To understand regulatory systems, it would be useful to uniformly determine how different components contribute to the expression of all other genes. We therefore monitored mRNA expression genome-wide, for individual deletions of one-quarter of yeast genes, focusing on (putative) regulators. The resulting genetic perturbation signatures reflect many different properties. These include the architecture of protein complexes and pathways, identification of expression changes compatible with viability, and the varying responsiveness to genetic perturbation. The data are assembled into a genetic perturbation network that shows different connectivities for different classes of regulators. Four feed-forward loop (FFL) types are overrepresented, including incoherent type 2 FFLs that likely represent feedback. Systematic transcription factor classification shows a surprisingly high abundance of gene-specific repressors, suggesting that yeast chromatin is not as generally restrictive to transcription as is often assumed. The data set is useful for studying individual genes and for discovering properties of an entire regulatory system.

We have designed a doxycycline-regulated form of the H1 promoter of RNA polymerase III that allows the inducible knockdown of gene expression by small interfering RNAs (siRNAs). As a proof-ofprinciple, we have targeted β-catenin in colorectal cancer (CRC) cells. T-cell factor (TCF) target-gene expression is induced by accumulated β-catenin, and is the main transforming event in these cells. We have shown previously that the disruption of β-catenin/TCF4 activity in CRC cells by the overexpression of dominant-negative TCF induces rapid G1 arrest and differentiation. Stable integration of our inducible siRNA vector allowed the rapid production of siRNAs on doxycycline induction, followed by specific downregulation of β-catenin. In these CRC cells, TCF reporter-gene activity was inhibited, and G1 arrest and differentiation occurred. The inhibition of two other genes using this vector system shows that it should be useful for the inducible knockdown of gene expression. EMBO reports 4, 609-615 (2003) doi:10.1038/sj.embor.embor865 INTRODUCTIONThe transactivation of T-cell factor (TCF) target genes induced by wingless-related (WNT) pathway mutations is the main transforming event in colorectal cancer (CRC; Kinzler & Vogelstein, 1996;Bienz & Clevers, 2000). We have recently studied the TCF target-gene programme by the inducible overexpression of dominant-negative versions of TCF1 and TCF4 in CRC cell lines (van de Wetering et al., 2002;Batlle et al., 2002). This overexpression disrupted endogenous β-catenin/TCF4 activity in CRC cells. Importantly, it induced rapid G1 arrest and blocked a genetic programme that is physiologically active in the proliferative compartment of colon crypts. Consequently, an intestinal differentiation programme was induced. We concluded that the β-catenin/TCF4 complex is the master switch that controls the decision between proliferation versus differentiation in healthy and malignant intestinal epithelial cells.As the overexpression of dominant-negative proteins might induce artefactual effects, we used a loss-of-function strategy to confirm our results. It is possible to carry out classical gene knockouts by homologous recombination in CRC cells (Shirasawa et al., 1993;Chan et al., 1999Chan et al., , 2002Kim et al., 2002;Sekine et al., 2002), but this technology is relatively time-consuming. Moreover, as we expected a growth-arrest phenotype, clones would either fail to develop or would be selected for other growth-promoting events. More recently, RNA interference (RNAi), a well-established method for gene knockdown in model organisms (Sharp, 2001), can also be used for gene knockdown in mammalian cells (Elbashir et al. 2001). So-called small interfering RNAs (siRNAs) have been introduced into mammalian cells by the transient transfection of synthetic doublestranded RNA. Alternatively, promoters of genes transcribed by RNA polymerase III have been used to drive the expression of hairpin RNAs, which are very similar to siRNAs (Brummelkamp et al., 2002;Miyagishi & Taira, 2002;Paul et al., 200...

Many biochemical, physiological and behavioural processes show circadian rhythms which are generated by an internal time-keeping mechanism referred to as the biological clock. According to rapidly developing models, the core oscillator driving this clock is composed of an autoregulatory transcription-(post) translation-based feedback loop involving a set of 'dock' genes. Molecular clocks do not oscillate with an exact 24-hour rhythmicity but are entrained to solar day/night rhythms by light. The mammalian proteins Cryl and Cry2, which are members of the family of plant blue-light receptors (cryptochromes) and photolyases, have been proposed as candidate light receptors for photoentrainment of the biological clock. Here we show that mice lacking the Cryl or Cry2 protein display accelerated and delayed free-running periodicity of locomotor activity, respectively. Strikingly, in the absence of both proteins, an instantaneous and complete loss of free-running rhythmicity is observed. This suggests that, in addition to a possible photoreceptor and antagonistic clock-adjusting function, both proteins are essential for the maintenance of circadian rhythmicity.

Summary Packaging of DNA into chromatin has a profound impact on gene expression. To understand how changes in chromatin influence transcription, we analyzed 165 mutants of chromatin machinery components in Saccharomyces cerevisiae. mRNA expression patterns change in 80% of mutants, always with specific effects, even for loss of widespread histone marks. The data is assembled into a network of chromatin interaction pathways. The network is function-based, has a branched, interconnected topology and lacks strict one-to-one relationships between complexes. Chromatin pathways are not separate entities for different gene sets, but share many components. The study evaluates which interactions are important for which genes and predicts new interactions, for example between Paf1C and Set3C, as well as a role for Mediator in subtelomeric silencing. The results indicate the presence of gene-dependent effects that go beyond context-dependent binding of chromatin factors and provide a framework for understanding how specificity is achieved through regulating chromatin.

therefore also play an important role in transmitting ac-2 Department of Medical Genetics tivator signals. Mediator can function as a scaffold for University Medical Center Utrecht repeated rounds of reinitiation by RNA polymerase II Universiteitsweg 100 (Yudkovsky et al., 2000), and a postinitiation role has 3584 CG Utrecht also recently been proposed (Wang et al., 2005). The The Netherlands importance of Mediator is exemplified by the immediate drop in virtually all transcripts upon inactivation of the Med17 subunit in the yeast Saccharomyces cere-Summary visiae (Holstege et al., 1998). Although Mediator is generally thought to play a pos-Mediator is an evolutionarily conserved coregulator itive role, several subunits have been implicated in of RNA polymerase II transcription. Microarray strucnegative regulation (Hampsey, 1998; Myers and Kornture-function analysis of S. cerevisiae Mediator reberg, 2000). Mediator from several organisms exists in veals functional antagonism between the cyclinat least two forms that differ mainly by the presence or dependent kinase (Cdk) submodule and components absence of a negative regulatory submodule (Liu et al., from the Tail (Med15, Med2, Med3), Head (Med20, 2001; Sato et al., 2004; Spahr et al., 2003). This negative Med18), and Middle (Med31). Certain genes exhibit regulatory submodule consists of a cyclin-dependent increased or decreased expression, depending on kinase (Cdk8) and its cyclin partner (CycC) as well as which subunit is deleted. Epistasis analysis with extwo additional subunits, Med12 and Med13, all of which pression-profile phenotypes shows that MED2 and are well conserved (Bourbon et al., 2004). Interestingly, MED18 are downstream of CDK8. Strikingly, Cdk8it has recently been shown that some mammalian Memediated modification of a single amino acid within diator complexes alternatively harbor Cdk11 (Sato et Mediator represses the regulon of a single transcripal., 2004). The negative role of Cdk8 is exemplified by tion factor, Rcs1/Aft1. Highly specific gene regulation upregulation of a significant subset of genes in yeast is thought to be determined by activators and combicells bearing kinase-defective Cdk8 (Holstege et al., natorial use of cofactors. Here, subtle modification of 1998). the general transcription machinery through one of Cdk8 is the most frequently studied Mediator subits own components is shown to determine highly unit, and several different models have been proposed specific expression patterns. Expression profiling for the mechanism of Cdk8 repression (Akoulitchev et can therefore precisely map regulatory cascades, and al., 2000; Chi et al., 2001; Fryer et al., 2004; Hengartner our findings support a role for Mediator as a direct et al., 1998; Nelson et al., 2003). Studies of mammalian processor of signaling pathways for determining transcription show that for some genes, Cdk8 is associspecificity. ated with inactive transcription complexes (Pavri et al., 2005). In other cases, Cdk8 is located within the initia-

The resting state of eukaryotic cells (G0) is relatively uncharacterized. We have applied DNA microarray expression profiling of S. cerevisiae to reveal multiple transitions during a complete 9-day growth cycle between stationary phase (SP) exit and entry. The findings include distinct waves of transcription after the diauxic shift (DS), identification of genes active in SP, and upregulation of over 2500 genes during the first minutes of lag phase. This provides a framework for analyzing large-scale reprogramming of gene expression. Despite global repression, the general transcription machinery is found to be present in quiescent cells but is largely inactive. Genome-wide location analysis by chromatin immunoprecipitation (ChIP on chip) reveals that RNA polymerase II is more predominantly bound at intergenic regions in SP, upstream of hundreds of genes immediately induced upon exit. In contrast to current models of activation-coupled recruitment, the results show that RNA polymerase II is located and maintained upstream of many inactive genes in quiescence.

SUMMARY To understand relationships between phosphorylation-based signaling pathways, we analyzed 150 deletion mutants of protein kinases and phosphatases in S. cerevisiae using DNA microarrays. Downstream changes in gene expression were treated as a phenotypic readout. Double mutants with synthetic genetic interactions were included to investigate genetic buffering relationships such as redundancy. Three types of genetic buffering relationships are identified: mixed epistasis, complete redundancy, and quantitative redundancy. In mixed epistasis, the most common buffering relationship, different gene sets respond in different epistatic ways. Mixed epistasis arises from pairs of regulators that have only partial overlap in function and that are coupled by additional regulatory links such as repression of one by the other. Such regulatory modules confer the ability to control different combinations of processes depending on condition or context. These properties likely contribute to the evolutionary maintenance of paralogs and indicate a way in which signaling pathways connect for multiprocess control.

Histone H3 di- and trimethylation on lysine 4 are major chromatin marks that correlate with active transcription. The influence of these modifications on transcription itself is, however, poorly understood. We have investigated the roles of H3K4 methylation in Saccharomyces cerevisiae by determining genome-wide expression-profiles of mutants in the Set1 complex, COMPASS, that lays down these marks. Loss of H3K4 trimethylation has virtually no effect on steady-state or dynamically-changing mRNA levels. Combined loss of H3K4 tri- and dimethylation results in steady-state mRNA upregulation and delays in the repression kinetics of specific groups of genes. COMPASS-repressed genes have distinct H3K4 methylation patterns, with enrichment of H3K4me3 at the 3′-end, indicating that repression is coupled to 3′-end antisense transcription. Further analyses reveal that repression is mediated by H3K4me3-dependent 3′-end antisense transcription in two ways. For a small group of genes including PHO84, repression is mediated by a previously reported trans-effect that requires the antisense transcript itself. For the majority of COMPASS-repressed genes, however, it is the process of 3′-end antisense transcription itself that is the important factor for repression. Strand-specific qPCR analyses of various mutants indicate that this more prevalent mechanism of COMPASS-mediated repression requires H3K4me3-dependent 3′-end antisense transcription to lay down H3K4me2, which seems to serve as the actual repressive mark. Removal of the 3′-end antisense promoter also results in derepression of sense transcription and renders sense transcription insensitive to the additional loss of SET1. The derepression observed in COMPASS mutants is mimicked by reduction of global histone H3 and H4 levels, suggesting that the H3K4me2 repressive effect is linked to establishment of a repressive chromatin structure. These results indicate that in S. cerevisiae, the non-redundant role of H3K4 methylation by Set1 is repression, achieved through promotion of 3′-end antisense transcription to achieve specific rather than global effects through two distinct mechanisms.