RNAP-II transcribes two small RNAs at the promoter and terminator regions of the RNAP-I gene in Saccharomyces cerevisiae


  • Maria D. Mayán

    Corresponding authorCurrent affiliation:
    1. Biomedical Research Centre (INIBIC), CHUAC, A Coruña, Spain
    • MRC Clinical Sciences Centre, Imperial College, London, UK
    Search for more papers by this author

M. D. Mayán, Rheumatology Division, Osteoarticular and Ageing Research Group, Biomedical Research Centre (INIBIC), CHUAC, Xubias de Arriba 84, 15006 A Coruña, Spain.

E-mail: MA.Dolores.Mayan.Santos@sergas.es and m.mayan@csc.mrc.ac.uk; http://www.inibic.es/


Three RNA polymerases coexist in the ribosomal DNA of Saccharomyces cerevisiae. RNAP-I transcribes the 35S rRNA, RNAP-III transcribes the 5S rRNA and RNAP-II is found in both intergenic non-coding regions. Previously, we demonstrated that RNAP-II molecules bound to the intergenic non-coding regions (IGS) of the ribosomal locus are mainly found in a stalled conformation, and the stalled polymerase mediates chromatin interactions, which isolate RNAP-I from the RNAP-III transcriptional domain. Besides, RNAP-II transcribes both IGS regions at low levels, using different cryptic promoters. This report demonstrates that RNAP-II also transcribes two sequences located in the 5′- and 3′-ends of the 35S rRNA gene that overlap with the sequences of the 35S rRNA precursor transcribed by RNAP-I. The sequence located at the promoter region of RNAP-I, called the p-RNA transcript, binds to the transcription termination-related protein, Reb1p, while the T-RNA sequence, located in the termination sites of RNAP-I gene, contains the stem–loop recognized by Rtn1p, which is necessary for proper termination of RNAP-I. Because of their location, these small RNAs may play a key role in the initiation and termination of RNAP-I transcription. To correctly synthesize proteins, eukaryotic cells may retain a mechanism that connects the three main polymerases. This report suggests that cryptic transcription by RNAP-II may be required for normal transcription by RNAP-I in the ribosomal locus of S. cerevisiae. Copyright © 2012 John Wiley & Sons, Ltd.


The nucleolus is the site of ribosomal RNA synthesis, which is performed by RNAP-I and -III. Although RNAP-II does not transcribe rRNA, it has been found bound to the intergenic non-coding regions (IGS1 and IGS2) of the RNAP-I locus in Saccharomyces cerevisiae (Mayan and Aragon, 2010) and in Schizosaccharomyces pombe (see Supporting information, Figure S1). In S. cerevisiae, intergenic non-coding RNAP-II molecules have been implicated in the activity of RNAP-I (Mayan and Aragon, 2010). However, transcriptions by RNAP-II and, in particular, high levels of RNAP-II cryptic transcripts have been associated with rDNA instability (Houseley et al., 2007; Kobayashi and Ganley, 2005). As in yeast, mammalian rDNA also contains intergenic non-coding regions between each rDNA repeat. In mammalian cells, the intergenic regions also contain coding regions for RNAP-II; however, the intergenic non-coding transcripts are transcribed by RNAP-I (McStay and Grummt, 2008) and have been implicated in both heterochromatin formation and the activity of the RNAP-I promoter (Mayer et al., 2006; McStay and Grummt, 2008).

We have previously reported that RNAP-II is found predominantly in a stalled conformation in budding yeast, transcribing both IGS regions at very low levels (Mayan and Aragon, 2010). Stalled RNAP-II is responsible for the chromatin interactions between the two intergenic non-coding regions, which isolate the RNAP-III from the RNAP-I transcription machinery. We have also reported that transcription by RNAP-II may be necessary for proper ribosomal rRNA transcription, because the inhibition of transcription by RNAP-II decreases transcription by RNAP-III while promoting interaction between the promoter and enhancer of the 35S rRNA transcribed by RNAP-I (Mayan and Aragon, 2010). Recent reports have highlighted the interplay between RNAP-III and RNAP-II in mammals, with RNAP-II cryptic transcription acting as an insulator in RNAP-III transcription elongation (Raha et al., 2010). However, little is known about the interplay between RNAP-I and RNAP-II. Using two inhibitors of RNAP-II transcription elongation, α-amanitin (AM) and 5,6-dichloro-1-β-d-ribobenzimidazole (DRB), together with a thermosensitive strain of RNAP-II, this report demonstrates that two sequences that overlap with the promoter and terminator regions of RNAP-I are also transcribed by RNAP-II.

Materials and methods

Strains and materials

C-terminal epitope tagging of proteins was performed using PCR allele replacement methods (BY4741, REB1-9MYC) (Janke et al., 2004). The anti-Myc 9E10 monoclonal antibody used was obtained from Roche. AM and DRB were obtained from Sigma. AM and DRB treatments were performed as previously described (Mayan, 2010). The Z118 strain (rpb1-1) was kindly provided by R. Young, and P-Met-RAT1 and the xrn1 strain were provided by Aziz El Hage and D. Tollervey. The J342 strain (PGALREB1) and the polyclonal antibody against biochemically purified Reb1p–LacZ was provided by J. R. Warner. Strains are available on request. The cells were harvested from exponential phase cultures growing and stored at –80°C. For all experiments, cells were resuspended in 50–100 µl IP lysis buffer and broken with glass beads in a Fast Prep machine (QBiogene) for two rounds of 20 s each, with intervening incubations on ice for 5 min. The primer sequences were kindly provided by D. Moazed (Huang and Moazed, 2003). All statistical calculations were performed using the GraphPad Prism v 5 statistical software package.

Chromatin immunoprecipitation

Standard DNA-ChIP was performed as described (Mayan and Aragon, 2010). For RNA immunoprecipitation (RNA-ChIP), RNase inhibitor (RNasin 60 U/ml; Promega) was added to the IP lysis buffer. The nuclei were isolated by centrifugation at 13 000 rpm for 1 min and collected as the pellet. After sonication, the extract was treated with DNase I (30 U/ml; Applied Biosystems) in 25 mm MgCl2, 5 mm CaCl2 and RNasin for 25 min at 37°C. Nuclease digestion was stopped by adding EDTA to 20 mm. Immunoprecipitation was performed by the addition of antibody to the DNase I-digested extract in IP lysis buffer and incubation at 4°C with low-power sonication for 30 min. Quantitative reverese transcription–polymerase chain reaction (qRT–PCR) was performed as in the ChIP assay (Mayan and Aragon, 2010).

RNA methods

Total RNA isolation was performed as described previously (Mayan and Aragon, 2010). Quantifications were performed using SYBr Green real-time PCR.


In budding yeast, both the 35S rRNA precursor (containing 28S, 18S and 5.8S RNAs) and the 5S rRNA gene are present in the same locus (Figure 1). The rDNA IGS regions are repressed for RNA polymerase II (RNAP-II) transcription, a phenomenon referred to as ‘rDNA silencing’ (Bryk et al., 1997; Smith and Boeke, 1997). However, a number of RNAP-II promoters have been identified within these regions, and several small non-coding transcripts (ncRNAs) have been detected between the two Reb1p-binding sites (Figure 1, yellow ovals) (Houseley et al., 2007; Kobayashi and Ganley, 2005; Mayan and Aragon, 2010; Santangelo et al., 1988).

Figure 1.

Representative diagram of Saccharomyces cerevisiae rDNA gene locus on chromosome XII. The locus contains multiple copies of ribosomal genes arranged as a 9.1 kb head-to-tail tandem repeat. Each ribosomal copy contains the 35S RNA and 5S rRNA genes separated by two intergenic non-coding regions (IGS1 and IGS2), which are transcribed by RNAP-II. Both IGS regions contain key elements for transcription regulation, DNA replication [replication fork barrier (RFB) and autonomously replicating sequence (ARS)] and recombination. Green line, binding pattern of RNAP-II; red lines, binding sequences and transcribed region of RNAP-I; yellow ovals, Reb1 DNA-binding sites; green ovals, location of several RNAP-II cryptic promoters: below, red line shows the RNAP-I primary transcript containing the 5′ external transcribed spacer (ETS) and 3′ ETS, which are part of the 35S rRNA precursor transcript and are required for transcription termination as well as processing of the 35S pre-rRNA. The termination sequences for RNAP-I are shown. The locations of the primers used in this study are shown. P13 primers solely amplified the RNAP-I primary transcript, containing the last sequence of RDN25 and the first sequences of the 3′ ETS. P14 primers contain 3′ ETS and IGS1 sequences. Primers P14–P22 correspond to previously reported IGS cryptic transcripts (Mayan and Aragon, 2010). The sequence amplified for P23 primers corresponding to the 5′ ETS, where the RNAP-I promoter is located

In a report published in 2010, we used qRT–PCR to detect several small-non coding transcripts between P14 and P22 primers (Mayan and Aragon, 2010) (Figure 1). Yeast tiling array assays have revealed both sense and antisense transcription at the IGS1-2 regions (Clemente-Blanco et al., 2011), confirming the previous results of others (Houseley et al., 2007). However, we were intrigued by the levels of transcription detected using P23 primers for the 35S promoter region (Mayan and Aragon, 2010). The P23 amplified sequence is located at the 35S rRNA gene; however, the transcript levels detected by qRT–PCR were always lower than the transcript levels detected for the rest of the primers used to study transcription by RNAP-I (P24 or P34) (Mayan and Aragon, 2010). In fact, P23 transcription levels were always similar to the RNAP-II small non-coding transcripts (P14–P22). In addition, the presence of a tiny but constant amount of RNAP-II bound to the promoter region of the 35S rRNA gene (Figure 1, see green line, P23 primers) caused us to investigate whether the P23 detected transcript is largely transcribed by RNAP-II rather than RNAP-I. To determine whether this is the case, three different mechanisms were used to inhibit transcription by RNAP-II (Figure 2A, B). AM binds to RNAP-II and inhibits the translocation step and consequently transcription elongation. In addition, DRB prevents transcription elongation by inhibition of the kinases that phosphorylate RNAP-II on CTD Ser2 [CDK9/cyclin T; positive transcription elongation factor b (P-TEFb)]. Finally, to study the function of the entire polymerase, we used mutants such as the rpb1-1 ts mutant, because the biggest subunit of polymerase II (RPB1) is degraded when the cells are incubated at 37°C (Figure 2B). Treatment of cells with a low concentration of AM (10 µg/ml) inhibits RNAP-II elongation without affecting RNAP-III or RNAP-I. The cells were grown to exponential phase in YPD medium and treated with 10 µg/ml AM or DRB (200 µm) or were incubated at 37°C for 1 h (Figure 2B, C). The results (shown in Figure 2B) confirmed the decrease in the levels of RNAP-II small non-coding transcripts after inhibition of RNAP-II elongation (P14–P22 primers). As previously reported (Mayan and Aragon, 2010), our results confirm that elongating RNAP-II is necessary for normal ribosomal RNA transcription, as the 5S rRNA transcript decreased significantly following AM or DRB treatment (Figure 2C), while 35S rRNA transcription increased in the absence of the RNAP-II complex (37°C, 1 h) and slightly increased following AM treatment (~1.5-fold) (Figure 2B). It should be noted that DRB is an inhibitor of CDK9; therefore, it is expected to affect transcription elongation by RNAP-I (Bouchoux et al., 2004; Grenetier et al., 2006) (see Figure 2B, DRB bars).

Figure 2.

RNAP-II transcribes P23- and P13-amplified sequences located within the 35S rRNA gene. (A) Quantification of RNA levels by real-time PCR, using the primers shown above. The calculated amplification efficiencies were > 95% for all primers used in this study (Huang and Moazed, 2003; Mayan and Aragon, 2010). Green circles show the termination sites for 35S rRNA and the location of the promoter at the 5′ ETS (23 primers). (B–D) Cells were treated as previously described (Mayan, 2010). IGSs, 35S rRNA precursor and 5S primary transcripts in wild-type and mutant rpb1-1 cells at 25 °C following incubation at 37 °C for 1 h (thus inactivating RNAP-II, grey bars) or at 25 °C following treatment with 10 µg/ml AM for 1 h (red bars) or 200 µm DRB for 4 h (blue bars). RNA levels are shown relative to each primer. To compare the four conditions, the data were normalized relative to the mature form of the U2 small nuclear RNA (Mayan and Aragon, 2010). All values are expressed as mean ± SEM, n = 2 or n = 3; **p < 0.005, *p < 0.05, Student's t-test, 25 °C vs 37 °C, 25 °C vs AM, 25 °C vs DRB. Green arrows show the decrease in expression for the two regions located within the 35S sequence under the three sets of conditions. The 5S primary transcript decreases significantly after inhibition of RNAP-II transcription. (E) RNAP-I transcription across the IGS regions: RNA levels were analysed after inhibition of RNAP-II transcription in a strain deleted for the 5′–3′ exonucleases, Rat1p and Xrn1p (using xrn1Δ cells and by inhibiting the expression of RAT1 by the addition of 8 mm methionine for 12 h). AM (10 µg/ml) was added 1 h before harvesting the cells

The wild-type (BY4741) strain was grown at 37°C for 1 h and tested to study the effect of temperature on 35S and 5S rRNA transcription (Figure 2D). Quantification of expression using qRT–PCR revealed that transcription by RNAP-I (P34 primers) increased ~ two-fold (Figure 2D). Therefore, the effect observed for 35S rRNA transcription in the rpb1-1 ts mutant at 37°C may be partly a result of the increase in temperature (Figure 2B, grey bars; see primers P11, P12, P24, P25, P32 and P34). However, the first transcript located at the RNAP-I transcription start site (P23 primers) decreased significantly after inhibition of RNAP-II elongation, using AM or DRB or incubating the rpb1-1 ts mutant cells at 37°C for 1 h (Figure 2B, grey bars). Surprisingly, the levels of a second transcript located in the 35S rRNA sequence (P13 primers) decreased following the addition of AM or DRB or in the absence of RPB1 (rpb1-1 ts mutant at 37°C) (Figure 2B). The results previously obtained in the transcription profiles of genome-wide, strand-specific tiling arrays illustrated that pRNA (P23 primers) and T-RNA (P13 primers), the first and last sequences of the 35S rRNA gene, both contained sense and antisense transcription (Clemente-Blanco et al., 2011). Furthermore, the detected transcription in both regions is modulated differently from the rest of the sequences of the 35S rRNA gene (Clemente-Blanco et al., 2011), supporting the current results.

The P13-amplified sequence (see sequence, Figure 3) is found at the 3′ end of the 35S rRNA gene sequence (Figure 1). It has been reported that the 3′ end of the 35S precursor, together with the first nucleotides of the IGS1 region, is implicated in the termination of RNAP-I. The 35S pre-rRNA is cleaved co-transcriptionally across a stem–loop structure by the RNase III-like endonuclease, Rnt1p, which initiates transcription termination of the 35S rRNA precursor (see Figure 1) (Kufel et al., 1999). For complete termination, the activity of the 5′ exonuclease, Rat1p, is required in the 3′ ETS region (Braglia et al., 2011; El Hage et al., 2008; Kawauchi et al., 2008). In fact, in the absence of Rat1p, RNAP-I spreads into the intergenic regions (Kawauchi et al., 2008). The exact mechanism of RNAP-I termination in yeast is largely unknown; however, in vitro and in vivo analyses have shown that most RNAP-I transcripts terminate near the Reb1p site located 93 nucleotides downstream from the 3′ end of the 35S sequence (T1). The RNAP-I molecules, which read through T1, stop at the downstream fail-safe terminator located near +250 from the 35S sequence (T2). P13-RNA contains the RNA sequence that is recognized by Rnt1p, cleavage of which begins the transcription termination of the 35S rRNA precursor. Therefore, the transcription of an identical transcript, P13-RNA (termination RNA: T-RNA), by RNAP-II may regulate RNAP-I termination. In fact, a decrease in the level of the RNAP-II cryptic transcripts directly affects RNAP-I termination at the ribosomal locus through the absence of transcription by RNAP-II, and RNAP-I complexes spread into the intergenic regions (Mayan and Aragon, 2010).

Figure 3.

Sequence of T-RNA and p-RNA amplified using primers P13 and P23, respectively. P13 and P23 sequences correspond to the ETS regions of the 35S rRNA precursor transcript (see Figure 1)

RNAP-I-transcripts in the IGS regions are constantly degraded by the 5′–3′ exonucleases, Rat1p and Xrn1p. Thus, RNAP-I transcripts within the IGS can only be only detected in the absence of Rat1p and Xrn1p, while RNAP-II transcripts within the IGS regions are normally detected. When the double mutant strain P-MET-RAT1/xrn1 cells were treated with AM to abolish RNAP-II transcription, elevated levels of IGS transcripts were detected (Figure 2E, red bars; primers 14–22). When RNAP-II transcription was inhibited using 10 µg/ml AM, the presence of detectable RNAP-I transcripts in both IGS regions suggested that cryptic transcription by RNAP-II may be necessary for proper termination of RNAP-I. Together with previously published results, our current results confirm that, in the absence of RNAP-II cryptic transcripts, RNAP-I does not terminate properly, leading to the transcription of the IGS regions.

In mice, it is well known that TTF1 cooperates with RNAP-I and the transcript release factor PTRF in conjunction with the T-rich DNA sequence to induce transcription termination and dissociation of RNAP-I, together with the dissociation of transcript from the template (Jansa and Grummt, 1999). So far, no equivalent transcript release factor has been identified in yeast. The budding yeast homologue of TTF-I, Reb1p, is thought to be involved in the termination of RNAP-I transcription (Grummt et al., 1986; Kawauchi et al., 2008; Kuhn et al., 1990; Lang and Reeder, 1993). In yeast, REB1 is essential for cell growth, and although it has two DNA binding sites in the rDNA locus, its roles in RNAP-I transcription and termination are not fully defined. Besides, in mammals, TIP (TTF-I interacting protein), the large subunit of the chromatin remodelling complex, NoRC, binds to a 150–300 nt IGS transcript whose sequence matches the small RNA at the promoter [promoter-associated RNA (pRNA); see Supporting information, Figure S2] TTF-I recruits NoRC to the rRNA promoter; therefore, this factor plays a critical role in triggering the chain of events by which the active or silent state of rRNA genes is established (Schmitz et al., 2010). The results presented in Figure 2 demonstrate the presence of a cryptic transcript in the promoter of the 35S rRNA gene (P23-pRNA) that is transcribed by RNAP-II. To study whether Reb1p interacts with the P23-pRNA sequence (see sequence, Figure 3), RNA-ChIP assays were performed, as described in Materials and methods. ACT1 ORF was amplified and used as a negative control for the binding of Reb1p (Figure 4). The DNA ChIP experiments confirmed that Reb1p binds predominantly to its DNA binding site, next to the promoter of the 35S gene (P22) (Mayan and Aragon, 2010). On the other hand, RNA-ChIP demonstrated that the P23-pRNA transcript binds to Reb1p (Figure 3A, red line). The results also showed that Reb1p binds to a transcript corresponding to the 5S rRNA sequence (see Figure 4A). As a second control, RNAP-II transcription was inhibited by treating the cells with 200 mm DRB for 4 h. In the absence of RNAP-II cryptic transcription, no interaction was detected (Figure 4B). This result confirmed that the transcript that interacts with Reb1p corresponds to a transcript transcribed by RNAP-II (P23 primers). No interaction was detected using the GAL-REB1 strain when REB1 expression was repressed by incubating cells in the presence of 2% glucose. Together, these results suggest that Reb1p in yeast could be playing a role similar to that of Reb1p/TTF-I or TIP5 in mammalian cells. However, in the case of budding yeast, pRNA is transcribed by RNAP-II (Figure 2; see also Supporting information, Figure S2).

Figure 4.

RNA ChIP analysis of Reb1p binding (REB1-9MYC) to IGS transcripts in wild-type cells (red line). (A) DNA ChIP is shown for comparison (grey). Before quantification by real-time PCR, cDNA was diluted 1:100. The ACT1 ORF was amplified as a negative control for both the DNA and RNA-ChIP assays. While Reb1p binds to its DNA binding site next to the 35S promoter (P22 primers), the small RNA associated with the promoter (P23-pRNA) interacts with Reb1p. (B) After ablation of RNAP-II transcription using DRB, Reb1p enrichment to its binding sites is significantly decreased. (C) RNA-ChIP analysis of Reb1p binding in a GAL-REB1 strain is shown. The cells were grown in 2% galactose/0.3% glucose medium (+REB1) and were shifted to 2% glucose to repress REB1 expression (–REB1). Polyclonal antiserum against a biochemically purified Reb1p–LacZ was used for RNA-ChIP analysis. All values are expressed as mean ± SEM, n = 2 or n = 3; **p < 0.005, *p < 0.05, Student's t-test, untreated vs DRB, +REB vs –REB


The results presented in this report demonstrate for the first time the presence of two RNAP-II cryptic transcripts located within the 35S rRNA sequence. The pRNA binds to Reb1p and likely performs a role similar to that of its mammalian counterpart. However, T-RNA could be implicated in the termination of RNAP-I as this sequence contains the substrate sequence for Rnt1p. How RNAP-II transcripts affect RNAP-I termination is an important question that will require further study.

Mammalian rDNA transcription units contain several consecutive termination sites, which are recognized by Reb1p/TTF-I. There is also a conserved binding site for an RNAP-I transcription terminator protein adjacent to the gene promoter. Indeed, in mammalian cells, it has been described that the binding of TTF-I to the promoter-proximal terminator stimulates RNAP-I transcription in vivo. Subsequent in vitro studies showed that TTF-I binding to the upstream terminator triggered structural alterations of the chromatin on preassembled nucleosomal templates, and these changes in chromatin structure correlated with the activation of RNAP-I transcription in vitro. Therefore, in higher eukaryotic cells, TTF-I is a multifunctional protein, having roles in termination, transcription regulation and chromatin structure (McStay and Grummt, 2008). So far, in budding yeast, none of these described activities in the promoter and terminator regions of the ribosomal RNA precursor have been attributed to Reb1p. Intriguingly, in both mammals and yeast, Reb1p and TTF-I are necessary for the DNA interaction between the promoter and enhancer of the 35S rRNA gene (Mayan and Aragon, 2010; Nemeth et al., 2008). In yeast, the interaction between the promoter and enhancer of 35S rRNA are partially mediated by Reb1p (Mayan and Aragon, 2010), and the results presented here demonstrate that the RNAP-II transcript located at the promoter region (pRNA) interacts with Reb1p. Further studies will be necessary to clarify the role of pRNA and Reb1p in chromatin silence or the promoter activity of the 35S rRNA gene.

Inhibition of RNAP-II transcription significantly decreased transcription by RNAP-III (5S rRNA) (Figure 2). Interestingly, other recent studies have highlighted the relationship between polymerase III and II, demonstrating the ability of RNAP-II to act as a mediator (Barski et al., 2010; Listerman et al., 2007; Oler et al., 2010; Raha et al., 2010; Zheng et al., 2009), with RNAP-II cryptic transcription likely playing an insulator role in RNAP-III transcription elongation. Together with our results, it is reasonable to suggest that cryptic RNAP-II transcripts at the IGS regions may play an insulator role in RNAP-III transcription at the rDNA locus. Crosstalk between RNAP-II and RNAP-III is not the only example of a relationship between polymerases. Zheng et al. (2009) have reported that adjacent intergenic non-coding transcripts made by RNAP-II play a central role in coordinating the other two polymerases in Arabidopsis.

The current results highlight the crosstalk between RNAP-I, -III and -II in budding yeast, where RNAP-II cryptic transcripts may regulate the activity of the promoter and termination of RNAP-I as well as transcription by RNAP-III. It will be interesting to further investigate the functions of RNAP-II cryptic transcripts at the rDNA locus and properly test whether cryptic transcription by RNAP-II also occurs at the rDNA locus in mammals.


This study was performed in Professor Luis Aragon's laboratory. I thank R. Young for the Z118 strain (rpb1-1) as well as Aziz El Hage and D. Tollervey for the P-Met-RAT1/xrn1Δ strain. Priscilla Branglia and J. Warner were instrumental in critically reading the manuscript, as was Alexandra MacAleenan for her critical reading and corrections. I also thank Adam Jarmuz for his helpful assistance. This work was supported by the Medical Research Council, UK. M.D.M. is currently funded by the Xunta de Galicia.