Multiplex sequencing of paired-end ditags (MS-PET): a strategy for the ultra-high-throughput analysis of transcriptomes and genomes

Abstract

The paired-end ditagging (PET) technique has been shown to be efficient and accurate for large-scale transcriptome and genome analysis. However, as with other DNA tag-based sequencing strategies, it is constrained by the current efficiency of Sanger technology. A recently developed multiplex sequencing method (454-sequencing) using picolitre-scale reactions has achieved a remarkable advance in efficiency, but suffers from short-read lengths, and a lack of paired-end information. To further enhance the efficiency of PET analysis and at the same time overcome the drawbacks of the new sequencing method, we coupled multiplex sequencing with paired-end ditagging (MS-PET) using modified PET procedures to simultaneously sequence 200,000 to 300,000 dimerized PET (diPET) templates, with an output of nearly half-a-million PET sequences in a single 4 h machine run. We demonstrate the utility and robustness of MS-PET by analyzing the transcriptome of human breast carcinoma cells, and by mapping p53 binding sites in the genome of human colorectal carcinoma cells. This combined sequencing strategy achieved an approximate 100-fold efficiency increase over the current standard for PET analysis, and furthermore enables the short-read-length multiplex sequencing procedure to acquire paired-end information from large DNA fragments.

Figure 1

Schematic overview of the MS-PET sequencing strategy. (A) The outline procedure showing the construction of diPETs, which were subjected to multiplex sequencing. (B) Structural details of a diPET. The numbers 5 and 3 represent bases within the 5′ and 3′ signatures, respectively, in each PET component. The orientations of cDNAs are indicated by the ‘AA’ remaining after poly(A) tail removal.

Figure 2

Validation of MS-PET-identified transcripts by quantitative real-time RT–PCR. Columns represent Mean Ct values (normalized against that of Actin) ±SD of each of 11 amplicons (n = 3). PET counts for each candidate transcript are shown in italics above each column. IFGBP (Interferon-gamma binding protein); SFXN4 (sideroflexin 4); TRUB2 (TruB pseudouridine synthase homolog 2); AARS (alanyl-tRNA synthetase); BRCA1 (breast cancer 1, early onset); PP5 (protein phosphatase 5, catalytic subunit); CBX3 (Chromobox protein homolog 3); SSR2 (signal sequence receptor); SET (SET translocation); CTSD (cathepsin D lysosomal aspartyl peptidase); TFF1 (Trefoil factor 1); Actin (reference), PET counts = 33.

Figure 3

Examples of genes identified by MS-PET analysis. (A) Novel gene discovery. PET_ID# 48 955.1 (green arrowed line) identifies a novel gene transcript on chromosome 4, and is verified by PCR [inset; F, flanking (primary); N, nested (secondary)] and DNA sequencing of the amplicon (black arrowed lines A07FF and A07FR). (B) Validation of a predicted gene. PET_ID# 282 423.1 (green arrowed line) identifies a predicted gene on chromosome 4, and is verified by PCR [inset; F, flanking (primary); N, nested (secondary)] and DNA sequencing of the amplicon (black arrowed blocks G06FF and G06FR).