MicroRNAs involved in cancer avoid secondary structure

B14

Control of gene expression is known to occur at any of the events from promotion of transcription to stabilization of the mature polypeptide product. mRNA secondary structures are known to be important in many gene regulation schemes. It has been observed that eukaryotic mRNAs have more secondary structure compared to randomized sequences (Seffens and Digby, 1999). Human mRNAs have even more excess secondary structure than lower animals. Excess stem-loop structure in human mRNAs may be related to miRNAs or other gene regulation. miRNAs are noncoding RNA molecules of about 21 nucleotides in length that regulate translational efficiency of target mRNAs involved in many critical cellular functions such as cell proliferation and apoptosis. These miRNAs bind to the 3’-UTR of mRNA resulting in translational inhibition but not mRNA degradation. miRNAs can contribute to cancer development and progression and are differentially expressed in normal tissues and cancers. From a large-scale miRnome analysis on samples including lung, breast, stomach, prostate, colon, and pancreatic tumors, a solid cancer miRNA signature was found to be composed of a large portion of overexpressed miRNAs. Among these miRNAs are some with well characterized cancer association, such as miR-17-5p, miR-20a, miR-21, miR-92, miR-106a, and miR-155. The predicted targets for the differentially expressed miRNAs are significantly enriched for tumor suppressors and oncogenes.
 We examined the structure of a set of these molecules involved in cancer by empirical and molecular dynamic calculations. Sequences were downloaded from the Sanger miRNA database, from which the secondary structure was determined using RNAstructure version 4.3. Folding energies to those calculated from sequences that were randomized 100 times were used to compute z-scores. The free energy of folding for the set of miRNAs examined partitions the miRNAs into two classes, those with or without secondary structure. We then compared the folding energies to those calculated from sequences that were randomized and computed z-scores. Results indicate that the miRNAs have less secondary structure than expected. These secondary structure values can be used to classify the miRNAs and to uncover correlations to functional categories such as disease outcome or expression levels in tissues. Tertiary structure determination and molecular dynamics calculations are CPU intensive. We examined two human miRNAs using routines in HyperChem to determine 3D structure in solution. The sequence for the 21 nt human miRNA, miR-23a, was obtained and a preliminary molecule model was created using HyperChem. Several cycles of geometry optimization and molecular dynamics (out to 300 ps) leads to two different configurations. A linear rod form and a condensed oblate form are persistent. Due to its small and unique structure, miRNAs may be particularly sensitive to physical perturbations compared to other biomolecules. Conformational free energies are available for correlational analysis to clinical or biochemical parameters. Together these results suggest the sequences of miRNAs avoid self-structures that may slow hybridization to target mRNAs. These results may be of use to design a diagnostic procedure based on the unique structure of theses molecules. This work was supported by NIH/NIGMS/MBRS/ SCORE/RISE (SCORE grant #S06GM08247).