NEW YORK (GenomeWeb News) – New research has used sequencing to provide a direct view of the complex alternative splicing patterns in the human genome.
In a paper published online today in Science Express, German researchers used the approach, called RNA-Seq, to map and compare the transcriptomes in two different human cell types. In the process, they uncovered new transcriptional units — and got an unprecedented picture of the type and frequency of alternative splicing events in these cells.
Senior author Marie-Laure Yaspo, a researcher at the Max Planck Institute for Molecular Genetics, told GenomeWeb Daily News that this improved view of alternative splicing was possible because the RNA-Seq approach gives a “really direct readout of the splicing event.”
It’s been long known that cells use alternative splicing to create alternative RNA transcripts from the same stretch of DNA. But, Yaspo said, until now there are no good tools to directly examine this alternative splicing. “What is known had been known by inference,” she said. For example, in the past, some researchers predicted splicing events from expressed sequence tag analysis and similar studies.
For this study, the researchers used RNA-Seq to investigate both gene expression and alternative splicing in the same data set. “That’s the beauty of this method,” Yaspo said. Indeed, several recently published studies have successfully employed RNA-Seq to decipher the transcriptomes of human cells, mice, Arabidopsis, and yeast.
Yaspo and her colleagues compared the gene expression and splicing profiles in two cell lines: human embryonic kidney cells and human B cells. They selected these cell lines because they were very distinct from one another and, therefore, expected to exhibit very different expression and splicing profiles, Yaspo explained. For each cell type, the researchers used the Illumina Genome Analyzer to sequence complementary DNA derived from polyadenylated RNA.
Using this approach, they detected roughly 10,000 to 12,000 transcribed genes in each cell type — about 25 percent more than detected by microarrays. Of these, about 66 percent corresponded to known genes while the remaining 34 percent represented non-annotated genes. RNA-Seq analysis revealed nearly 4,440 genes whose expression varied by cell type.
The team also identified 94,241 splice junctions — including more than 4,000 new splice junctions — in about 3,106 genes. In both cell types, they found that exon skipping was the most common form of alternative splicing, though they also detected lower levels of alternative 5’ and 3’ splicing.
“We were able to find novel junctions that weren’t seen before,” Yaspo said, something that would have been more difficult or impossible based on splicing predictions alone. “There’s a difference between predictions and a direct readout,” she emphasized.
In the future, Yaspo said, she and her co-workers plan to do similar experiments in additional cell types. Another step, she added, will be looking at gene expression and splicing data together, to get an idea of how the two are related in different cell types.
“I am intrigued and fascinated by this alternative splicing that we can see so clearly by this [RNA-Seq] method,” she said.