Integration of G2/M checkpoint, spindle assembly checkpoint,and Ran cycle regulators in the Saccharomyces cerevisiae DNA damage mitotic arrest response

Abstract

It is axiomatic that genomic stability is dependent upon regulatory pathways, termed checkpoints, which sense perturbations of cell cycle execution including damage to chromosomal DNA. In Saccharomyces cerevisiae, the principal DNA damage checkpoint is at G2/M. Heretofore, this and other checkpoints, such as the spindle assembly checkpoint, which is also operative at the metaphase/anaphase transition, have been viewed as essentially linear pathways, responding to a specific type of damage, signaling via sui generis proteins, and targeting a limited number of effectors for arrest. In a 1999 report, our laboratory reported bifurcation of the pathway downstream from Mec1 activation; this established the genetic basis of a previously unexplained phenotype: partial arrest defects of rad53 and pds1 strains. Moreover, the bifurcated pathway model established the framework for subsequent studies which determined the molecular targets of each. Here, I present evidence that the DNA damage and spindle checkpoint pathways are part of a network which is capable of bilaterally responding to damage. After DNA damage the Mec1-centric pathway is initially preeminent; the spindle pathway is redundant. After prolonged damage, however, the spindle checkpoint components become required for arrest. In studies of overexpression of the Mec1 homologue Tel1, I delineated the pathway responsible for the resultant constitutive delay; strikingly, the spindle components Mad1 and Mad2 are activated, not from the kinetochore, but from the nuclear periphery. This off-kinetochore pool of Mad proteins, anchored by the myosin-like proteins, Mlp1 and Mlp2, is likewise activated by the DNA damage response. Tel1 physically interacts with Xrs2 of the Mre11·Rad50·Xrs2 complex; evidence that Xrs2 participates in these same responses is also presented. Finally, the sensitivity of xrs2 to a microtubule poison, benomyl, suggests that M R·X may also participate in sensing spindle disruption. From a screen for novel checkpoint genes, I isolated Gtr1 (and later, Gtr2), which are negative regulators of the Ran cycle. Here, I provide evidence that deletion of either produces an identical partial arrest defect, which is independent of the Mec1-centric pathway. Because Gtr2 physically interacts with Esp1, I surmise that Gtr1/Gtr2 may enforce cytosolic localization of Pds1/Esp1 after DNA damage.