Knockdown of p53 levels in human keratinocytes accelerates Mcl-1 and Bcl-x L reduction thereby enhancing UV-light induced apoptosis Vijaya Chaturvedi 1 ,LeonidASitailo 1 , Jian-Zhong Qin 1 , Barbara Bodner 1 , Mitchell F Denning 1 , Jonathan Curry 1 , Wengeng Zhang 2 , Douglas Brash 2 , Brian J Nickoloff* ,1 1 Department of Pathology, Oncology Institute, Loyola University Medical Center, USA; 2 Department of Radiation Oncology, Yale University Medical School, USA Ultraviolet (UV) light exposure is a common cause of epithelial-derivedskincancers,andtheepidermalresponse toUV-lighthasbeenextensivelystudiedusingbothmouse models and cultured human keratinocytes (KCs). Elim- ination of cells with UV-induced DNA damage via apoptosis provides a powerful mechanism to minimize retention or expansion of genetically abnormal cells. This cell editing function has largely been ascribed to the biological role of the p53 tumor suppressor gene, as mutations or deletions involving p53 have been linked to skin cancer development. Rather than introducing muta- tions,orusingcellswithcompletelossofwild-typep53,we used an siRNA-based approach to knockdown, but not eliminate, p53 levels in primary cultures of human KCs followed by UV-irradiation. Surprisingly, when p53 levels were reduced by 50–80% the apoptosis induced by exposure to UV-light was accelerated and markedly enhanced(two-tothree-fold)comparedtocontrolsiRNA treated KCs. The p53 siRNA treated KCs were char- acterized by elevated E2F-1 levels accompanied by accelerated elimination of the Mcl-1 and Bcl-x L anti- apoptotic proteins, as well as enhanced Bax oligomeriza- tion. Forced overexpression of either Mcl-1 or Bcl-x L reduced the UV-light enhanced apoptotic response in p53 siRNA treated KCs. We conclude that p53 not only can provide proapoptotic signals but also regulates a survival pathway influencing Mcl-1 and Bcl-x L levels. This over- looked survival function of p53 may explain previous paradoxical responses noted by investigators using p53 heterozygous and knockout mouse models, and opens up thepossibilitythatnotallliaisonswithinthecellinvolving p53 necessarily represent fatal attractions. Oncogene (2005) 24, 5299–5312.doi:10.1038/sj.onc.1208650; published online 6 June 2005 Keywords: p53; UV-light; keratinocytes; apoptosis; E2F-1; Mcl-1; Bcl-x Introduction When epidermal keratinocytes (KCs) are exposed to ultraviolet (UV)-light, they can either become growth arrested, prematurely senesce, or undergo apoptosis (Chaturvedi et al.,1999;Qin et al.,2002).Theapoptotic response has been extensively studied because elimina- tionofDNAdamagedKCsisanefficientandpowerful protective mechanism for preventing development of skin cancer (Ziegler et al., 1994). While UV-light exposure triggers an assortment of cellular alterations involving multiple signaling pathways, dissection of the apoptoticresponserevealeditisprincipallymediatedby mitochondrialdysfunction(Jackson et al.,2000;Nickol- off et al., 2002; Bowden, 2004). Of all the signaling molecules impacted by UV-light exposure, perhaps most well studied is the tumor suppressor, p53. There are many reasons to link abnormalities in p53 signaling with development of UV-lightresponses,withanobviousrelationshiptoskin cancer. First, it was demonstrated early on that precancerous human skin lesions, and later invasive skincancers,frequentlyharboredp53mutationsbearing UV-light induced mutation signatures (Brash et al., 1991;Nelson et al.,1994;Ouhtit et al.,1998;Bolshakov et al., 2003). Second, use of p53 null mice, or mice engineered to overexpress mutated p53 proteins, were found to display increased incidences of malignant progression and conversion (Kemp et al., 1993; Wein- berg et al., 1994). Also, murine keratinocyte cultures derivedfromsuchmicewereobservedtohavedecreased apoptotic responses to UV-light. However, several paradoxical observations were also madewheninvestigatorsfurtherpursuedaroleforp53in cutaneouscarcinogenesisstudies(Greenhalgh et al.,1996; Jiang et al., 1999). When p53 heterozygote mice were studied, there were actually fewer skin cancers following various carcinogenesis protocols, and p53 null mice showed marked resistance to tumor formation when crossed with epidermal targeted oncogene-expressing mice (Greenhalgh et al., 1996; Jiang et al., 1999). Since one of the key differences between the expected and the unexpected in vivo results involved the relative levels of wild-type p53, we decided to explore the effect of only reducing, but not eliminating, wild-type p53 levels in cultured human KCs before and after UV-irradiation. Received 22 December 2004; revised 16 February 2005; accepted 23 February 2005; published online 6 June 2005 *Correspondence: BJ Nickoloff, Oncology Institute, Loyola Univer- sity Medical Center, Building 112, Room 301, 2160 S First Ave, Maywood, IL 60153, USA; E-mail: bnickol@lumc.edu Oncogene (2005) 24, 5299–5312 & 2005 Nature Publishing Group All rights reserved 0950-9232/05 $30.00 www.nature.com/onc