Genetic and Molecular Characterization of Treacher Collins Syndrome in Three Mexican Families

Abstract

Treacher Collins syndrome (TCS) is a rare disorder within the group of mandibulofacial dysostoses, occurring in 1 in 50,000 live births. It is characterized by anomalies in the maxillary, mandibular, and stapes bones, among others. TCS is caused by pathogenic variants in the TCOF1, POLR1D, POLR1C, and POLR1B genes with autosomal dominant or recessive inheritance patterns. Genetic data from Latin American populations remain scarce. Eleven patients from three different families were recruited. Whole-exome sequencing (WES) was performed on the probands to identify genetic variants, followed by Sanger sequencing for variant validation and familial segregation analysis. Finally, three-dimensional protein structures of wild-type and mutant proteins were predicted. In Family 1, a heterozygous pathogenic splice-site variant in the TCOF1 gene, c.4345 + 1 G > A, was identified and inherited from her mother. In Family 2, a heterozygous pathogenic variant in the TCOF1 gene, c.226_227insC (p.R77fs), was identified and inherited from the paternal lineage. In Family 3, a heterozygous pathogenic POLR1D variant, c.290_291delAG (p.G99fs), was identified among multiple affected relatives; direct parent-of-origin could not be established due to unavailability of one parent, but segregation supports autosomal dominant transmission across three generations. All findings were validated by Sanger sequencing. Our findings highlight the utility of WES for the molecular diagnosis of TCS and underscore the importance of including underrepresented populations in genetic studies to improve diagnosis, genetic counseling, and perinatal planning in at-risk pregnancies.

Keywords: POLR1B; POLR1C; POLR1D; TCOF1; TCS; Treacher Collins syndrome; ribosomopathies; whole exome sequencing.

Figure 1

Pedigree of the three families studied. (A) Family 1 includes three patients with different clinical manifestations. Patient 1 (III:4). Patient 2 (III:5). Patient 3 (II:4). (B) Family 2 comprises three patients with varied clinical manifestations. Patient 4 (IV:6). Patient 5 (III:8). Patient 6 (V:1). (C) Family 3 consists of several patients with clinical manifestations of TCS. Patient 7 (V:7). Patient 8 (V:8). Patient 9 (V:11). Patient 10 (V:16). Patient 11 (IV:26). WG, weeks of gestation; SAB, miscarriage; D&C: dilatation and curettage; ↗ indicate probands (P), and the symbol “?” denotes an unknown or unverified familial branch. Each colored box indicates different diseases presented by different members of each family. All vector graphics were created using Inkscape v1.4.3 (Inkscape Project, https://inkscape.org; accessed on 18 December 2025). All graphics were processed using GIMP v3.0.8 (GIMP Development Team, https://www.gimp.org; accessed on 18 December 2025).

Figure 2

Molecular and in silico splicing analysis of the TCOF1 c.4345 + 1 G > A variant. (A–C) Sanger sequencing electropherograms confirming the heterozygous canonical splice donor variant c.4345 + 1 G > A in intron 24 of TCOF1 in three affected individuals: (A) Patient 3 (mother of the proband, individual II:4), (B) Patient 1 (proband, individual III:4), and (C) Patient 2 (brother of the proband, individual III:5). (D) In silico predictions of the wild-type (WT) at left and mutant (MT) on right splice site of exon 24 generated with ESEFinder 3.0; possible 5′ splice sites are shown in the red bar graph, possible 3′ splice sites are shown in the pink bar graph; green arrows represent the canonical splice sites, in c.4345 marked with a blue arrow represents a possible loss of donor splice site. (E) Comparative representation of predicted splicing patterns based on in silico analysis. In the WT transcript, exon 24 is correctly spliced between exons 23 and 25 (constitutive splicing; green line). In the MT context, loss of canonical donor site recognition is strongly predicted by SpliceAI (donor loss score = 0.99), with exon 24 skipping as the most likely aberrant outcome (blue dashed line). Additional predicted alternative splicing events include partial intron 24 retention (purple line) and activation of a cryptic donor splice site approximately 229 bp upstream within exon 24 (red dashed line). Electropherograms generated and exported with SnapGene v8.2.2 (free Viewer mode; Dotmatics, https://www.snapgene.com/snapgene-viewer, accessed on 4 August 2025). All vector graphics were created using Inkscape v1.4.3 (Inkscape Project, https://inkscape.org). All raster graphics were processed using GIMP v3.0.8 (GIMP Development Team, https://www.gimp.org).

Figure 3

Analysis of the c.226_227insC (p.R77fs) genetic variant. Sanger sequencing results of exon 3: (A) Patient 4 (IV:6), (B) Patient 5 (III:8), and (C) Patient 6 (V:1). The heterozygous variant in TCOF1 gene was confirmed in Patients 4 and 5. (D) AlphaFold2-predicted structure of wild-type treacle. The mutation site (Arg77) is annotated in yellow (arrow). The LisH domain is shown in white, and the central Treacle repeat domain is shown in pink. (E) AlphaFold2-predicted structure of the mutant treacle (p.Arg77Ilefs*97). The frameshift at Arg77 is annotated in yellow (arrow) and is predicted to cause premature truncation at ~aa 173 in the reference transcript, with loss of the central Treacle repeat domain (pink). The resulting truncated C-terminal segment is rendered in red; the LisH domain (white) is partially retained. (F) Comparative analysis of mean raw read counts across exons 1–9 in POLR1C, showing exon-level coverage in the affected individual (Patient 6, red arrows) compared with unaffected controls and patients with Treacher Collins syndrome carrying variants in other genes. Bars represent the average read counts per exon. The red dashed horizontal line indicates the expected mean read depth calculated from healthy controls and TCS controls included in the study, serving as a reference for comparative exon coverage assessment. (D,E): yellow = mutation-site residue; white = LisH domain; pink = central Treacle repeat domain (present in WT, absent in mutant); red = truncated C-terminal segment in the mutant model. Note: These are predicted structural models (used for visualization/interpretation alongside genetic evidence), not experimental structures. Electropherograms generated and exported with SnapGene v8.2.2 (free Viewer mode; Dotmatics, https://www.snapgene.com/snapgene-viewer, accessed on 4 August 2025). Protein three-dimensional structures were predicted with AlphaFold2 (ColabFold v1.5.5, https://colab.research.google.com/github/sokrypton/ColabFold/blob/main/AlphaFold2.ipynb, accessed on 4 August 2025), an open-source tool for high-accuracy protein structure prediction [28]. Predicted structures were obtained from the AlphaFold Protein Structure Database [29]. Three-dimensional visualization and rendering were performed using UCSF ChimeraX v1.8 (https://www.cgl.ucsf.edu/chimerax/, accessed on 4 August 2025). R Core Team. (2024). R: A language and environment for statistical computing (Version 4.3.3). R Foundation for Statistical Computing. All vector graphics were created using Inkscape v1.4.3 (Inkscape Project, https://inkscape.org). All raster graphics were processed using GIMP v3.0.8 (GIMP Development Team, https://www.gimp.org).

Figure 4

Analysis of the variant c.290_291delAG (p.G99fs) in POLR1D gene. The heterozygous variant in POLR1D gene in exon 2 was confirmed. (A) Unaffected relative (IV:7), (B) Unaffected relative (IV:22), (C) Patient 11 (IV:26), (D) Patient 7 (V:7), (E) Patient 8 (V:8), (F) Unaffected relative (V:9), (G) Patient 9 (V:11), (H) Unaffected relative (V:12), (I) Patient 10 (V:16), and (J) Unaffected relative (VI:1). (K) AlphaFold2-predicted structure of wild-type POLR1D. The RBP11-like dimerization domain (aa 39–112) is shown in blue. The Gly99 residue (mutation site) is annotated in yellow (arrow). The distal C-terminal segment is rendered in green. (L) AlphaFold2-predicted structure of the mutant POLR1D (p.Gly99Ilefs*2). The frameshift at Gly99 is annotated in yellow (arrow) and introduces a premature stop two codons downstream (predicted total length ~100 aa), thereby truncating the C-terminus and disrupting the RBP11-like dimerization domain (blue). The residual terminal segment is rendered in red for clarity. (K,L): blue = RBP11-like dimerization domain (aa 39–112); yellow = mutation-site residue (Gly99); green = wild-type distal segment; red = residual truncated segment in the mutant model. Note: Orientation differences between WT and mutant renderings reflect visualization of truncation and do not imply experimentally validated dynamic conformational changes. Electropherograms generated and exported with SnapGene v8.2.2 (free Viewer mode; Dotmatics, https://www.snapgene.com/snapgene-viewer, accessed on 4 August 2025). Protein three-dimensional structures were predicted with AlphaFold2 (ColabFold v1.5.5), an open-source tool for high-accuracy protein structure prediction [28]. Predicted structures were obtained from the AlphaFold Protein Structure Database [29]. Three-dimensional visualization and rendering were performed using UCSF ChimeraX v1.8 (https://www.cgl.ucsf.edu/chimerax/, accessed on 4 August 2025). All vector graphics were created using Inkscape v1.4.3 (Inkscape Project, https://inkscape.org). All raster graphics were processed using GIMP v3.0.8 (GIMP Development Team, https://www.gimp.org).