The promoter function of the yeast forkhead transcription factors Fkh1 and Fkh2 is well established. But now, in Science, Jane Mellor and colleagues report a new role for Fkh1 and Fkh2 — coordinating early transcription elongation and pre-messenger RNA processing.

The authors used chromatin immunoprecipitation (ChIP) to investigate the binding of Fkh1 and Fkh2 to CLB2, which is part of the CLB2 cluster of genes that is expressed in early mitosis. They found that Fkh1 and Fkh2 not only bound the upstream activating sequence, but also the coding region, which indicated that they were involved in transcriptional elongation. Next, Mellor and colleagues looked at RNA polymerase II (RNAPII) distribution on CLB2. In wild-type yeast strains, RNAPII accumulated at the beginning of the coding region — this was also true in strains lacking Fkh2 (fkh2Δ), although there was an expected marked reduction in the amount of RNAPII bound. However, in strains lacking Fkh1 (fkh1Δ), RNAPII accumulated towards the end of the gene, indicating that there was an elongation block involving Fkh1.

Next, the authors used 6-azauracil (6AU), to see if Fkh1 and Fkh2 had opposing effects. 6AU depletes pools of GTP and prevents RNAPII from elongating efficiently, but wild-type strains survive because another protein, Imd2, replenishes GTP pools. fkh2Δ's sensitivity to 6AU, which was overcome with guanine, indicated a positive role in elongation for Fkh2. Conversely, fkh1Δ were resistant to 6AU and could suppress the sensitivity of fkh2Δ. This indicated that these Fkh factors do have opposing functions. Northern blot analysis showed that IMD2 is a target of Fkh regulation, and association of Fkh with the coding region, which paralleled that seen for CLB2, was observed when IMD2 transcription was induced by 6AU. RNAPII distribution at these loci was also Fkh-dependent.

So, how do Fkh1 and Fkh2 influence RNAPII? During elongation, serine (Ser) 5 and Ser2 of the heptad repeat (Tyr-Ser-Pro-Thr-Ser-Pro-Ser)27 of the C-terminal repeat domain (CTD) of RNAPII are differentially phosphorylated — phosphorylation of Ser5 increases and Ser2 decreases between the 5′ and 3′ end of genes — and RNAPII then recruits activities for processing transcripts. In fkh2Δ, phosphorylation of Ser5 and Ser2 was more constant across the gene, indicating that release of RNAPII into elongation phase was defective and that the normal temporal control of Ser phosphorylation might be disturbed. In fkh1Δ, Ser5 phosphorylation was reduced in the promoter region and low throughout the coding region, and Ser2 phosphorylation was almost undetectable.

Mellor and colleagues then decided to look at pre-mRNA 3′-end formation in fkh1Δ strains, because increased Ser2 phosphorylation in wild-type strains is associated with pre-mRNA 3′-end processing. Indeed, they found that 3′-end formation was defective in fkh1Δ. Then, using a transcription run-on assay, they showed that RNAPII in fkh1Δ was defective in pre-mRNA 3′-end formation and predicted that “...much of the RNAPII at the 3′ end is not actively engaged in transcription, a likely consequence of the lack of Ser5 and Ser2 phosphorylation.”

So, the authors suggest that the opposing actions of Fkh1 and Fkh2 at the beginning of genes might be part of an early elongation checkpoint mechanism that has been proposed to coordinate transcription and pre-mRNA processing through phosphorylation of the CTD of RNAPII. And the evolutionary conservation of the Fkh factors indicates that this “...may reflect a general feature of gene regulation in eukaryotes.”