Forefronts in Nephrology 1185

ipates in transcriptional regulation, as a component of are investigating whether or not this mechanism operatesin cycling mammalian cells in culture. In S. pombe, ge-transcription factor TFIIH. A physiological role for CDK7

in activating CDKs and promoting cell cycle progression netic studies by Beach et al identified a nonessentialkinase, Csk1, that seemed to stimulate the kinase associ-has now been firmly established by studies of a CDK7

homologue in Drosophila. In budding yeast, CAK activ- ated with Mcs2, the fission yeast homologue of cyclin H.We therefore constructed baculoviruses to overexpressity is supplied by a distinct single-subunit protein kinase,

Cak1, whereas Kin28, the closest homologue of CDK7, Mcs6 (the CDK7 homologue), Mcs2, and Csk1 (the po-tential CAKAK) for biochemical studies in vitro. Re-functions solely in transcription. A CAK in the fission

yeast, Schizosaccharomyces pombe, the Mcs6-Mcs2 com- markably, Csk1 is indeed a potent CAKAK in vitro. Wehave also combined temperature-sensitive mutations inplex, closely resembles the metazoan enzyme and so

presents a novel opportunity to study CDK activation mcs6 and mcs2 with a Dcsk1 null mutation; the resultinggrowth retardation (or arrest in the presence of an addi-in a genetically tractable organism. We are investigating

the regulatory pathways upstream of CAK by a combina- tional mutation, cdc2-3W) suggests more severe compro-mise of CDK activation when both functions are defec-tion of biochemical and genetic strategies in both mam-

malian cells and fission yeast. tive. We are also investigating how MAT1 influences theactivity of the trimeric CAK complex. We have estab-The two major forms of CDK7 in mammalian cells—

the dimer and the trimer—may be assembled by distinct lished stable mammalian cell lines that conditionally ex-press MAT1 in both wild-type and various mutant forms.molecular mechanisms. A stable association between

CDK7 and cyclin H in a binary complex requires phosphor- This will allow us to search for novel protein targets ofMAT1 that would be candidate substrates for CDK7.ylation of CDK7 by a CAK-activating kinase (CAKAK),

whereas ternary complex assembly and activation can Finally, we are comparing the substrate specificities andother enzymatic properties of pure dimeric and trimericoccur in the absence of phosphorylation through the

action of MAT1. Thus, to understand the mechanisms CAK generated in insect cells using baculoviruses, withseveral known substrates that may be important for cellcontrolling CAK function, we are pursuing two major

lines of investigation: the identification and characteriza- cycle control, transcription, or other processes.tion of CAKAKS, both in vitro and in vivo, and a system-atic search for proteins that interact with MAT1, either REFERENCESas substrates or regulators of its associated kinase activ- 1. Fisher RP, Jin P, Chamberlin HM, Morgan DO: Alternative mech-

anisms of CAK assembly require an assembly factor or an activatingity. To date, we have identified two distinct classes ofkinase. Cell 83:47–57, 1995CAKAKs: one in mammalian cells and one in S. pombe.

2. Fisher RP, Morgan DO: CAK in TFIIH: Crucial connection orHuman CDK7-cyclin H complexes are efficiently acti- confounding coincidence? Biochim Biophys Acta 1288:7–10, 1996

3. Fisher RP: CDKs and cyclins in transition(s). Curr Opin Genet Devvated in vitro by both CDC2 and CDK2, which them-7:32–38, 1997selves are critical targets for CAK. Thus, CDK activation

4. Larochelle S, Pandur J, Fisher RP, Salz HK, Suter B: Cdk7 isduring mammalian cell cycle progression may involve a essential for mitosis and for in vivo Cdk-activating kinase activity.

Genes Dev 12:370–381, 1998positive feedback loop of activating phosphorylation. We

Regulation of the yeast cell cycle by transcription andproteolysis of cyclin-dependent kinase regulators

FRED CROSS1 and KRISTI LEVINE

The Rockefeller University, New York, New York USA

Events in the budding yeast cell cycle are driven by through the cell cycle such that peak abundance of theregulators corresponds closely with their natural timecyclical accumulation and destruction of cyclin-depen-of function. There are at least three major points ofdent kinase (CDK) activators (cyclins) and inhibitors.transcriptional control: in postmitotic prereplicativeTranscription of most of these regulators is controlledcells, in cells committed to cell cycle initiation, and incells that have completed DNA replication and are pre-paring for mitosis. At all three points, positive feedback1 Speaker and participant

Forefronts in Nephrology1186

loops involving CDK activity and transcriptional regula- REFERENCEStion contribute to make these cell cycle positions semista- 1. Cross FR: Starting the cell cycle: What’s the point? Curr Opin

Cell Biol 7:790–797, 1995ble “states.” The expression of other genes is tied in to2. Cross F: Transcriptional regulation by a cyclin-cdk. Trends Genet

the same regulatory circuitry driving expression of these 11:209–211, 19953. Levine K, Tinkelenberg AH, Cross F: The CLN gene family:regulators, resulting in cell cycle regulation of genes in-

Central regulators of cell cycle Start in budding yeast. Prog Cellvolved in downstream aspects of cell cycle control. The Cycle Res 1:101–114, 19954. Nasmyth K: At the heart of the budding yeast cell cycle. Trendstranscriptional control of CDK regulators is not essential

Genet 12:405–412, 1996but, rather, probably serves to fine tune the biological 5. Oehlen L, Cross FR: The mating factor response pathway regu-lates transcription of TEC1, a gene involved in pseudohyphal dif-control of the system. Control by proteolytic and otherferentiation of Saccharomyces cerevisiae. FEBS Lett 429:83–88,inhibitory mechanisms is more central and essential for 1998

6. Oehlen L, Cross FR: The role of Cde42 in signal transductionappropriate cell cycle control.and mating of the budding yeast Saccharomyces cerevisiae. J BiolCyclins are absolutely required for enzymatic activa- Chem 273:8556–8559, 1998

7. Oehlen L, Cross FR: Potential regulation of Ste20 function bytion of CDKs. As in all eukaryotes, budding yeast con-the Cln1-Cdc28 and Cln2-Cdc28 cyclin-dependent protein kinases.tains multiple cyclin genes that appear to function to J Biol Chem 273:25089–25097, 1998

8. Joeung DI, Oehlen LJ, Cross FR: Cln3-associated kinase activitycontrol different cell cycle events. In order to begin disso-in Saccharomyces cerevisiae is regulated by the mating factor path-ciating the role of cyclins as simple enzymatic activators way. Mol Cell Biol 18:433–441, 1998

9. Cross FR, Levine K: Molecular evolution allows bypass of theof CDK as distinct from roles of cyclins as specific tar-requirement for activation loop phosphorylation of the Cdc28geting subunits, we have begun genetic approaches to cyclin-dependent kinase. Mol Cell Biol 18:2923–2931, 1998

10. Gartner A, Jovanovic A, Jeoung DI, Bourlat S, Cross FR,“protein engineering” a cyclin-independent version ofAmmerer G: Pheromone-dependent G1 cell cycle arrest requiresthe Cdc28 cyclin-dependent kinase. Thus far, the results Far1 phosphorylation, but may not involve inhibition of Cdc28-

suggest that some cell cycle events may occur simply as Cln2 kinase, in vivo. Mol Cell Biol 18:3681–3691, 199811. Oehlen LJ, Jeoung DI, Cross FR: Cyclin-specific START eventsa consequence of enzymatic activation of CDK, whereas and the G1-phase specificity of arrest by mating factor in budding

others may require specific targeting by cyclin. yeast. Mol Gen Genet 258:183–198, 1998

Mechanism and regulation of the SCF ubiquitination pathway

RAY DESHAIES

California Institute of Technology, Pasadena, California, USA

Entry into the S phase in budding yeast requires inacti- ize the localization, abundance, and composition of SCFcomplexes in vivo. Homologues of SCF components havevation of the S-phase cyclin-dependent kinase (CDK)

inhibitor Sic1. We have shown that this is achieved via been detected in other eukaryotes, and a human Cdc53homologue (CUL1) is able to complement a cdc53tsphosphorylation-triggered ubiquitin-dependent proteoly-

sis. During the early G1 phase, stable Sic1 prevents the mutant. Human CUL1, SKP1, and SKP2 (an F-box pro-tein that is presumed to serve as a Cdc4-like substrateprecocious activation of S-phase–promoting CDK com-

plexes. Late in the G1 phase, newly assembled G1 cyclin/ receptor) assemble into SCF-like complexes upon ex-pression in insect cells, but these complexes have notCDK complexes phosphorylate Sic1 (G1 cyclin/CDK

complexes are insensitive to Sic1 inhibition). Phospho- yet been shown to catalyze CDC34-dependent ubiquiti-nation of a defined substrate. However, human CUL1Sic1 is selectively recognized by a ubiquitin ligase (E3)

complex containing Cdc53, Cdc4, and Skp1. This assem- and human SKP1 can assemble into chimeric SCF com-plexes with yeast Cdc4, and these complexes can sustainblage has been named SCF for Skp1-Cdc53/CUL-F Box

receptor E3 complex. Upon binding to SCF, phospho- the Cdc34-dependent ubiquitination of phospho-Sic1,confirming that the yeast and human proteins possessSic1 is extensively ubiquitinated by the Cdc34 ubiquitin-

conjugating enzyme. Coprecipitation analyses have re- similar biochemical functions. We have recently devel-oped a biochemical assay that may allow us to identifyvealed that the Cdc4 and Skp1 subunits of SCF selectively

bind phospho-Sic1, thereby linking it to Cdc53 and substrates and/or regulators of human CUL1 complexes.The identification of CUL1 as a component of a ubiquitinCdc34. Efforts are currently underway to further subdi-

vide the functional organization of SCF and to character- ligase complex and the availability of biochemical assays