Biallelic germline DDX41 variants in a patient with bone dysplasia, ichthyosis, and dysmorphic features

Abstract

DDX41 (DEAD‑box helicase 41) is a member of the largest family of RNA helicases. The DEAD-box RNA helicases share a highly conserved core structure and regulate all aspects of RNA metabolism. The functional role of DDX41 in innate immunity is also highly conserved. DDX41 acts as a sensor of viral DNA and activates the STING-TBK1-IRF3-type I IFN signaling pathway. Germline heterozygous variants in DDX41 have been reported in familial myelodysplasia syndrome (MDS)/acute myeloid leukemia (AML) patients; most patients also acquired a somatic variant in the second DDX41 allele. Here, we report a patient who inherited compound heterozygous DDX41 variants and presented with bone dysplasia, ichthyosis, and dysmorphic features. Functional analyses of the patient-derived dermal fibroblasts revealed a reduced abundance of DDX41 and abrogated activation of the IFN genes through the STING-type I interferon pathway. Genome-wide transcriptome analyses in the patient's fibroblasts revealed significant gene dysregulation and changes in the RNA splicing events. The patient's fibroblasts also displayed upregulation of periostin mRNA expression. Using an RNA binding protein assay, we identified DDX41 as a novel regulator of periostin expression. Our results suggest that functional impairment of DDX41, along with dysregulated periostin expression, likely contributes to this patient's multisystem disorder.

Fig. 1

Clinical phenotype and identification of DDX41 variants. (A) Photographs of the patient, taken at age 20, show dysmorphic facial features. (i-iii) The patient was diagnosed with acromesomelic dysplasia, characterized by small hands and feet. Her right upper arm, from shoulder to elbow, measures 23.5 cm, and the elbow to ulnar head length is 15.9 cm. The left upper arm is slightly longer, with the shoulder to elbow length at 24.0 cm and the elbow to ulnar head length at 16.0 cm. The total length of the right hand is 9.8 cm with a palm length of 6.0 cm, while the left hand is slightly longer with a total length of 10.0 cm and the same palm length. Her total feet measures 15.0 cm for the right foot and 14.6 cm for the left. Clinical findings include idiopathic scleroderma-like ichthyosis with skin splitting and acanthosis nigricans (arrow). (B) Sanger sequencing chromatograms showing biallelic variants in the proband. The proband is a compound heterozygous for NM_016222.4 (DDX41): c.465G > A; p. Met155Ile (inherited from the mother, left) and NM_016222.4 (DDX41): c.1033G > A; p. Glu345Lys (inherited from the father, right). (C) Alignment of missense variants in DDX41 across multiple species including human (Homo sapiens), chimpanzee (Pan troglodytes), Rhesus macaque (Macaca mulatta), cat (Felis catus), mouse (Mus musculus), red junglefowl (Gallus gallus), Fugu rubripes (Takifugu rubripes), African clawed zebrafish (Danio rerio), and fruit fly (Drosophila melanogaster)

Fig. 2

Functional analysis of DDX41 in the proband fibroblasts. (A) Quantitative real-time PCR analysis showed no significant difference in mRNA expression between the between the proband and control fibroblasts (error bars indicate ± SD, n = 4). (B) Reduced abundance of DDX41 protein in the proband. (Left) Fibroblasts from controls and the proband were lysed in RIPA buffer. Levels of DDX41 were analyzed through immunoblotting of fibroblast lysates using an anti-DDX41 antibody. Immunoblotting of lysates with anti-vinculin confirmed equal loading of the samples. The positions of two molecular weight markers in kilodaltons are shown. (Right) Quantification shows fold change in protein abundance. The control value represents the mean of data from two controls (error bars indicate ± SD, n = 15 independent experiments, ***= P < 0.0001). (C) Reduced STING signaling in the proband. Control and the proband fibroblasts were stimulated with double-stranded DNA complex (Poly (dA: dT)/LyoVec) for 24 h and processed for protein and RNA extraction. (Left) Immunoblots of lysates with anti-phospho-TBK1, Total-TBK1, and anti-DDX41 are shown. (Right) quantitative analysis shows a significantly reduced phospho-TBK1/Total-TBK1 ratio in the proband’s fibroblasts compared to the control (error bars indicate ± SD, n = 4 independent experiments, *= P < 0.05). (D) Quantitative PCR analysis of cDNA shows significant downregulation of IFN response genes IFIT1, OAS1, and ISG15 in the proband fibroblasts. Error bars indicate ± SD, n = 5, *= P < 0.05)

Fig. 3

Transcriptome analysis. Analysis of paired-end RNAseq data obtained from 3 technical replicates of two controls (control I and control II) and the proband fibroblasts. In total, 65.4, 63.2, and 63.4 million aligned reads from RNA-Seq libraries made from control I, control II, and proband were obtained. The average base quality score per read (average Phred quality score) was 36.2, the average read length was 150 bp, and 100% of reads could be mapped to the human reference genome. (A) Volcano plot showing differentially expressed genes (DEGs) in the proband fibroblasts compared to controls. Red dots represent significantly up-regulated genes, and blue dots represent significantly down-regulated genes. (B) Hierarchical clustered heatmap showing the expression patterns of DEGs. Analysis of significant alternative splicing (differentially spliced) events in the proband compared to control fibroblasts using rMATs. (C) Distribution of 7349 RNA splicing events with significantly different frequencies between the proband and control cells: 4194 skipped exon (SE), 843 retained intron (RI), 1289 mutually exclusive exons (MXE), 590 alternative 5′ splice site (A5SS), and 433 alternative 3′ splice site (A3SS). (D) Hierarchical clustered heatmap showing all differentially spliced events in each sample. (E) Principal-component analysis (PCA) of differentially spliced event frequencies showed replicates of each sample that clustered

Fig. 4

DDX41 Regulates Periostin expression. (A) Quantitative real-time PCR analysis of Periostin (POSTN) gene expression using two different Taqman Probes shows significantly increased expression of POSTN in the proband-derived fibroblasts compared to controls (error bars indicate ± SD, n = 4–5 independent experiments, *= P < 0.05). (B) (Upper panel) Levels of periostin were analyzed through immunoblotting of fibroblast lysates using an anti-periostin antibody. (Middle panel) Levels of DDX41 were analyzed using an anti-DDX41 antibody. Arrowhead indicates a post-translationally modified form of DDX41. (Lower panel) immunoblotting of lysates with anti-vinculin confirmed equal loading of the samples. The positions of molecular weight markers in kilodaltons are shown. (C) The binding of p21 and POSTN mRNA to DDX41 protein was measured using a ribonucleoprotein immunoprecipitation assay. DDX41 was immunoprecipitated from total fibroblasts lysates with an anti-DDX41 antibody. Binding to normal rabbit IgG was used as a negative control and to calculate the relative binding of p21 and POSTN mRNA to DDX41 protein. The binding of p21 mRNA to DDX41 was used as a positive control (error bars indicate ± SD, n = 4–6 independent experiments, *= P < 0.05)