miRNA arm selection and isomiR distribution in gastric cancer

Abstract

Background: MicroRNAs (miRNAs) are small non-protein-coding RNAs. miRNA genes need several biogenesis steps to form function miRNAs. However, the precise mechanism and biology involved in the mature miRNA molecules are not clearly investigated. In this study, we conducted in-depth analyses to examine the arm selection and isomiRs using NGS platform.

Methods: We sequenced small RNAs from one pair of normal and gastric tumor tissues with Solexa platform. By analyzing the NGS data, we quantified the expression profiles of miRNAs and isomiRs in gastric tissues. Then, we measured the expression ratios of 5p arm to 3p arm of the same pre-miRNAs. And, we used Kolmogorov-Smirnov (KS) test to examine isomiR pattern difference between tissues.

Results: Our result showed the 5p arm and 3p arm miRNA derived from the same pre-miRNAs have different tissue expression preference, one preferred normal tissue and the other preferred tumor tissue, which strongly implied that there could be other mechanism controlling mature miRNA selection in addition to the known hydrogen-bonding selection rule. Furthermore, by using the KS test, we demonstrated that some isomiR types preferentially occur in normal gastric tissue but other types prefer tumor gastric tissue.

Conclusions: Arm selections and isomiR patterns are significantly varied in human cancers by using deep sequencing NGS data. Our results provided a novel research topic in miRNA regulation study. With advanced bioinformatics and molecular biology studies, more robust conclusions and insight into miRNA regulation can be achieved in the near future.

Figure 1

Mapping result of hsa-mir-101-1 in G1245N library. hsa-mir-101-1 encode minor miRNA (hsa-miR-101*) at the 5p arm, ranging from nt. 11 to 32 of pre-miRNA, and major miRNA (hsa-miR-101) at the 3p arm, ranging from nt. 47 to 67. The 5p and 3p miRNAs are presented in uppercase and they individually have 10 and 22 isomiR types, each of which is labeled with read count (before normalization) and location offset. The expression level of a mature miRNA is determined by summing the read counts of all its isomiRs.

Figure 2

Detection of 5p arm miRNA of hsa-mir-1307. According to miRBase annotation, hsa-mir-1307 encodes mature miRNA at only its 3p arm. We used stem-loop RT-PCR to succeed in detecting the additional opposite-arm miRNA in both G1245N and G1245T tissue.

Figure 3

The expression ratios of 5p arm miRNA to 3p arm miRNA. Several pre-miRNAs have significantly different ratios, demonstrating the arm selection preferences within the pre-miRNAs are significantly different between gastric normal and gastric tumor tissue.

Figure 4

IsomiR distribution patterns mature miRNAs between G1245N and G1245T library. The isomiR distribution patterns (left panel) and the corresponding KS-test comparison plot (right panel) are provided. (a) The most abundant isomiRs in two tissues are type four and type five, respectively. (b) The most abundant isomiR type in G1245T is 20% less in G1245N. (c) The most abundant isomiRs in two tissues are individually type 12 and type 13. (d) Even expression levels of hsa-miR-30b are dramatically different between tissues, the isomiR distribution patterns are still the same.