RNA polymerase II transcription attenuation at the yeast DNA repair gene DEF1 is biologically significant and dependent on the Hrp1 RNA-recognition motif

Abstract

Premature transcription termination (i.e. attenuation) is a potent gene regulatory mechanism that represses mRNA synthesis. Attenuation of RNA polymerase II is more prevalent than once appreciated, targeting 10-15% of mRNA genes in yeast through higher eukaryotes, but its significance and mechanism remain obscure. In the yeast Saccharomyces cerevisiae, polymerase II attenuation was initially shown to rely on Nrd1-Nab3-Sen1 termination, but more recently our laboratory characterized a hybrid termination pathway involving Hrp1, an RNA-binding protein in the 3'-end cleavage factor. One of the hybrid attenuation gene targets is DEF1, which encodes a repair protein that promotes degradation of polymerase II stalled at DNA lesions. In this study, we characterized the chromosomal DEF1 attenuator and the functional role of Hrp1. DEF1 attenuator mutants overexpressed Def1 mRNA and protein, exacerbated polymerase II degradation, and hindered cell growth, supporting a biologically significant DEF1 attenuator function. Using an auxin-induced Hrp1 depletion system, we identified new Hrp1-dependent attenuators in MNR2, SNG1, and RAD3 genes. An hrp1-5 mutant (L205S) known to impair binding to cleavage factor protein Rna14 also disrupted attenuation, but surprisingly no widespread defect was observed for an hrp1-1 mutant (K160E) located in the RNA-recognition motif. We designed a new RNA recognition motif mutant (hrp1-F162W) that altered a highly conserved residue and was lethal in single copy. In a heterozygous strain, hrp1-F162W exhibited dominant-negative readthrough defects at several gene attenuators. Overall, our results expand the hybrid RNA polymerase II termination pathway, confirming that Hrp1-dependent attenuation controls multiple yeast genes and may function through binding cleavage factor proteins and/or RNA.

Keywords: DEF1; CPF-CF; Hrp1; NNS; RNA polymerase II; RRM; attenuation; transcription termination.

Fig. 1.

Expression of the yeast DNA repair gene DEF1 is regulated by transcription attenuation and proteasome-mediated protein processing. a) Transcriptional regulation: in the absence of UV damage (−UV), Pol II transcription of DEF1 undergoes premature termination (attenuation) via a hybrid CF/Hrp1/Sen1/CPF pathway, resulting primarily in noncoding RNA. In the presence of UV damage (+UV), Pol II readthrough of the attenuator results in bypass of pA¹, leading to protein-coding mRNA and Def1 protein translation. b) Proteasomal regulation: in the absence of UV damage, Def1 protein remains cytoplasmic due to the presence of an NES. Following UV damage, the NES domain is removed by the proteasome, allowing processed Def1 (pr-Def1) to enter the nucleus. Nuclear pr-Def1 promotes ubiquitination (Ub) and degradation of Pol II stalled at UV-induced lesions (X), helping to promote DNA repair.

Fig. 2.

The def1 attenuator mutant (def1_atten) overexpresses mRNA and protein, exacerbating the cell toxicity of processed Def1 (def1_1–530). a) Schematic of def1 mutants (def1-atten and def1_1–530) along with RT-PCR primers (F1, R1, blue) used to detect the DEF1 mRNA readthrough product. Note: the drawing is not to scale. The R1 primer is specific to longer mRNA and not attenuated RNA. DNA positions are numbered relative to the +1 start codon of the DEF1 open reading frame. b) RT-PCR analysis of Def1 mRNA levels. 18S serves as a loading control for total RNA. Reverse Transcriptase (±RT) ensures that signal is dependent on RNA and not genomic DNA template. c) Western blot analysis of Def1 protein levels. Actin serves as a loading control for total protein. d) Spot test assay of def1 mutants on solid plate media. Liquid cultures were grown to saturation at 25°C, serially diluted, spotted onto YPAD, and grown at the temperatures indicated for 3–5 days. e, f) Growth of def1 mutants in liquid culture. Liquid yeast cultures were grown to saturation at 25°C, diluted back, and recovered to exponential phase prior to shifting to (e) 37°C or (f) 39°C for 6 h. Cell density was measured via OD₆₀₀ every hour, and doubling times were calculated from exponential lines of best fit for data between 60 and 360 min. Error bars represent SD from 3 biological replicates. Asterisks indicate statistical significance by Welch’s 2 sample t-test (*P ≤0.05, **P≤0.01, ***P≤0.001, ****P≤0.0001, ns—not significant).

Fig. 3.

Overexpression of def1_1–530 in a def1 attenuator mutant (def1_atten) exacerbates proteasomal degradation of Pol II subunit Rpb1. a, c) Western blot analysis of Rpb1 (Pol II) protein levels in def1 mutants (biological triplicate samples). Liquid yeast cultures were grown to saturation at 25°C, diluted back, and recovered to exponential phase prior to shifting to (a) 39°C or (c) kept at 25°C for 2 h. Actin serves as a loading control for total protein. The schematic below the blot indicates the approximate Def1 protein expression level and localization (cytoplasm or nucleus) based on observed activity of the mutants. b, d) The average Rpb1 (Pol II) protein levels (normalized to actin) were quantified from the 3 biological replicates in (a) and (c), and error bars represent the SD. Asterisks indicate statistical significance by Welch’s 2 sample t-test.

Fig. 4.

Hrp1 function in attenuator recognition is dependent on its interaction with CFIA and RNA and varies based on the Pol II terminator. a) Schematic of Hrp1 activity and hrp1 mutants expected to disrupt interaction with RNA (K160E), CFIA protein Rna15 (D193N), or CFIA protein Rna14 (L205S). b) Spot test growth assay of hrp1 mutants. A yeast shuffle strain [hrp1::KANMX, pRS316-HRP1 (URA3)] was transformed with HIS3-marked plasmids containing hrp1 mutants prior to 5-FOA shuffling. Serial dilutions were spotted on plates and grown 3 days at indicated temperatures. c) Schematic of attenuator-lacZ reporter gene system. d) Attenuator functionality assays using a lacZ reporter gene with hrp1 mutants. Yeast strains bearing HRP1 WT or hrp1 mutant plasmids were transformed with lacZ reporter genes containing DEF1 or HRP1 attenuators. The CYC1 terminator serves as a control for hybrid termination. Overnight cultures were grown to saturation at 30°C and recovered to exponential phase followed by a 2-h shift to nonpermissive temperature (37°C). Cells were lysed and β-galactosidase activity was measured to detect attenuator readthrough. Error bars represent SD of 3 biological replicates. Asterisks indicate statistical significance by Welch’s ANOVA.

Fig. 5.

The hrp1 RRM1 mutants W168A, F162W, and F204W are lethal and defective for attenuation. a) Protein–RNA interface for the Hrp1-EE complex (structure 2KM8; PyMOL). Relevant Hrp1 side chains (green) and RNA bases (red) are represented in stick format. b) Spot test growth assay of hrp1 mutants. A yeast shuffle strain [hrp1::KANMX, pRS316-HRP1 (URA3)] was transformed with plasmids containing hrp1 mutants prior to 5-FOA shuffling. Serial dilutions were spotted on plates and grown 3 days at indicated temperatures. c) Yeast strains containing osTIR1 and WT (HRP1) or an N-terminal degron tag (HRP1-N-AID*-Myc) were grown on YPAD ± auxin inducer (1 mM) for 3 days. d) Western blot analysis of Hrp1 depletion. Hrp1 degron strains (in biological triplicate) were grown in YPAD until exponential phase, followed by treatment with auxin (1 mM) or ethanol solvent control for 4 h. Hrp1 was detected from protein extracts with an anti-Myc antibody and normalized to actin as a loading control. e) Average Hrp1 protein levels were quantified from 3 biological replicates, and error bars represent SD. f) The Hrp1 degron strain (HRP1-N-AID*-Myc) was transformed with lacZ reporter genes and grown in selective media ± auxin for 4 h. Cells were lysed and β-galactosidase activity was measured to detect attenuator readthrough. Error bars represent SD of 3 biological replicates. g–i) The Hrp1 degron strain containing DEF1-lacZ was transformed with pRS313 plasmids containing empty vector, HRP1 WT, or hrp1 mutants. Auxin treatment and β-gal assays were performed as in (f). Asterisks indicate statistical significance by Welch’s 2 sample t-test.

Fig. 6.

IGB analysis of Hrp1 and pA occupancy at promoter-proximal regions. a–e) For genes with known and putative attenuators, the protein occupancy [Pol II (Rpb1), Hrp1, Nrd1, Nab3], TSS, and polyadenylation (pA) sites were aligned to the S. cerevisiae genome and visualized with the IGB. Horizontal axes indicate genomic coordinates, and vertical axes display relative factor/site levels. Open reading frames and transcription directionality (left or right; black arrows) are indicated. Putative attenuators (fuchsia dashed lines) are based on 5′-end Pol II peaks and other contextual information. f) Comparative analysis was performed for the highly transcribed RPS31 gene, with no evidence of attenuation.

Fig. 7.

Heatmap analysis of Hrp1 and pA promoter-proximal enrichment helps identify new attenuators, some of which are Hrp1-dependent. a) The pA signals and protein factor occupancy from published studies were used to generate a heatmap. Levels were normalized to local Pol II attenuator peaks to account for differences in transcription rate and then to the canonical NNS-dependent NRD1 attenuator. Blue shading indicates 1.5-fold enrichment or depletion relative to the NRD1 attenuator. Attenuators were cloned into the lacZ reporter gene, and terminator strength was determined by the decrease in β-gal signal compared with a control reporter plasmid lacking a terminator (red >90%; yellow >50%; green <50%). Pol II pausing strength and stress response association were taken from Cherry et al. (2012) and Collin et al. (2019), respectively. b, c) Representative lacZ data for candidate attenuators after a 1- or 2-h kinetic assay. A lacZ reporter lacking a terminator (No Term-lacZ) serves as a negative control. d, e) The Hrp1 degron protein was depleted as described in Fig. 5f, followed by detection of β-galactosidase activity. Asterisks indicate statistical significance by Welch’s 2 sample t-test.

Fig. 8.

The hrp1-F162W RRM1 mutant exhibits strong dominant-negative readthrough defects at a subset of Hrp1-dependent attenuators. The Hrp1 degron strain containing (a–c) MNR2- or (d–f) SNG1-lacZ reporter genes was transformed with plasmids containing empty vector, HRP1 WT, or mutant hrp1-F162W. Auxin treatment and β-gal assays were performed as in Fig. 5f. Asterisks indicate statistical significance by Welch’s 2 sample t-test or ANOVA. g) Western blot analysis of Hrp1 depletion ± auxin and Hrp1 plasmid-based expression. Following auxin treatment, the same MNR2-lacZ cultures used in (a) were used to harvest cells for protein extracts. AID-Hrp1 and untagged Hrp1 were detected with anti-Hrp1 antibody, and actin serves as a loading control. h) Protein–RNA interface of Hrp1-EE complex (structure 2KM8; PyMOL), highlighting Hrp1-F162 (WT) and RNA–A6 interaction. Hrp1 side chains (green) and RNA bases (red) represented in stick format, and the W162 mutant modeled using mutagenesis tool (PyMol).