her-2/neu on the AR pathway require AR-expression . Although the point where Her-2/neu and AR pathway intersects is still undefined, nuclear receptor coactivators might be potential targets since amplification of steroid receptor coactivator, AIB1, is documented in breast and ovarian cancer . Cross-talk between Her-2/neu and AR signaling pathways should provide a novel mechanistic insight into the development of androgen independence.
D. Gene Therapy Approaches with Adenovectors in Prostate Cancer
Recombinant Ad vectors are most commonly used for CaP because they have demonstrated the capacity to deliver genes intraprostatically in animal models . Hence, several ongoing human CaP clinical gene therapy trials are using Ad [171, 172]. With respect to these applications, several groups are developing transcriptionally targeted prostate-specific Ad [172-175]. These strategies are beneficial in gene therapy applications in that they potentially restrict the expression of cytotoxic therapeutic genes to the malignant cells. Most commonly, the kallikrein-protease prostate specific antigen (PSA) gene regulatory regions have been used to direct prostate-specific expression because prostate epithelia, normal or malignant, specifically express the PSA . Unfortunately, the transcriptional output from the native PSA enhancer and promoter (as from most highly regulated tissue-specific promoters) is much lower than from strong constitutive viral promoters such as CMV. For example, our studies suggest that the native PSA enhancer and promoter inserted into Ad can direct tissue-specific and androgen-inducible expression in LNCaP cells, but the transcriptional activity is 50-fold lower than the constitutive CMV promoter [Wu et al., unpublished data].
By exploiting the known properties of the native PSA control regions, we have improved the activity and specificity of the prostate-specific PSA enhancer (Wu et al. unpublished data). Previous studies had established that AR molecules bound cooperatively to AREs in the PSA enhancer core (—4326 to —3935) act synergistically with AR bound to the proximal promoter to regulate transcriptional output [177, 178]. To exploit the synergistic nature of AR action, we generated chimeric enhancer constructs by (1) insertion of a synthetic element containing four tandem copies of the proximal PSA promoter AREI (ARE4) element or (2) duplication of enhancer core and (3) removal of intervening sequences (—3744 to —2875) between the enhancer and the promoter. Each of these three strategies augments activity and androgen inducibility and retained a high degree of tissue discriminatory ability. As a result of these combined approaches, the two most active constructs are termed PSE-BC (duplication of core) and PSE-BAC (insertion of core and ARE4) are approximately 20-fold higher in activity than native PSA enhancer/promoter construct, PSE, composed of the PSA enhancer (-5322 to —2855) fused to the proximal promoter (—541 to +12). Most importantly, the enhanced activity and specificity of the new PSA-enhancer/promoter constructs are retained in an adenoviral vector. The recently developed human CaP xenografts should be excellent models to refine and evaluate this novel prostate-targeted gene therapy because their AR pathways are intact and their growth regulatory pathways bear close resemblance to clinical disease.
Was this article helpful?