- Email: [email protected]
Oscillatory dynamics of the extracellular signal-regulated kinase pathway
Available online at www.sciencedirect.com
Oscillatory dynamics of the extracellular signal-regulated kinase pathway Harish Shankaran1 and H Steven Wiley2 The extracellular signal-regulated kinase (ERK) pathway is a central signaling pathway in development and disease and is regulated by multiple negative and positive feedback loops. Recent studies have shown negative feedback from ERK to upstream regulators can give rise to biochemical oscillations with a periodicity of between 15 and 30 min. Feedback due to the stimulated transcription of negative regulators of the ERK pathway can also give rise to transcriptional oscillations with a periodicity of one to two hours. The biological significance of these oscillations is not clear, but recent evidence suggests that transcriptional oscillations participate in developmental processes, such as somite formation. Biochemical oscillations are more enigmatic, but could provide a mechanism for encoding different types of inputs into a common signaling pathway. Addresses 1 Computational Biology and Bioinformatics, Pacific Northwest National Laboratory, 902 Battelle Boulevard, Richland, WA 99352, USA 2 Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, 902 Battelle Boulevard, Richland, WA 99352, USA Corresponding author: Wiley, H Steven ([email protected])
Current Opinion in Genetics & Development 2010, 20:650–655 This review comes from a themed issue on Genetics of system biology Edited by Jeff Hasty, Alex Hoffmann and Susan Golden
0959-437X/$ – see front matter # 2010 Elsevier Ltd. All rights reserved. DOI 10.1016/j.gde.2010.08.002
One potential answer to how receptors generate a diversity of responses through common signaling pathways is through the relative magnitude by which different pathways are stimulated. Another potential answer is through both the magnitude and duration of signals generated through a single pathway. A good example of the latter process is the differential induction of cell proliferation in PC12 cells through the extracellular signal-regulated kinase (ERK) pathway. In this case, the addition of epidermal growth factor (EGF) induces transient activation of ERK whereas nerve growth factor (NGF) induces a prolonged activation. The ability of EGF to induce cell proliferation and NGF to induce cell differentiation has been ascribed to such divergent activation kinetics [2]. Although the underlying molecular mechanisms responsible for the cellular response are currently obscure, the source of the differential ERK activation profiles appears to be due to the pattern of feedback circuits used by the EGF versus NGF receptor systems [3]. The ERK pathway is a central regulator in a wide variety of different cellular responses, including proliferation, differentiation, migration and apoptosis. It was one of the first kinase signaling cascades identified and is a downstream mediator of Ras activity, which is frequently associated with many human cancers. The structure of the ERK cascade is that of a three-tier, high-gain amplifier with activation of the first kinase (Raf), stimulating the phosphorylation of a second kinase (MEK), leading to the activation of the mediator kinase ERK. Following its activation, ERK translocates to the nucleus where it phosphorylates transcription factors responsible for many of the biological effects of the associated receptor pathways (for a review of the ERK pathway, see [4]).
Introduction One of the most fundamental questions in cell and developmental biology is how information is encoded in the pattern of hormones, growth factors and environmental cues that a cell receives. Specific cell surface receptors are known to be the initial step by which cells detect many signals, but receptor-binding events must be translated into the downstream signals needed to generate appropriate cellular responses. Intensive molecular studies over the last several decades have shown that most receptors activate only a handful of different signaling pathways [1]. However, the responses resulting from the activation of different receptors are incredibly diverse. This gives rise to the obvious question of how a wide diversity of different responses can be generated by activating only a few pathways. Current Opinion in Genetics & Development 2010, 20:650–655
As one would expect for such a central player in signal transduction, the ERK pathway is regulated extensively, primarily through negative feedback mechanisms (see Figure 1). Early kinetic studies showed that the protein SOS, which couples the ERK pathway to upstream receptors through Ras, is rapidly phosphorylated and attenuated by ERK [5,6]. In the ensuing years, ERK has been shown to negatively regulate Raf by direct phosphorylation as well [7]. The activation of the ERK pathway also rapidly induces the expression of phosphatases that can directly inactivate ERK, notably DUSP1, DUSP 4 and DUSP6 [8]. ERK can also stimulate positive feedback of its pathway indirectly by either inactivating negative regulators of the pathway, such as RKIP [9], or by stimulating ligand shedding that can activate receptors www.sciencedirect.com
Oscillations in ERK pathway Shankaran and Wiley 651
Figure 1
Positive and negative feedback of the extracellular signal-regulated kinase (ERK) pathway in mammalian cells. The ERK pathway can be activated by multiple factors, such as hepatocyte growth factor (HGF) or transforming growth factor alpha (TGFa) which work through their cognate receptors (MET and the EGFR, respectively). The receptors phosphorylate and bind adaptor proteins, such as Gab1, Shc and Grb2, which can associate with Sos1 or Sos2. These complexes activate Ras, which then stimulate the activity of the Raf, MEK and ERK cascade. Negative feedback (red lines) can ensue from the phosphorylation of either Sos1 or Raf by ERK, or the induction of negative feedback inhibitors, such as the DUSPs through phosphorylation of transcription factors, such as AP1. Positive feedback (blue lines) can arise from the ERK-induced shedding of ligands, such as TGFa, the induced synthesis of the ligands, or the inactivation by phosphorylation of negative regulators of the ERK pathway, such as RKIP. The times of the different feedback processes (indicated above the lines) are distinct and can be used to identify which mechanisms are operating.
upstream of the ERK pathway [10]. ERK activity can also stimulate the transcription of those ligand genes, thus promoting long-term positive effects on ERK signaling [11].
transcriptional feedback effects. Here we refer to oscillations that are inherent to the ERK biochemical pathway as biochemical oscillations to distinguish them from transcriptional oscillations that require ERK-mediated gene expression.
Oscillatory behavior of the ERK pathway Because of the high gain of the central ERK pathway, it has the property of ultrasensitivity (nonlinear input– output behavior). An ultrasensitive system combined with negative feedback and time delays can give rise to negative feedback oscillations. The presence of both positive and negative feedback can result in relaxation oscillations. The numerous positive and negative feedback loops that act on the ERK pathway can thus potentially generate multiple types of oscillatory behavior. In addition to such feedback regulation, studies have also shown that sequestration of kinases and phosphatases by their substrates could potentially generate oscillations by creating ‘implicit’ positive and negative feedback loops within the multilevel MAPK cascade [12,13]. Theoretical studies have proposed that the central ERK pathway itself could exhibit both negative feedback [14] and relaxation oscillations [12,15] even in the absence of www.sciencedirect.com
Oscillatory behavior of the ERK pathway has been an intriguing possibility because of its potential to encode additional levels of information in this important signaling pathway. However, until recently, there had been no experimental evidence that the ERK pathway was capable of generating sustained oscillations. Several papers have demonstrated two successive peaks in the activation of the upstream mediator Ras due to feedback inhibition of SOS [16] but this did not cause oscillations of ERK, apparently because of the simultaneous strong induction of an ERK-specific phosphatase (DUSP1). A more recent paper has demonstrated several cycles of Ras activation accompanied by ERK phosphorylation and Hes1 gene expression in mouse cells following addition of FGF [17]. The authors suggest that these oscillations were due to negative feedback inhibition of SOS by phosphorylated ERK. However, the relatively long time Current Opinion in Genetics & Development 2010, 20:650–655
652 Genetics of system biology
Figure 2
Oscillations of nuclear ERK following stimulation of epithelial cells with EGF. Cells expressing an ERK-GFP fusion protein were stimulated with EGF at the indicated time and the levels of nuclear ERK were followed by automated image analysis at 1 min intervals, as described in [19]. Shown are the results from four individual cells expressing different levels of ERK-GFP.
period of the ERK phosphorylation cycles (approximately 100 min) are much longer than the periodicity described for SOS phosphorylation (<30 min) and are more consistent with negative feedback through the transcriptionally induced expression of feedback inhibitors, such as the DUSPs. The observed oscillations in SOS activity could thus be a secondary effect of the transcriptional oscillation in ERK activity.
its phosphorylation. Using a similar imaging approach, Cohen-Saidon et al. [20] observed two successive peaks of ERK nuclear translocation in a lung cancer cell line with a periodicity of approximately 30 min. Both studies observed variability in the fraction of cells within a population that underwent oscillations, but saw consistent properties of the waveform itself as well as its periodicity.
ERK oscillations mediated by transcriptionally induced feedback inhibitors have recently been convincingly demonstrated by Hilioti et al. [18], using yeast stimulated by mating pheromone. In this case, activation of a yeast homolog of ERK, Fus3, caused the synthesis of the negative regulators Sst2 and Msg5, resulting in oscillations of Fus3 phosphorylation, gene expression and mating projections. The authors proposed that oscillation-driven mating projections could facilitate yeast mating in shallow or noisy signal gradients.
By using an automated image analysis system to quantify the properties of the ERK translocation oscillations, Shankaran et al. were able to determine how modifying multiple parameters, such as EGF dose, cell density and ERK expression levels, changed oscillation characteristics. They found that the frequency of ERK oscillations was insensitive to experimental conditions, although the fraction of cells that oscillated was strongly dependent on cell density. Mathematical analysis showed that an explicit negative feedback loop between ERK and its upstream activators, such as SOS, was required to generate oscillations with the observed properties [19]. The analysis showed that the biochemical oscillations of ERK could be best characterized as negative feedback oscillations rather than relaxation oscillations.
Recently, compelling evidence of rapid biochemical oscillations in ERK activity was obtained through live cell imaging studies [19]. Instead of directly monitoring ERK phosphorylation, these studies followed the translocation of ERK from the cytoplasm into the nucleus following its activation. This allowed ERK dynamics to be continuously recorded for many hours. Using this approach, Shankaran et al. [19] demonstrated that the addition of EGF to epithelial cells resulted in the oscillation of ERK between the nucleus and cytoplasm with a periodicity of approximately 15 min (see Figure 2). Biochemical studies confirmed that the oscillations in ERK translocation were correlated with oscillations in Current Opinion in Genetics & Development 2010, 20:650–655
Relaxation oscillators are believed to be easier to evolve and provide the added benefit of a tunable frequency [21]. Why then does the ERK pathway employ a biochemical mechanism that results in a constant frequency? It may be possible that this feature is critical for the ability of the ERK pathway to encode and decode information. For example, the transcription systems that decode ERK oscillations may act as band-pass filters that operate www.sciencedirect.com
Oscillations in ERK pathway Shankaran and Wiley 653
optimally when stimulated with an appropriate frequency. Another interesting possibility is that frequency robustness is itself required to encode information when a system contains multiple levels of feedback control. For example, if transcriptional feedback mechanisms could change the frequency of the oscillation, then any potential information in the oscillation would be obliterated.
Potential functions of ERK oscillations ERK is a potent stimulator of gene transcription and so a possible function of slow-transcriptional ERK oscillations could be to drive periodic protein expression. Oscillations in the expression of transcription factors, such as Hes1, have been shown to be important in the maintenance of neural progenitors in the embryonic brain [22] and for controlling somite formation [23]. For example, it has been shown that in mouse, a segmentation clock with a periodicity of approximately two hours drives the pattern of the spine. The mechanism underlying the clock apparently involves coordinate regulated oscillations of the FGF, Notch and Wnt pathways [23]. This coordination most likely occurs through the induction of DUSPs that modulate the activity of the FGF-stimulated ERK pathway [24]. Postulating that oscillations arise from induced transcription of negative regulators has been an appealing concept because of the intrinsic time delay of protein translation [25]. However, multiple ERK-induced negative regulators have been shown to oscillate, making it difficult to determine whether they are causal to transcriptional oscillations of ERK activity or are a consequence. It is possible that rapid biochemical ERK oscillations are a means to selectively turn on a subset of genes. In the case of calcium oscillations, it has been shown that the proteins and genes that respond optimally to oscillations are the ones that can integrate the signal [26–29]. A target that integrates a signal should be induced in a sufficiently sensitive manner so that it can respond to small signal amplitudes, and should be stable enough to ‘remember’ the induction during the off-phase of the signal [30]. Several immediate-early-genes (IEGs) that respond to ERK signaling such as c-Fos, Fra-2, Fra-1 and c-Myc require long signal durations and sufficient signal strength for their induction [31,32]. The levels of activated ERK necessary to induce genes, such as c-Fos are considerably higher than those needed to activate oscillations [33]. Further, with regards to targets that would remain on during the signal’s off phase, most of the negative feedback regulators of cell signaling pathways tend to have extremely short mRNA and protein half-lives [34,35]. For example, DUSP6, which is induced by ERK activation, has an mRNA half-life of 18 min. It is possible that such negative regulators are not effectively induced in response to rapid biochemical ERK oscillations. This would allow the biochemical oscillation to persist for multiple hours, thereby providing sufficient signal www.sciencedirect.com
duration to activate other target genes. Whether there is a subset of ERK-responsive genes that are capable of integrating low amplitude oscillatory ERK signaling is not currently known. Oscillations could also be a mechanism for increasing the mean level of activated ERK in the cytoplasm while still retaining its ability to modulate gene transcription. ERK is capable of phosphorylating several cytoplasmic targets [3], but its phosphorylation results in rapid nuclear translocation [36]. Thus a single burst of ERK activation would only give rise to a transient pERK signal followed by a low steady-state level because of its rapid translocation to the nucleus. Because oscillations of ERK activity are accompanied by a regular egress of ERK from the nucleus [19], each wave of ERK activation would expose the cytoplasmic targets to a burst of ERK activity. The effect of Fus3 oscillations on cell morphogenesis is consistent with the hypothesis that the peak cytoplasmic activity of ERK during oscillations is necessary for at least a subset of cellular responses [18]. It has recently been shown that oscillations can also provide a mechanism to coordinately regulate the expression of sets of genes that are normally sensitive to the level of activated transcription factors [37]. In the case of the yeast transcription factor Crz1, extracellular calcium levels control the frequency at which it is translocated into the nucleus rather than its amplitude. Using a combination of experiments and modeling, the authors demonstrated that expression of multiple target genes was proportional to the frequency of Crz1 translocation and was independent of promoter characteristics [37]. This ‘frequency modulation’ mechanism of gene regulation requires oscillations whose frequency depends on the strength of the input signal, something that has not been observed in the case of ERK. However, it still remains possible that ERK oscillation frequencies could be modulated by secondary inputs, which in turn could affect the pattern of gene expression.
Implications of ERK oscillations on cross-talk with other signaling pathways Although receptor tyrosine kinases constitute the bestunderstood mechanism by which the ERK pathway is activated, it can also be activated in response to a variety of distinct stimuli including G-protein coupled receptors (GPCRs), oxidative stress, radiation and mechanical force. The intrinsic oscillatory dynamics of the ERK pathway would affect the manner in which it would respond to stimuli from these various sources, especially if they displayed pulsatile or oscillatory behavior as well. For example, the ERK pathway has been shown to respond to pulsatile gonadotrophin-releasing hormone (GnRH) secretion and calcium oscillations and that negative feedback control through ERK is involved in frequency decoding [38,39]. It has also been shown Current Opinion in Genetics & Development 2010, 20:650–655
654 Genetics of system biology
through a theoretical analysis that multilevel cascades such as the ERK pathway can act as band-pass filters that respond optimally to calcium signals of a particular frequency [40,41]. However, these studies have all considered an ERK pathway that is not in itself oscillatory. When we include the notion of ERK oscillations into this picture it raises the possibility that the ERK pathway could be optimally stimulated at its natural or resonant frequency. In this regard, calcium has been shown to oscillate with a time period of 10–30 min in mammalian oocytes undergoing fertilization [42,43]. Whether significant ERK activation occurs in this scenario is not known. It has also been shown that cross-talk between the ERK and other pathways occurs at multiple points [10]. If the activity of these other pathways also underwent oscillatory activity, then there would be the potential to conditionally integrate multiple signaling pathways.
Conclusions The study of the oscillatory properties of the ERK pathway has just begun and thus its significance in the broader context of cell signaling is not clear. It might seem remarkable that oscillatory behavior of such a well-studied pathway has not been characterized until recently, but this was because of the need for single cell measurements or biochemical techniques of sufficient specificity and temporal resolution that have only recently been developed. Oscillations clearly provide many opportunities for the selective regulation of gene expression and cell behavior. The ability of multiple signaling pathways to integrate information will clearly depend on both their oscillatory properties and how these oscillations can be modulated. Understanding these complex dynamic systems will require the use of realistic mathematical models, which will also drive developments in the emerging field of systems biology.
References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as: of special interest of outstanding interest 1.
Fambrough D, McClure K, Kazlauskas A, Lander ES: Diverse signaling pathways activated by growth factor receptors induce broadly overlapping, rather than independent, sets of genes. Cell 1999, 97:727-741.
2.
Marshall CJ: Specificity of receptor tyrosine kinase signaling: transient versus sustained extracellular signal-regulated kinase activation. Cell 1995, 80:179-185.
3.
von Kriegsheim A, Baiocchi D, Birtwistle M, Sumpton D, Bienvenut W, Morrice N, Yamada K, Lamond A, Kalna G, Orton R et al.: Cell fate decisions are specified by the dynamic ERK interactome. Nat Cell Biol 2009, 11:1458-1464. This study defined the proteins that differentially interacted with activated ERK when cells were stimulated with either EGF or NGF. They established that differences in signaling dynamics were due to a distributed control system rather than a single master switch. 4.
Shaul YD, Seger R: The MEK/ERK cascade: from signaling specificity to diverse functions. Biochim Biophys Acta 2007, 1773:1213-1226.
Current Opinion in Genetics & Development 2010, 20:650–655
5.
Langlois WJ, Sasaoka T, Saltiel AR, Olefsky JM: Negative feedback regulation and desensitization of insulin- and epidermal growth factor-stimulated p21ras activation. J Biol Chem 1995, 270:25320-25323.
6.
Dong C, Waters SB, Holt KH, Pessin JE: SOS phosphorylation and disassociation of the Grb2-SOS complex by the ERK and JNK signaling pathways. J Biol Chem 1996, 271:6328-6332.
7.
Dougherty MK, Muller J, Ritt DA, Zhou M, Zhou XZ, Copeland TD, Conrads TP, Veenstra TD, Lu KP, Morrison DK: Regulation of Raf1 by direct feedback phosphorylation. Mol Cell 2005, 17:215-224.
8.
Patterson KI, Brummer T, O’Brien PM, Daly RJ: Dual-specificity phosphatases: critical regulators with diverse cellular targets. Biochem J 2009, 418:475-489.
9.
Shin SY, Rath O, Choo SM, Fee F, McFerran B, Kolch W, Cho KH: Positive- and negative-feedback regulations coordinate the dynamic behavior of the Ras-Raf-MEK-ERK signal transduction pathway. J Cell Sci 2009, 122:425-435.
10. Joslin EJ, Shankaran H, Opresko LK, Bollinger N, Lauffenburger DA, Wiley HS: Structure of the EGF receptor transactivation circuit integrates multiple signals with cell context. Mol Biosyst 2010, 6:1293-1306. 11. Schulze A, Lehmann K, Jefferies HB, McMahon M, Downward J: Analysis of the transcriptional program induced by Raf in epithelial cells. Genes Dev 2001, 15:981-994. 12. Chickarmane V, Kholodenko BN, Sauro HM: Oscillatory dynamics arising from competitive inhibition and multisite phosphorylation. J Theor Biol 2007, 244:68-76. 13. Markevich NI, Hoek JB, Kholodenko BN: Signaling switches and bistability arising from multisite phosphorylation in protein kinase cascades. J Cell Biol 2004, 164:353-359. 14. Kholodenko BN: Negative feedback and ultrasensitivity can bring about oscillations in the mitogen-activated protein kinase cascades. Eur J Biochem 2000, 267:1583-1588. 15. Qiao L, Nachbar RB, Kevrekidis IG, Shvartsman SY: Bistability and oscillations in the Huang-Ferrell model of MAPK signaling. PLoS Comput Biol 2007, 3:1819-1826. 16. Foschi M, Chari S, Dunn MJ, Sorokin A: Biphasic activation of p21ras by endothelin-1 sequentially activates the ERK cascade and phosphatidylinositol 3-kinase. EMBO J 1997, 16:6439-6451. 17. Nakayama K, Satoh T, Igari A, Kageyama R, Nishida E: FGF induces oscillations of Hes1 expression and Ras/ERK activation. Curr Biol 2008, 18:R332-R334. This paper demonstrates that FGF can induce multiple waves of ERK activity and Hes1 expression in cultured mouse cells, although the mechanisms of the oscillations were not fully identified. 18. Hilioti Z, Sabbagh W, Paliwal S, Bergmann A, Goncalves MD, Bardwell L, Levchenko A: Oscillatory phosphorylation of yeast Fus3 MAP kinase controls periodic gene expression and morphogenesis. Curr Biol 2008, 18:1700-1706. This paper was the first to demonstrate oscillations in the activity of the ERK homolog Fus3 in yeast. Using different mutant strains, the authors showed that the oscillations required the transcription of multiple negative feedback regulators of the Fus3 pathway. 19. Shankaran H, Ippolito DL, Chrisler WB, Resat H, Bollinger N, Opresko LK, Wiley HS: Rapid and sustained nuclearcytoplasmic ERK oscillations induced by epidermal growth factor. Mol Syst Biol 2009:5. Using a fusion protein of ERK and GFP, the authors demonstrated sustained oscillations of actived ERK between the cytoplasm and nucleus of cells with a periodicity of approximately 15 min. This is the first study to demonstrate sustained ERK oscillations due to negative biochemical feedback control. 20. Cohen-Saidon C, Cohen AA, Sigal A, Liron Y, Alon U: Dynamics and variability of ERK2 response to EGF in individual living cells. Mol Cell 2009, 36:885-893. Using fluorescent tagging at the endogenous ERK chromosomal location, the authors showed multiple cycles of ERK cytoplasmic-nuclear translocation in a human non-small cell lung cancer cell line stimulated with EGF. www.sciencedirect.com
Oscillations in ERK pathway Shankaran and Wiley 655
21. Tsai TY, Choi YS, Ma W, Pomerening JR, Tang C, Ferrell JE Jr: Robust, tunable biological oscillations from interlinked positive and negative feedback loops. Science 2008, 321:126-129. 22. Shimojo H, Ohtsuka T, Kageyama R: Oscillations in notch signaling regulate maintenance of neural progenitors. Neuron 2008, 58:52-64. 23. Dequeant ML, Glynn E, Gaudenz K, Wahl M, Chen J, Mushegian A, Pourquie O: A complex oscillating network of signaling genes underlies the mouse segmentation clock. Science 2006, 314:1595-1598. 24. Niwa Y, Masamizu Y, Liu T, Nakayama R, Deng CX, Kageyama R: The initiation and propagation of Hes7 oscillation are cooperatively regulated by Fgf and notch signaling in the somite segmentation clock. Dev Cell 2007, 13:298-304. 25. Monk NA: Oscillatory expression of Hes1, p53, and NF-kappaB driven by transcriptional time delays. Curr Biol 2003, 13:1409-1413. 26. Colella M, Grisan F, Robert V, Turner JD, Thomas AP, Pozzan T: Ca2+ oscillation frequency decoding in cardiac cell hypertrophy: role of calcineurin/NFAT as Ca2+ signal integrators. Proc Natl Acad Sci USA 2008, 105:2859-2864. 27. Tomida T, Hirose K, Takizawa A, Shibasaki F, Iino M: NFAT functions as a working memory of Ca2+ signals in decoding Ca2+ oscillation. EMBO J 2003, 22:3825-3832. 28. Dolmetsch RE, Xu K, Lewis RS: Calcium oscillations increase the efficiency and specificity of gene expression. Nature 1998, 392:933-936. 29. De Koninck P, Schulman H: Sensitivity of CaM kinase II to the frequency of Ca2+ oscillations. Science 1998, 279: 227-230. 30. Salazar C, Politi AZ, Hofer T: Decoding of calcium oscillations by phosphorylation cycles: analytic results. Biophys J 2008, 94:1203-1215. 31. Murphy LO, MacKeigan JP, Blenis J: A network of immediate early gene products propagates subtle differences in mitogen-activated protein kinase signal amplitude and duration. Mol Cell Biol 2004, 24:144-153.
34. Legewie S, Herzel H, Westerhoff HV, Bluthgen N: Recurrent design patterns in the feedback regulation of the mammalian signalling network. Mol Syst Biol 2008, 4:190. 35. Yang E, van Nimwegen E, Zavolan M, Rajewsky N, Schroeder M, Magnasco M, Darnell JE Jr: Decay rates of human mRNAs: correlation with functional characteristics and sequence attributes. Genome Res 2003, 13:1863-1872. 36. Lenormand P, Sardet C, Pages G, L’Allemain G, Brunet A, Pouyssegur J: Growth factors induce nuclear translocation of MAP kinases (p42mapk and p44mapk) but not of their activator MAP kinase kinase (p45mapkk) in fibroblasts. J Cell Biol 1993, 122:1079-1088. 37. Cai L, Dalal CK, Elowitz MB: Frequency-modulated nuclear localization bursts coordinate gene regulation. Nature 2008, 455:485-490. Using a combination of modeling and experiments, the authors demonstrated that signal strength can be encoded in the frequency with which genes see a transcription factor and that this ‘frequency modulation’ mode of regulation can control the expression of hundreds of genes in a linearly proportional manner. 38. Kupzig S, Walker SA, Cullen PJ: The frequencies of calcium oscillations are optimized for efficient calcium-mediated activation of Ras and the ERK/MAPK cascade. Proc Natl Acad Sci USA 2005, 102:7577-7582. 39. Lim S, Pnueli L, Tan JH, Naor Z, Rajagopal G, Melamed P: Negative feedback governs gonadotrope frequency-decoding of gonadotropin releasing hormone pulse-frequency. PLoS One 2009, 4:e7244. Using a combination of experimental and computational methods, the study investigated the role of negative feedback regulation of ERK on the ability of oscillating levels of gonadotrophic-releasing hormone to stimulate specific gene expression. The authors established that negative feedback through ERK was an essential aspect of the frequency decoding mechanism. 40. Marhl M, Perc M, Schuster S: Selective regulation of cellular processes via protein cascades acting as band-pass filters for time-limited oscillations. FEBS Lett 2005, 579:5461-5465. 41. Marhl M, Perc M, Schuster S: A minimal model for decoding of time-limited Ca2+ oscillations. Biophys Chem 2006, 120:161-167.
32. Murphy LO, Smith S, Chen RH, Fingar DC, Blenis J: Molecular interpretation of ERK signal duration by immediate early gene products. Nat Cell Biol 2002, 4:556-564.
42. Campbell K, Swann K: Ca2+ oscillations stimulate an ATP increase during fertilization of mouse eggs. Dev Biol 2006, 298:225-233.
33. Mackeigan JP, Murphy LO, Dimitri CA, Blenis J: Graded mitogenactivated protein kinase activity precedes switch-like c-Fos induction in mammalian cells. Mol Cell Biol 2005, 25:4676-4682.
43. Saunders CM, Larman MG, Parrington J, Cox LJ, Royse J, Blayney LM, Swann K, Lai FA: PLC zeta: a sperm-specific trigger of Ca(2+) oscillations in eggs and embryo development. Development 2002, 129:3533-3544.
www.sciencedirect.com
Current Opinion in Genetics & Development 2010, 20:650–655
Oscillatory dynamics of the extracellular signal-regulated kinase pathway Harish Shankaran1 and H Steven Wiley2 The extracellular signal-regulated kinase (ERK) pathway is a central signaling pathway in development and disease and is regulated by multiple negative and positive feedback loops. Recent studies have shown negative feedback from ERK to upstream regulators can give rise to biochemical oscillations with a periodicity of between 15 and 30 min. Feedback due to the stimulated transcription of negative regulators of the ERK pathway can also give rise to transcriptional oscillations with a periodicity of one to two hours. The biological significance of these oscillations is not clear, but recent evidence suggests that transcriptional oscillations participate in developmental processes, such as somite formation. Biochemical oscillations are more enigmatic, but could provide a mechanism for encoding different types of inputs into a common signaling pathway. Addresses 1 Computational Biology and Bioinformatics, Pacific Northwest National Laboratory, 902 Battelle Boulevard, Richland, WA 99352, USA 2 Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, 902 Battelle Boulevard, Richland, WA 99352, USA Corresponding author: Wiley, H Steven ([email protected])
Current Opinion in Genetics & Development 2010, 20:650–655 This review comes from a themed issue on Genetics of system biology Edited by Jeff Hasty, Alex Hoffmann and Susan Golden
0959-437X/$ – see front matter # 2010 Elsevier Ltd. All rights reserved. DOI 10.1016/j.gde.2010.08.002
One potential answer to how receptors generate a diversity of responses through common signaling pathways is through the relative magnitude by which different pathways are stimulated. Another potential answer is through both the magnitude and duration of signals generated through a single pathway. A good example of the latter process is the differential induction of cell proliferation in PC12 cells through the extracellular signal-regulated kinase (ERK) pathway. In this case, the addition of epidermal growth factor (EGF) induces transient activation of ERK whereas nerve growth factor (NGF) induces a prolonged activation. The ability of EGF to induce cell proliferation and NGF to induce cell differentiation has been ascribed to such divergent activation kinetics [2]. Although the underlying molecular mechanisms responsible for the cellular response are currently obscure, the source of the differential ERK activation profiles appears to be due to the pattern of feedback circuits used by the EGF versus NGF receptor systems [3]. The ERK pathway is a central regulator in a wide variety of different cellular responses, including proliferation, differentiation, migration and apoptosis. It was one of the first kinase signaling cascades identified and is a downstream mediator of Ras activity, which is frequently associated with many human cancers. The structure of the ERK cascade is that of a three-tier, high-gain amplifier with activation of the first kinase (Raf), stimulating the phosphorylation of a second kinase (MEK), leading to the activation of the mediator kinase ERK. Following its activation, ERK translocates to the nucleus where it phosphorylates transcription factors responsible for many of the biological effects of the associated receptor pathways (for a review of the ERK pathway, see [4]).
Introduction One of the most fundamental questions in cell and developmental biology is how information is encoded in the pattern of hormones, growth factors and environmental cues that a cell receives. Specific cell surface receptors are known to be the initial step by which cells detect many signals, but receptor-binding events must be translated into the downstream signals needed to generate appropriate cellular responses. Intensive molecular studies over the last several decades have shown that most receptors activate only a handful of different signaling pathways [1]. However, the responses resulting from the activation of different receptors are incredibly diverse. This gives rise to the obvious question of how a wide diversity of different responses can be generated by activating only a few pathways. Current Opinion in Genetics & Development 2010, 20:650–655
As one would expect for such a central player in signal transduction, the ERK pathway is regulated extensively, primarily through negative feedback mechanisms (see Figure 1). Early kinetic studies showed that the protein SOS, which couples the ERK pathway to upstream receptors through Ras, is rapidly phosphorylated and attenuated by ERK [5,6]. In the ensuing years, ERK has been shown to negatively regulate Raf by direct phosphorylation as well [7]. The activation of the ERK pathway also rapidly induces the expression of phosphatases that can directly inactivate ERK, notably DUSP1, DUSP 4 and DUSP6 [8]. ERK can also stimulate positive feedback of its pathway indirectly by either inactivating negative regulators of the pathway, such as RKIP [9], or by stimulating ligand shedding that can activate receptors www.sciencedirect.com
Oscillations in ERK pathway Shankaran and Wiley 651
Figure 1
Positive and negative feedback of the extracellular signal-regulated kinase (ERK) pathway in mammalian cells. The ERK pathway can be activated by multiple factors, such as hepatocyte growth factor (HGF) or transforming growth factor alpha (TGFa) which work through their cognate receptors (MET and the EGFR, respectively). The receptors phosphorylate and bind adaptor proteins, such as Gab1, Shc and Grb2, which can associate with Sos1 or Sos2. These complexes activate Ras, which then stimulate the activity of the Raf, MEK and ERK cascade. Negative feedback (red lines) can ensue from the phosphorylation of either Sos1 or Raf by ERK, or the induction of negative feedback inhibitors, such as the DUSPs through phosphorylation of transcription factors, such as AP1. Positive feedback (blue lines) can arise from the ERK-induced shedding of ligands, such as TGFa, the induced synthesis of the ligands, or the inactivation by phosphorylation of negative regulators of the ERK pathway, such as RKIP. The times of the different feedback processes (indicated above the lines) are distinct and can be used to identify which mechanisms are operating.
upstream of the ERK pathway [10]. ERK activity can also stimulate the transcription of those ligand genes, thus promoting long-term positive effects on ERK signaling [11].
transcriptional feedback effects. Here we refer to oscillations that are inherent to the ERK biochemical pathway as biochemical oscillations to distinguish them from transcriptional oscillations that require ERK-mediated gene expression.
Oscillatory behavior of the ERK pathway Because of the high gain of the central ERK pathway, it has the property of ultrasensitivity (nonlinear input– output behavior). An ultrasensitive system combined with negative feedback and time delays can give rise to negative feedback oscillations. The presence of both positive and negative feedback can result in relaxation oscillations. The numerous positive and negative feedback loops that act on the ERK pathway can thus potentially generate multiple types of oscillatory behavior. In addition to such feedback regulation, studies have also shown that sequestration of kinases and phosphatases by their substrates could potentially generate oscillations by creating ‘implicit’ positive and negative feedback loops within the multilevel MAPK cascade [12,13]. Theoretical studies have proposed that the central ERK pathway itself could exhibit both negative feedback [14] and relaxation oscillations [12,15] even in the absence of www.sciencedirect.com
Oscillatory behavior of the ERK pathway has been an intriguing possibility because of its potential to encode additional levels of information in this important signaling pathway. However, until recently, there had been no experimental evidence that the ERK pathway was capable of generating sustained oscillations. Several papers have demonstrated two successive peaks in the activation of the upstream mediator Ras due to feedback inhibition of SOS [16] but this did not cause oscillations of ERK, apparently because of the simultaneous strong induction of an ERK-specific phosphatase (DUSP1). A more recent paper has demonstrated several cycles of Ras activation accompanied by ERK phosphorylation and Hes1 gene expression in mouse cells following addition of FGF [17]. The authors suggest that these oscillations were due to negative feedback inhibition of SOS by phosphorylated ERK. However, the relatively long time Current Opinion in Genetics & Development 2010, 20:650–655
652 Genetics of system biology
Figure 2
Oscillations of nuclear ERK following stimulation of epithelial cells with EGF. Cells expressing an ERK-GFP fusion protein were stimulated with EGF at the indicated time and the levels of nuclear ERK were followed by automated image analysis at 1 min intervals, as described in [19]. Shown are the results from four individual cells expressing different levels of ERK-GFP.
period of the ERK phosphorylation cycles (approximately 100 min) are much longer than the periodicity described for SOS phosphorylation (<30 min) and are more consistent with negative feedback through the transcriptionally induced expression of feedback inhibitors, such as the DUSPs. The observed oscillations in SOS activity could thus be a secondary effect of the transcriptional oscillation in ERK activity.
its phosphorylation. Using a similar imaging approach, Cohen-Saidon et al. [20] observed two successive peaks of ERK nuclear translocation in a lung cancer cell line with a periodicity of approximately 30 min. Both studies observed variability in the fraction of cells within a population that underwent oscillations, but saw consistent properties of the waveform itself as well as its periodicity.
ERK oscillations mediated by transcriptionally induced feedback inhibitors have recently been convincingly demonstrated by Hilioti et al. [18], using yeast stimulated by mating pheromone. In this case, activation of a yeast homolog of ERK, Fus3, caused the synthesis of the negative regulators Sst2 and Msg5, resulting in oscillations of Fus3 phosphorylation, gene expression and mating projections. The authors proposed that oscillation-driven mating projections could facilitate yeast mating in shallow or noisy signal gradients.
By using an automated image analysis system to quantify the properties of the ERK translocation oscillations, Shankaran et al. were able to determine how modifying multiple parameters, such as EGF dose, cell density and ERK expression levels, changed oscillation characteristics. They found that the frequency of ERK oscillations was insensitive to experimental conditions, although the fraction of cells that oscillated was strongly dependent on cell density. Mathematical analysis showed that an explicit negative feedback loop between ERK and its upstream activators, such as SOS, was required to generate oscillations with the observed properties [19]. The analysis showed that the biochemical oscillations of ERK could be best characterized as negative feedback oscillations rather than relaxation oscillations.
Recently, compelling evidence of rapid biochemical oscillations in ERK activity was obtained through live cell imaging studies [19]. Instead of directly monitoring ERK phosphorylation, these studies followed the translocation of ERK from the cytoplasm into the nucleus following its activation. This allowed ERK dynamics to be continuously recorded for many hours. Using this approach, Shankaran et al. [19] demonstrated that the addition of EGF to epithelial cells resulted in the oscillation of ERK between the nucleus and cytoplasm with a periodicity of approximately 15 min (see Figure 2). Biochemical studies confirmed that the oscillations in ERK translocation were correlated with oscillations in Current Opinion in Genetics & Development 2010, 20:650–655
Relaxation oscillators are believed to be easier to evolve and provide the added benefit of a tunable frequency [21]. Why then does the ERK pathway employ a biochemical mechanism that results in a constant frequency? It may be possible that this feature is critical for the ability of the ERK pathway to encode and decode information. For example, the transcription systems that decode ERK oscillations may act as band-pass filters that operate www.sciencedirect.com
Oscillations in ERK pathway Shankaran and Wiley 653
optimally when stimulated with an appropriate frequency. Another interesting possibility is that frequency robustness is itself required to encode information when a system contains multiple levels of feedback control. For example, if transcriptional feedback mechanisms could change the frequency of the oscillation, then any potential information in the oscillation would be obliterated.
Potential functions of ERK oscillations ERK is a potent stimulator of gene transcription and so a possible function of slow-transcriptional ERK oscillations could be to drive periodic protein expression. Oscillations in the expression of transcription factors, such as Hes1, have been shown to be important in the maintenance of neural progenitors in the embryonic brain [22] and for controlling somite formation [23]. For example, it has been shown that in mouse, a segmentation clock with a periodicity of approximately two hours drives the pattern of the spine. The mechanism underlying the clock apparently involves coordinate regulated oscillations of the FGF, Notch and Wnt pathways [23]. This coordination most likely occurs through the induction of DUSPs that modulate the activity of the FGF-stimulated ERK pathway [24]. Postulating that oscillations arise from induced transcription of negative regulators has been an appealing concept because of the intrinsic time delay of protein translation [25]. However, multiple ERK-induced negative regulators have been shown to oscillate, making it difficult to determine whether they are causal to transcriptional oscillations of ERK activity or are a consequence. It is possible that rapid biochemical ERK oscillations are a means to selectively turn on a subset of genes. In the case of calcium oscillations, it has been shown that the proteins and genes that respond optimally to oscillations are the ones that can integrate the signal [26–29]. A target that integrates a signal should be induced in a sufficiently sensitive manner so that it can respond to small signal amplitudes, and should be stable enough to ‘remember’ the induction during the off-phase of the signal [30]. Several immediate-early-genes (IEGs) that respond to ERK signaling such as c-Fos, Fra-2, Fra-1 and c-Myc require long signal durations and sufficient signal strength for their induction [31,32]. The levels of activated ERK necessary to induce genes, such as c-Fos are considerably higher than those needed to activate oscillations [33]. Further, with regards to targets that would remain on during the signal’s off phase, most of the negative feedback regulators of cell signaling pathways tend to have extremely short mRNA and protein half-lives [34,35]. For example, DUSP6, which is induced by ERK activation, has an mRNA half-life of 18 min. It is possible that such negative regulators are not effectively induced in response to rapid biochemical ERK oscillations. This would allow the biochemical oscillation to persist for multiple hours, thereby providing sufficient signal www.sciencedirect.com
duration to activate other target genes. Whether there is a subset of ERK-responsive genes that are capable of integrating low amplitude oscillatory ERK signaling is not currently known. Oscillations could also be a mechanism for increasing the mean level of activated ERK in the cytoplasm while still retaining its ability to modulate gene transcription. ERK is capable of phosphorylating several cytoplasmic targets [3], but its phosphorylation results in rapid nuclear translocation [36]. Thus a single burst of ERK activation would only give rise to a transient pERK signal followed by a low steady-state level because of its rapid translocation to the nucleus. Because oscillations of ERK activity are accompanied by a regular egress of ERK from the nucleus [19], each wave of ERK activation would expose the cytoplasmic targets to a burst of ERK activity. The effect of Fus3 oscillations on cell morphogenesis is consistent with the hypothesis that the peak cytoplasmic activity of ERK during oscillations is necessary for at least a subset of cellular responses [18]. It has recently been shown that oscillations can also provide a mechanism to coordinately regulate the expression of sets of genes that are normally sensitive to the level of activated transcription factors [37]. In the case of the yeast transcription factor Crz1, extracellular calcium levels control the frequency at which it is translocated into the nucleus rather than its amplitude. Using a combination of experiments and modeling, the authors demonstrated that expression of multiple target genes was proportional to the frequency of Crz1 translocation and was independent of promoter characteristics [37]. This ‘frequency modulation’ mechanism of gene regulation requires oscillations whose frequency depends on the strength of the input signal, something that has not been observed in the case of ERK. However, it still remains possible that ERK oscillation frequencies could be modulated by secondary inputs, which in turn could affect the pattern of gene expression.
Implications of ERK oscillations on cross-talk with other signaling pathways Although receptor tyrosine kinases constitute the bestunderstood mechanism by which the ERK pathway is activated, it can also be activated in response to a variety of distinct stimuli including G-protein coupled receptors (GPCRs), oxidative stress, radiation and mechanical force. The intrinsic oscillatory dynamics of the ERK pathway would affect the manner in which it would respond to stimuli from these various sources, especially if they displayed pulsatile or oscillatory behavior as well. For example, the ERK pathway has been shown to respond to pulsatile gonadotrophin-releasing hormone (GnRH) secretion and calcium oscillations and that negative feedback control through ERK is involved in frequency decoding [38,39]. It has also been shown Current Opinion in Genetics & Development 2010, 20:650–655
654 Genetics of system biology
through a theoretical analysis that multilevel cascades such as the ERK pathway can act as band-pass filters that respond optimally to calcium signals of a particular frequency [40,41]. However, these studies have all considered an ERK pathway that is not in itself oscillatory. When we include the notion of ERK oscillations into this picture it raises the possibility that the ERK pathway could be optimally stimulated at its natural or resonant frequency. In this regard, calcium has been shown to oscillate with a time period of 10–30 min in mammalian oocytes undergoing fertilization [42,43]. Whether significant ERK activation occurs in this scenario is not known. It has also been shown that cross-talk between the ERK and other pathways occurs at multiple points [10]. If the activity of these other pathways also underwent oscillatory activity, then there would be the potential to conditionally integrate multiple signaling pathways.
Conclusions The study of the oscillatory properties of the ERK pathway has just begun and thus its significance in the broader context of cell signaling is not clear. It might seem remarkable that oscillatory behavior of such a well-studied pathway has not been characterized until recently, but this was because of the need for single cell measurements or biochemical techniques of sufficient specificity and temporal resolution that have only recently been developed. Oscillations clearly provide many opportunities for the selective regulation of gene expression and cell behavior. The ability of multiple signaling pathways to integrate information will clearly depend on both their oscillatory properties and how these oscillations can be modulated. Understanding these complex dynamic systems will require the use of realistic mathematical models, which will also drive developments in the emerging field of systems biology.
References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as: of special interest of outstanding interest 1.
Fambrough D, McClure K, Kazlauskas A, Lander ES: Diverse signaling pathways activated by growth factor receptors induce broadly overlapping, rather than independent, sets of genes. Cell 1999, 97:727-741.
2.
Marshall CJ: Specificity of receptor tyrosine kinase signaling: transient versus sustained extracellular signal-regulated kinase activation. Cell 1995, 80:179-185.
3.
von Kriegsheim A, Baiocchi D, Birtwistle M, Sumpton D, Bienvenut W, Morrice N, Yamada K, Lamond A, Kalna G, Orton R et al.: Cell fate decisions are specified by the dynamic ERK interactome. Nat Cell Biol 2009, 11:1458-1464. This study defined the proteins that differentially interacted with activated ERK when cells were stimulated with either EGF or NGF. They established that differences in signaling dynamics were due to a distributed control system rather than a single master switch. 4.
Shaul YD, Seger R: The MEK/ERK cascade: from signaling specificity to diverse functions. Biochim Biophys Acta 2007, 1773:1213-1226.
Current Opinion in Genetics & Development 2010, 20:650–655
5.
Langlois WJ, Sasaoka T, Saltiel AR, Olefsky JM: Negative feedback regulation and desensitization of insulin- and epidermal growth factor-stimulated p21ras activation. J Biol Chem 1995, 270:25320-25323.
6.
Dong C, Waters SB, Holt KH, Pessin JE: SOS phosphorylation and disassociation of the Grb2-SOS complex by the ERK and JNK signaling pathways. J Biol Chem 1996, 271:6328-6332.
7.
Dougherty MK, Muller J, Ritt DA, Zhou M, Zhou XZ, Copeland TD, Conrads TP, Veenstra TD, Lu KP, Morrison DK: Regulation of Raf1 by direct feedback phosphorylation. Mol Cell 2005, 17:215-224.
8.
Patterson KI, Brummer T, O’Brien PM, Daly RJ: Dual-specificity phosphatases: critical regulators with diverse cellular targets. Biochem J 2009, 418:475-489.
9.
Shin SY, Rath O, Choo SM, Fee F, McFerran B, Kolch W, Cho KH: Positive- and negative-feedback regulations coordinate the dynamic behavior of the Ras-Raf-MEK-ERK signal transduction pathway. J Cell Sci 2009, 122:425-435.
10. Joslin EJ, Shankaran H, Opresko LK, Bollinger N, Lauffenburger DA, Wiley HS: Structure of the EGF receptor transactivation circuit integrates multiple signals with cell context. Mol Biosyst 2010, 6:1293-1306. 11. Schulze A, Lehmann K, Jefferies HB, McMahon M, Downward J: Analysis of the transcriptional program induced by Raf in epithelial cells. Genes Dev 2001, 15:981-994. 12. Chickarmane V, Kholodenko BN, Sauro HM: Oscillatory dynamics arising from competitive inhibition and multisite phosphorylation. J Theor Biol 2007, 244:68-76. 13. Markevich NI, Hoek JB, Kholodenko BN: Signaling switches and bistability arising from multisite phosphorylation in protein kinase cascades. J Cell Biol 2004, 164:353-359. 14. Kholodenko BN: Negative feedback and ultrasensitivity can bring about oscillations in the mitogen-activated protein kinase cascades. Eur J Biochem 2000, 267:1583-1588. 15. Qiao L, Nachbar RB, Kevrekidis IG, Shvartsman SY: Bistability and oscillations in the Huang-Ferrell model of MAPK signaling. PLoS Comput Biol 2007, 3:1819-1826. 16. Foschi M, Chari S, Dunn MJ, Sorokin A: Biphasic activation of p21ras by endothelin-1 sequentially activates the ERK cascade and phosphatidylinositol 3-kinase. EMBO J 1997, 16:6439-6451. 17. Nakayama K, Satoh T, Igari A, Kageyama R, Nishida E: FGF induces oscillations of Hes1 expression and Ras/ERK activation. Curr Biol 2008, 18:R332-R334. This paper demonstrates that FGF can induce multiple waves of ERK activity and Hes1 expression in cultured mouse cells, although the mechanisms of the oscillations were not fully identified. 18. Hilioti Z, Sabbagh W, Paliwal S, Bergmann A, Goncalves MD, Bardwell L, Levchenko A: Oscillatory phosphorylation of yeast Fus3 MAP kinase controls periodic gene expression and morphogenesis. Curr Biol 2008, 18:1700-1706. This paper was the first to demonstrate oscillations in the activity of the ERK homolog Fus3 in yeast. Using different mutant strains, the authors showed that the oscillations required the transcription of multiple negative feedback regulators of the Fus3 pathway. 19. Shankaran H, Ippolito DL, Chrisler WB, Resat H, Bollinger N, Opresko LK, Wiley HS: Rapid and sustained nuclearcytoplasmic ERK oscillations induced by epidermal growth factor. Mol Syst Biol 2009:5. Using a fusion protein of ERK and GFP, the authors demonstrated sustained oscillations of actived ERK between the cytoplasm and nucleus of cells with a periodicity of approximately 15 min. This is the first study to demonstrate sustained ERK oscillations due to negative biochemical feedback control. 20. Cohen-Saidon C, Cohen AA, Sigal A, Liron Y, Alon U: Dynamics and variability of ERK2 response to EGF in individual living cells. Mol Cell 2009, 36:885-893. Using fluorescent tagging at the endogenous ERK chromosomal location, the authors showed multiple cycles of ERK cytoplasmic-nuclear translocation in a human non-small cell lung cancer cell line stimulated with EGF. www.sciencedirect.com
Oscillations in ERK pathway Shankaran and Wiley 655
21. Tsai TY, Choi YS, Ma W, Pomerening JR, Tang C, Ferrell JE Jr: Robust, tunable biological oscillations from interlinked positive and negative feedback loops. Science 2008, 321:126-129. 22. Shimojo H, Ohtsuka T, Kageyama R: Oscillations in notch signaling regulate maintenance of neural progenitors. Neuron 2008, 58:52-64. 23. Dequeant ML, Glynn E, Gaudenz K, Wahl M, Chen J, Mushegian A, Pourquie O: A complex oscillating network of signaling genes underlies the mouse segmentation clock. Science 2006, 314:1595-1598. 24. Niwa Y, Masamizu Y, Liu T, Nakayama R, Deng CX, Kageyama R: The initiation and propagation of Hes7 oscillation are cooperatively regulated by Fgf and notch signaling in the somite segmentation clock. Dev Cell 2007, 13:298-304. 25. Monk NA: Oscillatory expression of Hes1, p53, and NF-kappaB driven by transcriptional time delays. Curr Biol 2003, 13:1409-1413. 26. Colella M, Grisan F, Robert V, Turner JD, Thomas AP, Pozzan T: Ca2+ oscillation frequency decoding in cardiac cell hypertrophy: role of calcineurin/NFAT as Ca2+ signal integrators. Proc Natl Acad Sci USA 2008, 105:2859-2864. 27. Tomida T, Hirose K, Takizawa A, Shibasaki F, Iino M: NFAT functions as a working memory of Ca2+ signals in decoding Ca2+ oscillation. EMBO J 2003, 22:3825-3832. 28. Dolmetsch RE, Xu K, Lewis RS: Calcium oscillations increase the efficiency and specificity of gene expression. Nature 1998, 392:933-936. 29. De Koninck P, Schulman H: Sensitivity of CaM kinase II to the frequency of Ca2+ oscillations. Science 1998, 279: 227-230. 30. Salazar C, Politi AZ, Hofer T: Decoding of calcium oscillations by phosphorylation cycles: analytic results. Biophys J 2008, 94:1203-1215. 31. Murphy LO, MacKeigan JP, Blenis J: A network of immediate early gene products propagates subtle differences in mitogen-activated protein kinase signal amplitude and duration. Mol Cell Biol 2004, 24:144-153.
34. Legewie S, Herzel H, Westerhoff HV, Bluthgen N: Recurrent design patterns in the feedback regulation of the mammalian signalling network. Mol Syst Biol 2008, 4:190. 35. Yang E, van Nimwegen E, Zavolan M, Rajewsky N, Schroeder M, Magnasco M, Darnell JE Jr: Decay rates of human mRNAs: correlation with functional characteristics and sequence attributes. Genome Res 2003, 13:1863-1872. 36. Lenormand P, Sardet C, Pages G, L’Allemain G, Brunet A, Pouyssegur J: Growth factors induce nuclear translocation of MAP kinases (p42mapk and p44mapk) but not of their activator MAP kinase kinase (p45mapkk) in fibroblasts. J Cell Biol 1993, 122:1079-1088. 37. Cai L, Dalal CK, Elowitz MB: Frequency-modulated nuclear localization bursts coordinate gene regulation. Nature 2008, 455:485-490. Using a combination of modeling and experiments, the authors demonstrated that signal strength can be encoded in the frequency with which genes see a transcription factor and that this ‘frequency modulation’ mode of regulation can control the expression of hundreds of genes in a linearly proportional manner. 38. Kupzig S, Walker SA, Cullen PJ: The frequencies of calcium oscillations are optimized for efficient calcium-mediated activation of Ras and the ERK/MAPK cascade. Proc Natl Acad Sci USA 2005, 102:7577-7582. 39. Lim S, Pnueli L, Tan JH, Naor Z, Rajagopal G, Melamed P: Negative feedback governs gonadotrope frequency-decoding of gonadotropin releasing hormone pulse-frequency. PLoS One 2009, 4:e7244. Using a combination of experimental and computational methods, the study investigated the role of negative feedback regulation of ERK on the ability of oscillating levels of gonadotrophic-releasing hormone to stimulate specific gene expression. The authors established that negative feedback through ERK was an essential aspect of the frequency decoding mechanism. 40. Marhl M, Perc M, Schuster S: Selective regulation of cellular processes via protein cascades acting as band-pass filters for time-limited oscillations. FEBS Lett 2005, 579:5461-5465. 41. Marhl M, Perc M, Schuster S: A minimal model for decoding of time-limited Ca2+ oscillations. Biophys Chem 2006, 120:161-167.
32. Murphy LO, Smith S, Chen RH, Fingar DC, Blenis J: Molecular interpretation of ERK signal duration by immediate early gene products. Nat Cell Biol 2002, 4:556-564.
42. Campbell K, Swann K: Ca2+ oscillations stimulate an ATP increase during fertilization of mouse eggs. Dev Biol 2006, 298:225-233.
33. Mackeigan JP, Murphy LO, Dimitri CA, Blenis J: Graded mitogenactivated protein kinase activity precedes switch-like c-Fos induction in mammalian cells. Mol Cell Biol 2005, 25:4676-4682.
43. Saunders CM, Larman MG, Parrington J, Cox LJ, Royse J, Blayney LM, Swann K, Lai FA: PLC zeta: a sperm-specific trigger of Ca(2+) oscillations in eggs and embryo development. Development 2002, 129:3533-3544.
www.sciencedirect.com
Current Opinion in Genetics & Development 2010, 20:650–655