Credit: Getty

The relevance of co-transcriptional versus post-transcriptional splicing of transcripts is debated. In a recent RNA sequencing (RNA-seq) study of the human inflammatory response, the authors found evidence for the widespread association of incompletely spliced transcripts with chromatin. Thus, this study supports post-transcriptional rather than co-transcriptional splicing as the predominant mechanism in mice.

Bhatt et al. investigated the transcriptional response to lipid A (LPA) addition in mouse macrophages — this stimulus induces an inflammatory response by means of a defined temporal cascade of genes. The authors carried out RNA-seq in a time course after LPA addition in three subcellular fractions: a chromatin-associated fraction, a nuclear fraction and a cytoplasmic fraction. They were thus able to identify a temporal sequence of transcriptional induction, to follow transport of RNAs to the cytoplasm and then to analyse promoter characteristics that are common between genes exhibiting similar responses.

However, analysis of the chromatin-associated RNA fraction in unstimulated cells along the gene nuclear factor kappa B1 (Nfkb1) gave an unexpected result. There was a consistent distribution of reads across the gene, rather than the expected read gradient from the 5′ end to the 3′ end, thus indicating that transcript release from the chromatin was slow relative to transcription. Furthermore, many of these transcripts were incompletely spliced. This was confirmed in the large majority of both inducible and constitutively expressed transcripts. The chromatin-associated transcripts were also depleted of reads after the poly(A) site, suggesting cleavage, and further analysis with sequenced oligo-dT-primed libraries showed that some of the unspliced chromatin-associated transcripts were polyadenylated, indicating a variable state of polyadenylation of the chromatin-associated transcripts.

Focusing on the transiently induced genes, the authors were able to follow the splicing progress in the time course. At early time points after induction, exons and introns were present in the chromatin-associated fraction, whereas at later time points introns were lost. This suggests a mechanism of post-transcriptional splicing in which the excised introns are released from the chromatin before the spliced transcript.

The authors thus show that, at least in these cells, post-transcriptional splicing is widespread.