Massively parallel resequencing of the isogenic Drosophila melanogaster strain w(1118); iso-2; iso-3 identifies hotspots for mutations in sensory perception genes

Abstract

We used the Illumina reversible-short sequencing technology to obtain 17-fold average depth (s.d. approximately 8) of approximately 94% of the euchromatic genome and approximately 1-5% of the heterochromatin sequence of the Drosophila melanogaster isogenic strain w(1118); iso-2; iso-3. We show that this strain has a approximately 9 kb deletion that uncovers the first exon of the white (w) gene, approximately 4 kb of downstream promoter sequences, and most of the first intron, thus demonstrating that whole-genome sequencing can be used for mutation characterization. We chose this strain because there are thousands of transposon insertion lines and hundreds of isogenic deficiency lines available with this genetic background, such as the Exelixis, Inc., and the DrosDEL collections. We compared our sequence to Release 5 of the finished reference genome sequence which was made from the isogenic strain y(1); cn(1) bw(1) sp(1) and identified 356,614 candidate SNPs in the approximately 117 Mb unique sequence genome, which represents a substitution rate of approximately 1/305 nucleotides ( approximately 0.30%). The distribution of SNPs is not uniform, but rather there is a approximately 2-fold increase in SNPs on the autosome arms compared with the X chromosome and a approximately 7-fold increase when compared to the small 4(th) chromosome. This is consistent with previous analyses that demonstrated a correlation between recombination frequency and SNP frequency. An unexpected finding was a SNP hotpot in a approximately 20 Mb central region of the 4(th) chromosome, which might indicate higher than expected recombination frequency in this region of this chromosome. Interestingly, genes involved in sensory perception are enriched in SNP hotspots and genes encoding developmental genes are enriched in SNP coldspots, which suggests that recombination frequencies might be proportional to the evolutionary selection coefficient. There are currently 12 Drosophila species sequenced, and this represents one of many isogenic Drosophila melanogaster genome sequences that are in progress. Because of the dramatic increase in power in using isogenic lines rather than outbred individuals, the SNP information should be valuable as a test bed for understanding genotype-by-environment interactions in human population studies.

Figure 1

Sequence coverage of 4^th Chromosome. (A) Polytene chromosome map of 4^th chromosome. (B) Coverage of 4^th chromosome in 100 nt increments.

Figure 2

Repetitive Region of 4^th Chromosome That is Not Sequenced. (A) Sequence coverage of region from 690 kb to 710 kb on the 4^th chromosome in 100 nt increments. (B) Sequence coverage of same region in 100 nt increments. (C) Genome map shows that this region contains a rover retrotransposon flanked by long terminal repeats (LTR). Notice that there is no sequence coverage of the repetitive sequence.

Figure 3

Sequence of region on X chromosome near the white gene. (A) Notice that there is a ~9 kb region of no sequence coverage (double arrow). (B) genome map shows that the first exon of the white gene is deleted in the w¹¹¹⁸; iso-2; iso-3 strain.

Figure 4

Sequence coverage in region immediately flanking the w¹¹¹⁸ deletion breakpoints. (A) 5′ end of ~9 kb deletion of the first exon of the white gene (See Fig. 3). Shown is the last sequenced nucleotide with a perfect match (arrow). (B) 3′ end of ~9 kb deletion. Shown is the first sequenced nucleotide with a perfect match (arrow). Notice that the sequence coverage decrease starting ~35 nt from the deficiency breakpoint (see text).

Figure 5

SNP distribution in the D. melanogaster genome. (A–F) SNPs/10 kb on the 2^nd (2L, 2R), 3^rd (3L, 3R), X, and 4^th chromosomes arranged in order along the chromosomes. Notice the dips near the ends of the chromosomes.

Figure 6

SNP-SNP Distances. (A) Distance between SNPs (nt) versus the number of genes. The first bar represents the number of genes with an average SNP-SNP distance between 33 nt and 100 nt (0 nt < SNP-SNp < 100 nt), the second bar represents 100 nt < SNP-SNp < 200 nt, etc. (B) Genes with SNP-SNP distances between 4 kb and 100 kb. The first bar represents 4 kb < SNP-SNp < 5 kb, etc.

Figure 7

SNP distribution near SNP hotspot and coldspot genes. (A and B) SNP hotspot genes Gr23a and Or67c, which are both involved in sensory transduction, and 10 genes on either side are shown. Notice the peak near the hotspot genes. (C and D) SNP coldspot genes cnc and rno, which are both developmental genes, and 10 genes on either side are shown. Notice the trough near the coldspot genes.

Figure 8

SNPs in the gene Obp57d and the number of SNPs per probeset. (A) Location of 9 SNPs in Obp57d. Synonymous mutations do not change the amino acid (L, Q, C, K) and non-synonymous mutations are indicated with an asterisk (I*, I*, T*, N*). There is also a t > c SNP in the intron (lower case letters).