Programas Científicos
Programa de Oncología Molecular
Grupo de Oncología Experimental
Presentación Última actualización 24/03/2009
Rasless cells: a genetic approach to identify novel regulators of Ras signalling
We have recently generated compound mice carrying H-Ras and N-Ras null loci (a gift from E. Santos), along with conditional (floxed) K-Ras alleles. These mice also carried an inducible CreERT2 recombinase knocked in at the locus encoding the large subunit of RNA II polymerase. Immortal H-Ras–/–; N-Ras–/–; K-Raslox/lox; RERTert/ert embryonic fibroblasts (K-Raslox MEFs) derived from these mice retain expression of the p16INK4a and p19ARF tumour suppressors as well as a functional p53 response as long as they are cultured in the presence of appropriate fetal bovine serum.
Exposure of these MEFs to the synthetic steroid 4-hydroxy-tamoxifen (4OHT) efficiently ablated K-Ras expression, hence completely eliminating Ras signalling. These cells (Rasless MEFs) underwent dramatic changes in morphology and ceased proliferation (Figure - A). However, they did not become senescent or undergo apoptosis even after long periods of time in culture.
To identify gene(s) that may revert the proliferative arrest caused by lack of Ras signalling, we submitted K-Raslox MEFs to a short hairpin RNA (shRNA) library (barcode) screen in collaboration with R. Bernard’s laboratory at the Netherland’s Cancer Institute (NKI), Amsterdam. Only one shRNA against the tumour suppressor and cell cycle inhibitor against p21Cip1 could be identified and subsequently validated. w p21Cip1 is a transcriptional target of p53, we examined whether shRNA-mediated downregulation of this tumour suppressor also allowed cell proliferation in the absence of Ras signalling. p53 depleted K-Raslox MEFs efficiently formed colonies in the presence of 4OHT. All colonies analysed lacked expression of Ras proteins and displayed efficient knockdown of p53 and p21Cip1 transcripts.
More importantly, Rasless MEFs which had been arrested for several days in culture, reverted their flat, enlarged senescencelike morphology and begun to proliferate upon infection with lentiviral vectors expressing shRNAs against p53 (Figure - B). Analysis of p53 mRNA and protein levels in Rasless cells revealed that they were not increased as compared to those present in proliferating K-Raslox or wild type MEFs. Moreover, there was no increase in phosphorylation of Ser18 or acetylation of Lys379; two post-translational modifications often associated with stress-activated p53. Thus, loss of Ras signalling did not induce a stress response that resulted in overt p53 activation.
In addition to its well-characterised role in sensing oncogenic stress and DNA damage, p53 also senses stress conditions caused by physiological cellular challenges. It is therefore possible that during evolution, multicellular organisms have developed safety systems to prevent cell proliferation under conditions of low mitogenic signalling, to help maintain quiescence during adult homeostasis, for example. Based on our observations, we would like to propose that the p53/p21 axis might be part of such a safety mechanism. If so, loss of p53 may also contribute to tumourigenesis by allowing cells to proliferate under suboptimal mitogenic conditions, in addition to its well known role as guardian of the genome.
Cdk1 is sufficient to maintain functional G1/S and G2/M checkpoints as well as DNA repair in mammalian cells
Last year we reported that interphase Cdks are dispensable for the mammalian cell cycle (Santamaría et al, Nature 448: 811, 2007). These findings, along with the essential requirements for the mitotic kinase Cdk1, led us to propose that the basic mechanisms driving the cell cycle have been conserved from yeast to mammalian cells. We have now investigated whether mammalian cells driven by Cdk1 alone retain functional checkpoints and efficient DNA repair mechanisms.
To this end, actively proliferating control and Cdk4–/–; Cdk6–/–; Cdk2–/– MEFs (TKO cells) were submitted to different doses of γ-irradiation and changes in cell cycle distribution analysed by FACS. Both cells showed a decrease in the percentage of S-phase with a concomitant increase in G2 a result which is indicative of a functional checkpoint response. To test the functionality of the G1 checkpoint, early passage MEFs were serum starved, γ-irradiated and serum stimulated. S-phase entry was monitored by the incorporation of BrdU. Both TKO and control cell populations displayed a functional G1/S checkpoint response as determined by a significant reduction in the number of BrdU positive cells when compared to the non-irradiated controls. These observations underscore the integrity of both G1 and G2 checkpoints in the absence of interphase Cdks.
To determine the integrity of DNA repair capacity in TKO cells they were exposed to a variety of chemicals including methyl methanesulfonate (MMS), neocarzinostatin (NCS) and aphidicolin. None of the treatments resulted in exacerbated sensitivity of the TKO MEFs when compared to control MEFs.
These observations indicate that Cdk1 is also sufficient to maintain efficient DNA repair activity in mammalian cells.
Mouse models of developmental syndromes induced by hyperactive Ras signalling
During 2008, we reported the generation and characterisation of a mouse model for Costello syndrome, a developmental disorder induced by germ line expression of an oncogenic H-RasG12V allele (Schuhmache et al. J. Clin Invest., 2008). To further analyse the role of abnormal Ras signalling in other developmental disorders, including Noonan and cardio-facio-cutaneous (CFC) syndromes, we have generated mice carrying hyperactivated K-Ras and B-Raf isoforms in the germ line.
To generate a model for Noonan syndrome, we have introduced a V14I mutation in mouse ES cells. V14I is one of the most frequent mutations observed in patients with Noonan syndrome. We have recenttly achieved germ line transmission of the mutated allele and we are currently eliminating the lox-STOP-lox cassette to allow its expression in germ line as well as in selective tissues to compare its potential neoplastic activity with that of the well characterised K-RasG12V allele.
To generate a mouse model for the cardiofacio- cutaneous syndrome we have taken advantage of mouse strain in which the highly oncogenic B-RafV600E allele is expressed at low levels due to transcriptional bypass of a lox-STOP-lox cassette inserted in the endogenous B-Raf locus. Preliminary characterisation of these mice in collaboration with the laboratories of X. Bustelo, Centro de Investigación del Cáncer (CIC), Salamanca, Spain, B. Kerr, and E. Burkitt-Wright, Manchester Children’s Hospital, UK, indicate that these mice share many of the phenotypic properties of CFC patients. Indeed, the mice develop severe neurological and cardiovascular problems and do not survive for more than six months. We hope that these, H-RasG12V, K-RasV14I and B-RafLSLV600E strains of mice will help us to better understand the molecular and physiopathological bases for developmental disorders induced by hyperactivation of Ras signalling pathways.
