Abstract 1673: In silico modeling of RNA binding domains (RBDs) of nucleolin and miRNA interactions

In eukaryotes, the stress-responsive RNA-binding phosphoprotein nucleolin (NCL) plays defined role/s in many critical cellular processes, including transcriptional and translational regulation of various non-coding and coding RNAs. The RNA binding domains (RBDs) in NCL varies in numbers (1 to 4) depending on the organism, and individual RBDs are more conserved in divergent species than within the same protein. NCL targets a diverse RNA species e.g. long-noncoding-RNA, rRNA, mRNA to control gene expression and is also implicated in dysregulation of miRNAs in a variety of cancers; NCL is often highly upregulated in tumors. Analyses of mRNA and miRNA transcriptomics suggest that NCL plays a direct role in biogenesis of certain oncogenic miRNA species; targeting NCL by aptamers reduces the expression of these miRNA as well as in vivo tumor load. However, there is a large gap in knowledge in understanding miRNA-NCL interactions. Full length NCL protein structure still remains unresolved limiting structural predictions. Available partial NCL-RBDs structures are inadequate in providing cues to the plethora of NCL functions, especially its ability to interact with many different species of RNA. Molecular phylogeny for NCL-RBDs remains elusive as does target specificity of NCL. The number of RBDs, which structural combinations of RBDs and the cellular conditions that trigger specific NCL-RNA interactions remain to be understood. We have used in silico tools to model NCL-RBDs to assess their interactions with different RNA species focusing on a subset of miRNAs whose biosynthesis is positively correlated with NCL levels in a variety of aggressive cancer types e.g. breast, lung, pancreas and prostate cancers. Here we present a combination of template-based and ab initio approaches to develop robust theoretical models of all 4 RBDs in tandem as well as 2 consecutive RBDs in different combinations, for human NCL. Development of such theoretical constructs is essential to address hypotheses pertaining to target specificity of NCL for different RNA species in the absence of complete structural information. Further, these refined models were used to probe binding affinities and specificities of various miRNAs using different docking algorithms that analyze protein-RNA interactions. Our results provide an initial glimpse into possible binding modes and RBD specificity of human NCL for different miRNA molecules. Our study is further enhanced by modeling and comparative analysis of RBDs from NCL homologs in different organisms that house a variable number of RBDs. The results from this molecular modeling study lay the groundwork for elaborating the detailed mechanism of NCL-miRNA interactions, the evolutionary basis of RNA-binding and need for the expansion of RBD units in higher organisms. Citation Format: Avdar San, Anjana D. Saxena, Shaneen Singh. In silico modeling of RNA binding domains (RBDs) of nucleolin and miRNA interactions [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1673.