The role of WNT signalling in urothelial cell carcinoma

Abstract

Urothelial cell carcinoma (UCC) of the bladder is one of the most common malignancies, causing considerable morbidity and mortality worldwide. It is unique among the epithelial carcinomas as two distinct pathways to tumourigenesis appear to exist: low grade, recurring papillary tumours usually contain oncogenic mutations in FGFR3 or HRAS whereas high grade, muscle invasive tumours with metastatic potential generally have defects in the pathways controlled by the tumour suppressors p53 and retinoblastoma. Over the last two decades, a number of transgenic mouse models of UCC, containing deletions or mutations of key tumour suppressor genes or oncogenes, have helped us understand the mechanisms behind tumour development. In this summary, I present my work investigating the role of the WNT signalling cascade in UCC.

The incidence of urothelial cell carcinoma (UCC) continues to rise, with significant implications for our healthcare systems. In the UK, it is the seventh most common cancer, accounting for 3% of all new cases, with an estimated 10,399 new cases diagnosed in the UK in 2011.¹

UCCs are predominantly transitional cell carcinomas (TCCs), arising from the transitional epithelium that lines the bladder. Of these, 70–80% of cases are low grade (LG), papillary, non-invasive tumours that tend to recur in over 30% of patients despite local excision.² There is controversy regarding whether these papillary tumours progress to invasive disease, with progression occurring in approximately 5% of cases.³ However, whether this is a true progression or development of de novo muscle invasive tumours remains unclear. Most morbidity and subsequent mortality secondary to UCC is due to the high grade (HG), non-papillary, muscle invasive form of the disease (20–30% of new TCC cases).⁴ These invasive tumours penetrate through the muscularis of the bladder and 50% of patients relapse with tumours that metastasise to distant sites despite treatment.⁵ The five-year survival rate for metastatic bladder cancer is approximately 7%.⁶

Clinical and pathological studies have found that these two forms of UCC arise through at least two separate mechanisms (Fig 1).^7,8 Recently, genomic profiling and expression studies have identified two separate signatures for different molecular subtypes of UCC that stratify into LG or HG disease and can independently predict the likelihood of development of metastasis and disease specific survival.^9–11

Figure 1 .

Important genetic defects that characterise the diverse pathways underlying urothelial cell carcinoma (UCC): Low grade, non-invasive papillary tumours are associated with mutations in either RAS pathway components or FGFR3, which are thought to be mutually exclusive. High grade, muscle invasive tumours are associated with loss of p53 or retinoblastoma protein function.

Multiple transgenic murine models that attempt to recapitulate the different subtypes of UCC have been developed to better understand the disease process (Table 1).

Table 1.

At a glance reference of published genetically engineered mouse models of UCC⁵⁷
[Matt to advise on whether permission is needed to reproduce.]

Mutation	Promoter	Phenotype	References
Hras+ (low copy) Hras+ (high copy)	UPII UPII	Low grade, non-invasive (long latency) Low grade, non-invasive (short latency)	Zhang, 200144 Zhang, 200144
Egfr+ Egfr+Hras+	UPII UPII	Hyperplasia Hyperplasia	Cheng, 200245 Cheng, 200245
Fgfr3+ Fgfr3+KrasG12D Fgfr3+β-cateninexon3/+ Fgfr3+Ptenfl/fl	UPII UPII UPII UPII	Hyperplasia Hyperplasia Hyperplasia Hyperplasia	Ahmad, 201146 Ahmad, 201146 Ahmad, 201146 Foth, 201447
Ncstnfl/fl	R26rtTA; tetO-	Invasive UCC	Rampias, 201448
p53fl/fl p53fl/fl p53fl/flHras+ p53fl/flPtenfl/fl	UPII UPKII UPII Adeno-Cre	Hyperplasia and dysplasia No phenotype High grade, invasive UCC Invasive, metastatic UCC	Gao, 200449 Ayala de la Peña, 201150 Gao, 200449 Puzio-Kuter, 200951
Ptenfl/fl Ptenfl/fl Ptenfl/fl Ptenfl/fl	Fabp-Cre Ksp-Cre Adeno-Cre UPII	Low grade UCC UCC of renal pelvis No phenotype No phenotype	Yoo, 2006,52 Tsuruta, 200653 Qian, 200954 Puzio-Kuter, 200951 Ahmad, 201134
Rb-/- Rb-/-Ptenfl/fl Rb-/-p53fl/fl	UPII Adeno-Cre UPII	No phenotype No phenotype Invasive UCC after chemical carcinogenesis	He, 200955 Puzio-Kuter, 200951 He, 200955
SV40 T antigen (low copy) SV40 T antigen (high copy) SV40 T antigen	UPII UPII UPKII	Carcinoma in situ Invasive, metastatic UCC Carcinoma in situ	Zhang, 199956 Zhang, 199956 Ayala de la Peña, 201150
β-cateninexon3/+ β-cateninexon3/+Ptenfl/fl β-cateninexon3/+HrasQ61L	UPII UPII UPII	Hyperplasia Low grade, non-invasive UCC Low grade, non-invasive UCC	Ahmad, 201134 Ahmad, 201134 Ahmad, 201135
UCC = urothelial cell carcinoma

The role of WNT signalling in urothelial cell carcinoma

Human urothelial cell carcinoma

The Wnt pathway is conserved widely throughout many species including Caenorhabditis elegans, Drosophila, Xenopus and higher order mammals. It was discovered originally in Drosophila where the gene wingless demonstrated the ability to influence segment polarity during larval development.¹² The WNT pathway is now known to control many events during embryonic development and regulates homeostatic self-renewal in a number of adult tissues. As a result, mutations in this pathway are associated with the onset of various cancers, owing to the influence of the WNT pathway on processes such as proliferation, motility and cell fate at the cellular level.¹³

In colorectal cancer, it is now well established that germline and somatic mutations of the APC gene, a key component of the WNT pathway, are found in most cases.^14–16 However, regarding UCC, there was still dubiety surrounding whether the disease was associated with somatic mutations of members of the WNT signalling pathway, including adenomatous polyposis coli (APC)^17–20 and β-catenin.^21–23 Loss of heterozygosity has been reported at the APC locus in 6–50% of UCC tumours analysed, with no APC mutations identified in tumours (n=24) or cell lines (n=4).^18–20 Subsequently, evidence of missense (13%) and frameshift (3%) deletions in the APC protein adjacent to the β-catenin binding sites was reported.²⁴ Interestingly, the authors were able to demonstrate that either APC mutations or β-catenin accumulation was associated with shorter disease free interval and shorter disease specific survival in multivariate analysis.

Earlier studies have demonstrated an upregulation of β-catenin by immunohistochemistry.^25–29 In addition, CpG hypermethylation silencing of the region encoding WNT inhibitory factor 1, an inhibitor of the WNT signalling pathway, has been reported as a frequent event in UCC.³⁰ In another study, epigenetic silencing of the four secreted frizzled receptor proteins, antagonists of the WNT signalling pathway, has been demonstrated as an independent predictor of invasive bladder cancer.³¹

Recent sequencing studies demonstrate alteration of WNT family members in up to 73% of 131 chemotherapy naive, muscle invasive, HG bladder tumours (Fig 2).³²

Figure 2.

Mutational profile of 131 individual cases of muscle invasive urothelial cell carcinoma (http://www.cbioportal.org/public-portal/)

It has also been demonstrated recently that urothelial cancer associated 1 ribonucleic acid (which actives WNT signalling in a WNT6 dependent manner) increases the cisplatin resistance of bladder cancer cells in vitro.³³

Indeed, in our own studies, we have demonstrated that WNT signalling is activated in approximately a third of clinical UCC samples.^34,35 A significant correlation was observed between activation of the WNT pathway and either phosphoinositide 3-kinase (PI3K) or mitogen activated protein kinase (MAPK) activation. Interestingly, both groups were mutually exclusive with almost no overlap between the WNT/PI3K and the WNT/MAPK tumours (p<0.0001), suggesting a different molecular basis for each tumour subtype.

Modelling mutations of the WNT pathway in mouse models

Our findings in the human tissue microarray led us to hypothesise that activation of WNT signalling could lead to human UCC but that it required activation of an additional second mitogenic pathway (either PI3K or MAPK), leading to distinct subsets of WNT driven tumours. Our investigation of the mechanistic basis of WNT pathway signalling in UCC prompted us to examine whether dysregulation of this pathway could induce disease in a murine model.

In the first instance, to activate Wnt signalling, we used transgenic mice that carry a dominant allele of the β-catenin gene in which exon 3 is flanked by loxP sequences.³⁶ When a urothelial specific Cre recombinase (UPII-Cre) is added, exon 3 is deleted, stopping β-catenin from being targeted for degradation, allowing accumulation in the urothelial cells and thus activating the Wnt signalling in a bladder specific manner.³⁷ Overexpression of nuclear β-catenin led to the formation of localised hyperproliferative lesions in the bladder epithelium by 3 months, which did not progress to malignancy despite the mice being aged to 18 months of age.³⁴ Furthermore, the UPII-Cre+β-catenin^exon3/+ mice showed a marked upregulation of the phosphatase and tensin homolog (Pten) tumour suppressor protein that appeared to be a direct consequence of activating Wnt signalling in the bladder, suggesting a potential compensatory mechanism that could prevent progression to invasive disease.

In order to investigate this potential block to tumourigenesis, we interbred the UPII-Cre+β-catenin^exon3/+ mice with mice engineered to activate either the PI3K (Pten^fl/fl) or MAPK (Hras^Q61L) pathways specifically in the urothelium. In both cases, this overcame the block to tumourigenesis and induced a non-invasive UCC phenotype in the murine bladders. The UPII-Cre+β-catenin^exon3/+Pten^fl/fl tumours contained increased activation of the PI3K pathway and their growth was dependent on mechanistic target of rapamycin (mTOR) signalling (ie regression occurred with rapamycin treatment, an inhibitor of mTOR signalling). In direct contrast, the tumours in UPII-Cre+β-catenin^exon3/+Hras^Q61L mice, although phenotypically similar, were driven by MAPK signalling, with no activation of the PI3K pathway (ie regression occurred with mitogen activated protein kinase kinase inhibition, an inhibitor of MAPK signalling, but not with rapamycin).^34,35

Consequently, we were able to demonstrate that by using transgenic models of either Wnt/PI3K or Wnt/MAPK pathway activation, we can induce non-invasive UCC in vivo, leading to tumours with differing molecular and treatment profiles (Fig 3). These tumours mirror their human counterparts with respect to pathway activation secondary to mutations associated with human UCC.

Figure 3.

Divergent pathways of WNT driven urothelial cell carcinoma: Our work has suggested that Wnt activated/Pten inactivated tumours are non-invasive tumours that fill the bladders of the mice. They are dependent on mechanistic target of rapamycin (mTOR) signalling, as highlighted by responsiveness to the mTOR inhibitor rapamycin. By contrast, Wnt activated/Ras activated tumours are non-invasive tumours that are dependent on mitogen activated protein kinase signalling, as demonstrated by their responsiveness to MEK inhibition but not rapamycin treatment.

Summary and outlook

Despite transgenic models providing researchers with new insights into the molecular mechanisms in UCC, many limitations remain. First, recapitulating murine metastatic models of UCC continues to be challenging with a lack of spontaneous osseous lesions. The reasons for this could be multiple, including differences in the pelvic venous drainage of the mouse, our inability to detect small metastasis in mouse’s skeleton or the fact that the combinations of genetic defects are insufficient to initiate bone metastasis. In addition, one should not disregard the relatively short lifespan of the mouse, which together with the aggressive nature of the tumours we are engineering may not allow enough time for metastatic lesions to develop before the animals must be sacrificed owing to primary tumour burden.

Surprisingly, there remains a paucity of preclinical trials using these mouse models.^35,38,39 As we discover additional biomarkers and ‘driver’ mutations through sequencing studies of human UCC, one would hope that we can develop new mouse models based on these findings.

An exciting development in the transgenic field involves using unbiased somatic mutagenesis to search for these ‘driver’ genes that contribute to the initiation and/or progression of UCC. These in vivo transposon-based methods could help us to identify biologically relevant genetic lesions in tumourigenesis, allowing identification of events that collaborate and/or synergise with the common mutations found in UCC.^40,41

Despite the emergence of clinical biobanks, next-generation sequencing and other ‘omic’ approaches, the heterogeneity and multifocal nature of clinical UCC means that only limited numbers of these studies using clinical materials have produced mature data to inform on disease outcome in bacillus Calmette–Guérin and chemotherapy resistant disease, with large sample numbers required to give statistical significance. However, the information revealed by cancer genome sequencing has also highlighted the simplicity of currently available mouse models: it is no longer enough to compare a mouse model and a primary human tumour at the level of tissue and cellular phenotypes. Nevertheless, mouse models of UCC will continue to provide invaluable information on the biology of the disease, with the hope that this can be ultimately translated to earlier diagnosis and therefore better treatment of patients with UCC.

Given the critical roles of WNT pathway activation in the pathophysiology of many human diseases including bladder cancer, interest in the development of WNT signalling inhibitors has intensified (reviewed by Kahn, and Anastas and Moon).^42,43 Different components of the WNT pathway can be targeted, with multiple agents (including small molecules, peptides and blocking antibodies) currently in clinical trials.

We have demonstrated that the WNT pathway is only active in a third of bladder cancers so patient stratification will be essential. Importantly, as with other cancers, inhibiting WNT signalling pathways in bladder cancer alone is unlikely to result in a substantial improvement in disease progression owing to the coactivation of additional oncogenic pathways. Studies aimed at identifying the genetic factors and biomarkers that can predict responses to treatment with WNT pathway modulators (either alone or in combination with other therapies) will be an important next step in determining the utility of these potential new therapies.