The DLEU2-miR-15a-16-1 Cluster Is a Determinant of Bone Microarchitecture and Strength in Postmenopausal Women and Mice

Abstract

This study explores how select microRNAs (miRNAs) influence bone structure in humans and in transgenic mice. In trabecular bone biopsies from 84 postmenopausal women (healthy, osteopenic, and osteoporotic), we demonstrate that DLEU2 (deleted in lymphocytic leukemia 2)-encoded miR-15a-5p is strongly positively associated with bone mineral density (BMD) at different skeletal sites. In bone transcriptome analyses, miR-15a-5p levels correlated positively with the osteocyte characteristic transcripts SOST (encoding sclerostin) and MEPE (Matrix Extracellular Phosphoglycoprotein), while the related miR-15b-5p showed a negative association with BMD and osteoblast markers. The data imply that these miRNAs have opposite roles in bone remodeling with distinct actions on bone cells. Expression quantitative trait loci (eQTL) variants confirmed earlier DLEU2 associations. Furthermore, a novel variant (rs12585295) showed high localization with transcriptionally active chromatin states in osteoblast primary cell cultures. The supposition that DLEU2-encoded miRNAs have an important regulatory role in bone remodeling was further confirmed in a transgenic mice model showing that miR-15a/16-1-deleted mice had significantly higher percentage bone volume and trabecular number than the wild type, possibly due to prenatal actions. However, the three-point mechanical break force test of mice femurs showed a positive correlation between strength and miR-15a-5p/miR-16-5p levels, indicating differential effects on cortical and trabecular bone. Moreover, these miRNAs appear to have distinct and complex actions in mice prenatally and in adult humans, impacting BMD and microstructure by regulating bone cell transcription. However, detailed interactions between these miRNAs and their downstream mechanisms in health and disease need further clarification.

Keywords: DLEU2 gene; bone; eQTL; miRNAs; osteoporosis; transgenic mice.

Figure 1

(A). Levels of presented miRNAs were analyzed in trans-iliac bone biopsies from healthy (n = 39), osteopenic (n = 18), and osteoporotic (n = 27) postmenopausal women, as defined by total hip BMD by a PCR-based method using TaqMan LDA arrays A and B (Applied Biosystems). (B). Correlations between total hip bone mineral density T-scores and the relative values of miRNAs in trans-iliac bone biopsies from healthy (n = 39), osteopenic (n = 18), and osteoporotic (n = 27) postmenopausal women. *: p < 0.05; ***: p < 0.001; ****: p < 0.0001.

Figure 2

Results of the hsa-miR15a-5p eQTL analysis. In the top part, the association results are shown (y-axis the −log¹⁰ of the respective p-values, x-axis the genomic position on chromosome 13). Deep-blue lines show recombination rates (cM/Mb). Black and red dashed lines represent the FDR thresholds for 0.05 and 0.1, respectively. In the middle part, genes and their respective positions are shown for the cis region surrounding DLEU2. In the bottom part, the E129 cell line 15-state chromatin model, as provided by ROADMAP [18], is overlaid.

Figure 3

(A,B) µCT images of trabecular tibia in WT (left) and knockout (KO) mice (right). (C,D) Percentage bone volume (BV) over total volume (TV) and trabecular numbers (Tb.N), respectively. Data were obtained from 4 knockout and 19 three-month-old mice. *: p < 0.05.

Figure 4

Genotyping of mouse bone DNA. The picture shows a typical 1.2% agarose gel for separation of PCR products, including molecular weight standards (Mw) and control (ctr). Text indicates genotype of the differentially sized product verified by Sanger sequencing.