Expression of PML3 facilitates localization of p53 and HIPK2 within nuclear bodies. Reproduced with permission © (2002) Nature Cell Biol.

According to Meatloaf, “2 out of 3 ain't bad”, but when it comes to p53 regulation, homeodomain-interacting protein kinase-2 (HIPK2) can improve on these odds. Two papers in the January issue of Nature Cell Biology describe how it is involved in three out of the four ways in which p53 is regulated.

Thomas G. Hofmann et al. had previously cloned human HIPK2, but Gabriella D'Orazi et al. isolated HIPK2 in a yeast two-hybrid screen, using a p53 mutant that lacked the transactivation domain as the bait. This provided the first insight into its function.

The HIPK2–p53 interaction was confirmed in vitro and in vivo. Hofmann et al. showed it to be independent of HIPK2's kinase activity; a point mutation (K221A) renders the kinase inactive, yet it can still bind p53. HIPK2 and p53 also co-localize in nuclear structures known as nuclear bodies, which is facilitated by PML3 (see picture). CREB-binding protein (CBP), which acetylates and activates p53, also localizes in nuclear bodies, indicating that they could be structures in which p53 is modified.

Could HIPK2 regulate p53's function as a transcriptional activator? Hofmann et al. transfected SAOS-2 cells with a luciferase reporter controlled by the promoter of the CDKN1A gene — which encodes WAF1, a cyclin-dependent kinase inhibitor and known p53 target. They showed that HIPK2 could induce luciferase expression in a p53-dependent manner, but only when its kinase activity was intact. D'Orazi et al. used a similar system, this time in H1299 cells, and found that although HIPK2 increased the luciferase expression from the MDM2 promoter in a dose-dependent manner, it had no effect on that driven by the CDKN1A promoter. HIPK2 therefore seems to be both cell-type and promoter specific.

So, does HIPK2 directly phosphorylate p53 to enhance p53-dependent transcription? HIPK2 could phosphorylate p53 on Ser46 in vitro and in vivo, and Hofmann et al. showed that a p53 mutant in which Ser46 was replaced with alanine (p53-S46A) was strongly reduced in HIPK2-mediated expression from the CDKN1A promoter.

p53's Ser46 residue is phosphorylated in response to ultraviolet (UV) irradiation, so does UV irradiation activate HIPK2? D'Orazi et al. showed that, following UV irradiation, HIPK2's expression levels increased, it co-precipitated with p53 and it could phosphorylate p53. Hofmann et al. also showed that UV irradiation induced co-localization of HIPK2, p53 and CBP in nuclear bodies. In fact, HIPK2 and CBP were shown to interact, and acetylation-specific antibodies indicated that expression of HIPK2, but not the kinase-deficient mutant, induces acetylation of p53's Lys373 and Lys382 residues. Insight into this mechanism was provided when the p53-S46A mutant was found to remain unacetylated when HIPK2 and CBP were co-expressed in H1299 cells. So, UV-irradiation-induced HIPK2-mediated phosphorylation of Ser46 is required for acetylation, and this further increases the transactivation activity of p53.

So what is the consequence of HIPK2 activation and the corresponding increase in p53 activity — apoptosis or cell-cycle arrest? D'Orazi et al. showed that the addition of antisense oligonucleotides to HIPK2 led to a fivefold decrease in apoptosis, as quantitated by the TUNEL assay. Both groups also showed that expression of HIPK2 suppresses growth in a colony formation assay, and D'Orazi et al. went on to attribute this to apoptosis: cells stained positive for TUNEL; a hypodiploid peak was present when cells were analysed for DNA content; and there was an increase in expression of the apoptotic genes BAX and PIG3 . Hofmann et al. confirmed that apoptosis was enhanced, but also showed that cells arrested in G1, due to an increase in WAF1 levels.

So, this newly identified p53 regulator is involved in localization, phosphorylation and acetylation of p53 in response to UV irradiation. Indeed, the only mechanism of p53 control that it doesn't seem to affect is degradation. Perhaps HIPK2, like so many other regulators of p53, will turn out to be a tumour suppressor in its own right.