Position effects at the FGF8 locus are associated with femoral hypoplasia

Magdalena Socha,Anna Sowińska-Seidler,Uirá Souto Melo,Bjørt K Kragesteen,Martin Franke,Verena Heinrich,Robert Schöpflin,Inga Nagel,Nicolas Gruchy,Stefan Mundlos,Varun K A Sreenivasan,Cristina López,Martin Vingron,Ewelina Bukowska-Olech,Malte Spielmann,Aleksander Jamsheer
DOI: https://doi.org/10.1016/j.ajhg.2021.08.001
2021-09-02
Abstract:Copy-number variations (CNVs) are a common cause of congenital limb malformations and are interpreted primarily on the basis of their effect on gene dosage. However, recent studies show that CNVs also influence the 3D genome chromatin organization. The functional interpretation of whether a phenotype is the result of gene dosage or a regulatory position effect remains challenging. Here, we report on two unrelated families with individuals affected by bilateral hypoplasia of the femoral bones, both harboring de novo duplications on chromosome 10q24.32. The ∼0.5 Mb duplications include FGF8, a key regulator of limb development and several limb enhancer elements. To functionally characterize these variants, we analyzed the local chromatin architecture in the affected individuals' cells and re-engineered the duplications in mice by using CRISPR-Cas9 genome editing. We found that the duplications were associated with ectopic chromatin contacts and increased FGF8 expression. Transgenic mice carrying the heterozygous tandem duplication including Fgf8 exhibited proximal shortening of the limbs, resembling the human phenotype. To evaluate whether the phenotype was a result of gene dosage, we generated another transgenic mice line, carrying the duplication on one allele and a concurrent Fgf8 deletion on the other allele, as a control. Surprisingly, the same malformations were observed. Capture Hi-C experiments revealed ectopic interaction with the duplicated region and Fgf8, indicating a position effect. In summary, we show that duplications at the FGF8 locus are associated with femoral hypoplasia and that the phenotype is most likely the result of position effects altering FGF8 expression rather than gene dosage effects.
What problem does this paper attempt to address?