Gain-of-function variants and overexpression of RUNX2 in patients with nonsyndromic midline craniosynostosis

Abstract

Craniosynostosis (CS), the premature fusion of one or more cranial sutures, is a relatively common congenital anomaly, occurring in 3-5 per 10,000 live births. Nonsyndromic CS (NCS) accounts for up to 80% of all CS cases, yet the genetic factors contributing to the disorder remain largely unknown. The RUNX2 gene, encoding a transcription factor critical for bone and skull development, is a well known CS candidate gene, as copy number variations of this gene locus have been found in patients with syndromic craniosynostosis. In the present study, we aimed to characterize RUNX2 to better understand its role in the genetic etiology and in the molecular mechanisms underlying midline suture ossification in NCS. We report four nonsynonymous variants, one intronic variant and one 18 bp in-frame deletion in RUNX2 not found in our study control population. Significant difference in allele frequency (AF) for the deletion variant RUNX2 p.Ala84-Ala89del (ClinVar 257,095; dbSNP rs11498192) was observed in our sagittal NCS cohort when compared to the general population (P = 1.28 × 10^-6), suggesting a possible role in the etiology of NCS. Dual-luciferase assays showed that three of four tested RUNX2 variants conferred a gain-of-function effect on RUNX2, further suggesting their putative pathogenicity in the tested NCS cases. Downregulation of RUNX2 expression was observed in prematurely ossified midline sutures. Metopic sites showed significant downregulation of promoter 1-specific isoforms compared to sagittal sites. Suture-derived mesenchymal stromal cells showed an increased expression of RUNX2 over matched unfused suture derived cells. This demonstrates that RUNX2, and particularly the distal promoter 1-isoform group, are overexpressed in the osteogenic precursors within the pathological suture sites.

Keywords: Birth defect; Craniofacial; Mesenchymal stromal cells; Nonsyndromic craniosynostosis; Osteogenesis; RUNX2.

Figure 1.. RUNX2 coding variants.

The diagram shows the schematic structure of the 7 coding isoforms of RUNX2 gene (source https://www.ensembl.org/; GRCh38), including the promoters, the coding and non-coding exons, the introns and the alternative untranslated regions at the 3’ end (3’UTRs). The large colored boxes span the regions amplified by the RT-qPCR primer sets (see Supplemental Table 1). The red dotted lines indicate the localization of 5 identified RUNX2 mutations (see Table 1 and text for details).

Figure 2.. Dual-luciferase assay of RUNX2 variants.

Control calvarial osteoblasts were transfected with (1) pOSE2-Luc reporter vector and (2) one of five RUNX2 variants in a pCMV6 vector with minimal promoter. Transfection of pOSE2-Luc (OSE2 binding element) alone resulted in negligible luciferase production. Six independent replicates were tested for each construct. Renilla vectors were used as internal transfection controls for all experiments. (*) represents significance at a level of < 0.05 compared to the wild-type (WT). Bars represent mean ± SE.

Figure 3.. RUNX2 expression in mNCS and sNCS.

The bar graphs show the relative expression of specific groups of RUNX2 transcript isoforms (i.e. total, P1- and P2-derived isoforms) in patient-matched samples of both mNCS and sNCS: A) expression levels analysed in suture tissues; B) expression levels analysed in suture-derived calvarial mesenchymal stromal cells (CMSC). *p≤0.05; **p≤0.01; ***p≤0.001; ****p≤0.0001.

Figure 4.. RUNX2 expression level comparison between different suture sites.

The histograms show the expression levels of specific groups of RUNX2 transcripts (i.e. total, P1- and P2-derived isoforms) in metopic and sagittal fused suture tissues and CMSC isolated thereof. *p≤0.05.

Figure 5.. RUNX2 expression in sagittal tissue samples.

The bar graphs display the expression levels of specific groups of RUNX2 transcripts (i.e. total, P1- and P2-derived isoforms) in both fused and unfused sagittal suture tissues. *p≤0.05; ***p≤0.001.