Genetically Modified Organisms in Urological Cancer

Abstract

Genetically modified organisms (GMOs) have become indispensable tools in pre-clinical research for urological cancer. Through genetic engineering techniques, researchers can modify the genetic composition of organisms, enabling the creation of appropriate experimental animal models that provide a deep insight into the mechanisms of tumorigenesis, progression, and potential therapeutic strategies for urological cancer. In this chapter, we provide a comprehensive overview of the current status of research utilizing GMOs in the investigation of prostate cancer, renal cancer, urothelial cancer, and other urological cancers. Topics covered the development of different genetically modified animal models, and the application of these models in urological cancer research. In addition, the limitations of GMOs in cancer research will be discussed.

1. Introduction

Urological malignancies, including cancers of the prostate, bladder, kidney, and other organs of the urinary system, pose a rapidly increasing global cancer burden [1]. In 2020, Global Cancer Observatory (GLOBOCAN) data indicated that over 2.5 million individuals were diagnosed with malignant tumors of the urinary system, resulting in nearly 800,000 deaths [2]. Prostate cancer accounts for 29% of all new cancer diagnoses among men, as per the American Cancer Society, with bladder, kidney, and renal pelvis cancers also ranking in the top 10 for incidence. Among women, renal cancer is the 9th most common cancer [3]. Despite advances in diagnostics and treatments, the outlook for patients with advanced urological cancers remains bleak [4, 5, 6, 7, 8, 9], underscoring the critical need for more research into their pathogenesis and treatment options.

The use of genetically modified organisms (GMOs), through transgenic technology, has revolutionized biology and medical research [10, 11]. Genetically modified animal models, in particular, have become indispensable for studying intricate biological processes, including development, aging, and disease progression [12, 13, 14, 15]. These models stand out from other animal models by more accurately mimicking the genetic and pathological aspects of human urological cancers. Notably, some of these models can spontaneously develop metastatic disease, facilitating study of the microenvironment at various stages of tumor metastasis. They also allow for the assessment of therapeutic effects and the validation of drug targets across different stages of the disease [16, 17, 18].

This chapter provides a comprehensive review of the utility of GMOs, particularly genetically modified animal models, in uncovering the pathogenesis of urological cancers and identifying new therapeutic strategies or drug targets.

2. GMOs as essential models in urological cancer research

2.1 Prostate cancer (PCa)

2.1.1 TRAMP model

The transgenic adenocarcinoma of the mouse prostate (TRAMP) model, first reported by N. M. Greenberg in 1995 [19], employs transgenic technology to introduce the androgen-dependent, prostate tissue-specific protein rat probasin (rPB) into fertilized mouse eggs. This process inactivates tumor suppressors p53, retinoblastoma 1 (Rb1), and protein phosphatase 2A (PP2A) in the prostate, leading to PCa development in mice [20, 21]. Characterized by its clinicopathological features and widespread metastasis, the TRAMP model is considered the most detailed PCa model available. By 8 weeks, TRAMP mice show abundant Simian Virus 40 (SV40) large T-antigen markers in the dorsolateral lobes of the prostate, progressing from low-grade to high-grade prostatic intraepithelial neoplasia by 10 weeks. At 12 weeks, distant metastasis begins, with all mice developing metastatic PCa by 28 weeks—a 100% tumor formation rate [22, 23]. Metastases occur in various locations, predominantly in the dorsolateral lobe of the prostate [24]. Due to its extensive metastatic spread [25, 26], the TRAMP model is pivotal for studying PCa’s pathogenesis and metastatic behavior [27].

2.1.2 LADY model

The LPB promoter driving the large T-antigen (LADY) model, using the large PB promoter (LPB) to express SV40 Large T-antigen (Tag) without the small t-antigen, contrasts with the TRAMP model to highlight neuroendocrine differences in metastatic lesions [28, 29]. It demonstrates progression from prostatic intraepithelial neoplasia (PIN) with neuroendocrine features to invasive neuroendocrine carcinoma [30]. Metastatic lesions appear at one site by 6 months and at multiple sites by 9–10 months, yet only 14% develop bone metastases. The model typically shows metastasis to lymph nodes, liver, lung, spleen, kidney, and occasionally bones [29, 30] but exhibits limited PCa cell proliferation and less invasiveness compared to TRAMP. It serves as an effective model for studying early PIN changes and tumor stroma [31, 32], making it suitable for exploring the mechanisms of neuroendocrine PCa (NEPC) [28].

2.1.3 Other SV40 T-antigen

Beyond the commonly used TRAMP and LADY models, other transgenic models have been developed using SV40 large T-antigen and small t-antigen, combined with various promoters. Jeffrey E. Green et al. introduced a C3(1) initiator to express SV40 large T-antigen, creating the first PCa genetically modified animal model [33, 34]. This model causes female mice to develop breast cancer at 6 months and male mice to develop prostate hyperplasia by 8 months, which can progress to PIN and PCa, occasionally metastasizing to the lung and bone [35]. This genetically modified animal model facilitates the study of hormone response components in vivo, as well as the multistage progression from normal tissue to PCa [36, 37].

Jeffrey I. Gordon et al. developed a model for metastatic PCa using the Cryptidin-2 gene promoter to direct expression of SV40 large T-antigen in prostate neuroendocrine cells, leading to neuroendocrine PCa [38]. The model mice developed PIN at 8 weeks, followed by invasive cancer within 2–4 weeks and metastasis to lymph nodes, liver, lung, brain, and bone within 16 weeks [39]. The tumors were androgen-independent, recapitulating multiple histopathological features of human PCa while avoiding androgen dependence on neuroendocrine cancer evolution [40, 41]. Compared to the TRAMP model, this model has a shorter tumor induction period and provides new biomarkers for clinical diagnosis of PCa neuroendocrine differentiation.

Jim W. Xuan et al. utilized the knock-in technique, using the prostate secretory protein 94 amino acid (PSP94) gene promoter/enhancer to target and regulate mouse prostate-specific expression of SV40 large T/small t-antigen, establishing a new PCa model, prostate secretory protein-knock-in mouse adenocarcinoma prostate (PSP-KIMAP) [42, 43]. This model recapitulates aspects of human PCa, with PIN observed around 10 weeks, progressing to invasive adenocarcinoma by 24 weeks and accompanied by distant metastases to lymph nodes, lungs, and liver [40]. Compared to TRAMP and LADY models, PSP-KIMAP more accurately simulates human PCa development, exhibiting more stable phenotypes and precise prostate tissue targeting [30, 44]. While the PSP-KIMAP study is not comprehensive, it is worth acknowledging for replicating PCa and serving as a valuable addition to other models, holding broad application prospects.

2.1.4 Pten

Phosphatase and tensin homolog deleted on chromosome ten (Pten) gene, a significant tumor suppressor, is often deficient in various human cancers [45]. In human PCa, Pten deletion is found in 23% of high-grade PIN (HGPIN), 69% of localized cases [46], and 86% of metastatic castration-resistant PCa (CRPC) [47]. This deficiency underscores Pten’s crucial role in PCa initiation, leading to the development of Pten-deficient mouse models. Shun you Wang et al. created a model with prostate-specific Pten deletion, revealing that heterozygous mice developed PIN at 10 months, while homozygous deletion resulted in invasive adenocarcinoma by 9 weeks, with metastasis by 12 weeks [48]. However, tumors also appeared in non-prostate tissues [49, 50, 51], which could limit the model’s specificity for PCa research.

Researchers have explored the role of interleukin-17 (IL-17) in PCa using these models, shedding light on the tumor microenvironment [52, 53]. They also investigated the interaction between tumor suppressors Rb and Pten using double mutant mice with cyclin-dependent kinase (Cdk) inhibitor p18Ink4c and Pten knockout [54]. Results show that double mutant mice have faster and more extensive tumor growth in the prostate anterior and dorsolateral lobes [54]. Additionally, loss of Nkx3.1, combined with Pten deficiency, significantly increases HGPIN incidence, mirroring early stage human PCa [55, 56]. Similarly, DePinho et al. found that prostate-specific deletion of Pten (Pten ^pc−/−) and prostate-specific deletion of Smad4 (Smad4 ^pc−/−) in mouse models exhibited highly invasive characteristics with deep lymphatic and pulmonary metastases [57]. Based on these findings, the Pten ^pc−/−/Smad4 ^pc−/− mouse model was constructed and applied to research on combinations of hypoxia-activated prodrug TH-302 and checkpoint blockade [58]. They found that the combination of TH-302 and checkpoint blockade significantly increased the survival of Pten ^pc−/−/Smad4 ^pc−/− mice [58]. Moreover, Xin Lu and colleagues also utilized this model to reveal potential pathways for improving immunotherapy in advanced prostate cancer through Pygo2 (Pygopus2 (Pygopus family plant homeo domain (PHD) finger 2))-targeted treatment [59]. Additionally, a model with Pten deletion and speckle-type pox virus and zinc finger protein (Spop) mutation demonstrated the role of Spop mutations in activating phosphatidylinositol-3-kinase/mammalian target of rapamycin (PI3K/mTOR) and androgen receptor (AR) signaling, advancing the understanding of PCa progression [60]. The Pten/Kras (Kirsten rat sarcoma viral oncogene homolog) model reported by David J. Mulholland et al. was found to accelerate prostate cancer progression due to Pten deficiency, with concomitant epithelial-mesenchymal transition (EMT) and extensive metastasis [61].

2.1.5 Myc

Previous studies have suggested that upregulation of Myc may be a critical driving event in the onset and progression of human PCa [62, 63]. Charles L. Sawyers and his team created transgenic mice expressing human c-Myc in the prostate [64], categorized into Hi-Myc and Lo-Myc groups based on androgen sensitivity, with Hi-Myc being androgen-sensitive. The Hi-Myc mice showed accelerated PIN progression compared to the Lo-Myc group [64]. Despite its advantages, the model’s lack of transferability is a significant limitation [64]. Interestingly, concurrent Myc overexpression and Pten deficiency in mice resulted in aggressive adenocarcinomas with distant metastases. The research on this animal model has confirmed that Homeobox protein Hox-B13 (Hoxb13) played a pivotal role in the causation of prostate cancer [65, 66]. Combining Hi-Myc mice with PB-Hepsin mice reduced adenocarcinoma progression from 24 to 12 weeks [67]. However, studies show that mTOR inhibitors are ineffective against Myc overexpression mice, suggesting that they may be contraindicated in Myc overexpression PCa patients [68]. Given Myc’s early amplification in PCa [62, 69], the Myc transgenic model serves as a foundational tool for studying genetic alterations in PCa progression, often used in conjunction with other PCa-related genes.

N-Myc and L-Myc, members of the Myc family, play roles in PCa development and progression. Mycl amplification is observed in precancerous lesions and early tumors [70], whereas N-Myc is associated with aggressive castration-resistant PCa (CRPC) and NEPC [71]. Etienne Dardenne et al. established a transgenic mouse model overexpressing N-Myc and identified it as an oncogenic driver of NEPC [72]. This model suggests that AR signaling is abolished and Polycomb Repressive Complex 2 (Ezh2) signaling is induced [72]. The model helps identify and validate potential therapeutic targets for treating NEPC, such as Ezh2 and Aurora A kinase.

2.1.6 Cdcp1

Abdullah Alajati et al. developed a mouse model of prostate cancer (PCa) that overexpresses CUB domain-containing protein 1 (Cdcp1, 73]. In this model, 50% of mice exhibited prostate hyperplasia within 4–6 months, which progressed to PIN by 7–9 months and to HGPIN by 14 months [73]. The simultaneous overexpression of Cdcp1 and knockdown of Pten markedly accelerated the development of metastatic PCa in these mice. Utilizing this model, the researchers identified promising drug treatment strategies for combating metastatic PCa [73].

2.2 Renal cancer

Research highlights the critical tumor-suppressing role of von Hippel-Lindau (Vhl) in clear cell renal cell carcinoma (ccRCC) development [74], notably at tumorigenesis’s initial stages [75]. While modern sequencing technologies corroborate this [76], studies indicate that Vhl loss alone does not suffice to induce renal cell carcinoma (RCC) [77, 78, 79, 80, 81]. This insight has led researchers to combine Vhl deletion with other oncogenic modifications, aiming to develop an optimal renal cancer model. Following Vhl, Polybromo 1 (Pbrm1) emerges as the second key suppressor gene in renal cancer, with about 40% of ccRCC cases involving Pbrm1 mutations [82, 83, 84]. Interestingly, single knockdown of Pbrm1 by some researchers through knockout technology did not result in the expected renal cancer model. However, discoveries of ccRCC with combined mutations in Vhl and Pbrm1 [83, 85], and metastatic ccRCC with mutations in both Brca1-associated protein 1 (Bap1) and Pbrm1 [85, 86], suggest ccRCC may require simultaneous knockdown of two or more genes.

2.2.1 Vhl/Pbrm1

Amrita M Nargund and colleagues have developed a ccRCC mouse model by knocking out both Vhl and Pbrm1 genes via the Ksp-Cre method [87]. This model not only led to renal cyst formation but also successfully induced ccRCC. Mice deficient in both Vhl and Pbrm1 showed a 30% incidence of preneoplastic polycystic kidney diseases by 6–9 months and a 50% ccRCC incidence by 10 months [87]. Similarly, Yi Feng Gu’s team created a double knockout model using Paired Box 8 (Pax8)-Cre, resulting in about 85% of mice developing extensive tumors by 9 months, escalating to 100% by 13 months [88], with tumors eventually nearly replacing the kidneys by 16 months [88]. The Vhl/Pbrm1 model serves as a valuable tool for exploring the molecular dynamics of Vhl and Pbrm1 mutations and for conducting drug efficacy tests.

2.2.2 Vhl/Bap1

Subsequent research identified the Bap1 gene as significantly associated with ccRCC, ranking it as the third most important ccRCC-associated gene with a mutation rate of approximately 15% [83, 85, 89]. Bap1 mutations are linked to higher-grade ccRCC, contrasting with the lower-grade associations of Pbrm1 mutations. Shan Shan Wang et al. initially attempted a simultaneous Vhl/Bap1 deletion using Six2-Cre, but mice with Bap1 deficiency died within a month of birth [89]. Later, Yi Feng Gu et al. successfully created a Vhl/Bap1 deletion model with Pax8-Cre, which, alongside the Vhl-Pbrm1 model, elucidates the role of these genes in ccRCC formation and tumor grading. Bap1-deficient tumors tend to be of higher-grade compared to Pbrm1-deficient tumors [88, 90].

2.2.3 Vhl/Trp53/Rb1

Sabine Harlander and colleagues developed a new ccRCC mouse model by deleting Vhl, transformation-related protein 53 (Trp53), and Rb1 in renal epithelial cells [91]. This model provides a basis for studying hypoxia-inducible factor-alpha (HIF-α) inhibition as a potential ccRCC treatment and offers a platform for drug screening and testing new therapies [91]. However, such triple gene inactivation is rare in human ccRCC, and the model does not exhibit metastases to lung, liver, bone, or brain, highlighting its limitations.

2.2.4 Myc/Vhl/Cdkn2a

Sean T. Bailey and his team created a metastatic renal cancer model by overexpressing Myc and deleting Vhl and cyclin-dependent kinase inhibitor 2A (Cdkn2a), mimicking human ccRCC [92]. Remarkably, about one-third of the mice developed liver metastases [92]. However, the combination of Vhl and Cdkn2a loss, along with Myc activation, is relatively rare in humans, limiting the model’s representativeness [92].

2.2.5 Flcn

The folliculin (Flcn) gene, identified from a Birt-Hogg-Dube syndrome (BHD) patient [93, 94], plays a role in causing BHD syndrome, which includes RCC among other diseases. Initially established whole-body Flcn knockout homozygote mice died at the embryonic stage, whereas heterozygous mice had a very late onset of disease with a low and erratic morbidity rate [95, 96, 97]. Using Ksp-Cre, researchers established an Flcn-deficient model that successfully mimics RCC [98, 99, 100], showing 100% renal cyst incidence and about 70% tumor incidence within 6–7 months [100]. Thus, the Flcn-deficient model can be utilized for molecular mechanism studies and drug testing and screening in renal cancer. Unfortunately, there is a diversity of histologic subtypes of RCC observed in this model, and it is not yet possible to determine whether there are other factors that determine the histologic type of RCC in this model, a question that remains to be investigated.

2.2.6 Other renal cancer models

Zachary S. Morris et al. developed a genetic mouse model with neurofibromatosis type 2 (Nf2) knockout using Villin-Cre [101], observing renal cell hyperplasia and cysts at 15 days, and small renal tumors by 3 months that progressed to invasive renal cancer by 6 to 10 months. This model also exhibited elevated epidermal growth factor receptor (Egfr) expression, indicating that Nf2 inactivation activates the Egfr signaling pathway, promoting renal cancer development [101]. Although Nf2 is not primarily associated with renal cancer, this model presents an early onset opportunity for studying Nf2-related renal cancer pathogenesis and drug screening.

Lorraine J. Gudas et al. reported a mouse model for Vhl renal cancer known as the TRAnsgenic model of Cancer of the Kidney (TRACK). This model is characterized by the specific expression of mutant hypoxia-inducible factor 1-alpha (Hif1a) in the renal proximal tubule cells, closely mirroring many early stage features of ccRCC [102].

Alessia Calcagnì et al. created the first genetic animal model of renal cancer with transcription factor EB (Tfeb) overexpression using cadherin 16 (CDH16)-Cre [103]. This model sheds light on the mechanism of TFE-fusion RCC and suggests a therapeutic approach targeting the Wingless/integrated (WNT) pathway [103]. However, the model mice developed severe renal impairment, an outcome not commonly seen in humans, which may limit the model’s utility for therapeutic exploration.

Qiang Hua Hu et al. produced a Wilms tumor (WT) mouse model by increasing Igf2 expression and eliminating WT1, making it the only WT transgenic mouse model closely resembling human tumors [104]. Visible tumors appeared at 9 weeks, providing a valuable tool for testing therapeutic strategies [105].

2.3 Urothelial cancer (UC)

The uroepithelium, one of the body’s slowest-renewing epithelia, undergoes unique biological transformations. Oncogene inactivation or activation plays a crucial role in bladder tumorigenesis [106]. With gradual clarification of UC-related molecular mechanisms and maturing molecular biology techniques, GMOs have been successfully established, better reproducing human UC biological behavior at the molecular level. GMOs are widely used to study specific gene functions of Hras (Harvey rat sarcoma viral oncogene homolog), P53, Pten, Rb, fibroblast growth factor receptor 3 (Fgfr3), and epidermal growth factor receptor (Egfr) in bladder cancer development [107].

2.3.1 SV40 T-antigen

Zhong Ting Zhang et al. developed a transgenic mouse model in 1999 using the uroplakin II (UpII) promoter to express SV40 T-antigen urothelially [108]. This model showed that both low and high copy number SV40T transgenic mice develop bladder carcinoma in situ (CIS), with high copy number mice progressing to invasive and metastatic transitional cell carcinoma (TCC) [108]. Francisco Ayala de la Peña et al. later created a model using UPK II-Cre that did not progress to invasive tumors [109].

2.3.2 Hras

Hras, the first oncogene identified in human UC [110], led Zhong Ting Zhang et al. in 2001 to establish a mouse model activating Hras in the uroepithelium. This resulted in urothelial hyperplasia and superficial papillary non-invasive bladder tumors [111]. They found that low copy number Hras transgenic mice developed non-invasive lesions, whereas high copy number mice died by 5 months, highlighting the RAS pathway’s role in developing low-grade, non-invasive papillary UC [111].

2.3.3 P53

P53, crucial for uroepithelial cell growth control [112], is frequently mutated or deleted in human UC [113]. Jing Gao and his team have demonstrated that the complete loss of p53 is a prerequisite for the activation of Hras to promote the generation of UC via a model constructed by hybridizing active Hras transgenic mice with p53-deficient mice [114]. In addition, Anna M. Puzio-Kuter et al. have shown that combined deletion of p53 and Pten in bladder epithelial cells leads to invasive cancer in a novel mouse model that provides a validating tool to study mTOR inhibitors for the treatment of invasive UC [115].

2.3.4 Pten

Chao Nan Qian and colleagues created a model by specifically knocking out Pten in kidney epithelial cells using Ksp-Cre [116]. About 57% of these mice developed UC of the renal pelvis by 12 months, with significantly increased phosphorylated mTOR levels, suggesting mTOR inhibitors as effective treatments [116].

2.4 GMOs of other urological cancers

Testicular germ cell tumors (TGCTs), the predominant form of testicular cancers among young men, originate from germ cells [117]. James A. Gill et al. developed TGCT in zebrafish by expressing SV40 large T-antigen under the pufferfish lymphocyte-specific protein tyrosine kinase (Flck) promoter, observing TGCT development after a latency of up to one year. Additionally, overexpression of the stem cell leukemia (Scl) gene in zebrafish testis also led to TGCT [118]. These findings demonstrate the viability of studying TGCT in vivo. Earlier mouse models were limited to benign tumors, not reflecting the malignant TGCT more commonly seen in young men [119, 120, 121]. Amy M. Lyndaker et al. constructed a mouse model of malignant TGCT by localizing Kras activation and Pten inactivation in mouse premeiotic germ cells [122, 123].

Penile cancer, a rare urinary system malignancy, is often preceded by penile intraepithelial neoplasia (PeIN), a precancerous lesion linked to human papillomavirus (HPV) [124, 125]. Beatriz Medeiros Fonseca and colleagues introduced the first mouse model mimicking HPV-related penile cancer [126]. This model was created by treating 10-week-old HPV16 transgenic mice with dimethylbenz(o)anthracene (DMBA) over a 16-week period, leading to the development of HPV-associated penile cancer traits similar to those observed in humans, such as condylomas and PeIN [126]. Their pioneering research has paved the way for further studies into the underlying mechanisms and potential treatment options for penile cancer.

3. Challenges and limitations of GMOs

Advances in genome editing, such as clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9 (CRISPR/Cas9) [127], face challenges like off-target effects and exogenous mutations, complicating the maintenance of transgenic models and impacting research. The complexity of gene regulation in urological cancer adds to the difficulty of creating accurate models. These challenges hinder the broader application of transgenic models in urological cancer research. Ethical and safety considerations of GMOs also demand careful attention.

4. Potential development trends of GMOs

Improvements in CRISPR/Cas9 and the advent of new editing techniques like the National Institute for Cancer Epidemiology and Research (NICER) and AsCas12f promise enhanced transgenic model efficiency and accuracy [128, 129]. Future urologic oncology research may benefit from interdisciplinary collaboration and integrating technologies like three-dimensional (3D) printing and artificial intelligence (AI), potentially revolutionizing our understanding of urological tumors.

5. Conclusions

In conclusion, these GMOs have greatly aided understanding of urological cancers’ etiology, progression, and metastasis, and are critical for testing novel drug targets, and assessing treatment responses (Table 1). They promote urological cancer research and allow researchers to choose suitable models for deep exploration in their fields.

Cancer type	Model	Time of penetrance, Pathology	Characteristics	References
Prostate cancer	TRAMP	10 weeks: LGPIN to HGPIN; 12 weeks: distant metastasis begins; 28 weeks: 100% metastatic tumor formation	First autochthonous mouse model of PCa	[19]
	LADY	6 months: metastatic tumor at one site; 9–10 months: metastatic tumor at multiple sites (14% bone)	Less invasiveness	[28]
	C3(1)	8 months: prostate hyperplasia even PIN and PCa	Hormone response components in vivo	[33]
	Cryptidin-2	8 weeks: PIN; 10–12 weeks: invasive cancer; Within 16 weeks: metastasis to lymph nodes, liver, lung, brain, and bone	Shorter tumor induction period	[38]
	PSP-KIMAP	10 weeks: PIN; 24 weeks: invasive AD; accompanied by distant metastases to lymph nodes, lungs, and liver	More accurately simulates human PCa	[42]
	Pten	Pten^+/−; 10 months: PIN; Pten^−/−; 9 weeks: invasive AD; 12 weeks: metastasis	Tumors also appeared in non-prostate tissues	[48]
	Pten/p18Ink4c	9 months: Pten^+/− /p18^−/−; HGPIN; 12 months: Pten^+/− /p18^+/−; HGPIN or carcinoma	Faster and more extensive PCa growth	[54]
	Pten/Nkx3.1	26–52 weeks: 60% HGPIN; 52 weeks: 100% HGPIN; >52 weeks: 84% AD, 25% lymph node metastasis	Mirror early stage human PCa	[55, 130]
	Pten/Smad4	7 weeks: LGPIN; 11 weeks: invasive PCa; 15 weeks: highly aggressive PCa; 32 weeks: 100% lymph node, 12% lung metastasis	Highly invasive	[57]
	Ptne/Kras	10 weeks: PIN; 20 weeks: AD; 40 weeks: Death	EMT and stem-like features	[61]
	Hi-Myc	>13 weeks: PIN; >26 weeks: AD	Accelerated PIN progression	[64]
	Hi-Myc/PB-Hepsin	4.5 months: invasive AD	Higher-grade AD	[67]
	N-Myc/Pten	NEPC	Abolish AR signaling	[72]
	Cdcp1	4–6 months: prostate hyperplasia; 7–9 months: PIN; 14 months: HGPIN	SRC/MAPK pathway activation	[73]
Renal cancer	Vhl/Pbrm1	Ksp-Cre; 6–9 months: 30% preneoplastic polycystic kidney; 10 months: 50% ccRCC; Pax8-Cre; 9 months: 85% ccRCC; 13 months: 100% ccRCC; 13 months: tumor nearly replace kidney	Valuable tool for ccRCC research	[87]
	Vhl/Bap1	Six2-Cre: die within a month of birth; Pax-8cre: different grades of ccRCC	Higher-grade tumors	[88]
	Vhl/Trp53/Rb1	10 of 25 develop a total of 64 tumors	No metastasis	[91]
	Myc/Vhl/Cdkn2a	one-third of the mice developed liver metastases	Rare in human ccRCC	[92]
	Flcn	6–7 months: 100% renal cyst, 70% tumor	Diversity of histologic subtypes	[100]
	Nf2	15 days: renal cell hyperplasia and cysts; 6–10 months: invasive renal cancer	Nf2-related renal cancer	[101]
	TRACK	ccRCC	Hif1a activation	[102]
	Tfeb	TFE-fusion RCC	First genetic animal; models of RCC	[103]
	Igf2	WT	The only WT transgenic mouse model	[104]
Urothelial cancer	SV40 T-Antigen	low copy number: CIS; high copy number: metastatic TCC	Develop bladder CIS and TCC	[108]
	Hras	low copy number: urothelial hyperplasia, superficial papillary non-invasive bladder tumors	RAS pathway activation	[111]
	P53/Pten	Invasive cancer	NA	[115]
	Pten	12 months: 57% UC	Develop UC of the renal pelvis	[116]
Testicular germ cell tumors	SV40 T-antigen	>1 year: TGCT	Develop TGCT in zebrafish	[118]
	Scl	TGCT	NA	[118]
	Kras/Pten	TGCT	NA	[122]
Penile cancer	HPV16	16 weeks: condylomas and PeIN	First mouse model mimicking HPV-related penile cancer	[126]

Table 1.

Commonly GMOs in urological cancers.

Abbreviations: PIN prostatic intraepithelial neoplasia; PCa prostate cancer; NEPC: neuroendocrine prostate cancer; LGPIN: low-grade prostatic intraepithelial neoplasia; HGPIN: high-grade prostatic intraepithelial neoplasia; AD: adenocarcinoma; AR: androgen receptor; ccRCC: clear cell renal cell carcinoma; WT: Wilms tumor; CIS: carcinoma in situ; TCC: transitional cell carcinoma; UC: urothelial cancer; TGCT: testicular germ cell tumor; NA: not applicable; PeIN: penile intraepithelial neoplasia.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (No. 82173345 and No. 82373154 to Le Qu).

Conflict of interest

The authors declare no conflict of interest.