Benchmarks

Evidence that the genome is pervasively transcribed into hundreds of thousands different RNAs (1,2) necessitates methods for independent high-throughput verifi-cation of transcriptional starting sites (TSSs) determined by genome-wide approaches. 5′-RACE PCR is a well-estab-lished and widely used method to specif-ically amplify the 5 ′ end of a transcript, facilitating mapping of the TSS and the approximate location of promoter elements (3). Conventionally, this mapping is done by cloning the 5′-RACE PCR product into a bacterial vector and sequencing a few clones by classic electrophoresis-based Sanger sequencing. A great challenge for transcriptional analysis in general is the recently recognized prevalence of mecha-nisms that expand the potential transcript repertoire, such as alternative splicing, alternative promoter usage, and multiple TSSs (1,4). In order for a 5′-RACE PCR assay to reflect such diversity, sequencing a fairly large number of clones is necessary, and when more than one transcript is to be analyzed by 5′-RACE PCR this procedure becomes comprehensive and time-consuming, not to mention costly. To address this drawback, we developed a simple assay in which 5′-RACE PCR products are subjected directly to high-throughput sequencing by the newly developed flow cell sequencing technol-ogies (5) that are currently revolution-izing DNA sequencing by enabling the sequencing of hundreds of millions bases in a single sequencing run. Based on clonal amplification of millions of single surface-immobilized DNA fragments (6,7), flow cell sequencing technologies have been recognized for their aptness for ultra-high throughput approaches such as whole-genome sequencing. However, since it eliminates the need for cloning steps and facilitates quantitative assessment of transcription, the high-throughput sequencing approach is also a very appealing option for single-gene analysis such as with 5′-RACE PCR. To explore the possibility of combining 5′-RACE PCR with high-throughput sequencing, we selected 17 genes or genomic loci displaying expression i