Granting an Extension: Phosphorylation of the Pol II CTD Regulates mRNAs Produced by Read-Through from Small Nuclear RNAs

The C-terminal domain (CTD) of RNA polymerase II (Pol II) extends from the catalytic core and consists of repeats of a seven-amino acid motif. The CTD functions in the regulation of Pol II and is subject to just about every protein modification you can think of, including methylation, acetylation, ubiquitination, phosphorylation, and O-GlcNAcylation (reviewed in Harlen and Churchman, 2017). These modifications affect mRNA capping, histone modifications, chromatin structure, and splicing. The various modifications alter the recruitment of RNA-processing enzymes, transcription factors, and chromatin-modifying proteins to Pol II during transcription.

Much to the relief of editors who can’t spell O-GlcNAcylation (not naming any names), most work on the Pol II CTD has examined phosphorylation. Phosphorylation of different residues of the Pol II CTD affects the transition from initiation of transcription (unmodified CTD, interacts with Mediator complex) to elongation (phosphorylated CTD). CTD phosphorylation is regulated by kinases and phosphatases, and perturbation of these enzymes affects the transcriptome. For example, Arabidopsis thaliana CTD PHOSPHATASE-LIKE4 (CPL4) RNA interference lines (CPL4_RNAi) show hyperphosphorylation of the CTD and changes in the levels of more than 200 transcripts. Fukudome et al. (2017) followed up on these studies, finding that GPI19, a gene strongly upregulated in CPL4_RNAi, has an intriguing link to small nuclear RNAs (snRNAs). GPI19, which encodes a phosphatidylinositol N-acetylglucosaminyltransferase, contains a U12 snRNA in its 5′ untranslated region. Wild-type plants produce a U12 pre-snRNA transcript, but CPL4_RNAi plants produce a U12-GPI19 fusion transcript (see figure). Mutation of a conserved element downstream of the snRNA increased expression of a luciferase reporter (U12-LUC) and mutation of an upstream element decreased reporter expression, indicating that read-through from transcription of the U12 snRNA affects regulation of the downstream GPI19 protein-coding gene.

RNA-sequencing shows accumulation of the U12-GPI19 fusion transcript in CPL4RNAi plants. Coverage of RNA-sequencing reads over the U12-GPI19 locus in wild-type and CPL4RNAi plants. Numbers above indicate genomic position and the structure of the U12-GPI19 locus is shown below. (Reprinted from Fukudome et al. [2017], Figure 1G.)

Rather than being an oddity restricted to GPI19, read-through also occurs at other loci: The authors used RNA-sequencing of CPL4_RNAi plants to show that many Pol II-dependent snRNA loci produce 3′ extensions in CPL4_RNAi plants. However, protein-coding mRNAs and Pol III-dependent snRNAs did not produce 3′-extended transcripts. Many of these 3′ extensions of snRNAs produced a fusion to a downstream protein-coding gene. For example, LONG-AFTER FAR-RED3 produces alternative transcripts, one initiating from an upstream snRNA and one initiating downstream of the snRNA; the former transcript accumulated in CPL4_RNAi plants.

Motifs associated with snRNA promoters also occur in transposable elements and these elements can affect the expression of nearby protein-coding genes. For example, SMALL SCP1-LIKE PHOSPHATASE14 (SSP14) has a snRNA-like promoter sequence in an upstream transposable element. In wild-type plants, the SSP14 locus produces only the 5′-part of coding region as an intermediate-length, unstable transcript terminated at an snRNA 3′-processing-signal-like sequence inside of the protein coding sequence. Also, the extended, full-length SSP14 transcript accumulates only in pollen, where high levels of read-through tend to occur, perhaps due to the low level of CPL4 expression. CPL4_RNAi induces ectopic production of the full-length SSP14. These results implicate this mechanism in tissue-specific regulation of gene expression and BLAST searches of ESTs from other plant species indicated that such fusions occur widely in plants, with some conserved combinations, such as U12-GPI19, and other combinations occurring by species-specific snRNA loci found in each genome.

Read-through occurs as part of normal snRNA production; for Pol II-transcribed snRNAs, the Pol II CTD recruits the Integrator complex, which removes the excess 3′ RNA. The observed accumulation of read-through transcripts requires the snRNA transcription and processing machinery, as a mutant affecting this machinery failed to accumulate snRNA fusion transcripts in CPL4_RNAi plants. Pol II CTD phosphorylation also changes in response to environmental factors and the authors found, by searching microarray databases and examining RNA-sequencing data, that salt stress induced the accumulation of fusion transcripts. Indeed, immunoblotting showed that salt-stressed plants accumulated Pol II with reduced CTD phosphorylation, in contrast to CPL4_RNAi plants, which accumulated Pol II with increased phosphorylation. A closer analysis, using antibodies that distinguish phosphorylation on different amino acids of the CTD repeat combined with locus-specific analyses by chromatin immunoprecipitation, showed that this apparent contradiction (since CPL4_RNAi plants and salt stress produce similar effects on transcripts) occurs at the whole cell level; however, Pol II CTD phosphorylation profiles at snRNA loci were consistent between CPL4_RNAi and salt stress and accumulated Pol II with increased phosphorylation.

Although further studies will be required to explore the full implications of this study, the identification of read-through transcripts fusing an upstream snRNA to a downstream protein-coding gene provides an intriguing indication that this mechanism may have key regulatory functions in development and stress responses. Moreover, transposon-mediated rearrangements of regulatory sequences show yet another way that transposons can generate novel effects on gene regulation.