Multicolor Whole-Genome Mapping in Nanochannels for Genetic Analysis

Abstract

Analysis of structural variations (SVs) is important to understand mutations underlying genetic disorders and pathogenic conditions. However, characterizing SVs using short-read, high-throughput sequencing technology is difficult. Although long-read sequencing technologies are being increasingly employed in characterizing SVs, their low throughput and high costs discourage widespread adoption. Sequence motif-based optical mapping in nanochannels is useful in whole-genome mapping and SV detection, but it is not possible to precisely locate the breakpoints or estimate the copy numbers. We present here a universal multicolor mapping strategy in nanochannels combining conventional sequence-motif labeling system with Cas9-mediated target-specific labeling of any 20-base sequences (20mers) to create custom labels and detect new features. The sequence motifs are labeled with green fluorophores and the 20mers are labeled with red fluorophores. Using this strategy, it is possible to not only detect the SVs but also utilize custom labels to interrogate the features not accessible to motif-labeling, locate breakpoints, and precisely estimate copy numbers of genomic repeats. We validated our approach by quantifying the D4Z4 copy numbers, a known biomarker for facioscapulohumeral muscular dystrophy (FSHD) and estimating the telomere length, a clinical biomarker for assessing disease risk factors in aging-related diseases and malignant cancers. We also demonstrate the application of our methodology in discovering transposable long non-interspersed Elements 1 (LINE-1) insertions across the whole genome.

Figure 1.

(A) De novo assembled optical maps of DLE–Cas9 labeled D4Z4 array on chromosome 4q in NA12878. On the top, 4qA haplotype is seen and on the bottom, 4qB haplotype can be seen. The wide green bar at the top denotes the hg38 reference. The wide blue bar below the reference represents consensus contigs from the de novo assembly. Individual molecules are represented by the thin yellow lines arranged under the consensus contigs. Dark blue vertical ticks on the single molecules (yellow lines) indicate labeled DLE sites and the red vertical ticks in the subtelomeric region indicate D4Z4 target-specific red labels. The figures show only a part of all labeled molecules aligned to 4qA and 4qB. (B) Graph of distances between the red labels plotted against their frequencies. Here, the X-axis indicates the distances between the two closest red labels which occurred along the length of the D4Z4 array of a molecule and the Y-axis indicates the frequency of the recorded distances across all mapped molecules.

Figure 2.

(A) De novo assembled optical maps of DLE–Cas9 labeled telomeric repeats array on chromosome 14q (top panel) and 20q (bottom panel) in NA12878. The wide green bar at the top denotes the hg38 reference. The wide blue bar below the reference represents consensus contigs from the de novo assembly. Individual molecules are represented by the thin yellow lines arranged under the consensus contigs. Dark blue vertical ticks on the single molecules (yellow lines) indicate labeled DLE sites and the red vertical ticks at the ends of single molecules indicate telomere red labels. Only a part of all aligned single molecules (yellow lines) are shown in the maps. (B) Plot with measured intensities of red labels at telomere-termini containing single molecules from 14q and 20q arms. Each filled circle represents the total red label intensity of a single molecule. The horizontal bar represents the average measured intensity.

Figure 3.

LINE-1 insertions detected in a Chr4 haplotype using our DLE–Cas9 approach. Both DLE and red labels are stretch-matched in the figure. (A) Haplotype with the 6 kbp line 1 insertion. (B) Second haplotype with no insertion at the same genomic region.