976 VOLUME 17 NUMBER 8 AUGUST 2010 NATURE STRUCTURAL & MOLECULAR BIOLOGY ARTICLES DNA replication and tRNA gene transcription are essential for eukaryotic life and occur at a high rate during active proliferation. Bulk DNA synthesis involves processive elongation by a large macro- molecular machine that assembles at the replication fork. The tRNA gene-transcription cycle involves assembly of a preinitiation complex (PIC) comprised of RNA polymerase III (Pol III) and two DNA-binding factors, TFIIIB and TFIIIC 1 . Assembly of the PIC is a stepwise process, with binding of TFIIIC to sequences in the coding region of the gene followed by sequential recruitment of TFIIIB and Pol III. Studies in budding yeast have revealed a surprising connection between replication elongation and tRNA gene transcription. That is, active tRNA genes act as ‘replication-fork barriers’ during normal proliferation because the Pol III PIC interferes with elongation by processive DNA polymerases 2–9 . Little is known about how the Pol III PIC impedes the replication fork. Although some tRNA genes show barrier activity only when they are oriented to fire Pol III into the leading edge of the elongating DNA polymerase complex 2 , pausing at tRNA genes that are co-directional with forks is now well documented 4,9 . The 274 tRNA genes in budding yeast exist in diverse chromosomal settings, raising the possibility that barrier activity is partly determined by the local chromatin environ- ment. However, a recent study of replication dynamics based on move- ment of the replisome-associated GINS complex did not reveal any link between fork pausing and either tRNA gene location or orientation (the tRNA genes were all equally difficult to replicate) 9 . Therefore, while a dose-dependent relationship between fork pausing and Pol III recruitment to tRNA genes is readily apparent 2 , the precise mechanism by which the Pol III PIC causes fork pausing remains unknown. The biological consequences of fork pausing, on the other hand, are relatively well understood. Perhaps most importantly, pausing predisposes forks to collapse 10 . And because the mechanisms engaged by the cell to rescue collapsed forks can cause undesirable genome rearrangements 11,12 , fork pausing poses an inherent risk to genome stability. There is good evidence that tRNA genes are naturally prone to replication-dependent instability. In normally cycling cells, sponta- neous chromosome breakage often occurs at sites that include fork- pausing tRNA genes 13 . Furthermore, breakage at these sites is elevated by treatment with hydroxyurea (HU), which causes fork movement to slow down by depleting cells of deoxyribonucleoside triphosphates 13 . Given the intimate connections between tRNA gene activity, repli- cation and genome stability control, it is not surprising that fork- associated proteins affect pausing at tRNA genes. The DNA helicase Rrm3 counteracts pausing at tRNA genes 3,14,15 , and Tof1, a protein of unknown function, helps to impose the pause 7 . Neither protein seems to modulate pausing at tRNA genes by regulating Pol III transcrip- tion 16 (B.W.C. and M.C.S., unpublished data). Although Rrm3 and Tof1 might collaborate to set the rate of fork progression through tRNA genes, there is no evidence that this rate is subject to physiological regulation by mechanisms that determine the balance of activity between Rrm3 and Tof1. Furthermore, tRNA gene regulation is not known to be tied to fluctuations in the rate of DNA replication other than by mechanisms that generally tune the proliferation rate to nutrient availability and overall cellular fitness for growth and division 17 . However, there is evidence that the appear- ance of double strand breaks (DSBs) in DNA can trigger repression of the tRNA genes. Specifically, tRNA gene transcription is actively Department of Biochemistry, School of Molecular and Systems Medicine, University of Alberta, Edmonton, Alberta, Canada. Correspondence should be addressed to M.C.S. (michael.schultz@ualberta.ca). Received 29 April; accepted 20 May; published online 18 July 2010; doi:10.1038/nsmb.1857 Replication stress checkpoint signaling controls tRNA gene transcription Vesna C Nguyen, Brett W Clelland, Darren J Hockman, Sonya L Kujat-Choy, Holly E Mewhort & Michael C Schultz In budding yeast, the transcriptional machinery at tRNA genes naturally interferes with replication in a way that can promote chromosome breakage. Here we show that a signaling module composed of core components of the replication stress checkpoint pathway represses this fork-pausing machinery in normally cycling and genotoxin-treated cells. Specifically, the sensor kinase Mec1, the signaling adaptor Mrc1 and the transducer kinase Rad53 relay signals that globally repress tRNA gene transcription during unchallenged proliferation and under conditions of replication stress. Repressive signaling in genotoxin-treated cells requires Rad53-dependent activation of a conserved repressor of tRNA gene transcription, Maf1. Cells lacking Maf1 are sensitive to replication stress under conditions of elevated tRNA gene transcription. We propose that checkpoint control of the fork-pausing activity of tRNA genes complements the repertoire of replisome-targeted mechanisms by which checkpoint proteins promote faithful DNA replication. © 2010 Nature America, Inc. All rights reserved.