Never say never. The NIMA-related protein kinases in mitotic control

Abstract

Mitosis sees a massive reorganization of cellular archi-tecture. The microtubule cytoskeleton is reorganized toform a bipolar spindle between duplicated microtubuleorganizing centers, the chromosomes are condensed,attached to the spindle at their kinetochores, and,through the action of multiple molecular motors, thechromosomes are segregated into two daughter cells.Mitosis also sees a substantial wave of protein phos-phorylation, controlling signaling events that coordinatemitotic processes and ensure accurate chromosome seg-regation. The key switch for the onset of mitosis is thearchetypal cyclin-dependent kinase, Cdc2. Under thedirection of Cdc2 is an executive of protein serine/threonine kinases that fall into three families: the Polokinases, Aurora kinases and the NIMA-related kinases(Nrk). The latter family has proven the most enigmaticin function, although recent advances from severalsources are beginning to reveal a common functionaltheme.Mitosisisthecrucialpointinthecellcycleatwhichonecelldivides to form two. The conservation of both the generalmechanisms and regulatory pathways controlling mitosisthroughout the eukaryotic kingdom highlights its ancientnature and importance. It is this very conservation thathas allowed the identification of a common retinue ofmitotic regulators from a diverse band of eukaryotesranging from yeasts to humans. Several key mitoticregulators are protein-serine/threonine kinases that areresponsible for modulating entry into, and progressionthrough, mitosis. Uppermost in the mitotic phosphoryl-ation cascade is the cyclin-dependent kinase Cdc2, which,upon activation late in G2 phase, induces the dramaticcellular reorganization that is associated with mitosis.Regulating the downstream events following Cdc2 acti-vation is a trio of protein kinase families: the Aurora, Poloand NIMA-related (Nrk) kinases. These are also highlyconserved in evolution, although, while yeasts have asingle member of each family, in humans these havegrown into multi-gene families encoding related proteins.The Aurora and Polo families have been best studied todate and regulate many events, from the initiationthrough to the exit of mitosis (for recent reviews onthese families, see [1–8]).The third family of mitotic kinases, the NIMA-relatedkinases (Nrk), is less well understood, and the purpose ofthis review is to summarize what is known about themembers of this enigmatic group and discusses theimplications for the regulation of mitosis.Aspergillus NIMA: discovery and early analysisThe early 1970s saw the beginnings of the genetic analysisof the cell cycle, with the seminal screens of Lee Hartwelland Paul Nurse in Saccharomyces cerevisiae and Schizo-saccharomyces pombe, respectively. At around the sametime,RonMorriscarriedoutascreenforcellcyclemutantsin the multicellular filamentous fungus Aspergillusnidulans (Fig. 1). Morris screened hundreds of tempera-ture sensitive mutants, with cell cycle mutants beingclassified as either bim mutants, for those blocked inmitosiswithcondensedchromosomesandmitoticspindles,ornimmutants,forthosethatwereneverinmitosisduetoan interphase arrest [9]. The bim collection eventuallyproved to include important mitotic genes such as spindlemotors and components of the anaphase-promoting com-plex (APC). The nim mutants curiously did not include acdc2 homolog, but did include many well-known cell cycleregulators such as a B-type cyclin, Cdc25 phosphatase andDNApolymerases.Theyalsoincludedfourallelesofagenedenoted as nimA [10].nimA mutants arrested late in G2, with duplicatedspindle pole bodies (SPBs), the fungal equivalent of thecentrosome [11]. By the mid-1980s, nimA was cloned andshown to encode a serine/threonine protein kinasedesignated NIMA [12,13] (see Fig. 2 for domain organiz-ation of NIMA and other Nrks). Further genetic analysesshowed that the interphase arrest of nimA5 mutants wasbypassed by the bimE7 mutation, with double mutantsentering an aberrant mitotic state characterized byabnormalities of the nuclear envelope, which remainsintact during mitosis in fungi, and the failure to form abipolar spindle [14]. This has been attributed to attempt-ing mitosis with suboptimal NIMA function, as the bimE7mutationdoesincreasethelevelandactivityofthemutantNIMA5 protein, although to only ,20% of wild-type levels[15]. Presumably, this activity is sufficient to overcome