Berit Lilje:
|
With the introduction of high throughput sequencing, it became possible to simultaneously study all transcribed genes in an unbiased manner. Sequence-based gene expression methods works by extracting and sequencing all transcripts present in a cell population. One such method is Cap Analysis of Gene Expression (CAGE) where the first ~20 base pairs of all capped transcripts are sequenced. This method has revealed that most genes have more than one transcription start site. The FANTOM consortium has used CAGE to map transcription start sites from a large collection of primary cells, tissues and cell-lines covering the entire human body. This provides a unique dataset to study gene expression, with promoter level precision. Here we use this large collection of data to study alternative transcription start site usage throughout the human body. We find that many alternative transcription start sites are downstream of annotated coding sequence, an in some cases even downstream of entire protein domains. If transcripts originating from these alternative transcription start sites are translated to protein, the resulting protein would lack amino acids and might have a different function from the full-length version. Our results suggest alternative transcription start site usage currently is underappreciated, since these start sites often show high expression compared to canonical start sites. We further show that the usage of alternative transcription start sites can both be tissue specific, as well as specific to certain differentiation states. By analysing CAGE data from two different inflammatory responses we also find that alternative transcription start site usage is an intrinsic part of the inflammatory response. CAGE was used to map transcription start sites in Caco-2 cells stimulated with tumour necrosis factor alpha, and on tissue from mice subjected to carbon nanotubes. Both systems show a strong response, with especially alternative transcription start sites being differentially regulated. Taken together this shows that alternative transcription start sites add an important layer to gene expression, both in healthy tissues and in response to inflammation.