Rebecca Perrett

Information transfer by leaky, heterogeneous, protein kinase signaling systems.

Significance Extracellular concentrations convey information to cells about their environment. To sense these signals, cells use biomolecular networks that exhibit inevitable cell-to-cell variability and basal activity. Basal activity is widespread under physiological conditions (with phenotypic consequences), is often raised in disease, and can eradicate the transfer of information. In an experimental study of ERK signaling by single cells exhibiting heterogeneous ERK expression and basal activity, we verify our central theoretical prediction: Negative feedback substantially increases information transfer to the nucleus by preventing a near-flat average response curve and reducing sensitivity to variation in the ERK expression level. Our results reveal an important role for negative feedback mechanisms in protecting information transfer by saturable cell signaling systems from basal activity. Cells must sense extracellular signals and transfer the information contained about their environment reliably to make appropriate decisions. To perform these tasks, cells use signal transduction networks that are subject to various sources of noise. Here, we study the effects on information transfer of two particular types of noise: basal (leaky) network activity and cell-to-cell variability in the componentry of the network. Basal activity is the propensity for activation of the network output in the absence of the signal of interest. We show, using theoretical models of protein kinase signaling, that the combined effect of the two types of noise makes information transfer by such networks highly vulnerable to the loss of negative feedback. In an experimental study of ERK signaling by single cells with heterogeneous ERK expression levels, we verify our theoretical prediction: In the presence of basal network activity, negative feedback substantially increases information transfer to the nucleus by both preventing a near-flat average response curve and reducing sensitivity to variation in substrate expression levels. The interplay between basal network activity, heterogeneity in network componentry, and feedback is thus critical for the effectiveness of protein kinase signaling. Basal activity is widespread in signaling systems under physiological conditions, has phenotypic consequences, and is often raised in disease. Our results reveal an important role for negative feedback mechanisms in protecting the information transfer function of saturable, heterogeneous cell signaling systems from basal activity.

Molecular Mechanisms of Gonadotropin-Releasing Hormone Signaling: Integrating Cyclic Nucleotides into the Network

Gonadotropin-releasing hormone (GnRH) is the primary regulator of mammalian reproductive function in both males and females. It acts via G-protein coupled receptors on gonadotropes to stimulate synthesis and secretion of the gonadotropin hormones luteinizing hormone and follicle-stimulating hormone. These receptors couple primarily via G-proteins of the Gq/ll family, driving activation of phospholipases C and mediating GnRH effects on gonadotropin synthesis and secretion. There is also good evidence that GnRH causes activation of other heterotrimeric G-proteins (Gs and Gi) with consequent effects on cyclic AMP production, as well as for effects on the soluble and particulate guanylyl cyclases that generate cGMP. Here we provide an overview of these pathways. We emphasize mechanisms underpinning pulsatile hormone signaling and the possible interplay of GnRH and autocrine or paracrine regulatory mechanisms in control of cyclic nucleotide signaling.

Dual specificity phosphatases 10 and 16 are positive regulators of EGF-stimulated ERK activity: Indirect regulation of ERK signals by JNK/p38 selective MAPK phosphatases

We have explored the possible role of dual specificity phosphatases (DUSPs) on acute EGF-mediated ERK signalling using high content imaging and a delayed MEK inhibition protocol to distinguish direct and indirect effects of the phosphatases on ERK activity. Using siRNAs, we were unable to find evidence that any of the MAPK phosphatases (MKPs) expressed in HeLa cells acts directly to dephosphorylate ppERK1/2 (dual phosphorylated ERKs 1 and/or 2) in the acute time-frame tested (0–14 min). Nevertheless, siRNAs against two p38/JNK MKPs (DUSPs 10 and 16) inhibited acute EGF-stimulated ERK activation. No such effect was seen for acute effects of the protein kinase C activator PDBu (phorbol 12,13 dibutyrate) on ERK activity, although effects of EGF and PDBu on ERK-dependent transcription (Egr-1 luciferase activity) were both reduced by siRNA targeting DUSPs 10 and 16. Inhibition of EGF-stimulated ERK activity by these siRNAs was reversed by pharmacological inhibition of p38 MAPK and single cell analysis revealed that the siRNAs did not influence the nuclear-cytoplasmic distribution of ppERK1/2. Thus, DUSPs 10 and 16 are positive regulators of activation, apparently acting by modulating cross-talk between the p38 and ERK pathways. A simplified mathematical model of this scenario accurately predicted the experimental data, supporting the conclusion that the major mechanism by which MKPs influence acute EGF-stimulated ERK responses is the negative regulation of p38, resulting in the positive regulation of ERK phosphorylation and activity.

Information Transfer in Gonadotropin-releasing Hormone (GnRH) Signaling: EXTRACELLULAR SIGNAL-REGULATED KINASE (ERK)-MEDIATED FEEDBACK LOOPS CONTROL HORMONE SENSING.

Cell signaling pathways are noisy communication channels, and statistical measures derived from information theory can be used to quantify the information they transfer. Here we use single cell signaling measures to calculate mutual information as a measure of information transfer via gonadotropin-releasing hormone (GnRH) receptors (GnRHR) to extracellular signal-regulated kinase (ERK) or nuclear factor of activated T-cells (NFAT). This revealed mutual information values <1 bit, implying that individual GnRH-responsive cells cannot unambiguously differentiate even two equally probable input concentrations. Addressing possible mechanisms for mitigation of information loss, we focused on the ERK pathway and developed a stochastic activation model incorporating negative feedback and constitutive activity. Model simulations revealed interplay between fast (min) and slow (min-h) negative feedback loops with maximal information transfer at intermediate feedback levels. Consistent with this, experiments revealed that reducing negative feedback (by expressing catalytically inactive ERK2) and increasing negative feedback (by Egr1-driven expression of dual-specificity phosphatase 5 (DUSP5)) both reduced information transfer from GnRHR to ERK. It was also reduced by blocking protein synthesis (to prevent GnRH from increasing DUSP expression) but did not differ for different GnRHRs that do or do not undergo rapid homologous desensitization. Thus, the first statistical measures of information transfer via these receptors reveals that individual cells are unreliable sensors of GnRH concentration and that this reliability is maximal at intermediate levels of ERK-mediated negative feedback but is not influenced by receptor desensitization.

Signaling to Extracellular Signal-regulated Kinase from ErbB1 Kinase and Protein Kinase C: FEEDBACK, HETEROGENEITY, AND GATING

Background: The mechanisms underlying acute ERK signaling are poorly understood. Results: Feedback influences basal and acutely stimulated ERK responses but does not render signaling kinetics robust to ERK concentration. Conclusion: Acute ERK response kinetics depend on ERK concentration and activation mechanism as well as feedback. Significance: ERK responses to transient stimulation can be gated by ERK concentration, and short-term activation appears distributive rather than processive. Many extracellular signals act via the Raf/MEK/ERK cascade in which kinetics, cell-cell variability, and sensitivity of the ERK response can all influence cell fate. Here we used automated microscopy to explore the effects of ERK-mediated negative feedback on these attributes in cells expressing endogenous ERK or ERK2-GFP reporters. We studied acute rather than chronic stimulation with either epidermal growth factor (ErbB1 activation) or phorbol 12,13-dibutyrate (PKC activation). In unstimulated cells, ERK-mediated negative feedback reduced the population-average and cell-cell variability of the level of activated ppERK and increased its robustness to changes in ERK expression. In stimulated cells, negative feedback (evident between 5 min and 4 h) also reduced average levels and variability of phosphorylated ERK (ppERK) without altering the “gradedness” or sensitivity of the response. Binning cells according to total ERK expression revealed, strikingly, that maximal ppERK responses initially occur at submaximal ERK levels and that this non-monotonic relationship changes to an increasing, monotonic one within 15 min. These phenomena occur in HeLa cells and MCF7 breast cancer cells and in the presence and absence of ERK-mediated negative feedback. They were best modeled assuming distributive (rather than processive) activation. Thus, we have uncovered a novel, time-dependent change in the relationship between total ERK and ppERK levels that persists without negative feedback. This change makes acute response kinetics dependent on ERK level and provides a “gating” or control mechanism in which the interplay between stimulus duration and the distribution of ERK expression across cells could modulate the proportion of cells that respond to stimulation.

Pulsatile Hormonal Signaling to Extracellular Signal-regulated Kinase EXPLORING SYSTEM SENSITIVITY TO GONADOTROPIN-RELEASING HORMONE PULSE FREQUENCY AND WIDTH

Background: Cellular decoding of stimulus dynamics is poorly understood. Results: GnRH pulses activate ERK, and response kinetics determine sensitivity to different pulse features. Conclusion: The system is sensitive to pulse frequency but robust to width; this distinction develops through the cascade and is dictated by response kinetics. Significance: We describe mathematical and biochemical “design features” for pulsatile hormonal signaling. Gonadotropin-releasing hormone (GnRH) is secreted in brief pulses that stimulate synthesis and secretion of pituitary gonadotropin hormones and thereby mediate control of reproduction. It acts via G-protein-coupled receptors to stimulate effectors, including ERK. Information could be encoded in GnRH pulse frequency, width, amplitude, or other features of pulse shape, but the relative importance of these features is unknown. Here we examine this using automated fluorescence microscopy and mathematical modeling, focusing on ERK signaling. The simplest scenario is one in which the system is linear, and response dynamics are relatively fast (compared with the signal dynamics). In this case integrated system output (ERK activation or ERK-driven transcription) will be roughly proportional to integrated input, but we find that this is not the case. Notably, we find that relatively slow response kinetics lead to ERK activity beyond the GnRH pulse, and this reduces sensitivity to pulse width. More generally, we show that the slowing of response kinetics through the signaling cascade creates a system that is robust to pulse width. We, therefore, show how various levels of response kinetics synergize to dictate system sensitivity to different features of pulsatile hormone input. We reveal the mathematical and biochemical basis of a dynamic GnRH signaling system that is robust to changes in pulse amplitude and width but is sensitive to changes in receptor occupancy and frequency, precisely the features that are tightly regulated and exploited to exert physiological control in vivo.

Signaling to ERK from ErbB1 and PKC: Feedback, Heterogeneity and Gating.

Background: The mechanisms underlying acute ERK signaling are poorly understood. Results: Feedback influences basal and acutely stimulated ERK responses but does not render signaling kinetics robust to ERK concentration. Conclusion: Acute ERK response kinetics depend on ERK concentration and activation mechanism as well as feedback. Significance: ERK responses to transient stimulation can be gated by ERK concentration, and short-term activation appears distributive rather than processive. Many extracellular signals act via the Raf/MEK/ERK cascade in which kinetics, cell-cell variability, and sensitivity of the ERK response can all influence cell fate. Here we used automated microscopy to explore the effects of ERK-mediated negative feedback on these attributes in cells expressing endogenous ERK or ERK2-GFP reporters. We studied acute rather than chronic stimulation with either epidermal growth factor (ErbB1 activation) or phorbol 12,13-dibutyrate (PKC activation). In unstimulated cells, ERK-mediated negative feedback reduced the population-average and cell-cell variability of the level of activated ppERK and increased its robustness to changes in ERK expression. In stimulated cells, negative feedback (evident between 5 min and 4 h) also reduced average levels and variability of phosphorylated ERK (ppERK) without altering the “gradedness” or sensitivity of the response. Binning cells according to total ERK expression revealed, strikingly, that maximal ppERK responses initially occur at submaximal ERK levels and that this non-monotonic relationship changes to an increasing, monotonic one within 15 min. These phenomena occur in HeLa cells and MCF7 breast cancer cells and in the presence and absence of ERK-mediated negative feedback. They were best modeled assuming distributive (rather than processive) activation. Thus, we have uncovered a novel, time-dependent change in the relationship between total ERK and ppERK levels that persists without negative feedback. This change makes acute response kinetics dependent on ERK level and provides a “gating” or control mechanism in which the interplay between stimulus duration and the distribution of ERK expression across cells could modulate the proportion of cells that respond to stimulation.

Decoding neurohormone pulse frequency by convergent signalling modules.

GnRH (gonadotropin-releasing hormone) mediates control of reproduction. It is secreted in pulses and acts via intracellular effectors to activate gene expression. Submaximal GnRH pulse frequency can elicit maximal responses, yielding bell-shaped frequency-response curves characteristic of genuine frequency decoders. GnRH frequency decoding is therapeutically important (pulsatile GnRH can drive ovulation in assisted reproduction, whereas sustained activation can treat breast and prostate cancers), but the mechanisms are unknown. In the present paper, we review recent work in this area, placing emphasis on the regulation of transcription, and showing how mathematical modelling of GnRH effects on two effectors [ERK (extracellular-signal-regulated kinase) and NFAT (nuclear factor of activated T-cells)] reveals the potential for genuine frequency decoding as an emergent feature of the GnRH signalling network, rather than an intrinsic feature of a given protein or pathway within it.

Information Transfer via Gonadotropin-Releasing Hormone Receptors to ERK and NFAT: Sensing GnRH and Sensing Dynamics

Information theoretic approaches can be used to quantify information transfer via cell signaling networks. In this study, we do so for gonadotropin-releasing hormone (GnRH) activation of extracellular signal-regulated kinase (ERK) and nuclear factor of activated T cells (NFAT) in large numbers of individual fixed LβT2 and HeLa cells. Information transfer, measured by mutual information between GnRH and ERK or NFAT, was <1 bit (despite 3-bit system inputs). It was increased by sensing both ERK and NFAT, but the increase was <50%. In live cells, information transfer via GnRH receptors to NFAT was also <1 bit and was increased by consideration of response trajectory, but the increase was <10%. GnRH secretion is pulsatile, so we explored information gained by sensing a second pulse, developing a model of GnRH signaling to NFAT with variability introduced by allowing effectors to fluctuate. Simulations revealed that when cell–cell variability reflects rapidly fluctuating effector levels, additional information is gained by sensing two GnRH pulses, but where it is due to slowly fluctuating effectors, responses in one pulse are predictive of those in another, so little information is gained from sensing both. Wet laboratory experiments revealed that the latter scenario holds true for GnRH signaling; within the timescale of our experiments (1 to 2 hours), cell–cell variability in the NFAT pathway remains relatively constant, so trajectories are reproducible from pulse to pulse. Accordingly, joint sensing, sensing of response trajectories, and sensing of repeated pulses can all increase information transfer via GnRH receptors, but in each case the increase is small.