In the current study, analysis of a cohort of aged male Men1
+/- mice showed a significant reduction in the number of mutant mice with normal prostate glands due to the occurrence of prostate lesions in these mice. Interestingly, six mutant mice developed prostate cancer, which was not observed in the age-matched wild-type littermates. The prostate lesions found in heterozygous Men1 mutant mice were characterised by a slow cancer development process, which ranged from in situ carcinoma to invasive adenocarcinoma. This cancer development pattern in aged male heterozygous Men1 mice is similar to other non SV40-TAg GEM prostate cancer models, likely reflecting the relative late and slow features of prostate cancer development in men. Although one prostate cancer case was previously documented in an independent heterozygous Men1 mutant mouse cohort , to our knowledge, the present study is the first systematic evaluation and characterisation of prostate cancer related to the inactivation of the Men1 gene.
Importantly, the menin protein was undetectable in cancerous cells, and partial LOH was found in three of five pre-and/or cancerous lesions from Men1
+/- mice. Menin loss in these mutant mice may indicate a close relationship between Men1 inactivation and the development of prostate cancer, suggesting that the Men1 gene may possess oncosuppressive activity and be involved in the control of cell proliferation in prostate epithelial cells. However, we noticed that the occurrence of prostate cancer in these mice can be seen only in aged mice with a low frequency, implying that Men1 inactivation may confer a relatively minor predisposition to prostate cancer development compared with the endocrine tissues affected in MEN1 pathology and that other factors may be involved in the development of this pathology. Interestingly, these results are similar to the findings with mice carrying either one mutated Pten or the Nkx3.1 allele. Male heterozygous Pten mice develop mPIN at ages older than 9 months , whereas Nxk3.1 mutant mice show only hyperplastic or dysplastic prostatic epithelium . However, the double mutant Pten
+/- mice display high grade PIN/early carcinoma lesions at a high frequency [24, 25]. Both genes have been found either mutated or down-regulated in human prostate cancer [19, 26]. The results of the present study suggest that the Men1 gene could be among the rare known tumour suppressors whose disruption leads to the development of prostate cancer in mice, albeit at low incidence and with a slow progression rate.
Intriguingly, a recent study reported that gain at the MEN1 locus was detected in a substantial proportion of human metastatic prostate cancers, and a trend toward increased menin expression in human prostate cancers and metastatic tissues was suggested by a meta-analysis of MEN1 expression data in prostate cancer . Similarly, Imachi et al. reported that menin expression in breast cancers could be used as a prognostic factor of worse outcome . Knowing the multifaceted role played by the MEN1 gene, menin may play an oncogenic role under certain circumstances in these tissues, particularly in recurring and aggressive cancers, in similarity to the well-known dual role played by the TGF-β pathway, whose several effectors are the protein partners of menin, in the process of tumourigenesis. However, further clinical studies are needed to validate these observations, as mentioned by the authors.
Our results suggest that the study of the potential link between menin and AR expression or activity in prostatic cells would be of great interest, as the latter is deregulated in the prostate cancers found in Men1 mutant mice. Generally considered a co-regulator of transcription, menin interacts physically and functionally with several nuclear receptors, such as ERα  and PPARγ , playing the role of a transcriptional coactivator via its LXXLL motif. Although a physical interaction between menin and AR has not been reported yet, the existence of cross-talk between menin and the AR pathways is possible. Indeed, menin and AR share common partners like Smad3, a downstream effector of the TGF-β signalling pathway, and β-catenin, an effector of WNT pathway [4, 29], but also common target genes important for cell cycle control, such as CDKN1B and cyclins D [4, 11, 12, 30, 31]. The observation of deregulated AR expression in the cancerous lesions found in Men1 mutant mice suggests that menin inactivation and subsequent AR deregulation in these cells may lead to the disturbance of the AR pathway, which could in turn promote the progression to cancer.
It is worth mentioning that the mutant mice enrolled in this study developed Leydig cell tumours with high frequency, which could suggest that the occurrence of these tumours plays a role in the tumorigenesis of prostate cells. However, the data from previously published mouse models of Leydig cell tumours did not offer any evidence supporting this hypothesis, as none reported prostate lesions [32–36]. Furthermore, AR signalling seems to play a growth suppressing function in the initial stage of tumour development . Nevertheless, it will be interesting in the future to investigate the possible interplay between the development of Leydig cell tumours and the occurrence of prostate cancer.
Prior studies have proposed that the growth inhibition function of AR signalling in normal prostatic luminal cells is mediated by its induction of CDKN1B expression . This indeed correlates with the downregulation of CDKN1B observed in prostate carcinomas developed in our Men1 mutant mice. CDKN1B, which is closely related to the tumorigenesis of prostate cancers as described above, is among the transcriptional targets of menin. The mouse models with CDKN1B disruption in combination with other genetic factors, such as Pten and Nkx3.1, have been shown to develop a range of prostate cancers, although CDKN1B inactivation alone displays no obvious neoplastic prostate lesions [19, 25]. The downregulation of CDKN1B found in prostate lesions in Men1 mutant mice suggests the in vivo importance of CDKN1B inactivation in the tumorigenesis related to Men1 inactivation, similar to observations in other mouse Men1 tumour models [12, 37–39]. Furthermore, it would be interesting to clarify the possible relation between menin and other factors involved in CDKN1B regulation in the prostate, such as Pten and Nkx3.1, in the future.