STRUCTURAL INSIGHTS INTO THE INTERACTION BETWEEN THE TANDEM PHD FINGER DOMAIN P5C5 OF NSD1 AND THE ZINC FINGER MOTIF C2HR OF NIZP1.

Point Mutations or translocation in NSD1 cause the overgrowth disorder Sotos syndrome and acute myeloid leukemia (AML), respectively (Berdasco M. et al, 2009, Wang G et al, 2003). NSD1 contains several chromatin related domains including a SET domain responsible for histone methyltranferases activity (H3K36 and H4K20), two nuclear receptor-interaction (NID) motifs, five zinc finger domains (PHD1-5), a variant PHD finger (C5HCH), two proline-tryptophan-proline-tryptophan (PWWP1-2) domains (Lucio-Eterovic AK, et al , 2011), suggesting a role in chromatin regulation and gene expression. 20 pathological Sotos mutations have been detected on the PHD tandem domain composed by NSD1-PHD5 and NSD1-C5HCH (NSD1-P5C5). The tandem domain is essential for the pathogenesis of acute myeloid leukaemia (AML) caused by the chimeric protein NUP98/NSD1 that forces the abnormal activation of Hox-A and Meis1 genes (Wang at al 2007). The deletion of this tandem domain is sufficient to abolish NUP98/NSD1 interaction with chromatin, preventing both the transcription activation of HOX genes and the immortalization of myeloid progenitors. The biological role of NSD1-P5C5 is still unclear. It was proposed that this tandem domain is involved in the recognition of both H3K4me3 and H3K9me3 histone marks, (Pasillas M et al. 2011). However, biophysical experiments in our laboratory did not confirm these results challenging the idea that this tandem domain can really work as epigenetic reader. Previous biochemical studies suggested that NSD1-P5C5 can also work as protein-protein interaction motif, being able to bind to the co-repressor Nizp1 by its C2HR zinc fingers motif (Nizp1-C2HR) thus mediating gene repression (Nielsen AL et al, 2004). The structural determinants of this interaction are still unknown and have been object of this thesis. In order to get more insights into the physiological and pathological role of NSD1-P5C5, we have solved its (i) solution structure by NMR spectroscopy and (ii) characterized its interaction with Nizp1-C2HR. NSD1-P5C5 folds as unique functional unit adopting a “face to side orientation”. In particular NSD1-PHD5 (or NSD1-P5) presents the canonical PHD finger fold, whereas the NSD1-C5HCH (or NSD1-C5) domain displays an atypical topology characterized by the presence of an additional two stranded β-sheet. In order to investigate the impact of Sotos point mutation on NSD1-P5C5 we expressed and purified seven mutants and analyzed them by NMR. The majority of them destabilize the fold, with the exception of the solvent exposed mutation Arg2152Gln and His2205Arg suggesting a functional role for these residues. We next solved the solution structure of the zinc finger Nizp1-C2HR, an atypical Cys2His2-type zinc finger in which the fourth zinc chelating residue is substituted by an arginine residue. Its fold consists of a short α-helix and of a short two-stranded β-sheet hold together by one zinc ion. Importantly, we showed that three zinc ligands are sufficient to maintain the protein domain fold and functionality. NMR titrations of 15N labelled NSD1-P5C5 with Nizp1-C2HR and 15N labelled Nizp1-C2HR with NSD1-P5C5 clearly show that the two proteins directly interact. Analysis of the chemical shift displacements upon complex formation allowed to identify the residues of the two protein domains involved in protein-protein interaction. The interaction surface is located on the interface between NSD1-P5 and NSD1-C5 and on the α-helix of Nizp1-C2HR, respectively. Based on this information using the software HADDOCK we have computed a data driven docking model of the protein complex. In the model Nizp1-C2HR places its α-helix in the groove at the interface between NSD1-P5 and NSD1-C5, creating both hydrophobic and polar intermolecular contacts. The thermodynamic parameters that govern complex formation were studied by ITC titrations: the binding reaction is entropy-driven, with a stoichiometry of 1:1 and a Kd of 3,80±0,66 μM. In order to solve the structure of the protein complex we performed crystallographic screenings, and we have found preliminary conditions for obtaining single crystals. In conclusion, the presented results provide novel information on the interaction between a tandem PHD finger domain and zinc finger motif. The results represent, to the best of our knowledge, the first biophysical characterization between two zinc binding domains. Most importantly, these data give the first molecular details of the interaction between NSD1 and Nizp1 and may provide useful insights into the function of NSD1 and its role in pathological conditions both in Sotos Syndrome and AML. Future work will be dedicated to the full three-dimensional characterization of the complex and to the analysis of Sotos mutations on complex formation.