Structure-Function Analysis Of Rpl18a, A Putative Binding Target Of Rigosertib

Rigosertib (ON 01910.Na; RGS) is a clinical stage anticancer agent that causes spindle abnormalities and mitotic arrest in neoplastic cells. The drug inhibits PI3K and PLK1 signaling pathways, down regulates cyclin D1 expression and induces apoptosis. Previously, we reported identification of RPL18A (L18A), a protein from the large ribosomal subunit, as a putative binding target of RGS [Proc. AACR 2014, #4595]. Knock-down of L18A with siRNA caused apoptosis in cancer cell lines. Role of L18A in the function of the ribosome is not known. Goal of this study was to conduct structure-function analysis of L18A and create deficiency mutant(s) for future experiments. Human L18A is predicted to adopt the following linear order of secondary structural elements - alpha helices (a) and beta strands (b) - when assembled into the ribosome - b1b2a1b3b4b5b6a2a3b7a4 (from www.rcsb.org/pdb/protein/Q02543). The predicted 3D structure of L18A suggests overall bilobal architecture with a flanking C-terminal tail. Packed in tandem, each lobe contains one alpha helix surrounded by a beta sheet formed by 4 (N-lobe) or 3 (C-lobe) beta strands. To map the potential binding site for RGS, we engineered multiple mutants of L18A including truncations at its N and C termini. Our approach involved successive removal of the structural elements and a few substitutions with alanine. Mutant proteins were expressed in E.coli as GST fusions, and derivative cell lysates were tested in pull-down assay with biotin-conjugated RGS (RGSbio) and avidin beads. Pulled-down protein-drug complexes were analyzed by Western blot with anti-GST antibody. At the C-terminus, abolishing a4 helix (C - a4) preserved RGSbio-binding activity comparable to wild type (WT) L18A. Further C-truncations resulted in partial (C - a3b7a4) or complete loss of activity (C - a2→a4; C - b6→a4 and C - b5→a4; latter represents N-lobe while missing entire C-lobe). Unlike N-lobe, independently expressed C-lobe had specific binding activity, albeit half that of WT. At the N-terminus, stepwise truncation of amino acids 6 through 12 in the first beta strand (b1) resulted in gradual reduction of specific binding activity. To verify that this was not due to structure misfolding we tested mutants with single Ala substitutions in b1 and obtained similar results. Such mutants will serve as RGS-non-binding controls in future studies. In the model, L18A packs in the ribosome in a way that its surface at the side of two alpha helices of both lobes is engaged in protein-RNA interactions. However, the outer surface of the two cross-braced beta sheets in both lobes is completely exposed to solvent and interestingly, there is a cavity formed at the junction of two lobes that appears to be deep enough to accommodate one molecule of RGS. Taken together, our data suggest that rigosertib approaches RPL18A at the exposed hinge region between the two lobes and may thus perturb ribosomal function in cancer cells. Citation Format: Irina A. Oussenko, Yogesh K. Gupta, Rodrigo Vasquez-Del Carpio, M.V. Ramana-Reddy, Aneel K. Aggarwal, E. Premkumar Reddy, James F. Holland, Takao Ohnuma. Structure-function analysis of RPL18A, a putative binding target of rigosertib. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 694. doi:10.1158/1538-7445.AM2015-694