Bistability in the JNK cascade

BACKGROUND: Important signaling properties, like adaptation, oscillations, and bistability, can emerge at the level of relatively simple systems of signaling proteins. Here, we have examined the quantitative properties of one well-studied signaling system, the JNK cascade. We experimentally assessed the response of JNK to a physiological stimulus (progesterone) and a pathological stress (hyperosmolar sorbitol) in Xenopus laevis oocytes, a cell type that is well-suited to the quantitative analysis of cell signaling. Our aim was to determine whether JNK responses are graded (Michaelian) in character; ultrasensitive in character, resembling the responses of cooperative enzymes; or bistable and all-or-none in character. RESULTS: The responses of JNK to both progesterone and sorbitol were found to be essentially all-or-none. Individual oocytes had either very high or very low JNK activities, with few oocytes possessing intermediate levels of JNK activity. Moreover, JNK activation was autocatalytic, indicating that the JNK cascade is either embedded in or downstream of a positive feedback loop. JNK also exhibited hysteresis, a form of biochemical memory, in its response to sorbitol. These findings indicate that the JNK cascade is part of a bistable signaling system in oocytes. CONCLUSIONS: In Xenopus oocytes, JNK responds to physiological and pathological stimuli in an all-or-none manner. The JNK response shows all the hallmarks of a bistable response, including strong positive feedback and hysteresis. Bistability is a recurring theme in the biochemistry of oocyte maturation and early embryogenesis; the Mos/MEK/p42 MAPK cascade also exhibits bistable responses, and the Cdc2/cyclin B system is hypothesized to be bistable as well. However, the mechanisms underpinning the positive feedback and bistability in the three cases are different, suggesting that evolution has repeatedly converged upon bistability as a way of producing digital responses.     

How robust are switches in intracellular signaling cascades?

Since all-or-none decisions of the cell are controlled by extracellular signals, cells have biochemical switches within their intracellular signaling networks. Central elements of these switches are multisite phosphorylation, enzymic saturation, and amplification by cascades. Moreover, positive feedback can contribute to switch-like behavior termed also ultrasensitivity. Here we analyse the robustness of these mechanisms exemplified by models of the three-molecule MAPK-cascade and the single-molecule Goldbeter-Koshland switch. We show that the ultrasensitivity in the MAPK-cascades is more robust against changes of the kinetic parameters than the Goldbeter-Koshland switch. If multiple parameters are changed randomly, the effects of parameter changes can compensate each other in the cascade leading to a remarkable robustness of the switch-like behavior. The different degrees of robustness can be traced back to the different mechanisms of generating ultrasensitivity. While in the Goldbeter-Koshland switch the saturation of the enzymes are crucial, in the MAPK-cascade the adjustment of working ranges determines the ultrasensitivity. Our results indicate that amplification of ultrasensitivity in cascades and multisite phosphorylation might be a design principle to achieve robust switches.     

Basic Properties of the p38 Signaling Pathway in Response to Hyperosmotic Shock

Some properties of signaling systems, like ultrasensitivity, hysteresis (a form of biochemical memory), and all-or-none responses at a single cell level, are important to understand the regulation of irreversible processes. Xenopus oocytes are a suitable cell model to study these properties. The p38 MAPK (mitogen-activated protein kinase) pathway is activated by different stress stimuli, including osmostress, and regulates multiple biological processes, from immune response to cell cycle. Recently, we have reported that activation of p38 and JNK regulate osmostress-induced apoptosis in Xenopus oocytes and that sustained activation of p38 accelerates cytochrome c release and caspase-3 activation. However, the signaling properties of p38 in response to hyperosmotic shock have not been studied. Here we show, using Xenopus oocytes as a cell model, that hyperosmotic shock activates the p38 signaling pathway with an ultrasensitive and bimodal response in a time-dependent manner, and with low hysteresis. At a single cell level, p38 activation is not well correlated with cytochrome c release 2 h after hyperosmotic shock, but a good correlation is observed at 4 h after treatment. Interestingly, cytochrome c microinjection induces p38 phosphorylation through caspase-3 activation, and caspase inhibition reduces p38 activation induced by osmostress, indicating that a positive feedback loop is engaged by hyperosmotic shock. To know the properties of the stress protein kinases activated by hyperosmotic shock will facilitate the design of computational models to predict cellular responses in human diseases caused by perturbations in fluid osmolarity.     

Mathematical modeling identifies inhibitors of apoptosis as mediators of positive feedback and bistability

The intrinsic, or mitochondrial, pathway of caspase activation is essential for apoptosis induction by various stimuli including cytotoxic stress. It depends on the cellular context, whether cytochrome c released from mitochondria induces caspase activation gradually or in an all-or-none fashion, and whether caspase activation irreversibly commits cells to apoptosis. By analyzing a quantitative kinetic model, we show that inhibition of caspase-3 (Casp3) and Casp9 by inhibitors of apoptosis (IAPs) results in an implicit positive feedback, since cleaved Casp3 augments its own activation by sequestering IAPs away from Casp9. We demonstrate that this positive feedback brings about bistability (i.e., all-or-none behaviour), and that it cooperates with Casp3-mediated feedback cleavage of Casp9 to generate irreversibility in caspase activation. Our calculations also unravel how cell-specific protein expression brings about the observed qualitative differences in caspase activation (gradual versus all-or-none and reversible versus irreversible). Finally, known regulators of the pathway are shown to efficiently shift the apoptotic threshold stimulus, suggesting that the bistable caspase cascade computes multiple inputs into an all-or-none caspase output. As cellular inhibitory proteins (e.g., IAPs) frequently inhibit consecutive intermediates in cellular signaling cascades (e.g., Casp3 and Casp9), the feedback mechanism described in this paper is likely to be a widespread principle on how cells achieve ultrasensitivity, bistability, and irreversibility.     

Compact Modeling of Allosteric Multisite Proteins: Application to a Cell Size Checkpoint

We explore a framework to model the dose response of allosteric multisite phosphorylation proteins using a single auxiliary variable. This reduction can closely replicate the steady state behavior of detailed multisite systems such as the Monod-Wyman-Changeux allosteric model or rule-based models. Optimal ultrasensitivity is obtained when the activation of an allosteric protein by its individual sites is concerted and redundant. The reduction makes this framework useful for modeling and analyzing biochemical systems in practical applications, where several multisite proteins may interact simultaneously. As an application we analyze a newly discovered checkpoint signaling pathway in budding yeast, which has been proposed to measure cell growth by monitoring signals generated at sites of plasma membrane growth. We show that the known components of this pathway can form a robust hysteretic switch. In particular, this system incorporates a signal proportional to bud growth or size, a mechanism to read the signal, and an all-or-none response triggered only when the signal reaches a threshold indicating that sufficient growth has occurred.     

Synthetic conversion of a graded receptor signal into a tunable,reversible switch

The ability to engineer an all‐or‐none cellular response to a given signaling ligand is important in applications ranging from biosensing to tissue engineering. However, synthetic gene network ‘switches’ have been limited in their applicability and tunability due to their reliance on specific components to function. Here, we present a strategy for reversible switch design that instead relies only on a robust, easily constructed network topology with two positive feedback loops and we apply the method to create highly ultrasensitive (n_H>20), bistable cellular responses to a synthetic ligand/receptor complex. Independent modulation of the two feedback strengths enables rational tuning and some decoupling of steady‐state (ultrasensitivity, signal amplitude, switching threshold, and bistability) and kinetic (rates of system activation and deactivation) response properties. Our integrated computational and synthetic biology approach elucidates design rules for building cellular switches with desired properties, which may be of utility in engineering signal‐transduction pathways.     

Robustness analysis identifies the plausible model of the Bcl-2 apoptotic switch

In this paper two competing models of the B-cell lymphoma 2 (Bcl-2) apoptotic switch were contrasted by mathematical modeling and robustness analysis. Since switch-like behaviors are required for models that attempt to explain the all-or-none decisions of apoptosis, ultrasensitivity was employed as a criterion for comparison. Our results successfully exhibit that the direct activation model operates more reliably to achieve a robust switch in cellular conditions. Moreover, by investigating the robustness of other important features of the Bcl-2 apoptotic switch (including low Bax basal activation, inhibitory role of anti-apoptotic proteins and insensitivity to small perturbations) the direct activation model was further supported. In all, we identified the direct activation model as a more plausible explanation for the Bcl-2 apoptotic switch.     

Polo-like kinase confers MPF autoamplification competence to growing Xenopus oocytes

During oogenesis, the Xenopus oocyte is blocked in prophase of meiosis I. It becomes competent to resume meiosis in response to progesterone at the end of its growing period (stage VI of oogenesis). Stage IV oocytes contain a store of inactive pre-MPF (Tyr15-phosphorylated Cdc2 bound to cyclin B2); the Cdc25 phosphatase that catalyzes Tyr15 dephosphorylation of Cdc2 is also present. However, the positive feedback loop that allows MPF autoamplification is not functional at this stage of oocyte growth. We report that when cyclin B is overexpressed in stage IV oocytes, MPF autoamplification does not occur and the newly formed cyclin B-Cdc2 complexes are inactivated by Tyr15 phosphorylation, indicating that Myt1 kinase remains active and that Cdc25 is prevented to be activated. Plx1 kinase (or polo-like kinase), which is required for Cdc25 activation and MPF autoamplification in full grown oocytes is not expressed at the protein level in small stage IV oocytes. In order to determine if Plx1 could be the missing regulator that prevents MPF autoamplification, polo kinase was overexpressed in stage IV oocytes. Under these conditions, the MPF-positive feedback loop was restored. Moreover, we show that acquisition of autoamplification competence does not require the Mos/MAPK pathway.     

Substrate competition as a source of ultrasensitivity in the inactivation of Wee1

The mitotic regulators Wee1 and Cdk1 can inactivate each other through inhibitory phosphorylations. This double-negative feedback loop is part of a bistable trigger that makes the transition into mitosis abrupt and decisive. To generate a bistable response, some component of a double-negative feedback loop must exhibit an ultrasensitive response to its upstream regulator. Here, we experimentally demonstrate that Wee1 exhibits a highly ultrasensitive response to Cdk1. Several mechanisms can, in principle, give rise to ultrasensitivity, including zero-order effects, multisite phosphorylation, and competition mechanisms. We found that the ultrasensitivity in the inactivation of Wee1 arises mainly through two competition mechanisms: competition between two sets of phosphorylation sites in Wee1 and between Wee1 and other high-affinity Cdk1 targets. Based on these findings, we were able to reconstitute a highly ultrasensitive Wee1 response with purified components. Competition provides a simple way of generating the equivalent of a highly cooperative allosteric response.     

Generation of Digital Responses in Stress Sensors

Ultrasensitivity, hysteresis (a form of biochemical memory), and all-or-none (digital) responses are important signaling properties for the control of irreversible processes and are well characterized in the c-Jun N-terminal kinase (JNK) system using Xenopus oocytes. Our aim was to study these properties in the AMP-activated protein kinase (AMPK) signaling system under stress conditions that could engage a cell death program, and compare them to the JNK responses. After characterization of Xenopus AMPK, we show here that the response to antimycin (nonapoptotic) was slightly cooperative and graded (analog) in individual oocytes, whereas the response to sorbitol (which induced cytochrome c release and caspase activation) was ultrasensitive, digital in single cells, and without hysteresis, hallmarks of a monostable system. Moreover, initial graded responses of AMPK and JNK turned into digital during a critical period for the execution of the cell death program, although single cell analysis did not show complete correlation between AMPK or JNK activation and cytochrome c release. We propose a model where the life or death decision in the cell is made by integration of multiple digital signals from stress sensors.The energy level in a cell/organism is probably the most remarkable vital constant and must be tightly regulated or cell death programs will otherwise be engaged. The AMP-activated protein kinase (AMPK)² is an energy sensor (activated by a high AMP/ATP ratio) and a homeostatic regulator of cellular ATP levels, taking central stage in orchestrating cell metabolism (1).Stress sensors should have, to function efficiently, ultrasensitive properties to respond to small changes in the important parameters of cell survival. In addition, all-or-none (digital) responses, at a single cell level, and sustained activation when the stimulus has disappeared (also named hysteresis) may be important properties of stress sensors in regulating cell death (2).Mammalian AMPK is a heterotrimeric complex consisting of a catalytic α-subunit and regulatory β- and γ-subunits. AMPK is activated by AMP in two ways, both antagonized by high concentrations of ATP. Binding of AMP to the γ-subunit induces: 1) allosteric activation that can account for a 5-fold increase in the activity, and 2) phosphorylation of the α-subunit at Thr-172 (in human), which is essential for activity and causes a much marked activation (3). The combination of the two effects causes a >1000-fold increase in kinase activity (4). The protein kinases LKB1 and CaMKKα/β have been reported to phosphorylate the AMPK α-subunit at Thr-172 (1, 5). This signaling system is a clear example of multistep sensitivity, which arises when a signal molecule (AMP) affects more than one step in a cascade (allosteric activation and phosphorylation). In addition, the AMPK upstream kinases have a very low K_m for AMPK (6), a phenomenon referred to as zero-order ultrasensitivity (7). In practice, the ultrasensitivity of a system is reflected in a stimulus/response curve with a very steep upstroke. These properties of the AMPK signaling system may be very important to maintain energy levels in the cell within narrow limits and can be useful for filtering noise. Although there is one report supporting that AMPK is an ultrasensitive system (6), no further research has considered its significance in the control of cell death as well as other important signaling properties associated with ultrasensitivity, such as hysteresis and all-or-none responses.It is well known that ultrasensitive systems embedded in a positive feedback loop have the potential to exhibit bistable behavior, switching between discrete stable steady states without being able to rest in intermediate states (8, 9). The three hallmarks of a bistable system are: 1) strong ultrasensitivity, 2) digital response at the individual cell level, and 3) hysteresis. Examples of such systems are JNK and the mitogen-activated protein kinase (MAPK) cascades, implicated in oocyte maturation and maybe in apoptosis (2, 10). The hallmarks of a monostable system are the same, but without hysteresis.Because AMPK is an energy sensor, it seems reasonable that the AMPK cascade would transmit at the individual cell level graded (analog) information about the energy status of the cell. In fact, AMPK can be considered part of a negative feedback loop because AMPK activation (by a high AMP/ATP ratio) regulates multiple steps in metabolism to restore ATP levels in the cell, which in turn down-regulates AMPK activity. However, sustained activation of AMPK by some stimuli has been considered a pro-apoptotic signal (11, 12). Because cell death is an all-or-none irreversible process, it may be more appropriate for the AMPK cascade in this situation to exhibit a digital (all-or-none) response to trigger cell death.Here, we analyze the sensitivity, the grade of hysteresis, and the single cell response of the AMPK system under different stress conditions that could engage a cell death program. We performed experiments with Xenopus oocytes, a suitable model for discerning the character of a signaling response, as has been previously reported for JNK and MAPK cascades. Based on our results, we propose that commitment to cell death will occur after integration of multiple digital responses from stress protein kinases.     

Two Distinct Mechanisms Control the Accumulation of Cyclin B1 and Mos in Xenopus Oocytes in Response to Progesterone

Progesterone-induced meiotic maturation of Xenopus oocytes requires the synthesis of new proteins, such as Mos and cyclin B. Synthesis of Mos is thought to be necessary and sufficient for meiotic maturation; however, it has recently been proposed that newly synthesized proteins binding to p34(cdc2) could be involved in a signaling pathway that triggers the activation of maturation-promoting factor. We focused our attention on cyclin B proteins because they are synthesized in response to progesterone, they bind to p34(cdc2), and their microinjection into resting oocytes induces meiotic maturation. We investigated cyclin B accumulation in response to progesterone in the absence of maturation-promoting factor-induced feedback. We report here that the cdk inhibitor p21(cip1), when microinjected into immature Xenopus oocytes, blocks germinal vesicle breakdown induced by progesterone, by maturation-promoting factor transfer, or by injection of okadaic acid. After microinjection of p21(cip1), progesterone fails to induce the activation of MAPK or p34(cdc2), and Mos does not accumulate. In contrast, the level of cyclin B1 increases normally in a manner dependent on down-regulation of cAMP-dependent protein kinase but independent of cap-ribose methylation of mRNA.     

Human aldosterone synthase: recombinant expression in E. coli and purification enables a detailed biochemical analysis of the protein on the molecular level

Recent experiments from our laboratory are consistent with the idea that hypothalamic astrocytes are critical components of the central nervous system (CNS) mediated estrogen positive feedback mechanism. The "astrocrine hypothesis" maintains that ovarian estradiol rapidly increases free cytoplasmic calcium concentrations ([Ca(2+)](i)) that facilitate progesterone synthesis in astrocytes. This hypothalamic neuroprogesterone along with the elevated estrogen from the ovaries allows for the surge release of gonadotropin-releasing hormone (GnRH) that triggers the pituitary luteinizing hormone (LH) surge. A narrow range of estradiol stimulated progesterone production supports an "off-on-off" mechanism regulating the transition from estrogen negative feedback to estrogen positive feedback, and back again. The rapidity of the [Ca(2+)](i) response and progesterone synthesis support a non-genomic, membrane-initiated signaling mechanism. In hypothalamic astrocytes, membrane-associated estrogen receptors (mERs) signal through transactivation of the metabotropic glutamate receptor type 1a (mGluR1a), implying that astrocytic function is influenced by surrounding glutamatergic nerve terminals. Although other putative mERs, such as mERβ, STX-activated mER-Gα(q), and G protein-coupled receptor 30 (GPR30), are present and participate in membrane-mediated signaling, their influence in reproduction is still obscure since female reproduction be it estrogen positive feedback or lordosis behavior requires mERα. The astrocrine hypothesis is also consistent with the well-known sexual dimorphism of estrogen positive feedback. In rodents, only post-pubertal females exhibit this positive feedback. Hypothalamic astrocytes cultured from females, but not males, responded to estradiol by increasing progesterone synthesis. Estrogen autoregulates its own signaling by regulating levels of mERα in the plasma membrane of female astrocytes. In male astrocytes, the estradiol-induced increase in mERα was attenuated, suggesting that membrane-initiated estradiol signaling (MIES) would also be blunted. Indeed, estradiol induced [Ca(2+)](i) release in male astrocytes, but not to levels required to stimulate progesterone synthesis. Investigation of this sexual differentiation was performed using hypothalamic astrocytes from post-pubertal four core genotype (FCG) mice. In this model, genetic sex is uncoupled from gonadal sex. We demonstrated that animals that developed testes (XYM and XXM) lacked estrogen positive feedback, strongly suggesting that the sexual differentiation of progesterone synthesis is driven by the sex steroid environment during early development. This article is part of a Special Issue entitled 'Neurosteroids'.     

Age-dependent changes in the hypothalamo-pituitary-ovarian axis of the laying hen.

The functional integrity of the components of the hypothalamo-pituitary-ovarian axis was examined in young and old laying hens. Ovarian function was tested by measuring the amount of progesterone released in response to an injection of LH, and pituitary function was investigated by measuring the increase in the plasma LH level after an injection of LH-RH. There were no differences between young and old birds in the response of the pituitary gland or the ovary to these stimuli. Hypothalamic function was investigated by studying the positive feedback action of a standard dose of progesterone on LH release; the positive feedback response was smaller (P less than 0.05) in old hens. It is suggested that the fall in the rate of lay in hens towards the end of their laying year is caused partly by a decrease in the response of the LH-positive feedback mechanism to progesterone.     

Negative feedback and ultrasensitivity can bring about oscillations in the mitogen-activated protein kinase cascades.

Functional organization of signal transduction into protein phosphorylation cascades, such as the mitogen-activated protein kinase (MAPK) cascades, greatly enhances the sensitivity of cellular targets to external stimuli. The sensitivity increases multiplicatively with the number of cascade levels, so that a tiny change in a stimulus results in a large change in the response, the phenomenon referred to as ultrasensitivity. In a variety of cell types, the MAPK cascades are imbedded in long feedback loops, positive or negative, depending on whether the terminal kinase stimulates or inhibits the activation of the initial level. Here we demonstrate that a negative feedback loop combined with intrinsic ultrasensitivity of the MAPK cascade can bring about sustained oscillations in MAPK phosphorylation. Based on recent kinetic data on the MAPK cascades, we predict that the period of oscillations can range from minutes to hours. The phosphorylation level can vary between the base level and almost 100% of the total protein. The oscillations of the phosphorylation cascades and slow protein diffusion in the cytoplasm can lead to intracellular waves of phospho-proteins.     

Modeling cell signaling networks

Cell signaling pathways interact with one another to form networks in mammalian systems. Such networks are complex in their organization and exhibit emergent properties such as bistability and ultrasensitivity. Analysis of signaling networks requires a combination of experimental and theoretical approaches including the development and analysis of models. This review focuses on theoretical approaches to understanding cell signaling networks. Using heterotrimeric G protein pathways an example, we demonstrate how interactions between two pathways can result in a network that contains a positive feedback loop and function as a switch. Different mathematical approaches that are currently used to model signaling networks are described, and future challenges including the need for databases as well as enhanced computing environments are discussed.     

The regulatory roles of the galactose permease and kinase in the induction response of the GAL network in Saccharomyces cerevisiae

The GAL genetic switch of Saccharomyces cerevisiae exhibits an ultrasensitive response to the inducer galactose as well as the "all-or-none" behavior characteristic of many eukaryotic regulatory networks. We have constructed a strain that allows intermediate levels of gene expression from a tunable GAL1 promoter at both the population and the single cell level by altering the regulation of the galactose permease Gal2p. Similar modifications to other feedback loops regulating the Gal80p repressor and the Gal3p signaling protein did not result in similarly tuned responses, indicating that the level of inducer transport is unique in its ability to control the switch response of the network. In addition, removal of the Gal1p galactokinase from the network resulted in a regimed response due to the dual role of this enzyme in galactose catabolism and transport. These two activities have competing effects on the response of the network to galactose such that the transport effects of Gal1p are dominant at low galactose concentrations, whereas its catabolic effects are dominant at high galactose concentrations. In addition, flow cytometry analysis revealed the unexpected phenomenon of multiple populations in the gal1delta strains, which were not present in the isogenic GAL1 background. This result indicates that Gal1p may play a previously undescribed role in the stability of the GAL network response.     

The JNK cascade as a biochemical switch in mammalian cells: ultrasensitive and all-or-none responses

JNK proteins are ubiquitously expressed, evolutionarily conserved MAP kinases that are involved in stress responses. Recently, it was shown that the JNK cascade in Xenopus oocytes exhibits sustained, all-or-none responses to graded, transient stimuli. Here, we have examined the character of the JNK cascade's response in mammalian cells. The steady-state responses of JNK to sorbitol and anisomycin were found to be highly ultrasensitive in HeLa cells, HEK 293 cells, and Jurkat T cells. The JNK responses were also reversible, not sustained, as was the case in oocytes. Jurkat cells activated their JNK in response to phorbol myristate acetate (PMA), and the response of the entire population of Jurkat cells was graded. However, analysis of subpopulations of the PMA-treated Jurkat cells revealed that the steady-state responses of both JNK and CD69, a T cell surface activation marker, were essentially all-or-none in character. These studies show that the JNK cascade commonly exhibits switch-like responses to a variety of stimuli.     

Robust,tunable genetic memory from protein sequestration combined with positive feedback

Natural regulatory networks contain many interacting components that allow for fine-tuning of switching and memory properties. Building simple bistable switches, synthetic biologists have learned the design principles of complex natural regulatory networks. However, most switches constructed so far are so simple (e.g. comprising two regulators) that they are functional only within a limited parameter range. Here, we report the construction of robust, tunable bistable switches in Escherichia coli using three heterologous protein regulators (ExsADC) that are sequestered into an inactive complex through a partner swapping mechanism. On the basis of mathematical modeling, we accurately predict and experimentally verify that the hysteretic region can be fine-tuned by controlling the interactions of the ExsADC regulatory cascade using the third member ExsC as a tuning knob. Additionally, we confirm that a dual-positive feedback switch can markedly increase the hysteretic region, compared to its single-positive feedback counterpart. The dual-positive feedback switch displays bistability over a 10⁶-fold range of inducer concentrations, to our knowledge, the largest range reported so far. This work demonstrates the successful interlocking of sequestration-based ultrasensitivity and positive feedback, a design principle that can be applied to the construction of robust, tunable, and predictable genetic programs to achieve increasingly sophisticated biological behaviors.