Recurrent Rare Copy Number Variants Increase Risk for Esotropia

Abstract

Purpose: To determine whether rare copy number variants (CNVs) increase risk for comitant esotropia.

Methods: CNVs were identified in 1614 Caucasian individuals with comitant esotropia and 3922 Caucasian controls from Illumina SNP genotyping using two Hidden Markov model (HMM) algorithms, PennCNV and QuantiSNP, which call CNVs based on logR ratio and B allele frequency. Deletions and duplications greater than 10 kb were included. Common CNVs were excluded. Association testing was performed with 1 million permutations in PLINK. Significant CNVs were confirmed with digital droplet polymerase chain reaction (ddPCR). Whole genome sequencing was performed to determine insertion location and breakpoints.

Results: Esotropia patients have similar rates and proportions of CNVs compared with controls but greater total length and average size of both deletions and duplications. Three recurrent rare duplications significantly (P = 1 × 10-6) increase the risk of esotropia: chromosome 2p11.2 (hg19, 2:87428677-87965359), spanning one long noncoding RNA (lncRNA) and two microRNAs (OR 14.16; 95% confidence interval [CI] 5.4-38.1); chromosome 4p15.2 (hg19, 4:25554332-25577184), spanning one lncRNA (OR 11.1; 95% CI 4.6-25.2); chromosome 10q11.22 (hg19, 10:47049547-47703870) spanning seven protein-coding genes, one lncRNA, and four pseudogenes (OR 8.96; 95% CI 5.4-14.9). Overall, 114 cases (7%) and only 28 controls (0.7%) had one of the three rare duplications. No case nor control had more than one of these three duplications.

Conclusions: Rare CNVs are a source of genetic variation that contribute to the genetic risk for comitant esotropia, which is likely polygenic. Future research into the functional consequences of these recurrent duplications may shed light on the pathophysiology of esotropia.

Figure 1.

Rare CNV burden in esotropia cases and controls. Esotropia cases have similar rates (number per person) and proportions (percent of people with at least one) of rare (<1% frequency), >10 kb deletions (A) and duplications (B) to controls. The total length and average size of each CNV, however, are larger in esotropia cases. (C–F) Odds ratios for esotropia given different CNV sizes (C, duplications, E, deletions) and frequencies (D, duplications, F, deletions). Duplications of 500 kb–1 MB and greater than 1 MB and deletions >1 MB were associated with higher risk of esotropia. Frequency of CNVs was not associated with esotropia. * P < 0.015, ** P < 0.0001.

Figure 2.

Segmental tests show three significant duplications. Manhattan plot of segmental association test results representing genome-wide corrected p values calculated at each CNV breakpoint. Black circles represent deletions, and gray circles represent duplications. Three duplications, on chromosomes 2, 4, and 10, were significant to P = 1 × 10⁻⁶. The three circles for chromosome 2 represent the same duplication, which has different breakpoints in different individuals. Some of the individual breakpoints have lower P values.

Figure 3.

Chromosome 2 duplication significantly enriched in esotropia cases. UCSC genome browser plot showing the region of duplication on 2p11.2 (2:87428677-87965359). Duplications across this region were present in 23 cases (light blue, n indicates number with each set of breakpoints) and four controls (dark blue). A nearby deletion was present in one control (dark red). RefSeq genes are listed underneath. Protein coding genes are denoted in blue, lncRNA genes in red, micro RNAs in green, and noncoding RNAs in orange. H3K27Ac mark indicates several putative regulatory regions fall within the duplication. This area is not well conserved over 100 vertebrates, but the genic and putative regulatory regions are well conserved in primates and other mammals. The blue vertical line indicates the position of the ddPCR probe used to confirm the duplication. The gap in the annotations indicates an unmappable area of the reference genome, usually because it is highly repetitive or of low complexity. At bottom are indicated repeats in the region identified by RepeatMasker: SINE, short interspersed nuclear elements; LINE, long interspersed nuclear elements; LTR, long terminal repeat elements; DNA, DNA repeat elements; SIMPLE, microsatellites, low complexity repeats, satellite repeats, RNA repeats, and other repeats.

Figure 4.

Chromosome 4 duplication significantly enriched in esotropia cases. UCSC genome browser plot showing the region of duplication on 4p15.2 (4:25554332-25577184). Duplications across this region were present in 27 cases (light blue, n indicates number with each set of breakpoints) and six controls (dark blue). The duplication includes exon 1 of LOC101929, a long noncoding RNA (orange). H3K27Ac marks show no putative regulatory regions within the duplication. This area is well conserved with monkeys, but not other mammals. There is no conservation with nonmammals. The blue vertical line indicates the position of the ddPCR probe used to confirm the duplication. At bottom are indicated repeats in the region identified by RepeatMasker: SINE, short interspersed nuclear elements; LINE, long interspersed nuclear elements; LTR, long terminal repeat elements; DNA, DNA repeat elements; SIMPLE, microsatellites, low complexity repeats, satellite repeats, RNA repeats, and other repeats.

Figure 5.

Chromosome 10 duplication significantly enriched in esotropia cases. UCSC genome browser plot showing the region of duplication on 10q11.22 (10:47049547-47703870). Duplications across the full 700kb were present in 36 cases (light blue, n indicates number with each set of breakpoints). A smaller, ∼300-kb duplication was present in 28 additional cases and 18 controls (dark blue). A nearby deletion was present in one case (red). RefSeq genes are listed underneath. Protein coding genes are in blue, long noncoding RNA genes are in red, and noncoding RNAs are in orange. H3K27Ac mark indicates several putative regulatory regions fall within the duplication. This area is not well conserved over 100 vertebrates. There is some conservation with monkey, but very little with other animals. The blue vertical line indicates the position of the ddPCR probe used to confirm the duplication. The gap in the annotations indicates an unmappable area of the reference genome, usually because it is highly repetitive or of low complexity. At bottom are indicated repeats in the region identified by RepeatMasker: SINE, short interspersed nuclear elements; LINE, long interspersed nuclear elements; LTR, long terminal repeat elements; DNA, DNA repeat elements; SIMPLE, microsatellites, low complexity repeats, satellite repeats, RNA repeats, and other repeats.

Figure 6.

Breakpoints of chromosome 4 duplication. Whole genome sequences from three unrelated individuals with esotropia who harbor the chromosome 4 duplication show increased sequence coverage across the duplication and split reads at the breakpoints. Top: Schematic of chromosome 4 region of duplication. Middle: Images from integrated genome viewer for each of three individuals. Coverage is indicated for each base pair by the height of the gray bar, split reads are shown in red and green below. Green reads indicate that the paired read maps further away than expected. Bottom: individual reads are shown across the breakpoints. The split reads (colored by base-pair that does not map to the reference sequence) were mapped back to indicate this is a tandem duplication. The left breakpoint (hg38: 4:25,554,985) is just upstream of exon 2 of LOC101929. The right breakpoint (hg38: 4:25,578,843) is in an intergenic region.

Figure 7.

Distribution of subtypes of esotropia. The overall esotropia cohort consists of 14% accommodative (light grey checkerboard), 20% infantile (dark gray stripes), and 66% nonaccommodative, noninfantile (black). Accommodative esotropia is underrepresented in the chromosome 2 and larger chromosome 10 duplications, and overrepresented in the chromosome 4 duplication. P = 0.069, chi square. Numbers of participants in each group are provided in the table.