Direct Analysis of Phosphorylation Sites on the Rpb1 C-Terminal Domain of RNA Polymerase II

Abstract

Dynamic interactions between RNA polymerase II and various mRNA-processing and chromatin-modifying enzymes are mediated by the changing phosphorylation pattern on the C-terminal domain (CTD) of polymerase subunit Rpb1 during different stages of transcription. Phosphorylations within the repetitive heptamer sequence (YSPTSPS) of CTD have primarily been defined using antibodies, but these do not distinguish different repeats or allow comparative quantitation. Using a CTD modified for mass spectrometry (msCTD), we show that Ser5-P and Ser2-P occur throughout the length of CTD and are far more abundant than other phosphorylation sites. msCTD extracted from cells mutated in several CTD kinases or phosphatases showed the expected changes in phosphorylation. Furthermore, msCTD associated with capping enzyme was enriched for Ser5-P while that bound to the transcription termination factor Rtt103 had higher levels of Ser2-P. These results suggest a relatively sparse and simple "CTD code."

Figure 1

msCTD supports cell growth and is functional for factor recruitment. (A) Amino acid sequences of wild-type CTD and msCTD shown with heptad repeats stacked vertically. msCTD has an additional N-terminal Prescission protease cleavage site (LEVLFQ/GP) followed by a 8xhistidine tag, and a C-terminal 3xFLAG tag. Mutated amino acids (K, R, or T incorporated at position 7 to generate tryptic peptides of unique mass) are highlighted in green. (B) Rpb1 with msCTD complements rpb1Δ. Plasmids carrying wild type RPB1 (WT) or RPB1-msCTD (MC) were introduced into a yeast strain lacking the chromosomal RPB1 by plasmid shuffling. Cell growth was assayed on rich media (YPD) or synthetic complete media (SC) plates. Three-fold serial dilutions of cultures were spotted onto plates and incubated at temperatures as labeled. (C) Immunoblotting of whole cell lysates from WT or MC strains. FLAG epitope is present only on Rpb1-msCTD. Rpb3 serves as a loading control for total RNApII. The asterisk (*) on the Ser7-P panel is a reminder that msCTD carries a reduced number of Ser7 residues due to S7K/R/T substitutions. (D) RNApII subunit Rpb3; CTD binding proteins Ceg1 and Rna15; and CTD Ser5, Ser7, and Ser2 phosphorylations were analyzed by chromatin immunoprecipitation (ChIP) along the PMA1 gene. These experiments all used the same chromatin preparations probed with specific antibodies. Schematic at the top shows PCR primer regions (numbered boxes) along the gene. Graphs underneath show the average of results from three independent ChIP experiments, with the y-axis indicating fold enrichment over a non-transcribed background signal and error bars showing standard error. See also Figure S1.

Figure 2

Analysis of msCTD phosphorylation on total RNApII extracted from yeast. (A) Schematic of msCTD purification for MS/MS analysis. RNApII is isolated from cell extracts using IgG-Sepharose binding to the TAP tag on Rpb3. Prescission protease releases msCTD, which is then further purified via binding of the 8x histidine tag to Ni-NTA agarose. Following elution, msCTD is digested by trypsin, subjected to desalting and analyzed by tandem mass spectrometry. (B) Box at left shows predicted tryptic peptides from msCTD. The three bar graphs in the center show the number of Peptide-Spectrum-Matches detected within peak reverse phase chromatographic fractions for each peptide in the un-, mono-, or diphosphorylated form. The number of phosphorylations divided by total PSMs was then normalized to the number of repeats in that peptide to estimate the average density of phosphorylations per heptad as shown in the right bar graph. (C) For each potential phosphorylation site in the heptad, the number of phosphorylations detected (middle line) was divided by the total occurrences of that amino acid (bottom line) to calculate the percentage of phosphorylation (top line). (D) The percentage of phosphorylation at each residue along the length of msCTD was calculated and graphed using a color scale as shown at bottom. The tryptic peptide and heptad numbers are shown at left shown from N-terminus to C-terminus. (E) Multiple phosphorylations detected on tryptic peptides. Only those that were confirmed by manual validation of MS2 spectra and detected in more than two different tryptic peptides are listed. See also Figure S3.

Figure 3

Effect of CTD phosphatase mutant fcp1-1 on msCTD phosphorylation. Rpb1-msCTD was introduced by plasmid shuffling into FCP1 and fcp1-1 strains. (A) Prior to trypsin digestion, a fraction of each sample was separated by SDS-PAGE and blots were probed with the indicated antibodies (Chapman et al., 2007). Protein size estimates (kDa) to the left are based on protein size markers. (B) The percentage of phosphorylation at each heptad residue was calculated as in Figure 2, with the number listed below the bar graph. Fold change was calculated by dividing %phosphorylation in the mutant by that in the isogenic wild-type strain. (C) The left panel shows percent phosphorylation for individual residues along the length of msCTD, with color code shown at the bottom. Fold change was calculated by log2(%phospho[mutant]/%phospho[wild-type]). As indicated by the color code at the bottom, purple signifies an increase and green a decrease. See also Figure S4.

Figure 4

Effect of CTD kinase mutants on msCTD phosphorylation. Rpb1-msCTD was introduced by plasmid shuffling into BY4741, bur2Δ, ctk1Δ strains. (A) Prior to trypsin digestion, a fraction of each sample was separated by SDS-PAGE and blots were probed with the indicated antibodies (Chapman et al., 2007). Protein size estimates (kDa) to the left are based on protein size markers. (B) The percentage of phosphorylation at each heptad residue was calculated as in Figure 2, with the number listed below the bar graph. Fold change was calculated by dividing %phosphorylation in the mutant by that in the isogenic wild-type strain. (C) The left panel shows percent phosphorylation for individual residues along the length of msCTD, with color code shown at the bottom. Fold change was calculated by log2(%phospho[mutant]/%phospho[wild-type]). As indicated by the color code at the bottom, purple signifies an increase and green a decrease.

Figure 5

Phosphorylations on msCTD associated with CTD binding proteins Ceg1 or Rtt103. (A) PMA1, and (B) ADH1 genes were analyzed for ChIP crosslinking of the TAP-tagged proteins indicated on top in strains carrying wild-type RPB1 (WT) or RPB1-msCTD (MC). Schematic at top shows the PCR primer regions (numbered boxes) along the gene. Graphs underneath show the average of quantified results from three independent ChIP experiments, with the y-axis representing fold enrichment over a nontranscribed background signal and error bars showing standard error. (C) Percent phosphorylation detected on each residue of the heptad was calculated as in Figure 2C. (D) Percent phosphorylation for each residue along the length of msCTD is graphed as in Figure 2D. (E) Listing of peptides carrying two phosphorylations, with phosphorylated sites marked in blue color. Only those that were detected in more than two different tryptic peptides are shown. See also Figure S5.