Pedigrees and genetic findings of patients with GTF3C3 variants. (A) Pedigrees of the three families. (B) Homozygosity mapping using the automap tool showing the identified regions of homozygosity on chromosome 2 encompassing the GTF3C3 gene in Families 1 and 2, in comparison to Family 3. (C) Schematic diagram showing the location of the identified GTF3C3 variants in this study (in red) and those described previously (in black) and their domains. (D) Portions of the sequencing electropherogram showing the segregation of the GTF3C3 variants in the three families. (E) The conservation of the three new missense variants p.Cys172Gly, p.Val427Phe and p.Ala509Thr.

Facial features and brain imaging of patients. (A) Facial photographs of four patients showing distinctive facial features: Patient 1—bifrontal narrowing, long eyelashes, upslanting palpebral fissures, hypertelorism, full cheeks, full nasal tip, tented upper lip, full lower lip; Patient 2—round face, frontal prominence, bifrontal narrowing, long eyelashes, hypertelorism, short nose, broad nasal bridge, full nasal tip, full cheeks, tented upper lip, full lower lip; Patient 3—long face, bifrontal narrowing, infraorbital crease, smooth philtrum, full lower lip; Patient 4—microcephaly, low anterior hairline, bifrontal narrowing, hypertrichosis over forehead, synophrys, long eyelashes, hypertelorism, upslanting palpebral fissures, epicanthus, full cheeks, short nose, full nasal tip, tented upper lip, full lower lip. (B) Brain MRI images: B1 and B2—sagittal T1-weighted image and coronal fluid-attenuated inversion recovery images demonstrating selective atrophy of the cerebellar cortex, involving both the cerebellar vermis (arrow, B1) and hemispheres (asterisk, B2); B3—axial T2WI showing relatively small frontal horns and hypoplastic frontal lobes (short arrows, B3) compared with the rest of the brain.

Modelling gtf3c3 KO in zebrafish recapitulates patients’ clinical features. (A) Morphology of zebrafish WT, and gtf3c3 F0 KO larvae at 3 dpf. Scale bars: 500 µm. Body length (B), head size (C) and eye size (D) of WT and gtf3c3 F0 KO larvae at 3 dpf (n = 13–15). (E) Schematic representation and image of transgenic (Tg) GFAP:GFP head in dorsal view of 3-dpf WT and gtf3c3 F0 KO larvae. Scale bar: 200 μm. (F) Quantification of brain size of gtf3c3 F0 KO larvae (3 dpf, n = 4) compared to WT (3 dpf, n = 5) showed a significant reduction that could reflect microcephaly. (G) Representative swimming tracks, and quantification of swimming and velocity of WT (n = 36) control and gtf3c3 F0 KO (n = 48) fish at 5 dpf. (H) Representative swim traces of control WT fish and gtf3c3 F0 KO followed through time with PTZ treatment at 4 dpf. Quantification of swimming distance with PTZ treatment normalized to fish without treatment showed that gtf3c3 F0 KO (n = 50) larvae exhibited seizure-like behaviour after 15 min of PTZ treatment compared to WT (n = 52). (I) Neuronal activity induced by PTZ treatment (3 mM; 0, 15, 30 and 45 min) in 4-dpf larvae analysed by imaging p-MAPK/ERK staining. Quantification of mean intensity fluorescence of p-MAPK/ERK staining in telencephalic region (surrounded in red) showed a significant increase in gtf3c3 F0 KO larvae compared to WT at 0-, 15- and 45-min treatment of PTZ. According to the neuronal activity, we can suggest that gtf3c3 F0 KO (0, 15, 30 and 45 min, n = 4–6) larvae exhibit seizure-like behaviour relative to the WT (n = 5). All data are represented as the mean ± SEM. Statistical significance was calculated by Student's t-test or Mann–Whitney test, or Sidak' multiple comparisons tests. *P < 0.05, **P < 0.01, #P < 0.005, ***P < 0.001, ****P < 0.0001; ns, not significant. n represents number of fish.

Effect of GTF3C3 mutations on POLR3 target gene expression. (A) Morphology of zebrafish WT, gtf3c3 F0 KO + GTF3C3^WT, gtf3c3 F0 KO+GTF3C3^C172G and gtf3c3 F0 KO + GTF3C3^V427F larvae at 3 dpf. Scale bars: 500 µm. Body length (B), eye size (C) and head size (D) of WT, gtf3c3 F0 KO + GTF3C3^WT, gtf3c3 F0 KO + GTF3C3^C172G and gtf3c3 F0 KO + GTF3C3^V427F at 3 dpf (n = 12–13). (E) Dorsal view of GFAP 3 dpf larvae WT, gtf3c3 F0 KO + GTF3C3^WT, gtf3c3 F0 KO + GTF3C3^C172G and gtf3c3 F0 KO + GTF3C3^V427F. Scale bars: 200 µm. (F) Head area of GFAP 3 dpf larvae WT, gtf3c3 F0 KO + GTF3C3^WT, gtf3c3 F0 KO + GTF3C3^C172G and gtf3c3 F0 KO + GTF3C3^V427F(n = 12–13). (G–J) Quantitative real-time PCR. (G) 7SL RNA, (H) tRNAleu, (I) pre-tRNAleu and (J) pre-tRNAile expression relative to the expression of the Pol II target gene hrpt, in pooled 3 dpf (n = 20 for each pool) gtf3c3 F0 KO compared to pooled WT larvae (N = 4–5). All data are represented as the mean ± SEM. Statistical significance was calculated by Mann–Whitney test, or Dunnett's multiple comparisons tests. *P < 0.05, **P < 0.01; #P < 0.005, ***P < 0.001; ****P < 0.0001; ns, not significant. n represents number of fish, N represents number of experimental repeats.