A novel hypervariable variable number tandem repeat in the dopamine transporter gene (SLC6A3)

Abstract

The dopamine transporter gene, SLC6A3, has received substantial attention in genetic association studies of various phenotypes. Although some variable number tandem repeats (VNTRs) present in SLC6A3 have been tested in genetic association studies, results have not been consistent. VNTRs in SLC6A3 that have not been examined genetically were characterized. The Tandem Repeat Annotation Library was used to characterize the VNTRs of 64 unrelated long-read haplotype-phased SLC6A3 sequences. Sequence similarity of each repeat unit of the five VNTRs is reported, along with the correlations of SNP-SNP, SNP-VNTR, and VNTR-VNTR alleles across the gene. One of these VNTRs is a novel hyper-VNTR (hyVNTR) in intron 8 of SLC6A3, which contains a range of 3.4-133.4 repeat copies and has a consensus sequence length of 38 bp, with 82% G+C content. The 38-base repeat was predicted to form G-quadruplexes in silico and was confirmed by circular dichroism spectroscopy. In addition, this hyVNTR contains multiple putative binding sites for PRDM9, which, in combination with low levels of linkage disequilibrium around the hyVNTR, suggests it might be a recombination hotspot.

Figure 1.. UCSC Genome Browser view of SLC6A3 with UCSC’s Simple Repeats track (TRF annotation of sequence) and a custom track showing the TRs detected by the TRAL with non-variable TRs in red and VNTRs in blue.

Figure 2.. Multi-color Mola charts illustrating the degree of variability of each repeat unit in the 5 VNTRs within SLC6A3.

(A) TR21, (B) TR17, (C) TR09, (D) TR22, and (E) TR30. Full names of haplotypes (each row) are available in Table S3B. Haplotypes are sorted based on the number of repeat units. The length of each unit varies because of insertions or deletions.

Figure 3.. Three-color Mola charts illustrating the modes of variability in the five VNTRs within SLC6A3.

(A) TR21, (B) TR17, (C) TR09, (D) TR22, and (E) TR30. Details are as in Fig 2, but the haplotypes are sorted based on similarity of the first repeat units in a haplotype, except for TR21 (A) and TR17 (B), which are sorted by number of repeat units due to their complexity. (Full names of the sequences and their alignment scores for each repeat are available in Table S3C).

Figure 4.. Heatmap of correlations between all SNPs and VNTRs in SLC6A3.

The gene is shown in the 5′ to 3′ orientation above the heatmap, with black lines indicating the location of each SNP. The five VNTRs are denoted with a blue asterisk above the heatmap, with TR09 on the far left and TR30 on the far right. On the heatmap, red indicates a stronger correlation and white indicates a weaker correlation.

Figure 5.. G4-Hunter results for SLC6A3 sequences from the human reference genome (GRCh38/hg38; yellow) and the alternate human reference genome (KI270791v1; purple).

(A) shows the G4 coverage percentage in each section of SLC6A3. (B) shows the G4 count in each section of SLC6A3.

Figure 6.. Circular dichroism results of the TR21 consensus sequence oligonucleotide.

The wavelength is shown on the x-axis with ellipticity on the y-axis. Three buffer solutions (pH 7.2) were used: 1 mM Na-phosphate (black), 1 mM Na-phosphate with 10 mM K-phosphate (orange), and 1 mM Na-phosphate with K-phosphate and K-chloride to 115 mM total K (blue). G4 signature peaks and troughs (210 and 260 nm peaks and a 240 nm trough) are displayed in all three buffers.