A common flanking variant is associated with enhanced stability of the FGF14-SCA27B repeat locus

Abstract

The factors driving or preventing pathological expansion of tandem repeats remain largely unknown. Here, we assessed the FGF14 (GAA)·(TTC) repeat locus in 2,530 individuals by long-read and Sanger sequencing and identified a common 5'-flanking variant in 70.34% of alleles analyzed (3,463/4,923) that represents the phylogenetically ancestral allele and is present on all major haplotypes. This common sequence variation is present nearly exclusively on nonpathogenic alleles with fewer than 30 GAA-pure triplets and is associated with enhanced stability of the repeat locus upon intergenerational transmission and increased Fiber-seq chromatin accessibility.

Extended Data Fig. 1 |. Relationship between 5′-flanking sequences and FGF14 GAA repeat lengths.

a, Schematic representation of the FGF14 gene, isoform 1b with the location of the (GAA)_n·(TTC)_n repeat locus in the first intron. The sequences of the reference 5′-flanking sequence (5′-RFS), the common 5′-flanking variant (5′-CFV; C4 variant), and the C1, C2, C3, C5, and short flanking variant sequences are shown. The sequences of the C1 through C5 variants are highlighted in blue. The sequences are presented relative to the positive strand (genomic context). b, Swarmplot related to Fig. 1c showing repeat lengths as estimated by PacBio HiFi sequencing for 4,382 alleles (from the Genomic Answers for Kids, Care4Rare-SOLVE, and All of Us cohorts) including each of the C1 through C5 sequence variants, separated into subgroups based on the presence of a single terminal adenine (A) or dual terminal adenines (AA). No alleles with C2AA or C5A 5′-flanking variants were found. Three C4 alleles not counted as part of the 5′-CFV group were observed with a single terminal adenine. This plot also extends the y-axis to show the two alleles of over 800 repeat units carrying the 5′-RFS that were not plotted in Fig. 1c for visual clarity. The color of the data points corresponds to the GAA repeat motif purity (a color legend is shown in the top right corner of the plot). The green dashed horizontal line indicates 30 GAA repeats. Abbreviations: 5′-CFV, common 5′-flanking variant; 5′-RFS, reference 5′-flanking sequence.

Extended Data Fig. 2 |. Relationship between 3′-flanking sequences and FGF14 GAA repeat lengths.

Distribution of repeat lengths estimated by PacBio HiFi sequencing for 4,382 alleles (data from the Genomic Answers for Kids, Care4Rare-SOLVE, and All of Us cohorts) in relation to variations observed in the 3′-flanking sequence of the FGF14 repeat locus. The distribution of FGF14 GAA repeat lengths in alleles with the common 3′-flanking sequence, alleles with the single nucleotide variation rs61965263, and alleles with other rare flanking variants is illustrated. The color of the data points corresponds to the GAA repeat motif purity (a color legend is shown in the top right corner of the plot).

Extended Data Fig. 3 |. Analysis of parent-offspring transmission of the FGF14 repeat according to the 5′-flanking sequence variant.

a, Analysis of GAA repeat size changes across 411 intergenerational transmissions (from the Genomic Answers for Kids cohort) as estimated by PacBio HiFi sequencing. Contractions are plotted below the dashed identity line while expansions are plotted above that line. b, Change in GAA repeat length across 411 intergeneration transmission (from the Genomic Answers for Kids cohort) as measured by PacBio HiFi sequencing separated by flanking variant group and parental allele size. The number of intergenerational transmission events in each group is indicated below the x-axis. The y-axis shows the change in repeat length from parent to child. Contractions are plotted below the dashed lines while expansions are plotted above them. Random noise was applied across the x-axis within each category to maximize data visualization. This panel extends Fig. 2b by plotting the 12 additional intergenerational events involving alleles carrying a C1, C2, C5, or other rare 5′-flanking sequence variant. In a and b, red dots are alleles passed from mother to child, while blue dots represent alleles passed from father to child. Abbreviations: 5′-CFV, common 5′-flanking variant; 5′-RFS, reference 5′-flanking sequence.

Extended Data Fig. 4 |. Haplotype analysis of the FGF14 flanking sequence variants.

Haplotype analysis of the FGF14 flanking sequence variants for 1,674 individuals (from the Genomic Answers for Kids and All of Us cohorts). a, Visualization of haplotypes for 1,674 individuals physically phased through the FGF14 repeat locus. Each row represents one of two alleles per individual. A color-coded legend adjacent to the dendrogram indicates the flanking variant group on each allele: green for 5′-CFV, red for 5′-RFS, blue for degenerate 5′-CFV sequences, and cyan for other flanking sequences. The 15 columns in the plot represent the variant status of each allele for 15 common SNVs derived from the 1000 Genomes and Human Genome Diversity Project (HGDP) dataset, with black indicating reference genotype and tan showing alternate genotype. A vertical red line marks the location of the GAA repeat locus. The heatmap on the right side of the plot displays the GAA repeat length of each allele, with yellow indicating larger repeats and dark blue smaller ones. b, Frequency distribution of the 10 major haplotype groups for each flanking sequence variant. The lower-left pie chart shows the distribution of the 10 major haplotype groups in the 1000 Genomes and HGDP dataset. The number of alleles plotted in each pie chart is given above the chart. Abbreviations: 1KG, 1000 Genomes; 5′-CFV, common 5′-flanking variant; 5′-RFS, reference 5′-flanking sequence; HGDP, Human Genome Diversity Project.

Fig. 1 |. A common 5′-flanking sequence variant is associated with smaller FGF14 GAA repeat sizes.

a, Schematic representation of the FGF14 gene, isoform 1b with the location of the (GAA)_n•(TTC)_n repeat locus in the first intron. The sequences of the reference 5′-flanking sequence (5′-RFS) and the common 5′-flanking variant (5′-CFV, in blue) are shown. The sequences are presented relative to the positive strand (genomic context). b, Swarmplot of repeat lengths as estimated by Sanger sequencing for 541 alleles shows that the 5′-CFV is consistently associated with alleles containing fewer than 30 GAA repeats, whereas the 5′-RFS is associated with larger alleles, including pathogenic ones. Each of the two alleles of patients with SCA27B is shown in either the 5′-RFS or 5′-CFV categories, even though only (GAA)_≥250 alleles are pathogenic (all carrying the 5′-RFS). Blue dots represent uninterrupted GAA repeats, whereas red dots represent interrupted GAA repeats. c, Swarmplot of repeat lengths as estimated by PacBio HiFi sequencing for 3,983 alleles (from the Genomic Answers for Kids, Care4Rare-SOLVE and All of Us cohorts) shows a similar pattern. Alleles possessing any other flanking sequences and two alleles of over 800 repeat units carrying the 5′-RFS were omitted for clarity (Extended Data Fig. 1). The color of the data points corresponds to the GAA repeat motif purity (a color legend is shown in the top right corner of the plot). In b and c, the green dashed horizontal line indicates 30 GAA repeats. Abbreviations: 5′-CFV, common 5′-flanking variant; 5′-RFS, reference 5′-flanking sequence; SCA27B, spinocerebellar ataxia 27B.

Fig. 2 |. Intergenerational instability of the FGF14 repeat locus.

a,b, Change in GAA repeat length across intergenerational events, grouped by parental allele size, 5′-flanking sequence group (5′-RFS or 5′-CFV), and repeat purity as estimated by Sanger sequencing for 67 intergenerational events (a) and PacBio sequencing for 399 intergenerational events (from the Genomic Answers for Kids cohort) (b). The y axis shows the change in repeat length from parent to child. Contractions are plotted below the dashed lines, whereas expansions are plotted above them. Random noise was applied across the x axis within each category to improve data visualization. Twelve alleles with flanks other than the 5′-RFS or 5′-CFV were omitted for clarity (Extended Data Fig. 3). Red dots are alleles transmitted from mother to child, whereas blue dots represent alleles transmitted from father to child. Abbreviations: 5′-CFV, common 5′-flanking variant; 5′-RFS, reference 5′-flanking sequence.