Identification of Alternative Polyadenylation in Cyanidioschyzon merolae Through Long-Read Sequencing of mRNA

Abstract

Alternative polyadenylation (APA) is widespread among metazoans and has been shown to have important impacts on mRNA stability and protein expression. Beyond a handful of well-studied organisms, however, its existence and consequences have not been well investigated. We therefore turned to the deep-branching red alga, Cyanidioschyzon merolae, to study the biology of polyadenylation in an organism highly diverged from humans and yeast. C. merolae is an acidothermophilic alga that lives in volcanic hot springs. It has a highly reduced genome (16.5 Mbp) and has lost all but 27 of its introns and much of its splicing machinery, suggesting that it has been under substantial pressure to simplify its RNA processing pathways. We used long-read sequencing to assess the key features of C. merolae mRNAs, including splicing status and polyadenylation cleavage site (PAS) usage. Splicing appears to be less efficient in C. merolae compared with yeast, flies, and mammalian cells. A high proportion of transcripts (63%) have at least two distinct PAS’s, and 34% appear to utilize three or more sites. The apparent polyadenylation signal UAAA is used in more than 90% of cases, in cells grown in both rich media or limiting nitrogen. Our documentation of APA for the first time in this non-model organism highlights its conservation and likely biological importance of this regulatory step in gene expression.

Spliceosome assembly and regulation: insights from analysis of highly reduced spliceosomes

“…Interestingly, some of these absent proteins and domains, such as the Prp18–Slu7 heterodimer, or the Toggle loop of the Prp8 RNase H domain, are proposed to act as throttles that normally allow the spliceosome to shift between two states: slow and accurate versus fast and inaccurate. The absence of such domains is consistent with the low fraction of spliced transcripts observed in Cm ( Stark et al 2015 ; Schärfen et al 2022 ), although we have not observed evidence of inaccuracy: while some cryptic splice site use is detected by Illumina and PacBio sequencing, there do not appear to be more of these in Cm than in other organisms ( Schärfen et al 2022 ).…”

“…We suggest Cm splicing—and by analogy that of other organisms with reduced spliceosomes—to be inefficient, inaccurate, and post-transcriptional. The inefficiency of Cm splicing contrasts strongly with splicing efficiency in other organisms including yeast ( Stark et al 2015 ; Schärfen et al 2022 ). Furthermore, Cm lacks the DEAD-box RNA helicase Prp28, which is not only associated with release of U1 snRNP, but also with proofreading the 5′SS and splicing fidelity ( Price et al 2013 ).…”

“…The simplified Cm spliceosome offers a glimpse into essential spliceosomal components ( Stark et al 2015 ). Cm lacks the cotranscriptional pre-mRNA splicing regulation offered by either Lea1/Msl1 ( Gunderson and Johnson 2009 ) or Prp45 ( Hálová et al 2017 ; Leung et al 2019 ), consistent with its inefficient splicing ( Stark et al 2015 ; Schärfen et al 2022 ). We assume the first transesterification is inefficient in Cm based on the splicing factors that are absent in the activated Cm spliceosome, particularly Prp17 and NTC proteins.…”

“…While two are ribosomal protein genes, and two others are also related to translation, the remainder have a wide variety of functions including protein folding, primary metabolism, and transport. No intron property yet tested correlates with the level of splicing across intron-containing genes, which is strikingly low ( Schärfen et al 2022 ).…”

Near chromosome–level genome assembly of the microsporidium Hamiltosporidium tvaerminnensis

“…For the analysis of splicing efficiency and APA, we followed the methodology described in Schärfen et al . (2022) .…”