Fig. 2: Schematic of KDM2B with pathogenic variants reported in this and previous studies.

Protein domains are reported according to UniProt entry Q8NHM5. Both the short (SF) and the full-length (FL) isoforms contain a DNA-binding domain (CxxC), a plant homeodomain (PHD), an F-box and a Leucine-rich repeat (LRR). The amino acids forming the CxxC zinc-finger domain are indicated with the single-letter amino acid code in the bottom panel. The CxxC domain contains two cysteine-rich clusters (CxxCxxCx4CGxCxxC and CxxRxC), with eight cysteines (indicated in red) coordinating two Zn2+ ions in a tetrahedral manner and a linker region containing the highly conserved motif KFGG (Lys-Phe-Gly-Gly) and the DNA-binding motif KQ (Lys-Gln) [5]. The latter, together with the upstream residue (Met), directly interacts with the CpG dinucleotide. The two variants identified in this study are indicated in bold, together with pathogenic variants from previous studies [2,3,4]. Family identifiers (square brackets) are reported according to the numbers used in previous publications: van Jaarsveld et al. [2] in purple, van Oirsouw et al. [3] in orange, Gomes et al. [4] in green, this study in black. Red dots indicate variants where at least one case exhibited the KDM2B episignature; black dots indicate variants for which methylation analyses were not performed.