Nội dung

MINIREVIEW Collective behavior in gene regulation: Post-transcriptional regulation and the temporal compartmentalization of cellular cycles Maria C. Palumbo1, Lorenzo Farina2, Alberto De Santis2, Alessandro Giuliani3, Alfredo Colosimo1, Giorgio Morelli4 and Ida Ruberti5 1 2 3 4 5 Department of Physiology and Pharmacology, Sapienza University of Rome, Italy Department of Computer and Systems Science ‘Antonio Ruberti’, Sapienza University of Rome, Italy Department of Environment and Primary Prevention, Istituto Superiore di Sanità (ISS), Rome, Italy National Research Institute for Food and Human Nutrition, Rome, Italy National Research Council, Institute of Molecular Biology and Pathology, Rome, Italy Keywords metabolic cycle; oscillations; posttranscriptional regulation; RNA-binding proteins Correspondence L. Farina, Dipartimento di Informatica e Sistemistica ‘Antonio Ruberti’, Via Ariosto 25, 00185 Rome, Italy Fax: +39 067 727 4088 Tel: +39 064 457 526 E-mail: lorenzo.farina@uniroma1.it (Received 10 December 2007, revised 31 January 2008, accepted 26 February 2008) doi:10.1111/j.1742-4658.2008.06398.x Self-sustained oscillations are perhaps the most studied objects in science. The accomplishment of such a task reliably and accurately requires the presence of specific control mechanisms to face the presence of variable and largely unpredictable environmental stimuli and noise. Self-sustained oscillations of transcript abundance are, in fact, widespread and are not limited to the reproductive cycle but are also observed during circadian rhythms, metabolic cycles, developmental cycles and so on. To date, much of the literature has focused on the transcriptional machinery underlying control of the basic timing of transcript abundance. However, mRNA abundance is known to be regulated at the post-transcriptional level also and the relative contribution of the two mechanisms to gene-expression programmes is currently a major challenge in molecular biology. Here, we review recent results showing the relevance of the post-transcriptional regulation layer and present a statistical reanalysis of the yeast metabolic cycle using publicly available gene-expression and RNA-binding data. Taken together, the recent theoretical and experimental developments reviewed and the results of our reanalysis strongly indicate that regulation of mRNA stability is a widespread, phase-specific and finely tuned mechanism for the multi-layer control of gene expression needed to achieve high flexibility and adaptability to external and internal signals. Self-sustained temporal structured activities displaying a clear and robust periodic time behavior are perhaps the most studied objects in science. In fact, the dawn of modern science is generally assumed to coincide with the development of general laws able to explain and predict the regular orbits of planets. On more empirical ground, the accomplishment of a task with a high degree of accuracy and robustness requires the presence of specific control mechanisms and strategies for actively sustaining the desired time profiles (i.e. a trajectory-tracking device) such as regular oscillations, in the presence of variable and largely unpredictable environmental stimuli and noise. Oscillations of the molecular components of the cell have been observed since the early days of molecular cell biology in the course of the study of protein levels Abbreviations OR, odds ratio; OX, oxidative; RB, reductive biosynthetic; RC, reductive charge; TF, transcription factor. 2364 FEBS Journal 275 (2008) 2364–2371 ª 2008 The Authors Journal compilation ª 2008 FEBS M. C. Palumbo et al. during the reproductive cell cycle. In fact, a major breakthrough occurring in the late 1980s was the discovery that the eukaryotic cell cycle is driven by stagespecific distinct waves of activation of cyclin-dependent kinases [1]. However, the study of the transcriptome in biological cycles has revealed that oscillations are present at the mRNA level and that a large number of genes are regulated in a periodic manner. Self-sustained oscillation of transcript abundance is actually widespread [2] and not limited to the reproductive cycle [3], but is also observed during circadian rhythms [4], metabolic cycles [5], the developmental cycles of parasitic protozoa [6] and somitogenesis [7], just to cite a few. Moreover, oscillations are also observed in response to external stimuli, as reported for mammalian cells after serum shock [8,9]. To date, much of the literature has focused on the transcriptional machinery underlying control of the basic timing of transcript abundance [10]. Key cycle regulators and their modes of action have been identified and networks of transcription factors (TFs) and their targets are currently available for many different cycles and organisms [11–13]. It is now widely accepted that periodic waves of transcription may be obtained by circular cascades of TFs [14] or by their combinatorial regulatory action allowing for sequential activation [15]. However, such a transcriptional control network is just part of the story because mRNA abundance is regulated at many levels, including the post-transcriptional regulation layer and the relative contribution of the two mechanisms to gene-expression programmes is currently a major challenge for molecular biology. Post-transcriptional regulation is due to different mechanisms such as processing, export, localization, decay and translation control but the recent discovery of micro-RNA has revitalized the study of the mRNA decay pathway and its regulation [16]. One key point is that the regulation of mRNA stability may act in concert with the transcriptional machinery [17] and significantly contribute to changes in gene-expression patterns in response to external stimuli. A question remains as to the extent that this type of regulation may occur under different biological scenarios like cellular cycles. In fact, the existence of such a regulatory layer may require re-evaluation of the common model of the control of gene expression which essentially invokes the turning on and off of gene transcription [18]. To address this issue on an experimental basis, a number of scientists have developed methods to simultaneously measure gene expression and mRNA stability during metabolic shifts [19] and cellular cycles [20]. As a general rule, there is often the need to combine both positive and negative control actions to keep a Regulation and compartmentalization of cell cycles stable and reliable trajectory over time. For example, in order to achieve fast, accurate and reliable vehicle dynamics, an integrated throttle and brake control system is usually required. This is the case, for example, of high-performance driving techniques, like heel-andtoe or left-foot-braking, which rely on simultaneous and ⁄ or alternate gas ⁄ brake pedal usage. Clearly, such a driving style is energy-consuming and is worth using only when precisely regulated levels are required in the face of sudden and unpredictable external events. In other words, fast, precise and robust behavior is very expensive both in terms of control strategies complexity and resource consumption. When dealing with gene-expression data, the ‘throttle pedal’ may correspond to the TF system, allowing for specific DNA sequences to be made accessible to polymerases and synthesizing the corresponding mRNA, whereas the ‘brake pedal’ may be associated with the so-called degradosome system, made up of those factors able to selectively degrade specific mRNA molecules. Indeed, there is growing interest in the study of post-transcriptional regulation [16] especially for the mechanisms leading to the modulation of mRNA turnover in response to environmental changes [19,21]. Such a dual mechanism for mRNA upregulation is present, for example, in MAPK signaling where MAPK-signaling-mediated phosphorylation activates a TF and ⁄ or a RNA-binding protein resulting in the upregulation of mRNA, via transcriptional activation or mRNA stabilization, respectively [22]. There is also growing evidence that mRNA decay regulation may play a fundamental role in cellular cycles. In fact, studies on the regulation of mRNA stability during the yeast reproductive cell cycle have been carried out on some specific transcripts, such as the histone mRNAs [23] or cyclin mRNAs [24] and proved that post-transcriptional control is important in establishing their periodicity [25]. Evidence of post-transcriptional regulation of the histone genes has also been shown in higher eukaryotes and mammalian cell types, including HeLa cells [26]. A fundamental role for mRNA degradation in the establishment and maintenance of oscillations has been reported in Drosophila [27], Arabidopsis [28] and mammalian [29,30] circadian rhythms and in mouse fibroblasts ultradian oscillations in response to serum [9]. On a genome-wide scale, a systematic study of the role exerted by the mRNA decay rate in cellular cycles is still lacking and it certainly is a topic worthy of further investigation. A recent experimental study reports a fundamental role for mRNA stability in global geneexpression regulation in the Plasmodium falciparuim FEBS Journal 275 (2008) 2364–2371 ª 2008 The Authors Journal compilation ª 2008 FEBS 2365 Regulation and compartmentalization of cell cycles M. C. Palumbo et al. intraerytrocytic developmental cycle [20] and a recent computational analysis [31] showed a specific and active role for transcript stability regulation in the yeast reproductive cell cycle. In the following section, we report a statistical reanalysis of the yeast metabolic cycle using gene expression data [5] and RNA-binding protein targets [32] aiming to show that the role of post-transcriptional regulation may be highly relevant for the regulation of transcripts oscillation on a global scale. Post-transcriptional regulation in the yeast metabolic cycle Saccharomyces cerevisiae (yeast) metabolic cycles can be considered as very reliable temporal structures in which an extremely complex apparatus – the yeast cell – strives to maintain self-sustained cycles at different functional levels, ranging from gene expression (recently assessed on a genome-wide scale by the McKnight group) [5] to metabolic activity (oxygen consumption). The accuracy and robustness of such cycles, in terms of frequency and amplitude, is much greater than any other known cycle at the cellular level, namely much greater than the reproductive cell cycle. For this reason, gene-expression dynamics of the yeast metabolic cycle is a natural candidate for studying the ‘basic principles’ of gene expression regulation. Recently, five members of the RNA-binding proteins PUF family (PUF1–PUF5), or PUMILIO-Fem3-binding proteins, have been studied for sequence specificity on a genome-wide scale [32], thus providing insight into one of the most important mechanisms of posttranscriptional regulation. The numbers of identified targets are 40 for PUF1, 146 for PUF2 and around 200 for PUF3, PUF4 and PUF5, indicating that the expression of a large number of genes can, in principle, be modulated by specific post-transcriptional events. PUF proteins, as mRNA-specific regulators of deadenylation, have been conserved throughout eukaryotes suggesting that they are likely to play a prominent role in the control of transcript-specific rates of deadenylation in yeast by interacting with the mRNA turnover machinery [33]. Recent integrated analysis by De Lichtenberg et al. [34] on yeast cell reproductive cycle data, did not show any relevant correlation between PUF family genes and the cell cycle. In fact, none of the PUF1–5 genes is present in their extended list of 1159 reproductive cell-cycle-regulated genes. Obviously, we do not exclude the possibility of a specific role for post-trascriptional regulation on a global scale, but no conclusions can be drawn basing upon current available data on PUF family targets. 2366 Gene transcription has received the most attention for historical and technical reasons, but transcription is just the first stage in the process of gene expression. From splicing to polyadenylation, every aspect of a transcript’s life is subject to elaborate control and it is therefore no surprise that many cellular factors and mechanisms are devoted entirely to modulating the rate of mRNA degradation [16]. Consequently, it is of paramount importance that this gap is filled and evidence provided of the mode of action of the other ‘arm’ of gene regulation during gene expression temporal programmes. Given the biological importance of the presence of links and coordinated action across multiple layers of control, understanding gene expression requires an integrated view by combining data from different aspects of regulation [35]. Although this approach holds great promise, there are currently few studies that take into account regulation at multiple levels [35]. The PUF control system: a statistical re-analysis In this section we investigate the role of the ‘PUF control system’ [32], in the temporally compartmentalized regulation of gene expression during the yeast metabolic cycle [5]. Our integrative statistical reanalysis demonstrates a clear phase specificity of PUF-mediated regulation in the yeast metabolic cycle, allowing us to hypothesize post-transcriptional regulation as a key player in coordinating the global timing of gene expression during specific phases of the metabolic cycle. The microarray gene-expression data relative to the Tu et al. study [5] were downloaded from http:// yeast.swmed.edu/cgi-bin/dload.cgi. Following the authors’ indication, we selected their sentinel genes as probes for the three phases in which the metabolic cycle was partitioned; these genes were MRPL10 [reductive biosynthetic (RB) phase], POX1 [reductive charge (RC) phase] and RPL17B [oxidative (OX) phase]. For each of the three sentinel genes we selected a cluster containing the 500 most correlated gene products to obtain a data set of 1500 genes. Each cluster belongs to one of the three phases of the metabolic cycle. The role of PUF genes during metabolic-cycle regulation was assessed by means of standard statistical tests, namely the correlation of PUF mRNA temporal variation with the centroid profile of the three clusters and the evaluation, by means of the odds ratio (OR) statistics, of the ‘enrichment’ of gene pairs sharing the same PUF within the same phase of the cycle. The binding specificity of PUF genes was assessed on FEBS Journal 275 (2008) 2364–2371 ª 2008 The Authors Journal compilation ª 2008 FEBS M. C. Palumbo et al. Regulation and compartmentalization of cell cycles the basis of Gerber et al.’s study [32]. The PUF-based results were compared with similar analysis based on TFs. The binding specificity of TFs was based upon McIsaac et al. [11] by considering a stringent P-value for DNA binding of 0.001. For each of the ‘sentinel genes’ of the three phases of the cycle, the 500 most correlated genes in terms of Pearson’s correlation coefficient along temporal profiles were selected so giving rise to three clusters. The temporal profile of each cluster is depicted in Fig. 1A in terms of cluster centroid activation together with the standard deviation. Data are expressed as Z-scores, that is, each gene-expression value is de-meaned and divided by its standard deviation. Genes pertinent to each cluster are correlated with the sentinel gene with Pearson’s correlation coefficient R ranging from 0.82 to 1 (RB), 0.87 to 1 (RC) and 0.85 to 1 (OX), giving a reliable picture of the three phases in time. The data are averaged over the three available cycles. Normalized fold induction A Phase centroids profiles 3 Normalized fold induction Fig. 1. Temporal profiles of the metabolic phases centroids (A) and of the five members of the PUF family (B) as obtained by the same experimental dataset. Data have been Z-normalized (zero mean, unit standard deviation). OX RB RC 2 1 0 –1 –2 B As is apparent from Fig. 1B, there is clear similarity between PUF time profiles and cluster centroids dynamics. Basically, PUF1 and PUF2 covary with the reductive charge phase and PUF4 and PUF5 go together with the oxidative phase, whereas PUF3 covaries with the reductive biosynthetic phase. Our statistical analysis provides empirical indication of temporal covariation between PUF family genes and the yeast metabolic cycle. In order to go into more depth, we need to discriminate between a pure episodical and a potentially biological significant link, and therefore we move a step further by computing the Pearson’s correlation coefficient for all the gene pairs resulting from the three clusters corresponding to the three metabolic phases. By doing so, we obtained more than one million (1 124 250) distinct correlation coefficients, that are distributed as in Fig. 2. As shown in Fig. 2, there is a clear bimodal distribution allowing for an unambiguous partition of the gene pairs into three classes: 0 50 100 150 Time (min) 200 250 PUFs mRNA profiles 3 PUF1 PUF2 PUF3 PUF4 PUF5 2 1 0 –1 –2 0 50 100 FEBS Journal 275 (2008) 2364–2371 ª 2008 The Authors Journal compilation ª 2008 FEBS 150 Time (min) 200 250 2367 Regulation and compartmentalization of cell cycles 8 x 10 3 M. C. Palumbo et al. Pearson correlation histogram 6 4 2 0 –1 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 0.8 1 Fig. 2. Pearson correlation histogram of all genes profiles pairs (1 124 250) obtained using the selected 1500 metabolic-cycle-regulated genes. 1 Positively correlated pairs (r > 0.6). 2 Linearly independent pairs (0.6 > r > )0.6). 3 Negatively correlated pairs (r < )0.6). This partition comes directly from the existence of three temporal clusters, so that gene pairs relative to the same cluster go into class 1, whereas genes in a pair coming from different clusters, go alternatively into class 2 and class 3, depending on the relative phase shift of the corresponding clusters. Such sharp distribution, together with the numerosity of the correlation coefficient data set allow us to consider two different filters on Fig. 2 to immediately highlight the role of post-transcriptional regulation compared with TF-based regulation. Consequently, from the initial population of Fig. 2 we selected only those pairs sharing at least one TF in common (Fig. 3A) and pairs sharing at least one PUF in common (Fig. 3B). As it is evident by comparison of Fig. 3A,B, pairs with a common PUF are far more rich in the posiA Transcriptionally co-regulated gene pairs 800 600 400 200 0 –1 B –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 0.8 1 Post-transcriptionally co-regulated gene pairs (PUF family proteins) 150 100 50 0 –1 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 0.8 1 Fig. 3. Pearson correlation histogram (A) of gene pairs sharing at least one transcription factor and (B) gene pairs sharing at least one mRNA binding protein of the PUF family (PUF1–5). 2368 tively correlated class, whereas pairs sharing the same TF are scattered along the distribution occupying all the three correlation classes. This points to the highest specificity of PUF-mediated post-transcriptional control with respect to TF control. Clearly, we must take into account the fact that we do have the binding specificity for many more TFs (118) than post-transcriptional regulators such as PUFs (5). In order to obtain an unbiased estimation of the different specificity of the two regulation systems, we computed the OR values for the TF and PUF commonality, respectively: OR (TF) = fraction of gene pairs having at least one common TF in the correlated subset with respect to all gene pairs having at least one common TF ⁄ fraction of pairs in the correlated subset w.r.t. all gene pairs. OR (PUF) = fraction of gene pairs having at least one common PUF in the correlated subset w.r.t. all gene pairs having at least one common PUF ⁄ fraction of pairs in the correlated subset w.r.t. all gene pairs. A value equal to 1.37 for transcription factor and 2.11 for PUF was obtained, again pointing to a greater specificity of PUF control. This was confirmed by the OR values computed for the anticorrelated pairs that was equal to 0.87 for TF and 0.28 for PUF. The regulation specificity of PUF family members is mainly driven by PUF3 having an OR equal to 2.49 as for the enrichment of correlated pairs and 0.04 as for the depletion in anticorrelated subsets. It is also worth noting that PUF2 it is mildly enriched for anticorrelated pairs (OR = 1.64) thus suggesting a regulatory role between oxidative and reductive charge phases, and such role is shared with some TFs. An overall pictorial representation of the PUF network is provided by Fig. 4. In Fig. 4 the red, blue and green circles correspond to genes of the different metabolic cycle phases; PUF family genes are represented by orange circles. The edges link each PUF with its targets. It is important to note the phase specificity of the PUFs together with the positioning of some PUFs between different metabolic cycle phases so that the possibility for posttrascriptional regulation working in the progression from one phase to another is further made clear. In particular, it is striking that PUF3 seems to be the main determinant of PUF family specificity in regulation of the yeast metabolic cycle. The reductive biosynthetic phase specificity of PUF3 strongly suggests a role for this regulator in the progression from the reductive biosynthetic to reductive charge phase. This is consistent with recent findings demostrating that PUF3 acts as a transcript-specific regulatory role of mRNA degradation in yeast. More precisely, PUF3 FEBS Journal 275 (2008) 2364–2371 ª 2008 The Authors Journal compilation ª 2008 FEBS M. C. Palumbo et al. Regulation and compartmentalization of cell cycles RB OX PUF4 PUF3 Fig. 4. Representation of the PUF network. Colours correspond to different phases of the metabolic cycle: red (oxidative phase), green (reductive biosynthetic phase) and blue (reductive charge phase). Orange circles denote PUF proteins and arrows point to their target genes. The picture makes very clear that the PUF3 gene and the most of PUF3 targets peak at the reductive biosynthetic phase thus showing the high specificity of the post-trascriptional regulation layer. has been shown to affect the stability of COX17 [33] and PET123 [36] transcripts. Taken together, these considerations suggest a role for PUF3 in the rapid downregulation of its targets at the end of the reductive biosynthetic phase and in the upregulation at the beginning of the same phase which is also consistent with a recent experimental study supporting for a role of PUF3 in the reduction of mitochondrial biogenesis during glucose repression [17] downstream of the TOR signaling pathway [37]. Next, we looked at TFs that, according to McIsaac et al. [11], have PUF genes as targets and we obtained the results reported in Table 1. The results regarding the PUF3 gene are particularly intriguing: both ABF1 and SWI4 have an expression peak in the reductive biosynthetic phase but it seems that transcriptional and post-transcriptional regulation are not simply related, given that PUF3 has only 10% of shared targets with ABF1 and virtually no common target with SWI4 and ROX1. Moreover, ABF1 is a very general TF, thus Table 1. Transcription factors binding PUF family gene promoters according to MacIsaac et al. [11] using a stringent P-value for binding of 0.001. PUF family member TFs PUF1 PUF2 PUF3 PUF4 PUF5 FKH1, FKH2, NDD1, SWI6, MCM1, UME6 SFP1, FHL1, RAP1 ABF1, ROX1, SWI4 No data available PHO2, MBP1, SWI4, SWI6 (YJR091C) (YPR042C) (YLL013C) (YGL014W) (YGL178W) PUF5 PUF1 RC PUF2 the ‘two arms’ (transcriptional and post-transcriptional) seem to be linked in a very complex way. This appears consistent, on the one hand, with the results of Cheadle et al. [18] where alternate regulation of mRNA transcription and mRNA stability was observed during human jurkat T-cell activation. On the other hand, the observed behavior after a carbon source shift in yeast is simultaneous regulation of transcription and degradation because a decrease in transcriptional activity and an increase in messenger stability result in an almost flat mRNA abundance time profile [19]. Conclusions The picture emerging from our integrative analysis of the yeast metabolic cycle is the presence of a very specific, finely wired, post-transcriptional PUF-mediated regulation. The interplay between transcriptional and post-transcriptional regulations observed during the metabolic cycle is consistent with the extreme robustness and reproducibility of such a self-sustained cycle. The two mechanisms are certainly coordinated, but from our computational analysis we could not deduce a simple relationship between the two. This picture has also an intriguing interpretation in terms of ‘car driving’. In fact, it is well known that high performances can be achieved only by actively operating the throttle and the brake pedal together, where the obvious biological counterparts are mRNA synthesis and degradation, respectively. Despite the FEBS Journal 275 (2008) 2364–2371 ª 2008 The Authors Journal compilation ª 2008 FEBS 2369 Regulation and compartmentalization of cell cycles M. C. Palumbo et al. fact that such a driving style is very expensive in terms of both control strategies complexity and resource consumption, it appears that, on specific occasions, the cell may prefer a nonoptimal behavior from an energetic point of view in order to accomplish other tasks efficiently. For example, it has been shown [38] that efficient re-entering into the cell-cycle from a nonproliferative state as terminal differentiation can be obtained by simply removing appropriate cyclin-dependant kinase inhibitors. Taken together, the results presented in this computational analysis performed using gene-expression time series from the yeast metabolic cycle and PUF family proteins binding data, indicate that the regulation of mRNA stability is a widespread, phase-specific and tightly regulated mechanism for the multilayer control of gene expression. Our analysis further supports the possibility that post-transcriptional regulation surpasses the richness and complexity of transcriptional regulation in many, if not all, physiological and developmental situations [35]. 7 8 9 10 11 12 13 Acknowledgement IR and GM would like to acknowledge support, in part, by a grant from ASI, Biotechnology Program. References 15 1 Evans T, Rosenthal ET, Youngblom J, Distel D & Hunt T (1983) Cyclin: a protein specified by maternal mRNA in sea urchin eggs that is destroyed at each cleavage division. Cell 33, 389–396. 2 Tsuchyia M, Wong ST, Yeo ZX, Colosimo A, Palumbo MC, Farina L, Crescenzi M, Mazzola A, Negri R, Bianchi MM et al. (2007) Gene expression waves – cell cycle independent collective dynamics in cutured cells. FEBS J 274, 2878–2886. 3 Spellman PT, Sherlock G, Zhang MQ, Iyer VR, Anders K, Eisen MB, Brown PO, Botstein D & Futcher B (1998) Comprehensive identification of cell cycle-regulated genes of the yeast Saccharomyces cerevisiae by microarray hybridization. Mol Biol Cell 9, 3273–3297. 4 Ueda HR, Matsumoto A, Kawamura M, Iino M, Tanimura T & Hashimoto S (2002) Genome-wide transcriptional orchestration of circadian rhythms in Drosophila. J Biol Chem 277, 14048–14052. 5 Tu BP, Kudliki A, Rowicka M & McKnight L (2005) Logic of the yeast metabolic cycle: temporal compartmentalization of cellular processes. Science 310, 1152– 1158. 6 Bozdech Z, Llinas M, Pulliam BL, Wong ED, Zhu J & De Risi JL (2003) The transcriptome of the intraerytro- 2370 14 16 17 18 19 20 21 cytic developmental cycle of Plasmodium falciparum. PLoS Biol 1, 85–100. Serth K, Schuster-Gossler K, Cordes R & Gossler A (2003) Transcriptional oscillation of Lunatic fringe is essential for somitogenesis. Genes Devel 17, 912–925. Balsalobre A, Damiola F & Schibler U (1998) A serum shock induces circadian gene expression in mammalian tissue culture cells. Cell 93, 929–937. Yoshiura S, Ohtsuka T, Takenaka Y, Nagahara H, Yoshikawa K & Kageyama R (2007) Ultradian oscillations of Stat, Smad and Hes1 expression in response to serum. Proc Natl Acad Sci USA 104, 11292–11297. Breeden LL (2003) Periodic transcription: a cycle within a cycle. Curr Biol 13, 31–38. MacIsaac KD, Wang T, Gordon DB, Gifford DK, Stormo GD & Fraenkel E (2006) An improved map of conserved regulatory sites for Saccharomyces cerevisiae. BMC Bioinformatics 7, 113–125. Ko CH & Takahashi JS (2006) Molecular components of the mammalian circadian clock. Hum Mol Genet 15, 271–277. Murray DB, Beckmann M & Kitano H (2007) Regulation of yeast oscillatory dynamics. Proc Natl Acad Sci USA 104, 2241–2246. Simon I, Barnett J, Hannett N, Harbison C, Rinaldi N, Velkert T, Wyrick J, Zeitlinger J, Gifford D & Jaakkola T (2001) Serial regulation of transcriptional regulators in the yeast cell cycle. Cell 106, 697–708. Kato M, Hata N, Banerjeee N, Futcher B & Zhang MQ (2004) Identifying combinatorial regulation of transcription factors and binding motifs. Genome Biol 5, R56. Garneau NL, Wilusz J & Wilusz CJ (2007) The highways and byways of mRNA decay. Nat Rev Mol Cell Biol 8, 113–126. Cereghino GP & Scheffler IE (1996) Genetic analysis of glucose regulation in Saccharomyces cerevisiae: control of transcription versus mRNA turnover. EMBO J 15, 363–374. Cheadle C, Fan J, Cho-Chung YS, Werner T, Ray J, Do L, Gorospe M & Becker KG (2005) Control of gene expression during T cell activation: alternate regulation of mRNA transcription and mRNA stability. BMC Genomics 6, 75–91. Garcia-Martinez J, Aranda A & Perez-Ortin JE (2004) Genomic run-on evaluates transcription rates for all yeast genes and identifies gene regulatory mechanisms. Mol Cell 15, 303–313. Shock JL, Fischer KF & De Risi J (2007) Whole genome analysis of mRNA decay in Plasmodium falciparum reveals a global lengthening of mRNA half-life during the intraerythrocytic development cycle. Genome Biol 8, R134. Wang Y, Liu CL, Storey JD, Tibshirani RJ, Herschlag D & Brown PO (2002) Precision and functional specific- FEBS Journal 275 (2008) 2364–2371 ª 2008 The Authors Journal compilation ª 2008 FEBS M. C. Palumbo et al. 22 23 24 25 26 27 28 29 30 ity in mRNA decay. Proc Natl Acad Sci USA 99, 5860– 5865. Sugiura R, Kita A & Kuno T (2004) Upregulation of mRNA in MAPK signaling. Cell Cycle 3, 286–288. Lycan DE, Osley MA & Hereford LM (1987) Role of transcriptional and posttranscriptional regulation in expression of histone genes in Saccharomyces cerevisiae. Mol Cell Biol 7, 614–621. Gill T, Cai T, Aulds J, Wierzbicki S & Schmitt E (2004) RNase MRP cleaves the CLB2 mRNA to promote cell cycle progression: novel method of mRNA degradation. Mol Cell Biol 24, 945–953. Xu H, Johnson L & Grunstein M (1990) Coding and noncoding sequences at the 3¢ end of yeast histone H2B mRNA confer cell cycle regulation. Mol Cell Biol 10, 2687–2694. Morris TD, Weber LA, Hickey E, Stein GS & Stein JL (1991) Changes in the stability of a human H3 histone mRNA during the HeLa cell cycle. Mol Cell Biol 11, 544–553. So WV & Rosbash M (1997) Post-transcriptional regulation contributes to Drosophila clock gene mRNA cycling. EMBO J 16, 7146–7155. Lidder P, Gutiérrez RA, Salomé PA, Robertson McClung C & Green PJ (2005) Circadian control of messenger RNA stability. Association with a sequence-specific messenger RNA decay pathway. Plant Physiol 138, 2374–2385. Kwak E, Kim TD & Kim KT (2006) Essential role of 3¢-untraslated region-mediated mRNA decay in circadian oscillations of mouse period3 mRNA. J Biol Chem 281, 19100–19106. Kim TD, Kim JS, Kim JH, Myung J, Chae HD, Woo KC, Jang SK, Koh DK & Kim KT (2005) Rhythmic Regulation and compartmentalization of cell cycles 31 32 33 34 35 36 37 38 serotonin N-acetyltrasnferase mRNA degradation is essential for the maintenance of its circadian oscillation. Mol Cell Biol 25, 3232–3246. Farina L, De Santis A, Morelli G & Ruberti I (2007) Dynamic measure of gene co-regulation. IET Syst Biol 1, 10–17. Gerber AP, Hershlag D & Brown PO (2004) Extensive association of functionally and cytotopically related mRNAs with Puf family RNA-binding proteins in yeast. PLoS Biol 2, 342–354. Olivas W & Parker R (2000) The Puf3 protein is a transcript-specific regulator of mRNA degradation in yeast. EMBO J 19, 6602–6611. De Lichtenberg U, Jensen TS, Jensen LJ & Brunak S (2003) Protein feature based identification of cell cycle regulated proteins in yeast. J Mol Biol 329, 663– 674. Mata J, Marguerat S & Bahler J (2005) Posttranscriptional control of gene expression: a genome-wide perspective. Trends Biochem Sci 30, 506–514. Garcia-Rodriguez LJ, Gay AC & Pon LA (2007) Puf3p a Pumilio family RNA binding protein, localizes to mitochondria and regulates mitochondrial biogenesis and motility in budding yeast. J Cell Biol 176, 197–207. Foat BC, Houshmandi SS, Olivas WM & Bussemaker HJ (2005) Profiling condition-specific, genome-wide regulation of mRNA stability in yeast. Proc Natl Acad Sci USA 102, 17675–17680. Pajalunga D, Mazzola A, Salzano AM, Biferi MG & De Luca G (2007) Critical requirement for cell cycle inhibitors in sustaining nonproliferative states. J Cell Biol 176, 807–818. FEBS Journal 275 (2008) 2364–2371 ª 2008 The Authors Journal compilation ª 2008 FEBS 2371