The yeast checkpoint kinase Mec1p functions in transcription termination by facilitating recruitment of Pcf11p and regulating the torpedo exonuclease Rat1p

Mec1p is required for recruitment of Pcf11p to 3’ ends of PMA1 and PYK1

Our previous results indicated that inactivation of checkpoint kinase Mec1p, even in the absence of exogenous genotoxic stress, results in slightly reduced efficiency of transcription termination and pre-mRNA cleavage at the pA sites²³. The defect in transcription termination was relatively mild and reproducibly observable only in cells expressing GAL1-ADH4 construct. The advantage of this construct is that there are no transcription units within 4.7 kb of 3’-UTR of ADH4, avoiding the interference from active transcription in close proximity. In addition, transcription of ADH4 is driven by strong inducible GAL1 promoter, allowing accurate measurements of RNAPII occupancies within 3’-UTR of ADH4⁸.

To test whether Mec1p promotes transcription termination by enhancing recruitment of the scaffold subunit of the CF IA complex Pcf11p to the 3’ ends of gene bodies, we performed a chromatin immunoprecipitation (ChIP) experiment to determine occupancy of Pcf11p within gene bodies and 3’-UTRs of PMA1 and PYK1 (Fig. 1A). PMA1 is especially suitable for this analysis, since there are no transcription units within 1 kb of 3’-UTR of PMA1 and the transcription termination of PMA1 can be analyzed in the absence of any interference from neighboring genes. Pcf11p is a key factor with roles in several nuclear processes: pre-mRNA 3’ processing, transcription termination, gene looping, and mRNA export^24–32. Pcf11p has distinct domains that can be functionally uncoupled³³. Pcf11p interacts with the C-terminal domain (CTD) of the largest RNAPII subunit Rpb1p through its CTD interaction domain (CID) in the N-terminal region. The CID domain is followed by sequences that mediate interaction with Rna14p and Rna15p subunits of the CF IA complex. The C-terminal region features two zinc-binding domains that bracket Clp1p interaction domain. As previously reported (14), the maximum occupancy of Pcf11p was found downstream of the pA sites of both PMA1 and PYK1. Importantly, the Pcf11p occupancies were significantly decreased in mec1∆sml1∆ cells (mec1Δ cells are viable only if harboring the sml1Δ mutation), primarily downstream of the pA sites (Fig. 1A).

Pcf11p interacts with the CTD of RNAPII through its CID domain. The affinity of this interaction is enhanced by phosphorylation of Ser2 of the CTD and the recruitment of Pcf11p to 3’ ends of genes is facilitated by Ser2 CTD phosphorylation^34–37. To explore the mechanism responsible for the reduced recruitment of Pcf11p to 3’ ends of PMA1 and PYK1 genes in mec1∆sml1∆ cells, we measured occupancy of RNAPII phosphorylated at Ser2 of the CTD (Ser2-P) and occupancy of the total RNAPII throughout PMA1 and PYK1 genes. To normalize the Ser2-P signal to the level of total RNAPII, we calculated Ser2-P/total RNAPII ratio (Fig. 1B,C). In agreement with previous reports^35,38–40 our results show that Ser2-P levels and Ser2-P/total RNAPII ratios are low at the 5’ ends of genes and significantly increase downstream of the 5’ ends of gene bodies of both PMA1 and PYK1 (Fig. 1B,C). Importantly, however, both Ser2-P levels and Ser2-P/total RNAPII ratios are slightly elevated in mec1∆sml1∆ cells, indicating, that the observed defect in Pcf11p recruitment is not due to the reduced level of RNAPII phosphorylated at Ser2 of the CTD in mec1∆sml1∆ cells. The decreased occupancy of Pcf11p downstream of the pA sites in mec1∆sml1∆ cells cannot be also attributed to the reduced Pcf11p level, since the cellular levels of Pcf11p do not significantly differ in wild-type and mec1∆sml1∆ cells (Fig. 1D).

Inactivation of MEC1 suppresses growth defect and pre-mRNA processing defects in cells with reduced expression of PCF11 and RNA14, respectively

To further explore the role of Mec1p in transcription termination and pre-mRNA processing, we introduced mec1Δsml1Δ mutations in strains expressing PCF11, RNA14, RNA15, and YSH1 from regulatable tetO₇ promoter. Expression of genes under the tetO₇ promoter is repressed by addition of doxycycline in a concentration-dependent manner, whereas doxycycline has no effect on gene expression in the wild-type cells^41,42. Together with PCF11, RNA14 and RNA15 are subunits of the CF IA complex and YSH1, subunit of the CPF complex, is the endonuclease responsible for pre-mRNA cleavage. We hypothesized that Mec1p inactivation will enhance the growth, transcription termination, and pre-mRNA processing defects caused by the reduced expression of the CF IA and/or CPF subunits. Contrary to our expectations, however, inactivation of Mec1p suppressed growth defect of tetO₇-PCF11 cells in the presence of doxycycline at 30 µg/ml or 100 µg/ml (Fig. 2A). Decreased growth rate of tetO₇-RNA14 or tetO₇-RNA15 cells was not suppressed by introducing mec1Δsml1Δ mutations (Fig. 2A).

To measure the efficiency of cleavage at the pA site of PMA1 and PYK1 pre-mRNAs, we measured the level of pre-mRNA transcripts that were not cleaved at the pA site by RT-qPCR with primer pairs that span the pA site (Fig. 2B,D). To account for the differences in the pre-mRNA levels among individual strains, we normalized the results with mRNA levels upstream of the pA site (B:A ratio, Fig. 2B,D). In addition, we used primers downstream of the pA site to assess the Rat1p-mediated degradation of the RNA created by the cleavage of the nascent transcript (C:A ratio, Fig. 2C,E). The results showed that the efficiencies of PMA1 and PYK1 pre-mRNA cleavage at the pA site and degradation of the RNA downstream of the pA site were significantly decreased in tetO₇-RNA14 and tetO₇-RNA15 cells and this effect was suppressed by Mec1p inactivation (Fig. 2B–E). These results were quite unexpected and possibly indicate that Mec1p inactivation promotes recruitment and/or activity of a protein that is recruited to the CF IA and/or CPF complexes and supports pA site cleavage and degradation of the RNA downstream of the pA site. Surprisingly, the cleavage at the pA site and degradation of the RNA downstream of the pA site were not decreased in tetO₇-PCF11 cells (Fig. 2B–E). These results indicate that unlike reduced expression of RNA14 and RNA15 in tetO₇-RNA14 and tetO₇-RNA15, the lower expression of PCF11 in tetO₇-PCF11 cells does not likely compromise the assembly and/or recruitment of the CF IA and CPF complexes to such an extend that would impact the efficiency of pA site cleavage (Fig. 2B–E). Since Mec1p appears to promote transcription termination by enhancing recruitment of Pcf11p to 3’ ends of gene bodies and 3’-UTRs (Fig. 1A), the finding that inactivation of Mec1p suppresses growth defect and reduced efficiency of cleavage at the pA site in cells with reduced expression of PCF11, RNA14 and RNA15, respectively, was quite unexpected.

Inactivation of MEC1 suppresses transcription termination defects in cells with reduced expression of PCF11 or RNA14

To determine whether inactivation of Mec1p in tetO₇-PCF11 or tetO₇-RNA14 cells also suppresses the defect in transcription termination, we measured RNAPII occupancies within gene bodies and 3’- UTR of PMA1 and PYK1 (Fig. 3A,C). To compare the transcription termination defects in different strains, we calculated the ratio of the RNAPII occupancy downstream of the pA site to RNAPII occupancy upstream of the pA site in PMA1 (6:3 ratio, Fig. 3B) and PYK1 (4:3 ratio, Fig. 3D). The results show that, as expected, the efficiency of transcription termination is significantly reduced in the tetO₇-PCF11 and tetO₇-RNA14 cells in comparison with the wild-type cells, as indicated by the elevated RNAPII occupancies downstream of the pA sites in the 3’-UTRs of both PMA1 and PYK1 (Fig. 3). This termination defect is more pronounced in tetO₇-PCF11 cells. Significantly, Mec1p inactivation suppresses the transcription termination defect in both tetO₇-PCF11 or tetO₇-RNA14 cells, as indicated by the reduced RNAPII occupancies downstream of the pA sites in mec1Δsml1ΔtetO₇-PCF11 and mec1Δsml1ΔtetO₇-RNA14 cells in comparison with tetO₇-PCF11 or tetO₇-RNA14 cells, respectively (Fig. 3).

These results show that the lower expression of PCF11 in tetO₇-PCF11 cells phenotypically separates two roles of Pcf11p: it does not affect the pA site cleavage, but significantly affects transcription termination. These results indicate that Mec1p inactivation likely suppresses the termination defect in tetO₇-PCF11 cells at a step subsequent to recruitment of the CF IA and CPF complexes to pre-mRNA and following the pA site cleavage. The suppression of the transcription termination defect (Fig. 3) provides the most likely molecular mechanism responsible for the suppression of the growth defect in tetO₇-PCF11 cells by Mec1p inactivation (Fig. 2A).

Inactivation of MEC1 does not suppress transcription termination defects of pcf11-2 and pcf11-9 cells

Since the suppression of growth and transcription termination defects in tetO₇-PCF11 cells by Mec1p inactivation was quite unexpected, we wanted to confirm this result using well characterized pcf11 mutations. To this end, we introduced pcf11-2 and pcf11-9 mutations in mec1Δsml1Δ cells. pcf11-2 mutation impairs pre-mRNA cleavage and pcf11-9 mutation impairs pre-mRNA cleavage as well as RNAPII CTD binding and transcription termination³³. Importantly, both pcf11-2 and pcf11-9 mutations impair recruitment of Rat1p to the pA sites⁸.

To address whether inactivation of Mec1p in pcf11-2 and pcf11-9 cells affects the efficiency of cleavage at the pA site of PMA1 and PYK1 pre-mRNAs, we measured the level of pre-mRNA transcripts that were not cleaved at the pA site (Fig. 4A,C) as well as the RNA levels downstream of the pA site (Fig. 4B,D). As expected, the results showed that the defect in PMA1 and PYK1 pre-mRNA cleavage at the pA site in pcf11-2 and pcf11-9 cells were significantly elevated in comparison with wild-type cells (Fig. 4A,C). Inactivation of Mec1p did not significantly affect the pre-mRNA cleavage at the pA sites and exonucleolytic degradation of RNA at 3’ of the pA site in mec1Δsml1Δpcf11-2 and mec1Δsml1Δpcf11-9 cells (Fig. 4A–D).

To determine whether inactivation of Mec1p in pcf11-2 and pcf11-9 cells suppresses the defect in transcription termination, similarly to the situation observed in mec1Δsml1ΔtetO₇-PCF11, we measured RNAPII occupancies within gene bodies and 3’ UTR of PMA1 and PYK1 (Fig. 4E–H). Surprisingly, the results showed that the transcription termination defect in pcf11-2 and pcf11-9 cells was not suppressed by Mec1p inactivation.

Why does Mec1p inactivation suppress the transcription termination defect in tetO₇-PCF11 cells but not in pcf11-2 and pcf11-9 cells? We hypothesized that a possible reason is the fact that while Pcf11p interacts with Rat1p and facilitates its recruitment to the pA sites⁸, pcf11-2 and pcf11-9 mutations disrupt this interaction and compromise the recruitment of Rat1p to the pA sites⁸. In this scenario, Pcf11p expressed in tetO₇-PCF11 cells is still competent for recruitment of Rat1p, and Mec1p inactivation promotes recruitment and/or activity of Rat1p and thus suppresses the termination defect in tetO₇-PCF11 cells. If Mec1p inactivation promotes Rat1p recruitment to the pA sites and/or activity of Rat1p, then we would expect that Mec1p inactivation suppresses also the pre-mRNA processing and transcription termination defect in rat1-1 cells.

Inactivation of MEC1 suppresses pre-mRNA processing and transcription termination defects of rat1-1 cells

To determine whether Mec1p affects the transcription termination and pre-mRNA processing function of the torpedo exonuclease Rat1p, we first measured the efficiency of cleavage at the pA site and degradation of the RNA downstream of the pA site in PMA1 and PYK1 (Fig. 5A–D). Since Rat1p and Xrn1p are two yeast 5’ to 3’ RNA exonucleases that display a significant sequence homology and partly functionally overlap^43,44, we included Xrn1p in these experiments. Rat1p and Xrn1p localize to different subcellular compartments: while Xrn1p is primarily cytosolic, Rat1p is located in the nucleus. When targeted to correct subcellular compartment, Xrn1p and Rat1p can functionally replace each other. Primary function of Xrn1p is 5’ to 3’ exonucleolytic decay of deadenylated mRNAs in the cytosol⁴⁵. However, Xrn1p shuttles between cytosol and nucleus^46,47, where it acts as a transcription activator^48,49 and is involved in rRNA and snoRNA processing⁴³.

The majority of the in vivo studies of Rat1p function are based on the rat1-1 mutant. The rat1-1 mutant was isolated in a screen for mutations that exhibit defect in export of mRNA from the nucleus⁵⁰. It contains a single mutation outside of the enzyme’s exonuclease domain (Y657C). Growth and transcription termination phenotypes of rat1-1 cells are not rescued by the coexpression of the catalytically inactive rat1-D325A mutant^8,13, suggesting that 5’-3’ exonuclease activity is essential for Rat1p function and that this activity is, at least partly, compromised in the rat1-1 cells.

Previous report demonstrated that the efficiency of the pA site cleavage is reduced in rat1-1 cells and that the RNA downstream of the pA site is degraded by both Rat1p and Xrn1p⁸. Consistently with this report, we observed a very significant defect in the pA site cleavage in rat1-1 cells (Fig. 5A,C) and great stabilization of the RNA downstream of the pA site in rat1-1xrn1∆ cells (Fig. 5B,D). As expected, our results showed that Mec1p inactivation suppressed the defects of rat1-1 cells in pre-mRNA cleavage at the pA site (Fig. 5A,C), as well as degradation of the RNA downstream of the pA site in both rat1-1 and rat1-1xrn1∆ cells (Fig. 5B,D). In agreement with the suppression of the pre-mRNA processing defect in rat1-1 cells by Mec1p inactivation, the transcription termination defect of rat1-1 cells of both PMA1 and PYK1 was suppressed in mec1∆sml1∆rat1-1 cells as indicated by significantly reduced RNAPII occupancies at positions 4, 5, and 6 of PMA1, and position 4 of PYK1 in mec1∆sml1∆rat1-1 in comparison with rat1-1 cells (Fig. 5E,G). Accordingly, the ratios of the RNAPII occupancies at positions 6 and 3 of PMA1 and at positions 4 and 3 of PYK1 were significantly reduced in mec1∆sml1∆rat1-1 in comparison with rat1-1 cells (Fig. 5F,H).

Inactivation of MEC1 suppresses ribosomal RNA and snoRNA processing defects of rat1-1 cells

In addition to its role in transcription termination, Rat1p plays number of other important roles in RNA metabolism, including processing rRNAs and small nucleolar RNAs (snoRNA)^51–53. 25S, 18S, and 5.8S rRNAs are derived from a large precursor, 35S pre-rRNA, by a series of endonucleolytic cleavage and exonucleolytic trimming steps. 18S, 5.8S, and 25S sequences are bracketed by external (ETS) and internal (ITS) transcribed spacers (Fig. 6A), which are removed during maturation of rRNAs. Precursors of 5.8S and 25S rRNA contain remaining sequences of ITS1 and ITS2, respectively. These 5’ extensions are removed by Rat1p and Xrn1p to yield mature 5.8 S and 25 S rRNAs^54–57. To address whether inactivation of Mec1p suppresses only the transcription termination defect of rat1-1 or whether the suppression extends to other Rat1p functions, we measured the level of unprocessed 5.8 S and 25 S rRNAs. Our results showed that the unprocessed 5.8 S and 25 S rRNAs greatly accumulated in rat1-1xrn1∆ cells and this accumulation was suppressed in mec1∆sml1∆rat1-1xrn1∆ cells (Fig. 6B,C).

Rat1p and Xrn1p are also involved in processing small nucleolar RNAs (snoRNAs). One of these snoRNAs, snR190, is transcribed as a larger precursor, pre-snR190, that contains 5’ extension. Rat1p and Xrn1p remove this extension to produce mature snR190⁵⁸ (Fig. 6D). Our results showed that this pre-snR190 accumulates in rat1-1xrn1∆ cells and this accumulation was suppressed in mec1∆sml1∆rat1-1xrn1∆ cells (Fig. 6E). We interpret these results to mean that Mec1p affects aspect(s) of Rat1p’s activity that are more general and not specific only to transcription termination.

Inactivation of MEC1 does not promote recruitment of Rat1p and Rat1-1p to 3’ ends of PMA1 and PYK1

The simplest mechanism that would explain the suppression of transcription termination defects in rat1-1 cells by Mec1p inactivation would be an increased occupancy of Rat1-1p downstream of the pA sites in mec1∆sml1∆ cells. To test this supposition, we determined the occupancy of Rat1p (the protein product of the wild-type RAT1 gene) and Rat1-1p (the protein product of the rat1-1 allele) in wild-type and mec1∆sml1∆ cells. In agreement with previous reports^13,40, we found maximum occupancy of Rat1p and Rat1-1p downstream of the pA sites of both PMA1 and PYK1 (Fig. 7A). Compared with wild-type cells, the occupancy of Rat1p in mec1∆sml1∆ cells was significantly reduced throughout the gene bodies and particularly downstream of the pA sites for both PMA1 and PYK1. The occupancy of Rat1-1p in mec1∆sml1∆ cells was significantly reduced only for PMA1, but not for PYK1. Importantly, Mec1p inactivation did not elevate occupancy of Rat1-1p downstream of the pA sites in both PMA1 and PYK1. We interpret these results to suggest that Mec1p inactivation suppresses the transcription termination defects in rat1-1 cells by enhancing the activity of Rat1-1p and not by elevating the recruitment of Rat1-1p downstream of the pA sites. These results are in agreement with the western blot analysis that showed that Mec1p inactivation did not increase cellular levels of Rat1-1p (Fig. 7B). Interestingly, the levels of Rat1p are higher than Rat1-1p in both wild-type and mec1∆sml1∆ cells, suggesting that the rat1-1 mutation (Y657C) might destabilize Rat1-1p; the phenotypes previously observed in rat1-1 cells^8,13,14,42 might be due to the lower level of Rat1-1p in comparison with Rat1p.

Suppression of transcription termination and ribosomal RNA and snoRNA processing defects in rat1-1 cells requires activation of Mec1p

During undisturbed cell cycle in the absence of genotoxic stress, the checkpoint signaling kinase Mec1p is activated during the S phase, primarily to elevate synthesis of dNTPs required for DNA replication^22,59. The canonical pathway for Mec1p activation depends on its recruitment to single-stranded DNA coated by the RPA complex. Recruitment and activation of Mec1p also requires factors that bind to ssDNA-dsDNA junctions, including Ddc2p, Dpb11p, Dna2p and PCNA-like clamp composed of Ddc1p, Mec3p, and Rad17p. This clamp, equivalent of the human 9-1-1 complex, is loaded onto DNA by the clamp loader Rad24p-RFC complex. The signal from activated Mec1p is transduced by checkpoint mediator Rad9p to checkpoint effector kinase Rad53p^22,59. To find out whether defects in Mec1p activation suppress the ribosomal RNA and snoRNA processing and transcription termination defects in rat1-1 and rat1-1xrn1∆ cells, we introduced rad17∆, rad24∆, and rad9∆ mutations in rat1-1 and rat1-1xrn1∆ cells, and measured the level of unprocessed 5.8 S and 25 S rRNAs in rat1-1xrn1∆rad17∆, rat1-1xrn1∆rad24∆, and rat1-1xrn1∆rad9∆ cells. Our results showed that the accumulation of unprocessed 5.8 S and 25 S rRNAs in rat1-1xrn1∆ cells was significantly reduced by introducing rad17∆, rad24∆, or rad9∆ mutations (Fig. 8A). Correspondingly, the defects in pre-mRNA cleavage at the pA site in rat1-1 cells (Fig. 8B), as well as degradation of the RNA downstream of the pA site in rat1-1xrn1∆ cells were partially suppressed in rat1-1rad17∆ and rat1-1xrn1∆rad17∆ cells (Fig. 8C). In addition, rad17∆ mutation partially suppressed the transcription termination defects in rat1-1 cells (Fig. 8D,E). These results suggest that defects in Mec1p activation suppress the ribosomal RNA and snoRNA processing and transcription termination defects in rat1-1 cells and imply that Mec1p activity downregulates Rat1p.

Yeast strains and media

All yeast strains are listed in Table 1. Standard genetic techniques were used to manipulate yeast strains and to introduce mutations from non-W303 strains into the W303 background⁷¹. Cells were grown in YPD medium (1% yeast extract, 2% Bacto peptone, 2% glucose).

Real-time RT-qPCR

The procedures to extract total RNA from yeast cells and perform real-time reverse transcription quantitative PCR were as previously described^72,73. The primers used for RT-qPCR are listed in Supplementary Table 1.

ChIP assays

In vivo chromatin crosslinking and immunoprecipitation was performed essentially as described⁷³. Each immunoprecipitation was performed four times using different chromatin samples, and the occupancy was calculated using not transcribed region on chromosome VII (nucleotides 17485–17569) as a negative control (NC). The results were calculated as fold increase in RNAPII, Pcf11p-myc, Rat1p-TAP, and Rat1-1p-TAP occupancy at the particular position in comparison with the NC. Immunoprecipitations were performed with the following antibodies: anti-RNAPII Rpb1p monoclonal antibody (8WG16; 664906, BioLegend), anti-RNAPII CTD (phospho S2) antibody (ab5095; Abcam), anti-myc monoclonal antibody (9E10; sc-40; Santa Cruz Biotechnology). Immunoprecipitation of Rat1p-TAP and Rat1-1p-TAP was performed with IgG Sepharose 6 Fast Flow (17096901, Cytiva). The primers used for qPCR are listed in Supplementary Table 2.

Western blotting

Western blotting was performed as described and samples of whole cell extracts corresponding to 25 µg of proteins were analyzed in each lane (72). Membranes were probed with anti-myc monoclonal antibody (9E-10; sc-40, Santa Cruz Biotechnology) at a dilution of 1:2000, anti-Pgk1p monoclonal antibody (22C5D8; 459250; Invitrogen) at a dilution of 1:5000, anti-TAP tag polyclonal antibody (CAB1001, Thermo Fisher Scientific) at a dilution of 1:3000. Signal was detected with ECL prime western blotting detection reagent (RPN2236, Amersham) using BioRad ChemiDoc imaging system.

Statistical analysis

All statistical analyses (one-way Anova, two-way Anova, paired t-test, and tests for normal distribution of data and statistical outliers) were performed in GraphPad Prism 10.4.1 package.

Competing interests

The authors declare no competing interests.