3.1.1. Non-Cancerous Prostate Epithelial Cell Lines

3.1.2. Prostate Cancer Cell Lines

Prostate cancer cell lines established from human patients are broadly categorized into two types (castration-sensitive and castration-resistant) depending upon their survivability under androgen-deprived conditions.

3.2.1. TRAMP

The transgenic adenocarcinoma of the mouse prostate (TRAMP) mice model was generated and characterized in 1996. The chloramphenicol acetyltransferase (CAT) gene was introduced into the germ line of mice under the control of the rat probasin (PB) promoter. In TRAMP mice, expression of both the large and small SV40 T antigens (TAG) is regulated by the prostate-specific rat PB promoter [154]. The PB-SV40 T antigen (PB-Tag) transgene is spatially restricted to the dorsolateral and ventral lobes of the prostate. The gene expression is male specific and restricted to the epithelial cells of the lateral, dorsal, and ventral prostatic lobes of the murine prostate [155]. TRAMP is a very useful model for studying the pathology of PCa as the progression occurs through PIN lesions to malignant disease, like human disease, in a predictable time. Epithelial hyperplasia develops by 10 weeks of age, PIN by 18 weeks of age, and lymphatic metastases after 28 weeks of age [154,156,157].

The TRAMP model has been used for PCa prevention and treatment studies [158,159]. It is also the first genetically engineered mouse model (GEMM) that displays castration-resistant disease progression [160]. One of the limitations of the TRAMP model, however, is that these mice often develop neuroendocrine PCa [161]. A simultaneous loss of Rb and p53 could be the reason for the development of neuroendocrine cancer [161,162]. Considering the higher chances of neuroendocrine disease, the TRAMP mouse model is clinically more relevant to study PCa of neuroendocrine origin.

3.2.2. LADY

The LADY PCa mouse model was developed in 1998 and is similar to the TRAMP model [163]. There are, however, a few key differences between the TRAMP and LADY. In the LADY, a larger fragment (12 kb) of the PB (a.k.a. LPB) promoter upstream of the SV40 T-antigen is used that contains additional androgen and growth factor-responsive sequences and thus allows consistently high transgene expression. Additionally, the LPB promoter is linked with a deletion mutant of the SV40 T-antigen (deleted small T-antigen) to allow the expression of large T-antigen, unlike small t-antigen in the TRAMP model. The purpose of deleting small t-antigen was to analyze the importance of neuroendocrine differences in metastatic lesions developed by LADY [164]. LADY model mice develop metastases to the liver, lymph nodes, and bones [164]. The metastases, however, primarily contain neuroendocrine cells, which is unlike the human metastasis [135,165]. Thus, the LADY mice are different from the most common type of human PCa from the perspective of rapid tumor growth and neuroendocrine tumor development. Nevertheless, the LADY model possesses the molecular changes similar to the human prostate, such as the multifocal nature of tumorigenesis, histopathologically changes from low- to high-grade dysplasia similar to PIN in humans, and the androgen-dependent growth of the primary tumors. Hence, the LADY model could be beneficial for investigating the stepwise mechanisms of PCa progression as well as therapeutic intervention [163].

3.2.3. Pten Deficient Mice

Loss of the PTEN tumor suppressor is a critical event in PCa initiation, as discussed above. However, homozygous knockout of Pten in mice embryonic stem cells through the deletion of the phosphatase domain led to embryonic lethality [166,167]. To overcome this limitation, Wang et al. generated Pten null mice by conditional deletion of Pten in the murine prostatic epithelium. They generated Pten ^loxp/loxp: PB-Cre4 mice in order to attain the prostate-specific Pten biallelic deletion. They showed that Pten null PCa progressed with a short latency of PIN formation by 6 weeks of age compared to heterozygous Pten deletion mice, which developed PIN by 10 months. Moreover, homozygous Pten deletion mice developed invasive adenocarcinoma by 9 weeks of age and metastasis to the lymph node and lung by 12 weeks of age. The effect of hormone ablation therapy on Pten null mice was evaluated by performing the castration of mice at week 16. The response of Pten null tumors at day 3 and day 6 post-castration was analyzed. In response to androgen abolition, the AR-positive prostatic epithelium showed an increase in the apoptosis leading to the decrease of prostate volume. Hence, these homozygous Pten mutant mice recapitulate the PCa by mimicking the histopathological features of human disease [40]. In contrast, heterozygous mutant (Pten^+/−) mice developed neoplasia in multiple tissues, including mammary glands, lymphoid cells, small intestines, thyroid, endometrial, and adrenal glands [166,168,169], further limiting the applicability of the heterozygous mutant over Pten null mice.

The Pten knockout model has been used to demonstrate the role of the tumor microenvironment, particularly interleukin-17 (IL-17), in the growth and progression of PCa [170,171]. To test how tumor suppressor Rb interacts with Pten, Bai et al. developed mice with double mutations in both the cyclin-dependent kinase (CDK) inhibitor p18Ink4c and Pten [172]. The double mutant mice develop a broader spectrum of prostate tumors in the anterior and dorsolateral lobes at an accelerated rate [172]. Loss of function of Nkx3.1 is crucial for PCa progression and has been associated with the development of prostatic epithelial hyperplasia, dysplasia, and PIN [30,67,173]. Nkx3.1 and Pten are shown to cooperate in prostate carcinogenesis in mice. Nkx3.1;Pten double mutant mice demonstrated an increased incidence of HGPIN, which resembles the early stages of human PCa [69].

3.2.4. Pten^pc−/−Smad4^pc−/−

To examine a cooperative action of Pten and Smad4 loss in PCa pathogenesis, De Pinho lab developed mice having prostate-specific genetic ablation of Smad4 in Pten-null mice. These mice were highly aggressive and exhibited profound lymph node and pulmonary metastasis [45]. The importance of Smad4 in PCa was further revealed by the development of metastatic and lethal PCa with 100% penetrance in Smad4 and Pten double knockout mouse prostate [45]. Pten^pc−/−Smad4^pc−/− has been used to analyze the efficacy of hypoxia-prodrug TH-302 and checkpoint blockade combination therapy. The combination of the hypoxia-prodrug and checkpoint blockade significantly extended the survival of Pten^pc-/-Smad4^pc-/- mice [174]. Furthermore, Wang and colleagues utilized the Pten^pc−/−Smad4^pc−/− mice model and identified that polymorphonuclear myeloid-derived suppressor cells (MDSCs) are one of the significant infiltrating immune cells in PCa and their depletion blocks PCa progression [175].

3.2.5. Hi/Lo-Myc

Two plasmids having a rat probasin (PB) promoter alone (PB-Mycfor lo-Myc) and PB coupled with a sequence of the ARR2 (ARR2PB for hi-Myc) were used to achieve prostate-specific overexpression of c-Myc. The ARR2PB promotor contained two additional androgen response elements that forced the development of invasive adenocarcinoma from prostatic intraepithelial neoplasia (mPIN) in about 26 weeks [27,176,177]. Hi-Myc mice also displayed a decreased expression of Nkx3.1 at both mRNA and protein levels [27]. The PB-Myc mice showed similar pathological changes, but a slower progression of 30 weeks (time to invasive PCa development from PIN lesions) [27]. The main differences between these two models are their androgen responsiveness. The Hi-Myc is androgen-responsive, while the Lo-myc model displays no such sensitivity [27]. The mice model generated by non-viral oncogene ARR2PB-Myc and PB-Myc develop invasive adenocarcinoma and offer advantages over those expressing SV40. However, they do not develop metastasis, which is a major drawback of this model. Hubbard et al. in 2016 showed that the combination of Myc overexpression and Pten loss in mice resulted in the development of lethal prostatic adenocarcinoma with distant metastases [29]. Moreover, homeobox protein Hox-B13 (HOXB13) was suggested to participate in the MYC activation and Pten loss genomic instability and aggressive prostate cancer [29,178].

3.2.6. MPAKT

The mouse prostate Akt (MPAKT) model is useful in studying the role of protein kinase B (Akt) in the transformation of prostate epithelial cells and in developing the biomarkers relevant to human PCa. This mouse model was developed by the introduction of Akt1 along with a myristoylation sequence (myr) and a hemagglutinin (HA) epitope in the form of the linearized rPb-myr-HA-Akt1. This insert was injected into the pronuclei of fertilized oocytes, and the friend leukemia virus B (FVB) mice founders were verified [179]. These mice exhibited the formation of PIN by 8 weeks. Immunohistochemistry analysis of the PIN lesions of MPAKT demonstrated numerous important findings such as Akt results in the activation of p70S6K and is associated with the development of PIN in MPAKT mice and Akt-induced PIN might be linked to neovascularization. Histological evaluation revealed that MPAKT mice had distinct phenotypic characteristics, including disorganized epithelial layers, loss of cell polarity, intraepithelial lumen formation, and nuclear atypia and apoptotic bodies. However, the MPAKT did not develop invasive carcinoma even after 78 weeks [180].

3.3.1. Three-Dimensional (3-D) Organoid Cultures

The transition from monolayer PCa cultures to the three-dimensional (3-D) cultures is a remarkable breakthrough in cancer research. Although culturing cancer cell lines is cost-effective and easy to handle, established cell lines do not carry the heterogeneity and genetic makeup of tumors from which they were initially derived [181,182]. These limitations are mostly overridden by the establishment of 3-D organoid culture models from the patient-derived tumors [183]. Dong et al. established the first PCa 3-D organoid culture from the biopsy of a patient in 2014 [184]. This organoid culture maintained the molecular signature of PCa, including TMPRSS2-ERG fusion, SPOP mutation, Chromodomain Helicase DNA Binding Protein 1 (CHD1) loss, and serine protease inhibitor Kazal-type 1 (SPINK1) overexpression. Further, whole-exome sequencing revealed mutations in several other genes, as well as the loss of the p53 and RB tumor suppressor pathway function [184]. Puca and colleges developed patient-derived organoids from needle biopsies of metastatic lesions from patients with neuroendocrine CRPC. These organoids showed genomic, epigenomic, and transcriptomic association with corresponding patient tumors [185]. 3-D models are thus beneficial for drug discovery and preclinical evaluation of therapeutic drugs for efficacy under in vitro setting that mimics the complex in vivo environment.

3.3.2. Patient-Derived Xenografts (PDX)

Patient-derived xenografts (PDXs) are essential tools in cancer research as the results obtained from these resources more accurately predict clinical responses in patients (Table 2). The reason is that these models retain the genetic diversity of patient tumors and maintain a closely resembling tumor microenvironment [186]. PDX grown in immunocompromised mice carry essential histological and molecular features of the patient tumors, including gene expression programs, mutations, epigenetic regulators, and structural genomic events that ultimately drive their 3D growth [187,188]. Recent technical advancements, including the co-injection of PCa tissues with extracellular matrix (ECM) and transplantation into renal capsules, have increased the success rate of PDX establishment in mice [189,190,191]. The first androgen-dependent PCa xenograft model, designated as PC-82, was developed in 1977 by Schröder and colleagues at Erasmus University Rotterdam [192]. For this, the patient prostatic tumor tissue was grafted into the shoulder of nude mice. Later, two more androgen-independent in vivo models, designated as PC-133 and PC-135, were developed [192]. In 1996, seven other PDX models were established [193]. During 1991-2005, numerous other PDX models were developed that carried the TMPRSS-ERG rearrangement, RB1 loss, AR amplification, PTEN deletion, SPOP mutation, Tp53 deletion and mutation, and BRCA2 loss [132,194,195]. The success rate of the localized PDX model has been increased in recent years due to the implantation of the chimeric graft with neonatal mouse mesenchyme. This method improved the survival rate and doubled the proliferation index of xenografted cancer cells [196]. The PDX models, however, have two significant limitations, i.e., the absence of functional human immunity and the lack of orthotopic modeling in the mice [197]. Further, the model takes a long time (about 8 months) for validation of detectable tumor growth in mice that limits its utility for the high-throughput drug screening [198].

3.4.1. Rat Models

Rat is one of the models for PCa research that was first established in the year 1937 by Moore and Melchionna after injecting the white rat prostate with benzpyrene. Following treatment, the columnar prostate epithelium underwent squamous metaplasia and also led to the induction of cancer in both the healthy and atrophic prostates [199]. These tumors spontaneously developed from a dorsal prostatic adenocarcinoma in an inbred Copenhagen rat and then were transplanted into a syngenic Copenhagen × Fischer F1 hybrid rat. These rat prostate tumors are well differentiated and slow growing [200]. The albino Lobund–Wistar (LW) rat model was first described by Pollard [201]. The LW rat developed spontaneous tumors at a mean age of 26 months. Moreover, a combination of N-methyl-N-nitrosourea (MNU) and testosterone treatments induced the development of prostate adenocarcinoma in the LW rat at a mean time of 10.5 months. The cancer of the LW rat resembles the human PCa in several aspects, including spontaneous development and progression to androgen independence and metastasis [201]. However, a major limitation of the rat models is that they have a long latency period for tumor development (2–3 years), have low tumor incidence, and lack spontaneous metastases.

3.4.2. Zebrafish Model

The zebrafish model for cancer research has been utilized by many to acquire information that is traditionally obtained by mice and cell culture systems, although there are limited studies on zebrafish in an in vivo model for PCa research. The zebrafish model is suitable for visual observation of labeled tumor cells through the imaging technique since they are transparent. Nevertheless, the limitation of orthotopic transplantation could be the hurdle owing to the anatomical difference between zebrafish and the human body such as the breast, prostate, or lung [202]. The cancer cells can be injected into a different site in the zebrafish embryos, such as the blastodisc region, the yolk sac, the hindbrain ventricle, and into the circulation via the duct of Cuvier [203,204]. Melong et al. inoculated androgen-sensitive LNCaP cells into zebrafish and observed the effect of testosterone on the growth. Administration of exogenous testosterone increased the proliferation of PCa cells [205]. Further, the growth-promoting effect of testosterone was reversed by the anti-androgen receptor drug, enzalutamide. The invasive potential of PC3 cells overexpressing the calcitonin receptor (CTR) has also been evaluated in the zebrafish model [206]. The zebrafish model has several advantages, including the fact that zebrafish are small and can generate a large number of offspring in a short time, and they are easy to maintain and observe owing to their transparency. Moreover, humans and zebrafish have 71% protein similarity, and, most importantly, zebrafish absorb molecules from water providing an additional route for drug administration.