Control of ribonucleotide reductase localization through an anchoring mechanism involving Wtm1

Abstract

The control of deoxyribonucleotide levels is essential for DNA synthesis and repair. This control is exerted through regulation of ribonucleotide reductase (RNR). One mode of RNR regulation is differential localization of its subunits. In Saccharomyces cerevisiae, the catalytic subunit hererodimer, Rnr2/Rnr4, is localized to the nucleus while its regulatory subunit, Rnr1, is cytoplasmic. During S phase and in response to DNA damage, Rnr2-Rnr4 enters the cytoplasm, where it presumably combines with Rnr1 to form an active complex. The mechanism of its nuclear localization is not understood. Here, we report the isolation of the WTM (WD40-containing transcriptional modulator) proteins as regulators of Rnr2/Rnr4 localization. Overproduction of Wtm2 increased Rnr2/Rnr4. Deletion of WTM1, a homolog of WTM2, leads to the cytoplasmic localization of Rnr2/Rnr4, and increased hydroxyurea (HU)-resistance in mec1 mutants. Wtm1 binds Rnr2/4 complexes and release them to the cytoplasm in response to DNA damage. Forced localization of Wtm1 to the nucleolus causes Rnr2/Rnr4 complexes to relocalize to the nucleolus. Thus, Wtm1 acts as a nuclear anchor to maintain nuclear localization of Rnr2/4 complexes outside of S phase. In the presence of DNA damage this association is disrupted and Rnr2/Rnr4 become cytoplasmic, where they join with Rnr1 to form an intact complex.

Figure 1.

WTM2 is a dosage suppressor of mec1Δ sml1Δ's HU sensitivity. (A) mec1Δ sml1Δ cells transformed with vector alone (pRS416) or pGAL1-WTM2 CEN6 URA3 (pYDL5) plasmid were struck on galactose plates (YPGal) with 20 mM of HU. Two independent transformants are shown for each plasmid. (B) mec1Δ sml1Δ cells transformed with vector alone (pRS416) or the pGAL1-WTM2 CEN6 URA3 (pYDL5) plasmid were first grown in liquid media and serially diluted 10× and spotted on either glucose or galactose plates at the HU concentrations indicated.

Figure 2.

Wtm2 interacts with Rnr2 and Rnr4, and its overproduction leads to the activation of RNR pathway by increasing Rnr2 and Rnr4 levels. (A) Log-phase whole-cell extracts from a wild-type and WTM2-3HA-tagged strain were immunoprecipitated with or without (no-IgG) anti-HA antibodies. Input (10% of immunoprecipitation samples) and immunoprecipitation samples were separated by 10% SDS-PAGE and blotted for HA, Rnr1, Rnr2, and Rnr4. (B) mec1Δ cells containing a MEC1 URA3 plasmid (pBAD045) were transformed with empty vector (pRS415) or the pGAL1::WTM2 overexpression plasmid (pYDL6). Cells were struck on YPGal plates for 2 d, then struck onto Gal-5-FOA plates and incubated for 4 d. (C) mec1Δ sml1Δ cells carrying either an empty vector (pRS416) or the pGAL1::WTM2 overexpression vector (pYDL5) were grown to log phase in galactose SC - URA media. Total protein was precipitated by TCA and resolved by 10% SDS-PAGE and blotted with anti-tubulin, anti-Rnr2, and anti-Rnr4 antibodies. (D,E) Cells were grown as in C and processed for indirect immunofluorescence. Samples were stained with anti-Rnr2 (D) or anti-Rnr4 (E) primary antibodies and FITC-conjugated secondary antibodies. The arrow indicates the cytoplasmic rod-shaped structure. All photomicrographs comparing Rnr levels of wild-type and wtm1Δ cells were taken with the same exposure.

Figure 3.

Nuclear localization of Rnr2 and Rnr4 is defective in wtm1Δ mutants. (A,B). Wild-type and wtm1Δ cells were first synchronized in G1 by α-factor, then stained with DAPI for DNA and anti-Rnr2 antibodies (A) and anti-Rnr4 antibodies (B) for immunofluorescence. (C) Wild-type and wtm1Δ cells were first synchronized in G1 by α-factor and fractionated into different subcellular compartments. Proteins extracted from whole-cell (W), cytoplasmic fraction (C), and nuclear fraction (N) were resolved by 10% SDS-PAGE and blotted with anti-Rnr2, anti-nuclear pore o-linked glycoprotein (Nuc Gly), and anti-phosphoglycerate kinase antibodies. (D) Log-phase wild-type, mec1Δ sml1Δ wtm1Δ, and dun1Δ wtm1Δ cells grown in YPD were stained with primary anti-Rnr2 and secondary FITC antibodies for immunofluorescence.

Figure 4.

Deletion of WTM1 increases HU resistance of mec1Δ sml1Δ mutants. (A) mec1Δ sml1Δ strains deleted for the WTM1 locus (wtm1Δ::TRP1) or WTM2 locus (wtm2Δ::TRP1) were serially diluted 10-fold and spotted on YPD plates with increasing concentrations of HU and grown for 2 d. A mec1Δ sml1Δ strain serves as a control. (B) Log-phase protein extracts of strains used in part A were resolved by 10% SDS-PAGE and blotted with anti-Rnr2 and anti-Tub1 antibodies as a loading control.

Figure 5.

Wtm1 interacts with Rnr2 and Rnr4 in a DNA damage-regulated manner. (A) Protein extracts from log-phase wild-type and 3MYC-WTM1 strains were immunoprecipitated with anti-MYC or anti-HA (control) antibodies. Input (10% of immunoprecipitation) and immunoprecipitation samples were resolved by 10% SDS-PAGE and blotted with anti-MYC, anti-Rnr2, and anti-Rnr4 antibodies. (B) 3MYC-WTM1-tagged and wild-type cells were arrested in G1 by α-factor, then half of each sample was maintained in α-factor (αF), while the other half was release into 200 mM HU. Samples were processed for indirect immunofluorescence and stained with anti-MYC and anti-Rnr2 antibodies, followed by Alexor-488 and Cy3-conjugated secondary antibodies. (C) Cells were grown as in B. Protein extracts were immunoprecipitated with anti-MYC antibodies. Input (equivalent to 10% immunoprecipitation) and immunoprecipitation samples were resolved by 10% SDS-PAGE and blotted with anti-MYC and anti-Rnr2 antibodies. (D) Wild-type cells arrested in G2/M by nocodazole were treated with or without phleomycin (100 μg/mL) for 90 min while maintained in nocodazole, then were stained with anti-Rnr2 antibodies for immunofluorescence. (E) Cells were grown and treated as in D. Wild-type and tagged 3MYC-WTM1 protein extracts were prepared by TCA precipitation, resolved by 8% SDS-PAGE, and blotted with anti-Rad53 antibodies. (F) Cells were grown and treated as in D, then fixed with formaldehyde and immunoprecipitated with anti-MYC antibodies, followed by the reversal of cross-linking. Input (equivalent to 10% immunoprecipitation) and immunoprecipitation samples were resolved by 10% SDS-PAGE and blotted with anti-MYC, anti-Rnr2, and anti-Rnr4 antibodies.

Figure 6.

Relocating Wtm1 to the nucleolus recruits Rnr2 and Rnr4 to the nucleolus. (A) A 3MYC-WTM1-his5⁺ fragment, constructed by fusion PCR, was transformed into the wtm1Δ haploid strain and recombined to the 3′ end of the endogenous NOP1 locus on chromosome IV to generate the NOP1-3MYC-WTM1 fusion. (B) Western blot analysis of protein extracts from the transformed cells blotted with anti-Nop1 antibodies shows the wild-type control (WT), truncated fusion (Trunc), and full-length fusion protein (FL). (C) For comparison, the top panel shows the localization of endogenous Wtm1 and Rnr2 relative to DAPI by immunofluorescence. wtm1Δ, NOP1-3MYC-WTM1, and truncation mutants were grown to log phase, and cells were stained with DAPI (blue), anti-Nop1/Alexor-488 (green), and anti-Rnr2/Cy3 (red) antibodies. The bottom panel shows immunofluorescence of log-phase NOP1-3MYC-WTM1 cells released into 200 mM HU for 2 h, then stained with DAPI, anti-Nop1/Alexor-488 (green), and anti-Rnr2/Cy3 (red) antibodies.