PTBP1 variants displaying altered nucleocytoplasmic distribution are responsible for a neurodevelopmental disorder with skeletal dysplasia

Abstract

Polypyrimidine tract-binding protein 1 (PTBP1) is a heterogeneous nuclear ribonucleoprotein primarily known for its alternative splicing activity. It shuttles between the nucleus and cytoplasm via partially overlapping N-terminal nuclear localization (NLS) and export (NES) signals. Despite its fundamental role in cell growth and differentiation, its involvement in human disease remains poorly understood. We identified 27 individuals from 25 families harboring de novo or inherited pathogenic variants - predominantly start-loss (89%) and, to a lesser extent, missense (11%) - affecting NES/NLS motifs. Affected individuals presented with a syndromic neurodevelopmental disorder and variable skeletal dysplasia with disproportionate short stature with short limbs. Intellectual functioning ranged from normal to moderately delayed. Start-loss variants led to translation initiation from an alternative downstream in-frame methionine, resulting in loss of the NES and the first half of the bipartite NLS, and increased cytoplasmic stability. Start-loss and missense variants shared a DNA methylation episignature in peripheral blood and altered nucleocytoplasmic distribution in vitro and in vivo with preferential accumulation in processing bodies, causing aberrant gene expression but normal RNA splicing. Transcriptomic analysis of patient-derived fibroblasts revealed dysregulated pathways involved in osteochondrogenesis and neurodevelopment. Overall, our findings highlight a cytoplasmic role for PTBP1 in RNA stability and disease pathogenesis.

Keywords: Bone development; Development; Genetic diseases; Genetics; RNA processing.

Figure 1. Clinical findings.

(A) Facial features of individuals with PTBP1 start-loss variants with a computed “average face” (red box “merge”). (B) Radiological and clinical skeletal findings. I 1 (5 days old): skull and limb X-rays showing acrocrania and advanced carpal bone maturation. I 2 (19 years old): hand X-ray and clinical images showing brachymetacarpia (except second ray) and brachydactyly (II and III phalanges). I 4 (27 years old): hand and foot images/X-rays showing brachymetacarpia, brachymetatarsia (digits I, III, and IV), brachytelephalangy, brachymesophalangy (digits II, V), and cone-shaped epiphysis (second phalanx, digit II). I 8 (11 years old): hand and foot images showing brachymetacarpia and brachytelephalangy (digits III–V), brachymetatarsia (digits III–V), fifth toe clinodactyly, sandal gap, and hallux valgus. I 11 (24 years old): hand and foot X-rays showing brachymetacarpia, brachytelephalangy, brachymesophalangy (digits II, V), cone-shaped epiphysis (second phalanx, digit II), and brachymetatarsia (digits II, III, V). I 18 (9 months old): hand and foot images showing short hand with brachymetacarpia. I 12 (1 year): hand X-ray showing advanced carpal maturation and cone-shaped distal epiphyses. I 13 (8 years old): hand and foot X-rays showing generalized brachymetacarpia, brachymesophalangy (digits II–V), brachytelephalangy with cone-shaped epiphyses, and brachymetatarsia. I 16 (hand: 7 years old; spine: 2 years old): hand X-ray showing brachymetacarpia, brachytelephalangy, and brachymesophalangy (digits II, V); cone-shaped epiphyses (middle phalanges of digit II, V; distal phalanges of digits I, IV); spine X-ray showing dysplastic lumbar vertebrae with anterior-posterior height disparity. I 19 (3 years old): spine X-ray showing dysplastic vertebral bodies with uneven anterior-posterior height.

Figure 2. Altered nucleocytoplasmic distribution of PTBP1 mutants.

(A) PTBP1 variants identified in our cohort (red) and a benign variant (green), mapped to functional domains. The untranslated region in start-loss variants is shaded gray. Schematics are not to scale. (B) Western blot of PTBP1 isoforms in fibroblast from controls and start-loss individuals (left); band interpretation shown on the right: band “a” corresponds to overlapping PTBP1-2 and PTBP1-4 isoforms; band “b” to PTBP1-1. Asterisked bands represent start-loss isoforms. Vinculin is the loading control. n = 4 independent experiments. (C) Quantification of PTBP1 bands from B; Mann–Whitney U test. (D) Cycloheximide chase assay in fibroblast at 0, 8, or 16 hours. n = 3 independent experiments. (E) Quantification of PTBP1 expression from D; Student’s t test. (F) Nuclear versus cytoplasmic PTBP1 (red) levels in immunostained fibroblasts shown in G; Hoechst (blue), actin (green). Sixty cells per condition were analyzed; Mann–Whitney U test. (H) PTBP1-4-tGFP localization (green) in NIH-3T3 cells transfected with WT, start-loss, or missense constructs; DAPI (blue). n = 3 independent experiments. (I) Quantification of the nuclear/cytoplasmic fluorescence in cells from (H). N = 30 cells/condition; Mann–Whitney U test. (J) Spearman’s correlation analysis of nuclear/cytoplasmic distribution from (I). WT: wild-type. Met1-sl: methionine 1 start-loss variant. CTCF: corrected total cell fluorescence. Scale bars: 20 μm. ****P ≤ 0.0001, ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05.

Figure 3. Molecular characterization of pathogenic PTBP1 variants.

(A) Heatmap and (B) multidimensional scaling (MDS) plot showing PTBP1 episignature clustering of start-loss (red), matched controls (blue), and missense cases (orange). Columns, PTBP1 individuals or controls; rows, selected probes. (C) Targeted cDNA amplicon sequencing of exon-skipping events in PTBP1 (exon 11), DLG4 (exon 18), and PBX1 (exon 7) transcripts from start-loss fibroblasts, untreated controls, and controls treated with either scramble (negative control) or PTBP1-specific siRNA (positive control). Mann–Whitney U test. n = 4 independent experiments.

Figure 4. Cytoplasmic colocalization of start-loss PTBP1 variants with P-bodies.

(A) Imaging of NIH-3T3 cells transfected with WT, start-loss or missense PTBP1-4-tGFP constructs (green), immunostained for DCP1A (red); DAPI (blue). n = 5 independent experiments. (B) Pearson’s correlation of DCP1A/PTBP1 colocalization. n = 30 cells/condition; Mann–Whitney U test. (C) Proximity ligation assay (PLA) of PTBP1–DCP1A interactions in PTBP1 start-loss fibroblasts and controls. n = 3 independent experiments. (D) Quantification of PLA puncta shown in C. n = 90 cells per condition; Student’s t test. Met1-sl, methionine 1 start-loss variant. Scale bars: 20 μm. ****P ≤ 0.0001, ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05.

Figure 5. Integrated analysis of patient-derived fibroblasts with start-loss PTBP1 variants.

(A) Workflow for RNA-seq, RIP-seq, and proteomic analyses in fibroblasts from controls and PTBP1 start-loss individuals. RoR = IP/input ratio in start-loss versus controls. (B) Hierarchical clustering of the top 5% most variable genes (n = 1,184) in input or IP fractions. (C) Volcano plots of differential gene expression in input (top) and PTBP1-IP (bottom) fractions. (D and E) Circular plots of deregulated mesodermal (D) and neurodevelopmental (E) pathways based on RNAseq (outer), RIP-seq (middle), and RoR (inner) Z-scores.

Figure 6. Proteomic dysregulation in fibroblasts from individuals with start-loss PTBP1 variants and associated biological pathways.

(A) Volcano plot of differential protein expression relative to controls. (B) Hierarchical clustering of the top 5% most variable proteins (n = 97). (C) Circular plot of pathways commonly deregulated at RNA and protein levels. From outer to inner: RNA-seq, RIP-seq, and RoR Z scores, and proteomics label-free quantification intensities. Red dots, significantly deregulated genes/proteins. Background colors indicate the normalized enrichment score (NES) direction from Gene/Protein Set Enrichment Analysis (G/PSEA).

Figure 7. Expression of human PTBP1 variants in zebrafish.

PTBP1 (red) immunofluorescence in 5 dpf zebrafish embryos injected at the one-cell stage with human wild-type (nuclear localization) or start-loss (cytoplasmic retention) mRNA. Hoechst (blue). Endogenous zebrafish ptbp1 is not detected. Representative images show samples under different magnifications. Original magnification, ×20 (columns I, II, and III); ×60 (column IV). In column IV, views of the areas outlined by dotted rectangles in column III are shown, highlighting finer structural details within the selected regions. n = 3 independent experiments.

Figure 8. Phenotypic analysis of zebrafish embryos injected with human PTBP1 mRNA at the one-cell stage.

(A) Lateral views of Alcian Blue-stained embryos at 5 dpf reveal caudal fin defects in WT and start-loss PTBP1-injected animals. The mild phenotype observed in embryos injected with Lys46Arg and Lys46Thr PTBP1 consist of slight or mild spine curvatures; severe cases show trunk and caudal fin deformities and ectopic staining (boxed). (B) Ventral view at 120 hpf showing craniofacial cartilage defects in severely affected embryos. (C) Stacked bar plot showing the phenotypic distribution of mild/severe defects across embryos injected with WT, benign (Lys46Arg), or pathogenic variants (start-loss and Lys46Thr). The number of the injected embryos per condition is indicated in the figure. χ² test. Met1-sl, methionine 1 start-loss variant. Dpf, days post fertilization. ****P ≤ 0.0001, ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05.