The Menin Tumor Suppressor Functions as a Molecular Adapter That Tethers LEDGF to MLL Proteins in Leukemogenesis.

Abstract

MLL gene rearrangements are present in 5 to 10% of acute leukemias, which are generally associated with a poor prognosis. Chromosomal translocations at the MLL locus create MLL fusion genes that constitute 5‘ portions of MLL and 3‘ portions of partner genes. The resultant MLL fusion proteins transform myeloid progenitors by inappropriately activating a HOX-associated self renewal program. However, the molecular mechanisms underlying MLL oncoprotein function are not fully understood. Previously we identified menin as a component of the MLL macromolecular complex. Menin is a product of the MEN1 tumor suppressor gene, whose loss of function causes multiple endocrine neoplasia type 1 (MEN1). Despite its tumor suppressor role in endocrine tissues, menin functions as an essential oncogenic cofactor for MLL fusion protein-dependent leukemogenesis. This notion was supported by three lines of evidence:MLL fusion proteins lacking the menin binding motif do not transform myeloid progenitors;inactivation of menin causes growth arrest and subsequent differentiation of MLL oncogene transformed cells; andmenin is required for MLL oncogene dependent transcriptional activation of HOX genes.These findings raised a fundamental question: how does menin contribute to MLL-dependent transcription? Because menin lacks known functional motifs, its molecular functions could not be deduced from its structure. We hypothesized that menin may function as an adapter that tethers MLL to unknown associated factors. To identify such associated factors, we performed affinity purification of the MLL-ENL/menin complex from nuclear extracts of cells that transiently over-expressed both MLL-ENL and menin. Mass spectrometry identified LEDGF, originally identified as a transcriptional co-activator, in the purified material as a novel associated factor. LEDGF associates conjointly with the MLL/menin complex but not with MLL or menin alone, supporting the hypothesis that menin plays an adapter role to bridge MLL and LEDGF. Further analysis revealed that LEDGF is critical for MLL fusion protein-dependent leukemogenesis. Fine mapping of the domain responsible for LEDGF binding determined MLL residues 109–153 as the LEDGF binding domain (LBD). Mutations in LBD resulted in loss of oncogenic activity of MLL fusion proteins. Moreover, knock down of LEDGF in MLL-transformed cells caused growth arrest and differentiation in the same manner as menin knock down. These results demonstrate that LEDGF is also required for the initiation and maintenance of MLL fusion protein-dependent transformation. In contrast to menin, LEDGF has a distinctive functional motif (the PWWP domain), which reportedly has chromatin binding activity. To further confirm that menin is an adapter that links MLL and LEDGF, we examined the oncogenic functions of an artificial MLL fusion protein whose menin binding motif is replaced by the PWWP domain of LEDGF. This PWWP-MLL-ENL fusion protein does not associate with menin because it lacks the menin binding motif, nevertheless transforms myeloid progenitors. Chromatin immunoprecipitation experiments show that the PWWP-MLL- ENL fusion protein localizes at the HOXA9 locus while menin is absent. Moreover, myeloid progenitors transformed by the PWWP-MLL-ENL fusion protein continue to proliferate after menin is genetically inactivated. Thus covalent tethering of the PWWP domain fully compensates for loss of menin's cofactor function. Therefore, menin's only role in MLL-associated leukemogenesis is to tether LEDGF to MLL fusion proteins. In summary, this study identifies a previously unknown essential oncogenic cofactor of MLL fusion proteins and proposes a stepwise association model in which the MLL fusion protein first associates with menin, then recruits LEDGF to its complex to become functionally active.