Targeted inactivation of Mdm2 RING finger E3 ubiquitin ligase activity in the mouse reveals mechanistic insights into p53 regulation.

It is believed that Mdm2 suppresses p53 in two ways: transcriptional inhibition by direct binding, and degradation via its E3 ligase activity. To study these functions physiologically, we generated mice bearing a single-residue substitution (C462A) abolishing the E3 function without affecting p53 binding. Unexpectedly, homozygous mutant mice died before E7.5, and deletion of p53 rescued the lethality. Furthermore, reintroducing a switchable p53 by crossing with p53ER(TAM) mice surprisingly demonstrated that the mutant Mdm2(C462A) was rapidly degraded in a manner indistinguishable from that of the wild-type Mdm2. Hence, our data indicate that (1) the Mdm2-p53 physical interaction, without Mdm2-mediated p53 ubiquitination, cannot control p53 activity sufficiently to allow early mouse embryonic development, and (2) Mdm2's E3 function is not required for Mdm2 degradation.

Regulation of the p53 tumor suppressor pathway: the problems and promises of studying Mdm2's E3 ligase function.

Mdm2 is a major negative regulator of the tumor suppressor p53 and has long been thought to inhibit p53 in two ways: by ubiquitinating p53 to signal for its degradation, and by binding to p53, masking its transactivation domain. Mdm2 is also believed to control its own levels by autoubiquitination. Despite the widespread acceptance of these hypotheses, the supporting data were drawn primarily from in vitro and ectopic expression studies, which have not always been corroborated when tested in the more physiologically relevant setting of a knock-in or knock-out mouse model. Recently, a mouse model was generated in which a single point mutation (C462A) in Mdm2s RING domain abrogated Mdm2s E3 activity while leaving Mdm2-p53 binding intact. This study called into question two major dogmas about Mdm2 by suggesting that when endogenously expressed, (1) Mdm2 cannot inhibit p53 sufficiently by binding without ubiquitination, and (2) Mdm2 may not be regulated by autoubiquitination. Two years later, we are still without definitive answers for why these results conflict with previous findings, but we have gained new insights from subsequent studies. Here, we discuss potential reasons for the discrepancies concerning Mdm2s functions and how they might be resolved, taking into account new research in the field.

Mdm2 RING mutation enhances p53 transcriptional activity and p53-p300 interaction.

The p53 transcription factor and tumor suppressor is regulated primarily by the E3 ubiquitin ligase Mdm2, which ubiquitinates p53 to target it for proteasomal degradation. Aside from its ubiquitin ligase function, Mdm2 has been believed to be capable of suppressing p53's transcriptional activity by binding with and masking the transactivation domain of p53. The ability of Mdm2 to restrain p53 activity by binding alone, without ubiquitination, was challenged by a 2007 study using a knockin mouse harboring a single cysteine-to-alanine point mutation (C462A) in Mdm2's RING domain. Mouse embryonic fibroblasts with this mutation, which abrogates Mdm2's E3 ubiquitin ligase activity without affecting its ability to bind with p53, were unable to suppress p53 activity. In this study, we utilized the Mdm2(C462A) mouse model to characterize in further detail the role of Mdm2's RING domain in the control of p53. Here, we show in vivo that the Mdm2(C462A) protein not only fails to suppress p53, but compared to the complete absence of Mdm2, Mdm2(C462A) actually enhances p53 transcriptional activity toward p53 target genes p21/CDKN1A, MDM2, BAX, NOXA, and 14-3-3σ. In addition, we found that Mdm2(C462A) facilitates the interaction between p53 and the acetyltransferase CBP/p300, and it fails to heterodimerize with its homolog and sister regulator of p53, Mdmx, suggesting that a fully intact RING domain is required for Mdm2's inhibition of the p300-p53 interaction and for its interaction with Mdmx. These findings help us to better understand the complex regulation of the Mdm2-p53 pathway and have important implications for chemotherapeutic agents targeting Mdm2, as they suggest that inhibition of Mdm2's E3 ubiquitin ligase activity may be sufficient for increasing p53 activity in vivo, without the need to block Mdm2-p53 binding.

ARF in the mitochondria: the last frontier?

The tumor suppressor ARF carries out different functions in different cellular compartments. In the nucleus, ARF interacts physically and functionally with Mdm2 to inhibit cell cycle progression through activation of p53. In the nucleolus, ARF interacts with B23/NPM to inhibit ribosomal biogenesis through control of rRNA processing. Recent studies have expanded ARF's territory into the mitochondria. New data have shown that ARF interacts with the mitochondrial protein p32/C1QBP and that the interaction is critical in order for ARF to localize to the mitochondria and induce apoptosis. Remarkably, the ARF-p32 interaction, and hence ARF's pro-apoptotic function, can be interrupted by human cancer-derived mutations in exon2 of the p14(ARF)-p16(INK4a) gene locus. Here, we discuss the implications of these studies and their potential relevance to human cancer.

Unlocking the Mdm2-p53 loop: ubiquitin is the key.

Mdm2 has been thought to regulate the tumor suppressor p53 in two ways: by masking p53's access to transcriptional machinery, and by ubiquitinating p53, targeting it for proteasomal degradation. This dogma was recently challenged by data generated from knockin mice in which Mdm2's RING E3 ubiquitin ligase activity was abrogated by a single point mutation. The RING mutant Mdm2 is fully capable of binding with p53, yet cannot suppress p53 activity, suggesting that Mdm2 cannot block p53 by binding alone, without ubiquitination. Data from the RING knockin mice also revealed that endogenous Mdm2 does not, as previously thought, regulate its own stability by self-ubiquitination. In this review, we will discuss these findings and their relevance to the field, including potential reasons for the discrepancies between previous data and that generated by our knockin mice, as well as the feasibility of targeting Mdm2's E3 ubiquitin ligase activity in cancer. We will also discuss additional research questions that may be addressed using our mouse model.

The human protein kinase C gamma gene (PRKCG) as a susceptibility locus for behavioral disinhibition.

This study explores the association between a highly heritable behavioral disinhibition phenotype and the protein kinase C gamma (PRKCG) gene in an ethnically diverse youth sample from Colorado, USA. The rationale for this study was based on the impulsive behavior and increased ethanol consumption observed in the protein kinase C gamma (PKC-gamma)-deficient mouse model. Two composite behavioral disinhibition phenotypes and their component behavioral scores [conduct disorder, attention-deficit hyperactivity disorder (ADHD), substance experimentation (SUB) and novelty-seeking] were examined for association with five independent PRKCG single nucleotide polymorphisms (SNPs). Association analysis for the five individual SNPs revealed modest genetic association of Exon 14 (rs2242244) and Upstream (rs307941) markers with the behavioral disinhibition composite variables in the combined, Hispanic and African-American samples. Additionally, haplotype-based association analysis for two SNPs located in Intron 3 (rs402691) and Exon 6 (rs3745406) indicated a significant overall association of the PRKCG locus with the ADHD-hyperactive subscale scores in the combined and Caucasian samples, supporting the relation between impulsive behaviors and the PRKCG gene. A significant haplotype association was also observed with SUB scores but only in the Hispanic ethnic group, highlighting the marker variability for each ethnic group. In conclusion, our results support the role of the PKC-gamma enzyme in behavioral impulsivity previously observed in mice. This study provides the first exploration of the PRKCG gene and its association with behavioral disinhibition and warrants further study in other larger population samples.

Association of the neuronal nicotinic receptor beta2 subunit gene (CHRNB2) with subjective responses to alcohol and nicotine.

Nicotine addiction and alcohol dependence are highly comorbid disorders that are likely to share overlapping genetic components. We have examined two neuronal nicotinic receptor subunit genes (CHRNA4 and CHRNB2) for possible associations with nicotine and alcohol phenotypes, including measures of frequency of use and measures of initial subjective response in the period shortly after first using the drugs. The subjects were 1,068 ethnically diverse young adults participating in ongoing longitudinal studies of adolescent drug behaviors at the University of Colorado, representing both clinical and community samples. Analysis of six SNPs in the CHRNA4 gene provided modest support for an association with past 6 month use of alcohol in Caucasians (three SNPs with P < 0.08), but no evidence for an association with tobacco and CHRNA4 was detected. However, a SNP (rs2072658) located immediately upstream of CHRNB2 was associated with the initial subjective response to both alcohol and tobacco. This study provides the first evidence for association between the CHRNB2 gene and nicotine- and alcohol-related phenotypes, and suggests that polymorphisms in CHRNB2 may be important in mediating early responses to nicotine and alcohol.