Interchromosomal template-switching as a novel molecular mechanism for imprinting perturbations associated with Temple syndrome

Abstract

Background: Intrachromosomal triplications (TRP) can contribute to disease etiology via gene dosage effects, gene disruption, position effects, or fusion gene formation. Recently, post-zygotic de novo triplications adjacent to copy-number neutral genomic intervals with runs of homozygosity (ROH) have been shown to result in uniparental isodisomy (UPD). The genomic structure of these complex genomic rearrangements (CGRs) shows a consistent pattern of an inverted triplication flanked by duplications (DUP-TRP/INV-DUP) formed by an iterative DNA replisome template-switching mechanism during replicative repair of a single-ended, double-stranded DNA (seDNA), the ROH results from an interhomolog or nonsister chromatid template switch. It has been postulated that these CGRs may lead to genetic abnormalities in carriers due to dosage-sensitive genes mapping within the copy-number variant regions, homozygosity for alleles at a locus causing an autosomal recessive (AR) disease trait within the ROH region, or imprinting-associated diseases.

Methods: Here, we report a family wherein the affected subject carries a de novo 2.2-Mb TRP followed by 42.2 Mb of ROH and manifests clinical features overlapping with those observed in association with chromosome 14 maternal UPD (UPD(14)mat). UPD(14)mat can cause clinical phenotypic features enabling a diagnosis of Temple syndrome. This CGR was then molecularly characterized by high-density custom aCGH, genome-wide single-nucleotide polymorphism (SNP) and methylation arrays, exome sequencing (ES), and the Oxford Nanopore long-read sequencing technology.

Results: We confirmed the postulated DUP-TRP/INV-DUP structure by multiple orthogonal genomic technologies in the proband. The methylation status of known differentially methylated regions (DMRs) on chromosome 14 revealed that the subject shows the typical methylation pattern of UPD(14)mat. Consistent with these molecular findings, the clinical features overlap with those observed in Temple syndrome, including speech delay.

Conclusions: These data provide experimental evidence that, in humans, triplication can lead to segmental UPD and imprinting disease. Importantly, genotype/phenotype analyses further reveal how a post-zygotically generated complex structural variant, resulting from a replication-based mutational mechanism, contributes to expanding the clinical phenotype of known genetic syndromes. Mechanistically, such events can distort transmission genetics resulting in homozygosity at a locus for which only one parent is a carrier as well as cause imprinting diseases.

Keywords: Absence of heterozygosity (AOH); Complex genomic rearrangement; DUP-TRP/INV-DUP; ES; Inter-homologous chromosomal template switch; MMBIR; Replicative-based mechanism; Runs of homozygosity (ROH); Triplication.

Fig. 1

Triplication followed by ROH shows DUP-TRP/INV-DUP pattern. a Agilent aCGH (V10.1) plot of chromosome 14q23.2q23.3 region harboring a 2.2-Mb triplication (TRP; log2 ratio 1). b Agilent tiling high-density aCGH revealed a triplication (TRP; blue box) flanked by small duplications (DUP; red boxes) at both centromeric (Jct2) and telomeric junctions (Jct1) (c). d Color-matched sequence alignment of Jct1 and Jct2. Sequencing data of long-range PCR product was obtained by standard Sanger. Structure is now defined as DUP-TRP/INV-DUP. Genomic reference segments are displayed in blue (PROX, proximal) and red (DIST, distal). Strand orientation is indicated in parentheses. Microhomology at the junction is represented in black. e Read depth coverage plot of 17.7  Mb segment of 14q (chr14: 54,179,300-71,874,900) that includes the DUP-TRP/INV-DUP region. Normalized read depth coverage confirms the relative copy-number differences along the CGR consistent with a triplication (31.8/15.8 = 2) flanked by duplications (DUP1 = 20.1/15.8 = 1.3; DUP2 = 18.3/15.8 = 1.2). Y-axis: read depth coverage; X-axis: 100 bp bins at 14q. Genomic coordinates are in hg19

Fig. 2

SNP analysis enabled AOH/ROH detection in addition to a copy-number neutral junction supporting underlying replication-based mechanism. a Agilent aCGH (V10.1) SNP array data (B-allele frequency, BAF) for chromosome 14 indicates a 42.2 Mb of AOH. b Top: pedigree structure of family HOU2583; Bottom: Illumina SNP array HumanOmniExpress-24 Beadchip B-allele frequency (BAF) plot of ~ 42 Mb telomeric segment spanning 14q (chr14:62584057-107287663) reveals a de novo complex genomic rearrangement (CGR) and ROH/AOH exclusively present in BAB7004 (please refer to Additional file 1: Figure S1 for SNP array results in other family members). BAF revealed that the TRP segment presents two distinct genotypes: a small one (92 kb) with unequal allele dosage, followed by a larger one with equal allele dosage (2.2 Mb). Unequal allele dosage TRP, light blue rectangle; equal allele dosage TRP, dark blue rectangle; flanking small DUP, red rectangle; AOH, orange rectangle. c Expected BAF genotypes for distinct copy-number states. d Color-matched schematic model of the 14q23.2q23.3 CGR formation in BAB7004. CGR presents at least three breakpoint junctions, i.e. Jct1, Jct2, Jct3, which are hypothesized to be generated by template switches during replication-based repair (see main text for discussion). Top: genomic coordinates of junctions inferred from multiple technical approaches. Bottom: representation of the SNP allele dosage in each segment involved in this CGR. A, B: SNP alleles. Genomic coordinates are in hg19

Fig. 3

Methylation profile consistent with maternal imprinting in UPD region. a Ideogram of chromosome 14q displays location of DUP-TRP/INV-DUP and AOH segment (top). Two out of three differential methylated regions (DMR) located at 14q32.2 (bottom) were assayed by Infinium Human Methylation450 Beadchip (Illumina). *DMR not assayed. b Beta value (β, i.e., percentage of methylation of a given cytosine) plots: β > 0.8 methylated, 0.2 > β < 0.8 partially methylated, β < 0.2 unmethylated. X-axis indicates genomic coordinates for each CpG dinucleotide assayed by the array for MEG3-DMR and MEG8-DMR (hg19). Color-matched dashed lines display β values for three groups of assayed individuals: blue, UPD(14)pat individuals (BAB7706); green, UPD(14)mat individuals (BAB489, BAB7704, BAB7705); red, controls without UPD(14) (BAB7085, BAB7086, BAB7087, BAB7088, BAB7291, BAB7292). Black line displays BAB7004 results. In both MEG3-DMR and MEG8-DMR, BAB7004 displays methylation profile consistent with UPD(14)mat. c Plot displays the methylation log-likelihood ratio (LLR) for the CpGs spanning the MEG3-DMR locus calculated using Nanopolish [36] in BAB7004 and control NA12878. Positive LLR indicates methylated CpG. Haplotypes are color-matched. For BAB7004, both haplotype alleles show skewed lack of methylation across MEG3-DMR consistent with maternal imprinting whereas control NA12878 shows a haplotype pattern expected for a non-UPD sample. Genomic coordinates are in hg19