The Dun1 checkpoint kinase phosphorylates and regulates the ribonucleotide reductase inhibitor Sml1

Abstract

Cell cycle checkpoints are evolutionarily conserved surveillance systems that protect genomic stability and prevent oncogenesis in mammals. One important target of checkpoint control is ribonucleotide reductase (RNR), which catalyzes the rate-limiting step in dNTP and DNA synthesis. In both yeast and humans, RNR is transcriptionally induced after DNA damage via Mec1/Rad53 (yeast) and ATM/CHK2 (human) checkpoint pathways. In addition, yeast checkpoint proteins Mec1 and Rad53 also regulate the RNR inhibitor Sml1. After DNA damage or at S phase, Mec1 and Rad53 control the phosphorylation and concomitant degradation of Sml1 protein. This new layer of control contributes to the increased dNTP production likely necessary for DNA repair and replication; however, the molecular mechanism is unclear. Here we show that Dun1, a downstream kinase of Mec1/Rad53, genetically and physically interacts with Sml1 in vivo. The absence of Dun1 activity leads to the accumulation of Sml1 protein at S phase and after DNA damage. As a result, dun1Delta strains need more time to finish DNA replication, are defective in mitochondrial DNA propagation, and are sensitive to DNA-damaging agents. Moreover, phospho-Sml1 is absent or dramatically reduced in dun1Delta cells. Finally, Dun1 can phosphorylate Sml1 in vitro. These results suggest that Dun1 kinase function is the last step required in the Mec1/Rad53 cascade to remove Sml1 during S phase and after DNA damage.

Figure 1

sml1Δ suppresses synthetic growth defects of dun1Δ mec1 and dun1Δ rad53 strains and the increased petite formation in dun1Δ mec1 and dun1Δ strains. (a) Three tetrads (displayed horizontally) are shown for each diploid. The genotype of inviable spores is deduced from those of the sister spore clones and are shown dun1Δ rad53-1 Left and dun1Δ rad53-17 Right. The arrows indicate dun1Δ rad53-1 sml1Δ and dun1Δ rad53-17 sml1Δ triple mutants. Spore clones containing other genotypes grow equally well. (b) The slow-growing spore clones indicated by the arrows are dun1Δ mec1-3 and dun1Δ mec1-8. The faster growing triple mutants (dun1Δ mec1-3 or -8 sml1Δ) are indistinguishable from wild-type or the single mutants. (c) The averages and SDs of the frequency of petite formation are plotted for the four strains indicated. (d) Spore clones from b were streaked on yeast extract/peptone/dextrose (YPD) medium and then replica-plated onto YPGlycerol medium. Petite cells grow only on YPD but not on YPGlycerol medium. The strains depicted are (1) sml1Δ, (2) dun1Δ mec1-8 sml1Δ, (3) dun1Δ mec1-8, (4) dun1Δ mec1-3, and (5) dun1Δ mec1-3 sml1Δ. Cells in sectors 3 and 4 are mostly petite.

Figure 3

After DNA damage, Sml1 protein accumulates in dun1Δ strains whereas the phosphorylated forms of Sml1 are absent. (a) Wild-type, dun1Δ, and mec1Δ strains were treated with 0.05% methyl methanesulfonate for 1 h or were irradiated by UV light (120 J/m²) or γ-rays (30 krads). Sml1 protein levels were examined by protein blot with anti-Sml1 Ab. Arrowheads indicate the position of Sml1 protein. The band above Sml1 crossreacts with anti-Sml1 serum and serves as a loading control. The mec1Δ strain, which is also defective in Sml1 regulation, is shown for comparison. Because the mec1Δ strain is unable to grow in the presence of Sml1, its viability was maintained by a 2 μm-RNR1 plasmid (24). (b) dun1Δ strains containing either GST-Dun1 or GST-Dun1-D328A were irradiated by γ-rays (30 krads). The levels of the GST-fusion proteins as well as Sml1 protein were examined by protein blot with anti-GST Ab and anti-Sml1 Ab, respectively. (c) Wild-type, dun1Δ, and mec1Δ strains, all containing a RNR1 plasmid, were treated as in a. Proteins were extracted by trichloroacetic acid methods (5), and phosphorylated Sml1 bands (bracket) were visualized by immunoblot with anti-Sml1 Ab. Note that the crossreacting band above Sml1 in a does not appear under these conditions.

Figure 4

Dun1 interacts with and phosphorylates Sml1. (a) Plasmids containing different GBD- and GAD-fusion proteins were cotransformed into a two-hybrid strain (PJ69-4A, 14), and the transformants were selected on medium lacking tryptophan and leucine (-T-L). Interactions between fusion proteins were indicated by growth on histidine-less (-T-L-H) and adenine-less media (-T-L-A) to examine the activation of the two-hybrid reporters. The interaction between GBD-Rnr1 and GAD-Sml1 (3) was used as a positive control. (b) dun1Δ sml1Δ strains that overexpress GST-Dun1 and GST-Dun1-D328A fusion proteins were irradiated with γ-rays (30 krads). Before and after DNA damage, the fusion proteins were purified by using glutathione beads and then incubated with Sml1 protein in kinase reactions. The levels of the fusion proteins and Sml1 were visualized by immunoblot (Left). The brackets indicate the phosphorylated forms of Sml1, and the arrows below show the position of unphosphorylated Sml1 protein. The phosphorylation levels (³²P) were visualized after autoradiography (Right).

Figure 5

A model for the role of Dun1 in regulation of RNR activity after DNA damage. DNA lesions provoke the activation of Mec1/Rad53/Dun1 via the bifurcated Rad9 and Rad17/Rad24/Mec3 pathways (5). Dun1 kinase phosphorylates Sml1 directly leading to its degradation. We suspect that the unbound form of Sml1 is targeted for destruction (6). At the same time, Dun1 activates the transcriptional induction of RNR genes although its direct substrate remains unknown. The transcriptional induction likely causes increased production of RNR proteins (dashed arrow). The dual mode of Dun1 function outlined here ensures a rapid yet lasting increase in RNR activity likely necessary for efficient DNA repair.