Background: Pim-1, 2 and 3 are a group of enzymes related to the calcium calmodulin family of protein kinases. Over-expression of Pim-1 and Pim-2 in mice promotes the development of lymphomas, and up-regulation of Pim expression has been observed in several human cancers. Results: Here we show that the pim kinases are constitutively active when expressed in HEK-293 cells and are able to phosphorylate the Bcl-2 family member Bad on three residues, Ser112, Ser136 and Ser155 in vitro and in cells. In vitro mapping showed that Pim-2 predominantly phosphorylated Ser112, while Pim-1 phosphorylated Ser112, but also Ser136 and Ser155 at a reduced rate compared to Ser112. Pim-3 was found to be the least specific for Ser112, and the most effective at phosphorylating Ser136 and Ser155. Pim-3 was also able to phosphorylate other sites in Bad in vitro, including Ser170, another potential in vivo site. Mutation of Ser136 to alanine prevented the phosphorylation of Ser112 and Ser155 by Pim kinases in HEK-293 cells, suggesting that this site must be phosphorylated first in order to make the other sites accessible. Pim phosphorylation of Bad was also found to promote the 14-3-3 binding of Bad and block its association with Bcl-X_L. Conclusion: All three Pim kinase family members predominantly phosphorylate Bad on Ser112 and in addition are capable of phosphorylating Bad on multiple sites associated with the inhibition of the pro-apoptotic function of Bad in HEK-293 cells. This would be consistent with the proposed function of Pim kinases in promoting cell proliferation and preventing cell death.

In eukaryotes, DNA damage repair is implemented by a host of proteins that are coordinated by defined molecular signals. One such signal that transpires during the Fanconi Anemia (FA) - interstrand crosslink (ICL) repair pathway is the site-specific monoubiquitination of FANCD2 and FANCI proteins by a large, multi-protein FA core complex. The mechanics for this exquisitely specific monoubiquitin signal has been elusive. Here we show FANCL, the RING E3 module of the FA core complex, allosterically activates its cognate E2 Ube2T for monoubiquitination by a mechanism distinct from the typical RING-based catalysis. FANCL triggers intricate re-wiring of Ube2T’s intra-residue network thus activating the E2 for precision targeting. This network is intrinsically regulated by conserved gates and loops which can be engineered to yield Ube2T variants that enhance FANCD2 ubiquitination by ~30-fold without compromising on target specificity. Finally, we also uncover allosteric networks in other ubiquitin E2s that can be leveraged by RING E3 ligases to drive specific ubiquitination.

The DNA damage response depends on ubiquitin signalling to orchestrate DNA repair. The Fanconi Anemia pathway for interstrand crosslink repair, and the translesion synthesis pathway for DNA damage tolerance, both require cycles of monoubiquitination and deubiquitination. The ubiquitin specific protease USP1 regulates both these pathways by deubiquitinating monoubiquitinated PCNA, FANCD2 and FANCI. Loss of USP1 activity gives rise to chromosomal instability. While many USPs hydrolyse ubiquitin-ubiquitin linkages, USP1 targets ubiquitin-substrate conjugates at specific sites. The molecular basis of USP1’s specificity for multiple substrates is poorly understood. Here we show that the molecular determinants for substrate deubiquitination by USP1 reside within the highly conserved and extended N-terminus. We find that the N-terminus of USP1 harbours a FANCD2-specific binding sequence required for deubiquitination of K561 on FANCD2. In contrast, the N-terminus is not required for PCNA or FANCI deubiquitination. Furthermore, we show that the N-terminus of USP1 is sufficient to engineer specificity in a more promiscuous USP.