This page has been archived and is no longer updated
Tuning gene expression to changing environments: from rapid responses to evolutionary adaptation.
Author: Luis López-Maury
Keywords
Keywords for this Article
Add keywords to your Content
Save
|
Cancel
Share
|
Cancel
Revoke
|
Cancel
Rate & Certify
Rate Me...
Rate Me
!
Comment
Save
|
Cancel
Flag Inappropriate
The Content is
Objectionable
Explicit
Offensive
Inaccurate
Comment
Flag Content
|
Cancel
Delete Content
Reason
Delete
|
Cancel
Close
Full Screen
"Cells show a remarkable regulatory flexibility that allows them to thrive under different external conditions and to survive harsh situations. The ability to constantly sense and adapt to environmental changes is important for all organisms to maintain cellular functions (home- ostasis), but is especially acute for plants and microor- ganisms; their sessile lifestyle leave them more exposed to the environment than animals. Modulation of gene expression has a central role in cellular adaptation to short- or long-term environmental changes, with exten- sive regulation occurring at both the transcriptional and post-transcriptional level. Signal-transduction pathways can translate extracellular signals into specific intracel- lular responses, including the launch of alternative gene expression programmes to cope with new conditions. In addition to these ?hard-wired? responses, gene expres- sion networks show considerable plasticity to adapt to a wide range of challenges, including those not encoun- tered during evolutionary history (for example, ectopic gene expression). The application of genome-wide approaches is now providing a global view on gene expression responses to many different stress conditions, leading to exciting recent advances in our understanding of the cellular strategies that are used to stay in tune with environ- mental conditions. Many of these concepts have been developed in microorganisms, most notably yeast cells, which finely balance energy-efficient growth with the ability to rapidly adapt to sudden external challenges, and which provide ideal models to study gene expres- sion under tightly controlled conditions. Research in yeast and in other organisms is uncovering conserved principles for regulatory strategies in response to chang- ing environments. This Review will highlight some of the emerging principles underlying gene expression responses to environmental factors. The emphasis will be on transcriptional mechanisms, which have been most intensely studied for technical and histori- cal reasons, and for which several recent papers have greatly advanced our understanding. The details of regulatory pathways or specific cellular responses to different stresses have been reviewed elsewhere 1?4 . We focus on the general principles by which cells are able to adjust their gene expression programmes to respond to changing environments, both in the short-term and over evolutionary timescales. Emerging data reveal that cells finely balance the expression of stress-related and growth-related genes, which are antagonistic pro- grammes distinguished by distinct regulatory mecha- nisms. Stress-related genes generally contain TATA boxes, a promoter element that seems to promote both short-term variability and long-term evolvability of transcriptional responses. Maintaining cellular func- tionality under variable conditions thus enhances gene expression variability and is both a constraint and a driving force for evolution. Department of Genetics, Evolution and Environment, and UCL Cancer Institute, University College London, London WC1E 6BT, UK. Correspondence to J.B. e-mail: jurg@sanger.ac.uk doi:10.1038/nrg2398 Published online 1 July 2008 Tuning gene expression to changing environments: from rapid responses to evolutionary adaptation Luis L�pez-Maury, Samuel Marguerat and J�rg B�hler Abstract | Organisms are constantly exposed to a wide range of environmental changes, including both short-term changes during their lifetime and longer-term changes across generations. Stress-related gene expression programmes, characterized by distinct transcriptional mechanisms and high levels of noise in their expression patterns, need to be balanced with growth-related gene expression programmes. A range of recent studies give fascinating insight into cellular strategies for keeping gene expression in tune with physiological needs dictated by the environment, promoting adaptation to both short- and long-term environmental changes. Not only do organisms show great resilience to external challenges, but emerging data suggest that they also exploit these challenges to fuel phenotypic variation and evolutionary innovation. NATURE REVIEWS | GENETICS VOLUME 9 | AUGUST 2008 | 583 REVIEWS Core stress response Involves genes with expression levels that are regulated in a stereotypical manner in all (or most) of the environmental stress conditions tested in yeasts. This response is also known as the environmental stress response (ESR), common environmental response (CER) or core environmental stress response (CESR). Chemostat A fermenter that is operated in continuous-culture mode and is used in microbiology for growing and harvesting microorganisms at defined growth rates and under tightly controlled conditions. Below, we will first give an overview of stress-related gene expression programmes, which are controlled antagonistically to growth-related gene expression pro- grammes, followed by a description of recent insights into the mechanistic principles of gene expression tuning. We finish with an evolutionary perspective on the adaptation of gene expression to both known and previously unencountered environmental factors, which in turn seem to promote regulatory variation and evolution. Global stress-response programmes Cells respond to abrupt environmental changes by launching gene expression programmes that help to adjust the cellular physiology and metabolism to the new conditions and that protect against cell damage or death. The use of DNA microarrays for expression profiling has provided comprehensive insights into tran- scriptional responses to a wide range of stress conditions in several organisms, including yeasts 5?8 , plants 9 , flies 10,11 and humans 12 . These and similar studies have uncovered hundreds of genes, the transcripts of which are either induced or repressed in response to stress. Most of these gene expression responses are transient and, even with persistent stress, gene expression returns after some time to new steady-state levels that are close to those in unstressed cells. Stress responses show consider- able sophistication and fine-tuning: the magnitude and duration of the response is proportional to the dose or severity of the perturbation, and different perturbations result in distinct expression signatures 7,13,14 . Moreover, when exposed to different simultaneous stresses, the cel- lular response approximates the sum of the responses for each individual stress, indicating that gene expression programmes can be combined for a precise response 7 . Besides specific responses to different stresses, both Saccharomyces cerevisiae and Schizosaccharomyces pombe show a large response that is common to all or most stress conditions, and this core response is largely conserved between the two distantly related yeasts 5?7 . Stress-induced genes are enriched for heat-shock and antioxidant func- tions as well as for carbohydrate metabolism and energy- generation functions, whereas most stress-repressed genes have growth-related functions, such as transla- tion and ribosome biogenesis, reflecting a redirection of resources from rapid proliferation to stress protection. The core stress response could explain the phenomenon of cross-protection, in which cells treated with low levels of one stress become also more resistant to other stresses 15 . Other organisms seem to launch much smaller, if any, core stress responses. For example, the pathogenic yeast Candida albicans shows only a limited core response to different stresses 8 . It is possible that this difference reflects the different lifestyles: unlike free-living yeasts, C. albicans proliferates in human hosts where it is largely shielded from environmental variations. In multicellu- lar organisms, the core stress response seems to include only a small set of genes, which is enriched for regulatory functions 9?12 . Unlike in free-living yeasts, ribosome- and other growth-related genes are not 10,11 or are only weakly 9,12 downregulated in multicellular organisms. Some of these differences could reflect the fact that many cell types are present in multicellular organisms, most of which are differentiated, whereas the experiments in yeast have been performed with homogeneous, rapidly growing cells. In fact, quiescent S. cerevisiae cells (that resemble the differentiated G0 state) launch a stress response that overlaps only partially with the response in rapidly growing cells and that does not involve repression of growth-related genes 16 . Moreover, stress responses in multicellular organisms seem to differ between different cell lines and tissues 9,12,17 . Balanced control of growth and stress response The balance between energy-efficient growth and the ability to rapidly respond to fluctuating environments is a fundamental physiological challenge, most notably for microorganisms. Cellular functions, such as metabolism, stress protection, and growth and proliferation, reflect external factors and can be dynamically adjusted to both transient and long-term environmental changes. Rapidly growing S. cerevisiae cells spend most of their tran- scriptional energy on ribosome synthesis, requiring the coordinated activity of all three RNA polymerases, and cellular economy dictates that ribosome synthesis is in tune with the growth rate 18,19 . Nutrient availability in the environment limits growth and proliferation and accord- ingly has a great influence on global gene expression. S. cerevisiae cells can survive starvation by undergoing differentiation, including quiescence 20 , filamentation 21 or sporulation 22 , which are all triggered by specialized gene expression programmes and culminate in highly stress-resistant cells or spores, and which can therefore be regarded as sophisticated stress responses. Rapid growth and high stress resistance seem to be mutually exclusive. As described above, the core stress response in S. cerevisiae redirects resources from growth to stress functions, and stress-resistant cells are non-growing. The degree of stress resistance is inversely correlated with growth rate 23 . Thus, yeast cells need to balance rapid growth and increased stress-resistance according to nutrient availability and stress conditions (FIG. 1). This balance between maximal physiological activity and stress resistance might be more universally valid. For example, nutrient limitation (caloric restriction) leads to increased stress resistance and lifespan in organisms ranging from yeast to man 24 . Moreover, cancer cells sus- tain rapid proliferation by upregulating ribosome- and other growth-related genes, but can also enter dormant stages with increased resistance to drugs 25 . The expression of growth- and stress-related genes as a function of growth rate has recently been studied in S. cerevisiae grown under steady-state conditions in chemostats 26?28 . Chemostats maintain constant growth conditions, such as oxygen levels, nutrient concentra- tions and pH values, while tightly controlling growth at defined rates. Different nutrient limitations have been used to distinguish the effects of specific nutrients from general growth-related effects on gene expres- sion. Hundreds of genes are correlated with growth rate, both positively and negatively, revealing striking overlaps with the core stress response: genes that are R E V I E W S 584 | AUGUST 2008 | VOLUME 9 www.nature.com/reviews/genetics Nature Reviews | Genetics ? ? Nutrients and/or growth factors Phenotypic balance Growth signalling (TOR) Stress signalling (SAPK) Stress Stress Growth Growth-related genes ? No TATA box ? Robust expression Stress-induced genes ? TATA box ? Noisy expression highly expressed during rapid growth are enriched for genes that are repressed during stress, whereas genes that are expressed at a low level during rapid growth (but highly expressed during slow growth) are enriched for genes that are induced during stress. These findings raise the possibility that many of the stress-response genes do not respond directly to stress but to a reduction in growth rate caused by stress 26,27 . In principle, cells could use properties of their internal state, such as growth rate, metabolite concentrations or energy levels, to coordinate gene expression with physi- ological needs. An alternative strategy would be to rely on signalling pathways that are responding to external condi- tions to deduce internal needs. To distinguish between these two possibilities, a recent study in S. cerevisiae asked whether the expression of ribosome-biogenesis genes and stress-related genes responds to internal or external fac- tors 29 . Using a mutant that grows faster on non-fermentable carbon sources than on glucose (the opposite of wild-type cells, which grow faster on glucose), the actual growth rate was uncoupled from the environmentally expected growth rate shown by wild-type cells. Under these condi- tions, the correlation between growth rate and expression of ribosome-biogenesis genes was lost: unlike wild-type cells, the mutant cells expressed ribosome-biogenesis genes at low levels and stress-related genes at high lev- els during rapid growth. Thus, the mutant cells tuned gene expression to the environment rather than to their growth rate. Moreover, changes in ribosome-biogenesis gene expression and stress-related gene expression preceded changes in growth rates in chemostat cultures that were adapting to environmental perturbations, and growth rates did not correlate with ribosome-biogenesis gene expression in a large set of deletion mutants 14,29 . Together, these findings indicate that the expression of ribosome-biogenesis genes and stress-related genes mainly respond to environmental signals rather than to any internal growth-dependent feedback mechanism. Cells use multiple signalling pathways to respond to environmental changes and to coordinate growth with stress responses. Two notable examples are the stress- activated protein kinase (SAPK) and target of rapamycin (TOR) pathways, which are conserved from yeast to man 1,2,4 . Both pathways integrate multiple functions by regulating gene expression at several levels and also by controlling proteins post-translationally. The SAPK and TOR pathways have central roles in balancing stress resistance with growth (FIG. 1). The TOR pathway pro- motes growth and negatively regulates stress-response genes, and possibly also promotes the SAPK path- way when activated by nutrients or growth factors 2,4 . Lowering TOR signalling by mutations or by drugs is sufficient to increase stress resistance and lifespan in S. cerevisiae 30 . The SAPK pathway promotes stress resist- ance when activated by environmental perturbations 1,3 ; in S. pombe, it plays a central part both for the induction of stress-related genes and for the repression of growth- related genes 6,13 . It is possible that special regulatory proteins can integrate inputs from SAPK, TOR and other signalling pathways to coordinate the expression of stress- and growth-related genes. Mechanisms of transcriptional tuning Clearly, gene expression is regulated at multiple levels, with post-transcriptional levels of control having impor- tant roles in the cellular response to environmental factors (BOX 1). Below, we will focus on transcriptional mechanisms that allow cells to rapidly switch between growth- and stress-related gene expression programmes. Figure 1 | Balancing the expression of growth- and stress-related genes. The simplified scheme shows the interplay between two major signalling pathways that antagonistically regulate the expression of growth- and stress-related genes. The target of rapamycin (TOR) pathway is upregulated in response to nutrients and growth factors, whereas the stress-activated protein kinase (SAPK) pathway is upregulated in response to stress 1,2,4 . TOR promotes growth-related genes and inhibits stress- related genes, whereas SAPK promotes stress-related genes and inhibits growth-related genes. In addition, it is possible that the TOR and SAPK pathways can also negatively regulate each other. Growth-related genes (green) are characterized by robust expression levels and the absence of TATA boxes in their promoters, whereas stress-related genes (red) are characterized by noisy transcription and the presence of TATA boxes in their promoters. Environmental conditions influence the balance between TOR and SAPK signalling, which in turn determines the balance between the expression of growth- and stress-related genes, and thus the cellular phenotype with regard to growth rate and stress resistance. R E V I E W S NATURE REVIEWS | GENETICS VOLUME 9 | AUGUST 2008 | 585 TFIID A complex that is composed of TBP and TAFs. Binding of TFIID to DNA is necessary but not sufficient for transcription initiation from most promoters. SAGA A large multi-protein complex involved in the regulation of transcription that possesses histone acetyltransferase and TBP-binding activities. The budding-yeast complex includes Gcn5, several proteins of the Spt and Ada families, and several TAFs; analogous complexes in other species have analogous compositions, and usually contain homologues of the yeast proteins. The TATA box and stress-related genes. Most eukaryo- tic protein-coding genes are transcribed by RNA polymerase II (pol II). Accurate initiation of tran- scription requires a core promoter region, as well as flanking regulatory sequences. Accessibility to pro- moter sequences is regulated by chromatin structure, and several steps are required to initiate transcription. First, sequence-specific DNA-binding proteins coordi- nate the remodelling of chromatin and the recruitment of the transcription machinery. Second, the general transcription factors and pol II assemble into a pre- initiation complex. Third, pol II and its associated elon- gation factors produce an mRNA transcript. The TATA box is a core promoter motif, present in approximately 20% of all S. cerevisiae genes 31 , that is involved in the assembly of the transcriptional machinery. Interestingly, TATA-containing genes are enriched in stress-related genes, and they are extensively regulated compared with TATA-less genes in S. cerevisiae 31 . These data sug- gest that TATA boxes are associated with promoters of genes that require rapid and variable regulation, which seems to be true also in humans and plants 32,33 . The TATA box not only promotes short-term regulatory tuning but also noisy transcription (BOX 2) and the evolu- tion of gene expression control (see below). Conversely, TATA-less genes are enriched among ?housekeep- ing? or growth-related genes 31 . These findings reveal a bipolar transcriptome with distinct types of core pro- moters being used to control growth- or stress-related genes (FIG. 1). Distinct regulatory strategies for stress- and growth- related genes. Switching between growth and stress programmes requires the regulation of numerous genes, and cells use two separate mechanisms involving differential use of the core transcriptional machinery and TATA boxes to achieve this task. Recent studies in S. cerevisiae contribute to our understanding of such a global mode of regulation. An early step in transcription is the recruitment of the TATA-binding protein (TBP) to core promoters, which requires interaction with co- activator complexes such as TFIID and SAGA. Although TFIID and SAGA contribute to the expression of nearly all genes, the regulation of any given gene is usually dominated by one or the other factor. SAGA-dominated genes are enriched in stress-related functions and are highly regulated compared with TFIID-dominated genes 34 . Accordingly, TATA-containing genes show a stronger dependence on SAGA, whereas TATA-less genes are more dependent on TFIID 31 . An analysis of the dynamics of co-activator complexes during heat stress revealed that the TFIID complex disassembles from growth-related genes that are repressed, while compo- nents of the SAGA complex assemble at stress-related genes 35 . These results provide a mechanistic insight into a bipolar regulation of the transcriptome by showing that growth- and stress-related gene expression pro- grammes rely on distinct sets of co-activators. SAGA- dominated genes are more extensively regulated than TFIID-dominated genes. First, they make abundant use of negative regulators: Bdf1 seems to block the assembly Box 1 | Examples of post-transcriptional controls to balance growth and stress programmes RNA processing in P-bodies and stress granules Following stress exposure, yeast cells reduce the production of growth-related proteins while increasing the production of stress-related proteins. Besides transcript synthesis and degradation, the reversible storage of transcripts provides an economical and rapid mechanism to regulate protein levels in response to stress. Transcripts that are not, or are no longer, translated can join big cytoplasmic structures called P-bodies 91,92 . P-bodies contain, among many other proteins, the components of the RNA-decay machinery together with transcripts targeted for degradation. As well as being a site of RNA degradation, P-bodies are a site of RNA storage. As an example, this system is used by quiescent yeast cells or during growth in human cells, which assemble P-bodies containing transcripts that can then re-enter translation when environmental conditions change 93,94 . A different type of structure, called a stress granule, is present in plant and mammalian cells. Stress granules assemble in response to environmental stress and disperse after stress recovery. Unlike P-bodies, these structures contain stalled translation initiation complexes. Stress granules are closely associated with P-bodies and might be ?triage centres? that sort, remodel and export specific transcripts for translation, decay or storage 95 . These data highlight the complex cellular strategies used to distribute transcripts in different cellular structures that control the fate of these transcripts according to environmental conditions. Translational control of gene expression during stress Translation provides another important regulatory layer to tune protein levels in response to environmental factors. Changes in transcription are often potentiated by homodirectional changes in translation, although some genes go against this trend 96 . Stress triggers a general translational inhibition of most transcripts, including growth-related transcripts such as those involved in ribosome biosynthesis 19,97 . This general inhibition is mediated by general control nonderepressible 2 (GCN2)-like proteins that are activated during stress, leading to inhibitory phosphorylation of the eukaryotic translation initiation factor 2 (eIF2). Additional layers of regulation are mediated by the growth and stress signalling pathways: target of rapamycin (TOR) signalling stimulates translation by activating S6 kinase (S6K) 4 , and by inhibiting eIF4E-binding protein (4EBP) and, in Saccharomyces cerevisiae, Gcn2 (REF. 2), whereas stress-activated protein kinase (SAPK) signalling transiently inhibits translation in mammalian cells by phosphorylating the translational initiation factor eIF4E 98 . Specific transcripts that are important for the stress response, including yeast GCN4 or human activating transcription factor 4 (ATF4) transcripts, escape this general translational repression by relying on special regulatory sequences upstream of their coding regions 97 . R E V I E W S 586 | AUGUST 2008 | VOLUME 9 www.nature.com/reviews/genetics Nature Reviews | Genetics Le v e l of r e p o r t e r p r o t e i n 1 ( a l l e l e A ) Level of reporter protein 2 (allele B) Extrinsic noise Intrinsic noise Single cell of SAGA on stress-related genes in the absence of stress, whereas Mot1 is recruited with the active SAGA com- plex and might control the transient induction that is typical of stress-induced genes in S. cerevisiae 35 (see previous section). Second, SAGA-dominated genes are regulated by the coordinate action of several chromatin regulators (FIG. 2). The S. cerevisiae heat-shock response provides an additional example for the use of the core transcrip- tional machinery to optimize adaptation to changing environments 36 . During temperature stress, cells redistribute the core transcriptional machinery to heat-response genes. Simultaneously, a large number of partial pre-initiation complexes are assembled at a different set of genes. These partial complexes contain all basal transcription factors except for pol II and TFIIH. Interestingly, some of these complexes are converted into full pre-initiation complexes following exposure to oxidative stress 36 . These findings could help to explain cross-protection 15 . It is not known whether this global mechanism is conserved for the stress responses in other organisms. However, in the next section we will see that principles similar to those described above are applied by multicellular eukaryotes during development and cell differentiation. Related regulatory strategies for stress response and differentiation. Metazoan cells that differentiate into a mature tissue downregulate many genes that are involved in normal proliferation while inducing tissue- specific transcripts. This process is reminiscent of yeast cells deciding between growth programmes and specialized stress programmes, including quiescence, filamentation or sporulation. A recent study describes a global mechanism whereby muscle cells manage to downregulate their growth programme during dif- ferentiation into mature myotubes 37 . In these cells, differentiation is accompanied by a dramatic down- regulation at the protein level of TBP and associated factors, suggesting a loss of the TFIID complex. This process could provide an effective mechanism for a glo- bal repression of growth-related genes, as their func- tions are needed only in undifferentiated cells. This loss of core factors is selective, leaving proteins such as TBP- associated factor 3 (TAF3) and TBP-related factor 3 (TRF3), a vertebrate-specific TBP homologue, unaf- fected. These proteins interact with each other and form an alternative core promoter-recognition com- plex that binds the myogenin promoter and activates transcription. The authors propose a model in which core promoter switching mediates the global down- regulation of TFIID-dependent genes while allowing for selective activation of genes that are required for myogenesis, which depend only on the TRF3?TAF3 complex 37 . Interestingly, TRF3 is essential for embryo- genesis in Xenopus laevis and in zebrafish 38,39 , and it controls the expression of a master regulator for zebrafish haematopoiesis 40 . These findings suggest that the use of alternative core promoter-binding com- plexes might not be restricted to mouse myogenesis. Together, these data highlight how yeast and verte- brates exploit similar strategies, involving differential use of the core transcriptional machinery, to achieve massive reprogramming of their transcriptomes in response to external stimuli. An important component of the core transcrip- tional machinery is pol II, and its recruitment to the promoter is often the limiting step for transcription. However, pol II sometimes stalls at the start of genes ? this was first described for heat-shock genes in flies and later for several viral and mammalian genes 41 . This phenomenon is now widely documented also at genome-wide levels, showing that pol II stalling is widely used to regulate genes during stress response and development 42?45 . In quiescent S. cerevisiae cells, Box 2 | The TATA box and gene expression noise Whereas global studies uncovered the association of TATA boxes with stress- related genes, single- cell studies provided a mechanistic insight into this association. Such studies are based on the measurement of variation (that is, noise) in expression levels of proteins in single cells: the expression of two different reporter genes are put under the control of two identical genomic loci in diploid cells, followed by detection of expression levels of the reporters in single cells using flow cytometry or other methods. The within-cell and between-cell variability in protein expression is then analysed, leading to estimates of intrinsic and extrinsic noise, respectively 99 (see figure). Intrinsic noise is more formally defined as the variation in expression of identical proteins in the same cell (in the figure, this is seen where data points for individual cells do not lie on the diagonal, solid arrow), whereas extrinsic noise is the variation in expression of identical proteins owing to differences between cells (in the figure, this is represented by the total range of expression levels along the diagonal, dashed arrow) 100 . Unlike extrinsic noise, intrinsic noise is promoter specific 67 . Stress-related proteins that are expressed from genes with a TATA box are characterized by high levels of intrinsic noise, whereas growth-related proteins exhibit low noise 65 . This finding suggests that the growth- and stress-response programmes are characterized by different levels of intrinsic noise. Mutations that are introduced into the sequence of the TATA box decrease the intrinsic noise levels 66,67 . This behaviour could be modelled by increasing the dissociation rate of TATA-binding protein (TBP) from promoters 66 . Low TBP-dissociation rates favour the formation of a stable transcriptional scaffold, enabling repeated recruitment of RNA polymerase II ? a phenomenon known as re-initiation 101 ? culminating in high noise levels and sustained transcriptional ?bursts?. Conversely, high TBP-dissociation rates destabilize the transcriptional scaffold, resulting in shorter bursts and lower noise levels 66 . Chromatin regulation has a role in these differential noise patterns 102,103 . Interestingly, sustained transcriptional bursts can confer a fitness benefit in cases in which the expressed gene is necessary to combat environmental stress 66 . These data provide a mechanistic model of how the TATA box can confer a specific advantage to stress-related genes by increasing transcriptional noise and thus increasing expression variability. R E V I E W S NATURE REVIEWS | GENETICS VOLUME 9 | AUGUST 2008 | 587 Nature Reviews | Genetics SAGA-dominated genes ? ~10% of genome ? Stress response TFIID-dominated genes ? ~90% of genome ? Housekeeping Rpd3 Mot1 Bur6 Srb10 Ssn6 Gene-specific repressors Stress-induced activators Activators Hda1 Tup1 TATA TATA-less promoter H3 Ac H4 Ac Bdf1 Spt3 TBP TBP SAGA TFIID Gcn5 H4 Ac Low stress High stress For example, Msn2, 4 and Yap1 TAFs Bdf1 for example, inactive pol II is bound upstream of hundreds of genes that are rapidly induced following exit from the quiescent state 44 . Thus, yeast seems to prepare for a rapid response by having pol II ?ready to go? on the relevant genes. Alternatively, or in addition, DNA binding of minimal amounts of pol II, which is required to exit quiescence, could pro- tect pol II from degradation. Pol II stalling is also widespread during fly development, in which pol II is stalled on genes that are involved in development and in the response to stimuli 43,45 . A similar picture emerges in human stem cells, in which pol II is stalled on genes encoding developmental regulators 42 . These findings suggest that strategies applied by yeast to respond to changing environments are also applied by metazoans to regulate stress responses or cell differentiation. Interplay between stress and evolution This section deals with stress-response strategies from an evolutionary perspective. We will explore how gene expression adapts to changing environments, and how changing environments themselves provide a driving force for evolution by triggering variability in gene expression. The resulting plasticity and stochasticity of transcriptional responses in turn are instrumental in the adaptation to unforeseen challenges not previously encountered during evolutionary history. Long-term adaptation of gene expression. Genetic varia- tion in regulatory DNA elements provides an important basis for phenotypic variation from yeast to man, lead- ing to heritable regulatory changes 46?51 . Transcription- factor binding sites show remarkable plasticity and rapid divergence 52?55 , promoting evolvability of regulatory processes while maintaining functional robustness 56,57 . Microorganisms have been widely used to study long- term adaptation of gene expression to environmental challenges, because they can be grown under tightly regulated conditions for hundreds or even thousands of generations 58 . For example, Escherichia coli cells were grown under glucose limitation 59 or temperature stress 60 to analyse changes in gene expression over extended times. In both cases, the regulation of a few dozen genes was reproducibly modulated in repeated experiments, and this adaptation of gene expression led to increased fit- ness relative to the ancestor cells. Another study showed that protein expression from the E. coli lac operon is precisely optimized during evolution, reflecting the costs and benefits for different environments 61 . Experiments in S. cerevisiae grown under nutrient-limiting conditions also revealed characteristic gene expression patterns 62 as a long-term adaptation to the selective pressure. These studies demonstrate that long-term environmental changes lead to persistent changes in gene expression, which can occur within a short evolutionary time to sup- port cellular adaptation to the environmental challenge. Figure 2 | Stress- and growth-related genes show distinct regulatory features. In budding yeast, stress-related genes tend to be extensively regulated, their control is dominated by the co-activator complex SAGA and their promoters often contain TATA boxes, whereas growth-related and housekeeping genes tend to have TATA-less promoters and their control is dominated by the co-activator complex TFIID. Factors depicted in red are negative regulators, whereas those in green are positive regulators. Stress-related genes tend to be positively regulated by the SAGA complex and by acetylation (Ac) of histones H3 and H4, and negatively regulated by histone de-acetylases (Hda1 and Rpd3), repressors and repressor complexes (Ssn6-Tup1 complex, Mot1, Bur6 and Bdf1), and a member of the mediator complex (Srb10). Growth-related genes, on the other hand, are positively regulated by histone H4 acetylation and Bdf1. The dashed arrows show that SAGA also contributes to the expression of genes targeted by TFIID, and vice versa. Gcn5, general control nonderepressible 5; TAF, TBP-associated factor; TBP, TATA-binding protein. This figure is modified, with permission, from REF. 34 ? (2004) Cell Press. R E V I E W S 588 | AUGUST 2008 | VOLUME 9 www.nature.com/reviews/genetics Stress promotes regulatory variation and evolvability of gene expression. Changes in external conditions trigger adaptive variation in intracellular regulation, but an excess of unpredictable regulatory variation can interfere with the robustness of cellular functions. The interplay between variability and robustness ? that is, between promoting and buffering intracellular changes ? is fundamental for evolution 48,57,63,64 . As dis- cussed above, stress-related genes tend to have more noisy expression characteristics than growth-related genes, and noisy expression is associated with TATA boxes (BOX 2). Several recent studies indicate that noise levels are tuned by evolution to balance fidelity and variation of gene expression; although too much noise with respect to the regulation of processes such as cell proliferation or development could be detrimental for fitness, variable regulation of environmental responses could actually increase the chance of survival dur- ing stress 63,65?67 . Stress-related genes are also more volatile with respect to gene duplications and losses than growth-related genes in yeasts 68 , and duplicated stress-related genes show higher expression divergence than developmental genes in Arabidopsis thaliana 69 . Between-cell variation in the expression of stress- related genes might be beneficial by enabling popula- tions to sample multiple phenotypes, thus increasing the chance that some cells survive adverse conditions. This strategy should be particularly effective for dealing with unpredictable and rapid changes in the environ- ment. For example, phenotypic heterogeneity within clonal bacterial populations promotes stress tolerance: E. coli can switch stochastically into a ?persistent? state, characterized by slow growth and increased stress tol- erance, which allows it to survive antibiotic treatment 70 . A recent study in S. cerevisiae suggests that cells might tune the rate of stochastic switching between differ- ent phenotypes to the frequency of environmental fluctuations in order to optimize survival 71 . Variable gene expression that is triggered as an adap- tation to stress in turn increases the evolvability of gene expression. Long-term evolution of gene expression changes between yeast species is correlated with short- term regulatory changes to environmental stress: genes that show high environmental responsiveness within one species also tend to show high expression divergence between species 72 . Accordingly, stress- related genes that are associated with TATA boxes show exceptionally rapid regulatory evolution, which also is true in worms, flies, plants and mammals 72 . These findings raise the possibility that the regulatory characteristics of TATA-box promoters encourage the evolvability of gene expression. Indeed, the presence of a TATA box leads to increased sensitivity of gene expression to both mutations and environmental stress 73 , and TATA-containing genes tend to be regu- lated by more transcription factors than TATA-less genes 72,73 . TATA-containing genes are also highly variable between natural isolates 46 and experimentally evolved S. cerevisiae strains 31 . Together, these data not only highlight the importance of TATA-box genes for both short- and long-term regulatory adaptation, but also suggest mechanisms for their increased regulatory evolvability. Besides promoting noisy and divergent gene expres- sion, stress can trigger phenotypic variation and can speed up evolution by a range of additional processes (BOX 3). These two findings indicate that changing environments and stressful conditions keep organisms ?on their toes?, and stress not only promotes short-term adaptations but seems to be a major driving force for evolutionary innovation. Unspecific components of transcriptional responses to stress. As described above, environmental changes activate signal-transduction pathways to trigger exten- sive cellular responses in gene expression, including hundreds of genes that are either up- or downregu- lated. This raises questions about specificity 74 : do these global responses reflect adaptations to environmental challenges or do they include large stochastic com- ponents? Many of the genes that are regulated dur- ing stress do not seem to have any direct functional relevance to the specific perturbation. Accordingly, functional profiling in S. cerevisiae indicates an over- all poor correlation between the stress sensitivity of mutants and the regulation of corresponding genes in the same stress 75?77 . Similarly, out of 16,000 A. thaliana genes that are regulated during temperature stress, only 43 genes seem to have substantial adaptive value 78 . These findings suggest that a large proportion of the gene expression response to a specific stress is not adaptive for this stress. This apparent lack of specifi- city could simply be an indication that a large propor- tion of the regulated genes are not specific for any given stress but form part of a core stress response as described above. Indeed, large stereotypical regulation of the growth- and stress-related modules dominates the gene expression responses to a wide range of chemical or genetic perturbations in S. cerevisiae 14,79 . A general, unspecific stress response has the advantage that it can cross protect against multiple environmental conditions, which might frequently occur together and Box 3 | Additional examples of interplay between stress and evolution The examples given in this box illustrate the versatile interplay between stress and evolution. Nutrient stress in Saccharomyces cerevisiae triggers a large gene expression programme controlling sexual differentiation 22 , which in turn might accelerate adaptation to harsh environments by increasing genetic variation 104 . Transposition is another mechanism to create potentially useful genetic variation 105,106 ; as an example, oxidative stress in Schizosaccharomyces pombe switches Tf2 transposons from a silenced to an activated state 6,107 . Stress is also known to induce genomic rearrangements in plants 106 . A striking study in Arabidopsis thaliana indicates that the capacity for genetic change after stress treatment persists over successive generations, even in the absence of stress; this memory effect, which is based on an epigenetic mechanism, could increase the potential for adaptation 108 . Intriguingly, and somewhat controversially, several studies in microorganisms suggest that stress can directly induce mutations that in turn could accelerate evolution 109 . Moreover, environmental stress can also affect evolution by aggravating or alleviating the deleterious effects of mutations 110 and, in the long term, by increasing the robustness to the effect of mutations 57,111 , as well as by unmasking silent genetic variation when the buffering capacity of the heat-shock protein 90 chaperone is compromised 112 . R E V I E W S NATURE REVIEWS | GENETICS VOLUME 9 | AUGUST 2008 | 589 in combination with stresses that the cell cannot sense. It is also possible, however, that a portion of the stress- response programme reflects neutral evolutionary drift or large-scale connectivity and dynamic compensatory adjustment of the transcriptional regulatory network. Gene expression tuning to unknown challenges. Some environmental changes, such as those caused by day?night cycles, are predictable, and organisms have evolved specialized periodic gene expression programmes to prepare for these expected fluctua- tions; these circadian-clock or rhythm programmes are maintained for some time even in the absence of the environmental cues (BOX 4). Unpredictable but known challenges, on the other hand, trigger large gene expression responses as and when required. We have seen above that these gene expression responses show limited specificity for any particular stress. This is perhaps not surprising given that the number of potential environmental conditions encountered during the lifetime of an organism and, even more so, during evolutionary history can be expected to greatly exceed the repertoire of signalling pathways and other gene regulatory processes. In this respect, it is particularly insightful to study how gene expression responds to unknown perturbations that cells could not have encountered during evolution, thus elimi- nating the possibility of ?pre-selected? gene expression programmes. Recent studies are starting to address this intriguing issue. In an elegant approach in S. cerevisiae, the essen- tial HIS3 gene was placed under ectopic GAL control so that changing to a glucose-containing medium led to repression of HIS3 (FIG. 3a), resulting in a severe chal- lenge to which the cells adapted over several genera- tions 80,81 . Although global gene expression responses were evident during adaptation to the unforeseen challenge, a large proportion of these responses was not reproducible in repeated biological experiments. This finding is surprising given that the initial changes in gene expression apparently did not reflect a popula- tion selection process but transcriptional reprogram- ming of most cells 81 ; the differences in gene expression responses might therefore reflect physiological differ- ences in the repeated experiments. The transient gene expression responses are not explained by metabolic logic, as would be expected for responses that had not been specifically selected for during evolution. Thus, it seems that unknown challenges can trigger stochastic, unspecific and transient gene expression responses on the basis of regulatory network plastic- ity, followed by selection for adaptive portions of the response 74,80 . A study on bacterial adaptation to different media led to similar conclusions: although independent E. coli cultures showed the same growth and metabolic phenotypes after many generations, the underly- ing gene expression states differed, including a large number of compensatory expression changes and only a few adaptive changes 82 . However, a related study of long-term adaptation to glucose limitation reported highly similar gene expression responses in parallel E. coli cultures 83 . This discrepancy could indicate that variation in glucose concentration is a familiar situation during bacterial evolution, which the cells could have adapted to by using a pre-selected response. Two papers provide insight into a possible mecha- nism for the tuning of adaptive gene expression in the absence of any hard-wired and pre-selected regulatory responses. Using a synthetic bistable gene switch with mutually inhibitory operons directing the expression of genes that are required for alternative nutrients (FIG. 3b), E. coli cells reliably selected the adaptive state among two stable attractor states without the help of signal transduction 84 . The authors propose that this selection indicates the presence of stochastic gene- network dynamics: in the non-adaptive state, low cellular activity leads to high gene expression noise, making this state less stable than the adaptive state. This prediction was tested by mathematical model- ling of the bistable gene switch 84 . A recent theoretical study expanded the model to a more general situation in which adaptive states with optimal growth rates are spontaneously selected, this is because cells are easily ?kicked out? of slow-growth states owing to higher gene expression noise in these states 85 . Gene expression changes by attractor selection are less efficient and slower than changes by signal transduction, but this simple and robust principle could be an ancient mech- anism for gene expression adaptation that precedes the evolution of dedicated regulatory mechanisms. Moreover, it seems likely that attractor selection still contributes to cellular robustness and to the survival of rare or unknown challenges. Box 4 | Anticipating environmental changes: circadian clocks The rotation of the Earth produces one of the most influential environmental fluctuation organisms have to face. Day?night oscillations not only influence light availability but are often also accompanied by fluctuations in temperature, nutrients or other factors. These constant and predictable fluctuations led to the widespread evolution of circadian clocks (from the Latin circa diem, meaning ?about a day?), which control daily rhythms in gene expression and other regulatory aspects, thus permitting organisms to anticipate and prepare for predictable environmental changes 113,114 . Circadian mechanisms are not homologous and seem to have evolved independently several times in bacteria, fungi, plants and animals. Cyanobacteria and animals with altered circadian rhythms or arrhythmic mutations are outcompeted in changing environments 114 . Circadian clocks are driven by oscillators (also called pacemakers), which are regulated by transcriptional and post-transcriptional feedback loops 115 ? these loops continue to fluctuate even in the absence of the environmental cues. For fine-tuning, circadian clocks are set to the actual environmental time cues (or ?zeitgebers?). Microorganisms, plants and flies have circadian pacemakers that can be reset by environmental cues, whereas in mammals and birds only the pacemaker in the central nervous system can be entrained by environmental cues, and they then synchronize peripheral oscillators by secreting uncharacterized molecules 113 . The output from circadian clocks is the periodic transcription of clock-controlled genes. The extent of circadian transcriptional control varies from organism to organism, and also between different tissues, ranging from the entire genome in cyanobacteria to ~10% of all genes in mammals. In all cases, transcriptional oscillations then influence key physiological aspects that are defined by the specific needs of organisms, including metabolism, photosynthesis, fungal spore formation and/or germination, or mammalian behaviour 113,114 . R E V I E W S 590 | AUGUST 2008 | VOLUME 9 www.nature.com/reviews/genetics Nature Reviews | Genetics a Wild type b Reprogrammed GAL HIS3HIS3 promoter GAL HIS3 HIS3 GAL promoter GAL promoter GAL promoter HIS3 promoter Operon 2 Operon 1 Transcription induced State 1: Operon 1 expressed Transcription repressed State 2: Operon 2 expressed Conclusions ?Adversity has the effect of eliciting talents, which in prosperous circumstances would have lain dormant?. Horace?s quote very much applies to cells exposed to stress, which use impressive strategies to ensure that global gene expression meets environmental challenges. To stay competitive, yeasts and other free-living microor- ganisms balance maximal cell growth with stress protec- tion, depending on a multitude of environmental factors. This balance requires the interplay between regulatory mechanisms controlling growth- and stress-related gene expression programmes. Variable gene expression, either driven by hard-wired signalling pathways or by stochas- tic transcriptional plasticity, helps to cope with variable environments in the short term but might also trigger longer-term adaptations during evolution by enhanc- ing phenotypic variability and robustness. Consistent with this idea, intriguing parallels exist between cel- lular stress-response and differentiation programmes: different stresses trigger yeast cells to differentiate into different stress-resistant states; cellular stress responses and differentiation are both regulated by conserved signalling pathways in all eukaryotes; and the differential use of the core transcriptional machinery for gene expression reprogramming during metazoan differ- entiation is reminiscent of the transcriptional mechanisms used during the yeast stress response. These parallels raise the possibility that stress responses are primordial processes for the evolution of cellular differentiation. Many of these concepts have first been developed in yeast and other simple model organisms, but the available studies in multicellular organisms indicate that they reflect fundamental principles with broad implications and rel- evance for more complex processes. The recent findings reviewed here therefore provide a valuable basis to better understand gene expression programmes in human cells, including the varied reprogramming of gene expression accompanying ageing, cancer and other diseases. Compared with transcriptional mechanisms, post- transcriptional mechanisms for gene expression tuning are relatively poorly understood, although they clearly have important roles 86 . Issues that are yet to be resolved are the relative contribution of post-transcriptional con- trol in tuning gene expression to environmental change, and how transcriptional and post-transcriptional layers of control are integrated for coordinated cellular responses. Finding the answers will require multi-dimensional approaches to sample different regulatory levels 87 . Future work will deepen both our understanding of the impact that variable environments have on evolution and our insight into the contribution of gene expres- sion to short- and long-term adaptation to environ- mental factors, for which cells need to juggle regulatory plasticity with robust transcriptional responses. The findings reviewed here provide the foundation for a systems-level understanding of gene expression tun- ing to changing environments. This ambitious goal will require the integration of diverse and comprehensive data sets, and the application of interdisciplinary approaches to model regulatory networks that are triggered by environ- mental factors 88,89 and, eventually, to predict quantitative cellular responses to different environments 90 . Such deep understanding will ultimately enable us to re-engineer regulatory circuits and design the cellular responses we want, thus having gene expression play to our tunes. Figure 3 | Synthetic regulatory circuits to study responses to unknown challenges. a | Schematic representation of relevant expression profiles in the reprogrammed yeast strain that was used to study novel challenges 80 . Wild-type cells express HIS3 under its normal promoter, and expression is independent of glucose but activated by the absence of histidine. This simple circuit has been re-wired in the reprogrammed yeast cells, which express HIS3 under the control of the GAL promoter, leading to HIS3 repression in the presence of glucose. Thus, when the reprogrammed cells are grown in the presence of glucose and absence of histidine, they encounter a serious challenge, because HIS3 expression is downregulated under conditions in which its expression is essential. b | An illustration of a simple genetic switch, which was used in Escherichia coli to study attractor selection, consisting of two operons that mutually inhibit each other 84 . Operon 1 codes for a transcriptional repressor of operon 2, whereas operon 2 codes for a transcriptional repressor of operon 1. If operon 1 is expressed above a threshold, it will repress the expression of operon 2 and vice versa. High expression of operon 1 and 2 are mutually exclusive, thus defining two stable states. Both operons also encode condition-specific selectable markers and different fluorescent proteins for gene expression monitoring at the single-cell level, enabling the study of adaptation to different cellular states in the absence of dedicated signalling pathways. 1. Bahn, Y. S. et al. Sensing the environment: lessons from fungi. Nature Rev. Microbiol. 5, 57?69 (2007). 2. De Virgilio, C. & Loewith, R. Cell growth control: little eukaryotes make big contributions. Oncogene 25, 6392?6415 (2006). 3. Roux, P. P. & Blenis, J. ERK and p38 MAPK-activated protein kinases: a family of protein kinases with diverse biological functions. Microbiol. Mol. Biol. Rev. 68, 320?344 (2004). 4. Wullschleger, S., Loewith, R. & Hall, M. N. TOR signaling in growth and metabolism. Cell 124, 471?484 (2006). 5. Causton, H. C. et al. Remodeling of yeast genome expression in response to environmental changes. Mol. Biol. Cell 12, 323?337 (2001). 6. Chen, D. et al. Global transcriptional responses of fission yeast to environmental stress. Mol. Biol. Cell 14, 214?229 (2003). 7. Gasch, A. P. et al. Genomic expression programs in the response of yeast cells to environmental changes. Mol. Biol. Cell 11, 4241?4257 (2000). This pioneering paper presents a thorough global analysis of gene expression programmes in multiple stress conditions, and defines a core stress response in S. cerevisiae. 8. Enjalbert, B. et al. Role of the Hog1 stress-activated protein kinase in the global transcriptional response to stress in the fungal pathogen Candida albicans. Mol. Biol. Cell 17, 1018?1032 (2006). 9. Ma, S. & Bohnert, H. J. Integration of Arabidopsis thaliana stress-related transcript profiles, promoter structures, and cell-specific expression. Genome Biol. 8, R49 (2007). 10. Girardot, F., Monnier, V. & Tricoire, H. Genome wide analysis of common and specific stress responses in adult Drosophila melanogaster. BMC Genomics 5, 74 (2004). R E V I E W S NATURE REVIEWS | GENETICS VOLUME 9 | AUGUST 2008 | 591 11. Sorensen, J. G., Nielsen, M. M. & Loeschcke, V. Gene expression profile analysis of Drosophila melanogaster selected for resistance to environmental stressors. J. Evol. Biol. 20, 1624?1636 (2007). 12. Murray, J. I. et al. Diverse and specific gene expression responses to stresses in cultured human cells. Mol. Biol. Cell 15, 2361?2374 (2004). 13. Chen, D. et al. Multiple pathways differentially regulate global oxidative stress responses in fission yeast. Mol. Biol. Cell 19, 308?317 (2008). 14. Hughes, T. R. et al. Functional discovery via a compendium of expression profiles. Cell 102, 109?126 (2000). 15. Kultz, D. Molecular and evolutionary basis of the cellular stress response. Annu. Rev. Physiol. 67, 225?257 (2005). 16. Aragon, A. D., Quinones, G. A., Thomas, E. V., Roy, S. & Werner-Washburne, M. Release of extraction-resistant mRNA in stationary phase Saccharomyces cerevisiae produces a massive increase in transcript abundance in response to stress. Genome Biol. 7, R9 (2006). 17. Dinneny, J. R. et al. Cell identity mediates the response of Arabidopsis roots to abiotic stress. Science 320, 942?945 (2008). 18. Jorgensen, P., Tyers, M. & Warner, J. R. in Cell growth (eds Hall, M. N., Raff, M. & Thomas, G.) 329?370 (Cold Spring Harbor Laboratory Press, New York, 2004). 19. Warner, J. R. The economics of ribosome biosynthesis in yeast. Trends Biochem. Sci. 24, 437?440 (1999). 20. Gray, J. V. et al. ?Sleeping beauty?: quiescence in Saccharomyces cerevisiae. Microbiol. Mol. Biol. Rev. 68, 187?206 (2004). 21. Truckses, D. M., Garrenton, L. S. & Thorner, J. Jekyll and Hyde in the microbial world. Science 306, 1509?1511 (2004). 22. Chu, S. et al. The transcriptional program of sporulation in budding yeast. Science 282, 699?705 (1998). 23. Elliott, B. & Futcher, B. Stress resistance of yeast cells is largely independent of cell cycle phase. Yeast 9, 33?42 (1993). 24. Bishop, N. A. & Guarente, L. Genetic links between diet and lifespan: shared mechanisms from yeast to humans. Nature Rev. Genet. 8, 835?844 (2007). 25. Aguirre-Ghiso, J. A. Models, mechanisms and clinical evidence for cancer dormancy. Nature Rev. Cancer 7, 834?846 (2007). 26. Castrillo, J. I. et al. Growth control of the eukaryote cell: a systems biology study in yeast. J. Biol. 6, 4 (2007). This paper, together with references 27 and 28, analyzes the relationships between growth rate and global gene expression in S. cerevisiae. 27. Brauer, M. J. et al. Coordination of growth rate, cell cycle, stress response, and metabolic activity in yeast. Mol. Biol. Cell 19, 352?367 (2008). 28. Regenberg, B. et al. Growth-rate regulated genes have profound impact on interpretation of transcriptome profiling in Saccharomyces cerevisiae. Genome Biol. 7, R107 (2006). 29. Levy, S. et al. Strategy of transcription regulation in the budding yeast. PLoS ONE 2, e250 (2007). This paper asks how transcriptional responses to changing environments are coordinated with the actual physiological needs of the cell. Experimental conditions that decouple actual from environmentally expected growth rates reveal that the transcriptional response is determined by environmental rather than by intrinsic conditions. 30. Kaeberlein, M. et al. Regulation of yeast replicative life span by TOR and Sch9 in response to nutrients. Science 310, 1193?1196 (2005). 31. Basehoar, A. D., Zanton, S. J. & Pugh, B. F. Identification and distinct regulation of yeast TATA box-containing genes. Cell 116, 699?709 (2004). This paper identifies the TATA box consensus sequence and provides a list of the TATA-containing genes in S. cerevisiae. The authors show that TATA- containing and TATA-less promoters are differentially regulated and associated with stress- response or housekeeping functions, respectively. 32. Walther, D., Brunnemann, R. & Selbig, J. The regulatory code for transcriptional response diversity and its relation to genome structural properties in A. thaliana. PLoS Genet. 3, e11 (2007). 33. Yang, C., Bolotin, E., Jiang, T., Sladek, F. M. & Martinez, E. Prevalence of the initiator over the TATA box in human and yeast genes and identification of DNA motifs enriched in human TATA-less core promoters. Gene 389, 52?65 (2007). 34. Huisinga, K. L. & Pugh, B. F. A genome-wide housekeeping role for TFIID and a highly regulated stress-related role for SAGA in Saccharomyces cerevisiae. Mol. Cell 13, 573?585 (2004). 35. Zanton, S. J. & Pugh, B. F. Changes in genomewide occupancy of core transcriptional regulators during heat stress. Proc. Natl Acad. Sci. USA 101, 16843?16848 (2004). This paper shows that heat stress causes disassembly of the TFIID pathway at genes that are inhibited by stress, and assembly of the SAGA pathway genes that are induced by stress, and it studies the regulation of these two pathways in response to heat stress. 36. Zanton, S. J. & Pugh, B. F. Full and partial genome-wide assembly and disassembly of the yeast transcription machinery in response to heat shock. Genes Dev. 20, 2250?2265 (2006). 37. Deato, M. D. & Tjian, R. Switching of the core transcription machinery during myogenesis. Genes Dev. 21, 2137?2149 (2007). This paper reports that differentiation of myoblasts to myotubes involves the disruption of the canonical TFIID complex and its replacement by a novel TRF3?TAF3 complex. It proposes core promoter-complex switching as an effective process to selectively turn on one transcriptional programme while silencing others. 38. Bartfai, R. et al. TBP2, a vertebrate-specific member of the TBP family, is required in embryonic development of zebrafish. Curr. Biol. 14, 593?598 (2004). 39. Jallow, Z., Jacobi, U. G., Weeks, D. L., Dawid, I. B. & Veenstra, G. J. Specialized and redundant roles of TBP and a vertebrate-specific TBP paralog in embryonic gene regulation in Xenopus. Proc. Natl Acad. Sci. USA 101, 13525?13530 (2004). 40. Hart, D. O., Raha, T., Lawson, N. D. & Green, M. R. Initiation of zebrafish haematopoiesis by the TATA-box-binding protein-related factor Trf3. Nature 450, 1082?1085 (2007). 41. Saunders, A., Core, L. J. & Lis, J. T. Breaking barriers to transcription elongation. Nature Rev. Mol. Cell Biol. 7, 557?567 (2006). 42. Guenther, M. G., Levine, S. S., Boyer, L. A., Jaenisch, R. & Young, R. A. A chromatin landmark and transcription initiation at most promoters in human cells. Cell 130, 77?88 (2007). 43. Muse, G. W. et al. RNA polymerase is poised for activation across the genome. Nature Genet. 39, 1507?1511 (2007). 44. Radonjic, M. et al. Genome-wide analyses reveal RNA polymerase II located upstream of genes poised for rapid response upon S. cerevisiae stationary phase exit. Mol. Cell 18, 171?183 (2005). 45. Zeitlinger, J. et al. RNA polymerase stalling at developmental control genes in the Drosophila melanogaster embryo. Nature Genet. 39, 1512?1516 (2007). 46. Townsend, J. P., Cavalieri, D. & Hartl, D. L. Population genetic variation in genome-wide gene expression. Mol. Biol. Evol. 20, 955?963 (2003). 47. Fay, J. C., McCullough, H. L., Sniegowski, P. D. & Eisen, M. B. Population genetic variation in gene expression is associated with phenotypic variation in Saccharomyces cerevisiae. Genome Biol. 5, R26 (2004). 48. Gerhart, J. & Kirschner, M. The theory of facilitated variation. Proc. Natl Acad. Sci. USA 104 (Suppl 1), 8582?8589 (2007). 49. Wray, G. A. The evolutionary significance of cis- regulatory mutations. Nature Rev. Genet. 8, 206?216 (2007). 50. Rockman, M. V. & Kruglyak, L. Genetics of global gene expression. Nature Rev. Genet. 7, 862?872 (2006). 51. Stranger, B. E. et al. Population genomics of human gene expression. Nature Genet. 39, 1217?1224 (2007). 52. Borneman, A. R. et al. Divergence of transcription factor binding sites across related yeast species. Science 317, 815?819 (2007). 53. Odom, D. T. et al. Tissue-specific transcriptional regulation has diverged significantly between human and mouse. Nature Genet. 39, 730?732 (2007). 54. Tanay, A., Regev, A. & Shamir, R. Conservation and evolvability in regulatory networks: the evolution of ribosomal regulation in yeast. Proc. Natl Acad. Sci. USA 102, 7203?7208 (2005). 55. Tirosh, I., Weinberger, A., Bezalel, D., Kaganovich, M. & Barkai, N. On the relation between promoter divergence and gene expression evolution. Mol. Syst. Biol. 4, 159 (2008). 56. Carroll, S. B., Grenier, J. K. & Weatherbee, S. D. From DNA to Diversity (Blackwell Science, Malden, Massachusetts, 2001). 57. Wagner, A. Robustness and Evolvability in Living Systems (Princeton Univ. Press, Princeton, 2005). 58. Elena, S. F. & Lenski, R. E. Evolution experiments with microorganisms: the dynamics and genetic bases of adaptation. Nature Rev. Genet. 4, 457?469 (2003). 59. Cooper, T. F., Rozen, D. E. & Lenski, R. E. Parallel changes in gene expression after 20,000 generations of evolution in Escherichia coli. Proc. Natl Acad. Sci. USA 100, 1072?1077 (2003). 60. Riehle, M. M., Bennett, A. F., Lenski, R. E. & Long, A. D. Evolutionary changes in heat-inducible gene expression in lines of Escherichia coli adapted to high temperature. Physiol. Genomics 14, 47?58 (2003). 61. Dekel, E. & Alon, U. Optimality and evolutionary tuning of the expression level of a protein. Nature 436, 588?592 (2005). 62. Ferea, T. L., Botstein, D., Brown, P. O. & Rosenzweig, R. F. Systematic changes in gene expression patterns following adaptive evolution in yeast. Proc. Natl Acad. Sci. USA 96, 9721?9726 (1999). 63. Barkai, N. & Shilo, B. Z. Variability and robustness in biomolecular systems. Mol. Cell 28, 755?760 (2007). 64. Lenski, R. E., Barrick, J. E. & Ofria, C. Balancing robustness and evolvability. PLoS Biol. 4, e428 (2006). 65. Newman, J. R. et al. Single-cell proteomic analysis of S. cerevisiae reveals the architecture of biological noise. Nature 441, 840?846 (2006). 66. Blake, W. J. et al. Phenotypic consequences of promoter-mediated transcriptional noise. Mol. Cell 24, 853?865 (2006). This paper models and experimentally verifies the effects of TATA sequences on transcriptional ?bursts? and accompanying noise, which provide a benefit for the response to acute environmental stress. 67. Raser, J. M. & O?Shea, E. K. Control of stochasticity in eukaryotic gene expression. Science 304, 1811?1814 (2004). 68. Wapinski, I., Pfeffer, A., Friedman, N. & Regev, A. Natural history and evolutionary principles of gene duplication in fungi. Nature 449, 54?61 (2007). 69. Ha, M., Li, W. H. & Chen, Z. J. External factors accelerate expression divergence between duplicate genes. Trends Genet. 23, 162?166 (2007). 70. Balaban, N. Q., Merrin, J., Chait, R., Kowalik, L. & Leibler, S. Bacterial persistence as a phenotypic switch. Science 305, 1622?1625 (2004). 71. Acar, M., Mettetal, J. T. & van Oudenaarden, A. Stochastic switching as a survival strategy in fluctuating environments. Nature Genet. 40, 471?475 (2008). This study uses an engineered S. cerevisiae strain that stochastically switches between two phenotypes, revealing that survival is increased if the inter-phenotype switching rate is in tune with the frequency of environmental changes. 72. Tirosh, I., Weinberger, A., Carmi, M. & Barkai, N. A genetic signature of interspecies variations in gene expression. Nature Genet. 38, 830?834 (2006). This comparative genomic study reveals that TATA-containing genes show higher interspecies variability and evolutionary divergence in gene expression across a wide spectrum of eukaryotes. 73. Landry, C. R., Lemos, B., Rifkin, S. A., Dickinson, W. J. & Hartl, D. L. Genetic properties influencing the evolvability of gene expression. Science 317, 118?121 (2007). 74. Koonin, E. V. Chance and necessity in cellular response to challenge. Mol. Syst. Biol. 3, 107 (2007). 75. Giaever, G. et al. Functional profiling of the Saccharomyces cerevisiae genome. Nature 418, 387?391 (2002). 76. Thorpe, G. W., Fong, C. S., Alic, N., Higgins, V. J. & Dawes, I. W. Cells have distinct mechanisms to maintain protection against different reactive oxygen species: oxidative-stress-response genes. Proc. Natl Acad. Sci. USA 101, 6564?6569 (2004). 77. Warringer, J., Ericson, E., Fernandez, L., Nerman, O. & Blomberg, A. High-resolution yeast phenomics resolves different physiological features in the saline response. Proc. Natl Acad. Sci. USA 100, 15724?15729 (2003). 78. Swindell, W. R., Huebner, M. & Weber, A. P. Plastic and adaptive gene expression patterns associated with temperature stress in Arabidopsis thaliana. Heredity 99, 143?150 (2007). R E V I E W S 592 | AUGUST 2008 | VOLUME 9 www.nature.com/reviews/genetics 79. Tanay, A., Steinfeld, I., Kupiec, M. & Shamir, R. Integrative analysis of genome-wide experiments in the context of a large high-throughput data compendium. Mol. Syst. Biol. 1, 2005.0002 (2005). 80. Stern, S., Dror, T., Stolovicki, E., Brenner, N. & Braun, E. Genome-wide transcriptional plasticity underlies cellular adaptation to novel challenge. Mol. Syst. Biol. 3, 106 (2007). This paper analyzes the global transcriptional reprogramming underlying the adaptation to a novel challenge in S. cerevisiae. The findings suggest that a nonspecific transcriptional response reflecting the natural plasticity of the regulatory network can support cellular adaptation. 81. Stolovicki, E., Dror, T., Brenner, N. & Braun, E. Synthetic gene recruitment reveals adaptive reprogramming of gene regulation in yeast. Genetics 173, 75?85 (2006). 82. Fong, S. S., Joyce, A. R. & Palsson, B. O. Parallel adaptive evolution cultures of Escherichia coli lead to convergent growth phenotypes with different gene expression states. Genome Res. 15, 1365?1372 (2005). 83. Pelosi, L. et al. Parallel changes in global protein profiles during long-term experimental evolution in Escherichia coli. Genetics 173, 1851?1869 (2006). 84. Kashiwagi, A., Urabe, I., Kaneko, K. & Yomo, T. Adaptive response of a gene network to environmental changes by fitness-induced attractor selection. PLoS ONE 1, e49 (2006). This theoretical study predicts a general mechanism for selecting an adaptive state to different environments in the absence of any specific signalling pathways. The adaptive state shows a higher growth rate and less stochastic gene expression, which makes it inherently more stable than non-adaptive states. 85. Furusawa, C. & Kaneko, K. A generic mechanism for adaptive growth rate regulation. PLoS Comput. Biol. 4, e3 (2008). 86. Mata, J., Marguerat, S. & B�hler, J. Post-transcriptional control of gene expression: a genome-wide perspective. Trends Biochem. Sci. 30, 506?514 (2005). 87. Lackner, D. H. et al. A network of multiple regulatory layers shapes gene expression in fission yeast. Mol. Cell 26, 145?155 (2007). 88. Ishii, N. et al. Multiple high-throughput analyses monitor the response of E. coli to perturbations. Science 316, 593?597 (2007). 89. Luscombe, N. M. et al. Genomic analysis of regulatory network dynamics reveals large topological changes. Nature 431, 308?312 (2004). 90. Bonneau, R. et al. A predictive model for transcriptional control of physiology in a free living cell. Cell 131, 1354?1365 (2007). 91. Eulalio, A., Behm-Ansmant, I. & Izaurralde, E. P bodies: at the crossroads of post-transcriptional pathways. Nature Rev. Mol. Cell Biol. 8, 9?22 (2007). 92. Parker, R. & Sheth, U. P bodies and the control of mRNA translation and degradation. Mol. Cell 25, 635?646 (2007). 93. Bhattacharyya, S. N., Habermacher, R., Martine, U., Closs, E. I. & Filipowicz, W. Relief of microRNA- mediated translational repression in human cells subjected to stress. Cell 125, 1111?1124 (2006). 94. Brengues, M., Teixeira, D. & Parker, R. Movement of eukaryotic mRNAs between polysomes and cytoplasmic processing bodies. Science 310, 486?489 (2005). 95. Anderson, P. & Kedersha, N. RNA granules. J. Cell Biol. 172, 803?808 (2006). 96. Preiss, T., Baron-Benhamou, J., Ansorge, W. & Hentze, M. W. Homodirectional changes in transcriptome composition and mRNA translation induced by rapamycin and heat shock. Nature Struct. Biol. 10, 1039?1047 (2003). 97. Holcik, M. & Sonenberg, N. Translational control in stress and apoptosis. Nature Rev. Mol. Cell Biol. 6, 318?327 (2005). 98. Mathews, M. B., Sonenberg, N. & Hershey, J. W. B. (eds) Origins and Principles of Translational Control (Cold Spring Harbor Laboratory Press, New York, 2007). 99. Swain, P. S., Elowitz, M. B. & Siggia, E. D. Intrinsic and extrinsic contributions to stochasticity in gene expression. Proc. Natl Acad. Sci. USA 99, 12795?12800 (2002). 100. Maheshri, N. & O?Shea, E. K. Living with noisy genes: how cells function reliably with inherent variability in gene expression. Annu. Rev. Biophys. Biomol. Struct. 36, 413?434 (2007). 101. Dieci, G. & Sentenac, A. Detours and shortcuts to transcription reinitiation. Trends Biochem. Sci. 28, 202?209 (2003). 102. Choi, J. K. & Kim, Y. J. Epigenetic regulation and the variability of gene expression. Nature Genet. 40, 141?147 (2008). 103. Tirosh, I. & Barkai, N. Two strategies for gene regulation by promoter nucleosomes. Genome Res. 30 Apr 2008 (doi:10.1101/gr.076059.108). 104. Goddard, M. R., Godfray, H. C. & Burt, A. Sex increases the efficacy of natural selection in experimental yeast populations. Nature 434, 636?640 (2005). 105. Wessler, S. R. Turned on by stress. Plant retrotransposons. Curr. Biol. 6, 959?961 (1996). 106. McClintock, B. The significance of responses of the genome to challenge. Science 226, 792?801 (1984). 107. Cam, H. P., Noma, K., Ebina, H., Levin, H. L. & Grewal, S. I. Host genome surveillance for retrotransposons by transposon-derived proteins. Nature 451, 431?436 (2008). 108. Molinier, J., Ries, G., Zipfel, C. & Hohn, B. Transgeneration memory of stress in plants. Nature 442, 1046?1049 (2006). 109. Rosenberg, S. M. & Hastings, P. J. Microbiology and evolution. Modulating mutation rates in the wild. Science 300, 1382?1383 (2003). 110. Kishony, R. & Leibler, S. Environmental stresses can alleviate the average deleterious effect of mutations. J. Biol. 2, 14 (2003). 111. Kaneko, K. Evolution of robustness to noise and mutation in gene expression dynamics. PLoS ONE 2, e434 (2007). 112. Sangster, T. A., Lindquist, S. & Queitsch, C. Under cover: causes, effects and implications of Hsp90-mediated genetic capacitance. Bioessays 26, 348?362 (2004). 113. Bell-Pedersen, D. et al. Circadian rhythms from multiple oscillators: lessons from diverse organisms. Nature Rev. Genet. 6, 544?556 (2005). 114. Wijnen, H. & Young, M. W. Interplay of circadian clocks and metabolic rhythms. Annu. Rev. Genet. 40, 409?448 (2006). 115. Gallego, M. & Virshup, D. M. Post-translational modifications regulate the ticking of the circadian clock. Nature Rev. Mol. Cell Biol. 8, 139?148 (2007). Acknowledgements We apologize to colleagues in the field for not citing all rele- vant papers owing to space constraints. We thank C. Wilkinson, J. Mata, D. Lackner, V. Pancaldi and F. Schubert for comments on the manuscript. Research in our laboratory is supported by Cancer Research UK. L.L.-M. and S.M. are supported by postdoctoral fellowships from FEBS and the Swiss National Science Foundation, respectively. DATABASES Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene HIS3 UniProtKB: http://ca.expasy.org/sprot Bdf1 | Mot1 | TAF3 | TBP | TRF3 FURTHER INFORMATION B�hler laboratory: http://www.sanger.ac.uk/PostGenomics/S_pombe ALL LINKS ARE ACTIVE IN THE ONLINE PDF R E V I E W S NATURE REVIEWS | GENETICS VOLUME 9 | AUGUST 2008 | 593 "
Add Content to Group
|
Bookmark
|
Keywords
|
Flag Inappropriate
share
Close
Digg
Facebook
MySpace
Google+
Comments
Close
Please Post Your Comment
*
The Comment you have entered exceeds the maximum length.
Submit
|
Cancel
*
Required
Comments
Please Post Your Comment
No comments yet.
Save Note
Note
View
Public
Private
Friends & Groups
Friends
Groups
Save
|
Cancel
|
Delete
Please provide your notes.
Next
|
Prev
|
Close
|
Edit
|
Delete
Genetics
Gene Inheritance and Transmission
Gene Expression and Regulation
Nucleic Acid Structure and Function
Chromosomes and Cytogenetics
Evolutionary Genetics
Population and Quantitative Genetics
Genomics
Genes and Disease
Genetics and Society
Cell Biology
Cell Origins and Metabolism
Proteins and Gene Expression
Subcellular Compartments
Cell Communication
Cell Cycle and Cell Division
Scientific Communication
Career Planning
Loading ...
Scitable Chat
Register
|
Sign In
Visual Browse
Close
Comments
CloseComments
Please Post Your Comment