Cold reshapes RNA splicing while plant genes are still being copied

Editing RNA as it emerges

When a gene is transcribed, RNA polymerase II travels along DNA and produces an RNA copy. In many plant genes, that initial copy contains introns: stretches that must be removed as the RNA is processed. Splicing can begin before polymerase has finished copying the gene, but observing several splicing events on the same still-forming molecule has been challenging.

Researchers at the John Innes Centre developed an approach called plant polymerase intact nascent transcript sequencing, or pPOINT-seq, to examine that relationship in Arabidopsis thaliana. Their preprint reports that splicing patterns are associated with where the transcription machinery is found and that those patterns change after prolonged cold exposure. The findings describe associations, not a demonstrated molecular mechanism for the changes.

A snapshot of polymerase-associated RNA

The team isolated RNA associated with a particular modified form of polymerase II, marked by phosphorylation at a site known as Ser5 on its tail. They then used long-read nanopore sequencing to examine individual RNA molecules. The position of an RNA's 3′ end served as a proxy for where the polymerase was on the gene, while the read could also reveal whether multiple introns had been removed.

In three warm-condition biological replicates, the method captured transcripts from 25,010 genes. The researchers focused on reads without a substantial poly(A) tail as a way to study nascent, rather than already cleaved and polyadenylated, transcripts. They also filtered out reads ending near 5′ splice sites to reduce the contribution of splicing intermediates.

The resulting profiles are snapshots of a population, not movies of individual polymerases. Even so, they showed that splicing often occurred while transcription was still under way. More than 40% of introns in one analysis were already spliced when the inferred polymerase position had reached 10% of the first exon downstream; the proportion was approximately 55% by the end of that exon. An exon is a sequence retained when the surrounding introns are removed.

Distant introns, linked patterns

Splicing was not simply a matter of how far polymerase had travelled beyond an intron. Upstream introns were more frequently spliced in transcripts whose inferred polymerase position was in an exon than in those associated with the immediately following intron. The authors propose that different transcription or RNA-processing states could contribute, but the measurements do not establish why the pattern occurs.

The long reads also allowed the team to compare introns far apart on the same RNA. Among 168,501 reads spanning both the first and last introns across 9,250 genes, those introns tended to share a splicing state. Of reads in which the first intron had been removed, 85.9% also had the last intron removed. When the first was retained, 72.9% also retained the last. This concordance suggests transcript-wide coordination, although it does not show that removing one intron causes removal of the other.

In another comparison, the researchers examined 606 genes with enough coverage near and beyond their main transcription end sites. RNAs extending beyond those sites had lower last-intron splicing ratios than RNAs ending within the end-site region. That links incomplete splicing with transcriptional readthrough in the sampled molecules; it does not establish which process influences the other.

What changed in the cold

For the cold experiment, seedlings were grown for 10 days at 20 °C and then transferred to 5 °C for 14 days. Across the polymerase-position categories examined, the cold-exposed samples had higher intron splicing ratios. Of the introns meeting the study's criteria for a significant change, 787 showed increased splicing ratios and 156 showed decreases.

Consecutive unspliced introns also became less common. They appeared in 7.63% of informative reads under warm conditions and 4.09% after cold exposure. Among reads retaining at least one intron, the corresponding proportions were 30.97% and 22.56%. The order inferred for splicing adjacent introns, however, remained largely similar between conditions.

These comparisons have important limits. The cold-treated plants had grown for an additional two weeks, so temperature and developmental age are not cleanly separated by the described comparison. The method also samples one modified polymerase population, and RNA degradation, cleavage or incomplete reverse transcription can affect the observed 5′ ends. Because the measurements are snapshots, higher spliced fractions do not on their own prove faster splicing. The authors suggest several possible explanations, including changes in transcription dynamics or RNA-processing complexes, but testing those possibilities will require further experiments.