miRNA-dependent target regulation: functional characterization of single-nucleotide polymorphisms identified in genome-wide association studies of Alzheimer’s disease

Biological validation of the PolymiRTSs identified in silico

We next looked at whether allelic variations in miRNA binding sequences might affect the impact of the miRNAs being studied on target regulation. As seen from the sequence alignment of FERMT2 3?-UTR and miR-4504, the minor rs7143400-T allele creates an illegitimate canonical binding site compared with the major rs7143400-G allele (Fig. 3b). The presence of rs7143400-G is thus expected to decrease miR-4504’s downregulating effect on the FERMT2 3?-UTR reporter levels (Fig. 3a). However, the differences in luciferase activity for HEK293 cells cotransfected with miR-4504 (relative to SCR control miRNA) were similar for rs7143400-T- and rs7143400-G-bearing constructs (i.e., decreases of 22.6?±?7.1 % and 22.3?±?5.8 %, respectively) (Fig. 3c). Nevertheless, we observed decreases in luciferase activity of 23.6?±?6.6 % and 24.4?±?7.5 % between rs7143400-G and rs7143400-T when coexpressed with either SCR control miRNA or miR-4504, indicating an allelic effect independent of miRNA expression. In HeLa cells, this allelic difference was not found. However, miR-4504 decreased the expression of the reporter gene carrying the rs7143400-T allele (by 28.7?±?3.3 %) but not that carrying the rs7143400-G allele (as expected on the basis of our in silico analysis) (Fig. 3d). In summary, and although the observations in HeLa and HEK293 cells were not identical, our data suggest that the rs7143400-T (minor) allele triggers FERMT2 downregulation (compared with the rs7143400-G [major] allele) and that this downregulation might partially be linked to the presence of miR-4504.

As shown in Fig. 3b, the canonical binding site for miR-3945 on MS4A2 corresponds to the major rs2847655-T allele, whereas the minor rs2847655-C allele disrupts perfect complementarity. As expected, cotransfection of HEK293 and HeLa cells with the MS4A2:rs2847655-T 3?-UTR luciferase constructs and miR-3945 resulted in a decrease in luciferase activity (by 38.4?±?6.7 % and 28.4?±?9.3 %, respectively, compared with SCR control miRNAs) (Fig. 3e, f). However, when miR-3945 was coexpressed with MS4A2:rs2847655-C 3?-UTR luciferase constructs, similar decreases in luciferase activity in MS4A2:rs2847655-T-expressing cells were observed in HEK293 and HeLa cells (39.4?±?23.5 % and 18.8?±?8.9 %, respectively) (Fig. 3e, f). Hence, miR-3945 appears to regulate MS4A2 expression, but its function is probably not affected by rs2847655.

For NUP160:miR-1185-1-3p, the canonical binding site corresponds to the major rs9909-G allele, whereas the minor rs9909-C allele disrupts perfect complementarily. Accordingly, we expected that the minor allele would limit or abolish the effects of miR-1185-3p on the expression of NUP160 3?-UTR luciferase constructs. This prediction was confirmed in both cell lines. In fact, the rs9909-G allele was associated with significant decreases in luciferase activity in HEK293 and HeLa cells cotransfected with miR-1185-3p (by 52.8?±?4.1 % and 52.6?±?2.8 %, respectively, compared with SCR control miRNAs). In contrast, the rs9909-C allele completely disrupted this regulation in HEK293 and HeLa cells cotransfected with miR-1185-3p compared with SCR control miRNAs (Fig. 3g, h). It is noteworthy that the luciferase activity was significantly greater in HEK293 cells (by 91.9?±?13.0 %) and HeLa cells (by 205?±?31.1 %) cotransfected with the rs9909-C luciferase constructs and SCR control miRNAs compared with the rs9909-G allele cotransfected with the same miRNAs.

In summary, our data indicate that (1) compared with the rs9909-G allele, the rs9909-C allele may increase NUP160 expression; and (2) at least part of this increase may be due to the disruption of a miR-1185-3p binding site.