Current Biology Vol 24 No 21 R1044
8. Emery, N.J., and Clayton, N.S. (2004). The mentality of crows: convergent evolution of intelligence in corvids and apes. Science 306, 1903–1907. 9. Homberg, U. (2008). Evolution of the central complex in the arthropod brain with respect to the visual system. Arthropod Struct. Dev. 37, 347–362.
10. Pfeiffer, K., and Homberg, U. (2014). Organization and functional roles of the central complex in the insect brain. Annu. Rev. Entomol. 59, 165–184. 11. Whitney, D. (2009). Neuroscience: toward unbinding the binding problem. Curr. Biol. 19, 251–253.
Neuronal Plasticity: How Do Neurons Know What To Do? A recent study confirms activity-dependent co-regulation of membrane conductances as a mechanism underlying homeostatic regulation of neuronal properties. How multiple cellular and synaptic homeostatic mechanisms interact in a neuronal circuit is best studied with a combination of experimentation and modeling. Astrid A. Prinz Nervous systems face two challenges: to be plastic and able to change, adapt, and learn, while at the same time functioning reliably to ensure an animal’s survival in an ever-changing environment. A growing body of experimental and computational work indicates that to do so brains rely on multiple plasticity and homeostasis mechanisms that act on both synaptic and cell-intrinsic parameters. While the triggers and molecular mechanisms of some prominent forms of synaptic plasticity are increasingly well understood [1], what factors govern — and what mechanisms underlie — the plasticity and stability of neuronal properties is less clear. Recent work reported in Current Biology by Schulz and colleagues [2] shows that electrical activity plays a role in regulating correlated expression levels of ionic membrane channels that had previously been found to also depend on the presence of neuromodulators [3,4]. This warrants a brief review of various triggers and regulators of neuronal plasticity and stability. An important finding in the area of neuronal plasticity and homeostasis is that electrophysiologically relevant parameters of neurons — for example, the magnitudes of different ionic conductances in a neuron’s membrane — can vary widely between different neurons of the same type [5]. Such variability of parameters that support similar and physiological network output has also been confirmed in computational models
[6]. This introduces the notion of a ‘solution space’, i.e. the idea that instead of being limited to a particular and narrowly defined combination of cellular and synaptic parameters, networks can achieve functional output in an often extensive subset of their high-dimensional space of cell and synapse parameters [7]. Despite this variability in individual parameters, the electrical activity produced by neurons of the same type can be highly stereotyped. One mechanism through which neurons appear to achieve reliable and functional activity is the imposition of constraints on how cellular and synaptic parameters can vary. Such constraints often take the form of linear relationships between pairs or higher numbers of parameters [3,8], thus reducing the dimensionality of the solution space occupied by the biological system. Such linear correlations between cellular and synaptic parameters have been independently demonstrated at the level of electrophysiological properties such as ionic membrane conductances [3,9] and synapse strengths [10], and at the level of mRNA copy numbers for ion channels, like in the recent paper by Schulz and Colleagues [2]. Why might correlations between neuronal parameters be important for the ability of neurons and networks to generate and maintain proper biological output? Computational models of neurons and networks show that imposing pairwise correlations on neuronal or synaptic parameters can increase the likelihood
Department of Biology, Lund University, So¨lvegatan 35, 22362 Lund, Sweden. E-mail: [email protected]
http://dx.doi.org/10.1016/j.cub.2014.09.041
that a given neuron or network generates functional activity despite variability in individual parameters [11]. In some cases, sets of cellular parameters that are found to be correlated in biological neurons appear to be functionally tied into ‘modules’ [12]. For example, in bursting neurons, the conductances of the slow inward current IB and the potassium current IA constitute a burst generation module, while conductances for the transient calcium current ICaT and the delayed rectifier current IKd, if co-regulated, determine the peak and duration of the slow voltage oscillations underlying bursts [12]. Imposing correlations between the conductances within a given module will therefore help ensure proper neuronal behavior. Intriguingly, the recent Schulz et al. paper [2] and previous reports by this and other groups [3,4] show that correlations between cellular parameters that appear to support functional network output in an intact and unperturbed circuit are sometimes abandoned when the circuit is exposed to — and has to overcome — massive perturbation or injury. For example, in the stomatogastric nervous system of crabs, rhythmically active central pattern-generating neurons exhibit several pairwise linear correlations between mRNAs coding for various ion channels when the circuit is intact, under the influence of neuromodulators, and generating appropriate motor patterns [3,8]. In contrast, most of these pairwise correlations disappear when the circuit is challenged to produce its motor pattern in the absence of neuromodulation and after a period of quiescence [3,4]. It is as if the circuit ‘knows’ that it needs to abandon its previously implemented correlation rules in order to explore a larger swath of its parameter space and find a new solution to generating functional activity under perturbed conditions. What tells a neuron whether — and how — to adjust its properties in order to maintain proper function? Figure 1
Dispatch R1045
illustrates some of the characteristics of neuronal circuits that have been implicated as potentially involved in sensing whether a network functions properly, and in adjusting cellular and/or synaptic parameters if it doesn’t. The well-studied phenomenon of synaptic scaling [1] is thought to be triggered and regulated primarily by the overall level of activity in a neuronal network, and is thus often hypothesized to be a mechanism that supports firing rate homeostasis [13]. In contrast, some forms of synaptic plasticity, most notably spike-timing-dependent plasticity (STDP), are thought to be sensitive to the precise relative timing of action potentials in pre- and postsynaptic neurons, and thus appear to depend on the details of network activity, beyond average firing rates [14]. In some systems, synaptic plasticity and homeostasis have also been shown to be triggered and regulated by the level of neurotransmission present in the system [15], or by tissue-wide concentrations of neuromodulators or growth factors, whose release can itself be dependent on neuronal activity levels [13]. What does the recent study by Schulz and colleagues [2] add to this already complex — albeit likely far from complete — picture of the factors that can trigger and regulate plasticity and homeostasis in neuronal systems? By carefully decoupling electrical activity, neurotransmission, and modulatory state, Schulz and colleagues show that in stomatogastric neurons, the maintenance of correlations between mRNA levels for different ion channel types is activity-dependent. This is in keeping with previous predictions from computational work by various groups that indicated that homeostatic regulation of cellular properties could in principle rely on the sensing of electrical activity [16,17], most likely through calcium-based activity sensors and intracellular regulatory cascades [18]. This notion is furthermore supported by experiments that demonstrated that stomatogastric neurons, when isolated from their network, can cell-autonomously regulate their properties in an activity- and calcium-dependent manner [19]. What initially makes Schulz and colleagues finding surprising is that the
Modulation Activity
Timing Transmission Current Biology
Figure 1. Potential triggers and regulators of cellular and synaptic plasticity. Neuronal network characteristics potentially involved in triggering and regulating cellular and synaptic plasticity include the overall level of electrical activity, its precise timing within and between neurons, the presence of neuromodulatory substances, and the presence of neurotransmitters.
same and other groups have argued in earlier work that the maintenance of correlation rules between ionic conductances and mRNA levels in stomatogastric neurons does in fact depend on the presence of neuromodulators, and not the electrical activity generated by the network. How do these seemingly discrepant findings go together? Schulz and colleagues point out that their recent work examines channel mRNAs that are distinct from those whose correlations were previously found to be modulator-dependent. They further argue that they were able to tease apart the effects of neuromodulation from those of electrical activity more thoroughly in the newer work than in previous experiments. It is tempting to speculate that a further explanation of the seemingly different results lies in the fact that activity- and modulator-dependent homeostasis mechanisms need not be mutually exclusive. More likely, neuronal circuits, especially those that have evolved to reliably produce stereotyped activity patterns under various conditions, employ multiple, parallel homeostatic processes that are triggered by different neuronal network characteristics (such as those illustrated in Figure 1), target multiple network parameters, and serve to maintain various aspects of network activity [20]. Clearly, much further work, ideally in a continuing back-and-forth between experiment
and modeling, will be necessary to fully understand the complement of plasticity and homeostasis mechanisms used by neurons and neuronal networks, and how multiple mechanisms might interact to achieve and maintain proper network function. References 1. Turrigiano, G.G. (2008). The self-tuning neuron: synaptic scaling of excitatory synapses. Cell 135, 422–435. 2. Temporal, S., Lett, K.M., and Schulz, D.J. (2014). Activity-dependent feedback regulates correlated ion channel mRNA levels in single identified motor neurons. Curr. Biol. 24, 1899–1904. 3. Khorkova, O., and Golowasch, J. (2007). Neuromodulators, not activity, control coordinated expression of ionic currents. J. Neurosci. 27, 8709–8718. 4. Temporal, S., Desai, M., Khorkova, O., Varghese, G., Dai, A., Schulz, D.J., and Golowasch, J. (2012). Neuromodulation independently determines correlated channel expression and conductance levels in motor neurons of the stomatogastric ganglion. J. Neurophysiol. 107, 718–727. 5. Swensen, A.M., and Bean, B.P. (2005). Robustness of burst firing in dissociated Purkinje neurons with acute or long-term reductions in sodium conductance. J. Neurosci. 25, 3509–3520. 6. Prinz, A.A., Bucher, D., and Marder, E. (2004). Similar network activity from disparate circuit parameters. Nat. Neurosci. 7, 1345–1352. 7. Prinz, A.A. (2010). Computational approaches to neuronal network analysis. Philos. Trans. R. Soc. Lond. B Biol. Sci. 365, 2397–2405. 8. Schulz, D.J., Goaillard, J.-M., and Marder, E.E. (2007). Quantitative expression profiling of identified neurons reveals cell-specific constraints on highly variable levels of gene expression. Proc. Natl. Acad. Sci. USA 104, 13187–13191. 9. MacLean, J.N., Zhang, Y., Johnson, B.R., and Harris-Warrick, R.M. (2003). Activityindependent homeostasis in rhythmically active neurons. Neuron 37, 109–120.
Current Biology Vol 24 No 21 R1046
10. Goaillard, J.-M., Taylor, A.L., Schulz, D.J., and Marder, E. (2009). Functional consequences of animal-to-animal variation in circuit parameters. Nat. Neurosci. 12, 1424–1430. 11. Hudson, A.E., and Prinz, A.A. (2010). Conductance ratios and cellular identity. PLoS Comput. Biol. 6, e1000838. 12. Nair, S.S., Ball, J.M., Ransdell, J., and Schulz, D.J. (2011). Potassium current co-regulation preserves bursting properties in a computational model of the crustacean cardiac ganglion. Society for Neuroscience Abstract Viewer and Itinerary Planner 41. 13. Turrigiano, G.G., and Nelson, S.B. (2004). Homeostatic plasticity in the developing nervous system. Nat. Rev. Neurosci. 5, 97–107. 14. Feldman, D.E. (2012). The spike-timing dependence of plasticity. Neuron 75, 556–571.
15. Lindsly, C., Gonzalez-Islas, C., and Wenner, P. (2014). Activity blockade and GABA(A) receptor blockade produce synaptic scaling through chloride accumulation in embryonic spinal motoneurons and interneurons. PLoS One 9, e94559. 16. Liu, Z., Golowasch, J., Marder, E., and Abbott, L.F. (1998). A model neuron with activity-dependent conductances regulated by multiple calcium sensors. J. Neurosci. 18, 2309–2320. 17. Olypher, A.V., and Prinz, A.A. (2010). Geometry and dynamics of activity-dependent homeostatic regulation in neurons. J. Comput. Neurosci. 28, 361–374. 18. Gunay, C., and Prinz, A.A. (2010). Model calcium sensors for network homeostasis: sensor and readout parameter analysis from a database of model neuronal networks. J. Neurosci. 30, 1686–1698.
Centriole Duplication: When PLK4 Meets Ana2/STIL Polo-like kinase 4 is known to drive centriole duplication, but the relevant substrate remains elusive. A new study shows that PLK4 phosphorylates a key centriolar component, Ana2/STIL, to initiate centriole assembly. Minhee Kim1,2, Chii Shyang Fong1, and Meng-Fu Bryan Tsou1,2,* Animal centrioles serve as the basal body for cilia assembly, and form the core of the centrosome for microtubule organization. To faithfully execute these functions, centriole numbers are strictly regulated in cycling cells, primarily through the precise control of duplication and segregation [1,2]. Duplication involves the doubling of centrioles in S phase in which exactly one new (or daughter) centriole forms in close proximity to the pre-existing (or mother) centriole. For more than a decade, the serine/threonine-protein kinase PLK4 (ZYG-1 in worms; SAK in flies) has been demonstrated as the master kinase driving centriole biogenesis [3–5], as PLK4 associates with mother centrioles where its level profoundly affects the number of daughter centrioles assembled [3]. Conversely, loss of PLK4 abolishes all signs of centriole duplication, revealing PLK4 as one of the most upstream regulators in the pathway. Thus, there has been a strong interest in understanding how PLK4 catalyzes centriole assembly, and in identifying the relevant substrates. The search has finally led to an exciting discovery by the group of David Glover,
published in this issue of Current Biology [6]. Centriole integrity is established through a stepwise assembly of a series of structural components, many of which serve as the building block, scaffold, or stabilizing factor. While self-assembly is a key feature of centriole duplication, the initiation step is critically guarded by enzymatic regulators to allow quantity/quality control. Centriolar proteins SAS-6 and STIL (human Ana2) are the first two components recruited to the normally single assembly site specified at the periphery of each mother centriole [7,8]. SAS-6 and likely together with STIL form the primary backbone of the cartwheel [9–11], which is the geometric scaffold promoting 9-fold symmetrical assembly of the centriole. How the assembly site is chosen or limited to one per mother centriole is not fully understood, but in vertebrate cells, the initial loading of SAS-6 or STIL is strictly dependent on the kinase PLK4 [8,12]. Moreover, when PLK4 is overexpressed, additional assembly sites can form around the mother centriole [3], where SAS-6 and STIL are loaded to promote extra daughter centriole formation. A few PLK4 substrates have been reported [13–16], but a direct link to the
19. Golowasch, J., Abbott, L.F., and Marder, E. (1999). Activity-dependent regulation of potassium currents in an identified neuron of the stomatogastric ganglion of the crab Cancer borealis. J. Neurosci. 19, RC33. 20. Marder, E., and Tang, L.S. (2010). Coordinating different homeostatic processes. Neuron 66, 161–163.
Department of Biology, Emory University, O. Wayne Rollins Research Center, Room 2105, 1510 Clifton Road, Atlanta, GA 30322, USA. E-mail: [email protected]
http://dx.doi.org/10.1016/j.cub.2014.09.064
initiation step of centriole assembly remains obscure. Finding the relevant substrates of PLK4, and knowing how the phosphorylation triggers centriole assembly are no doubt two of the most urgent goals in centrosome biology. Using in vitro kinase assay with purified Drosophila components, Dzhindzhev et al. found that Ana2 (fly STIL), but not SAS-6, can be strongly phosphorylated by PLK4, consistent with the physical interaction between STIL and PLK4 reported recently [6,17]. Four phosphorylation sites (S318, S365, S370 and S373) were identified, and all of them are clustered within the STAN motif [18], a conserved domain found in all Ana2/STIL orthologues essential for centriole duplication. Importantly, the same phosphorylation sites can also be detected in vivo in cells overexpressing Ana2 and PLK4. Moreover, the authors showed that these phosphorylation sites are functionally relevant, as the non-phosphorylatable Ana2 in which the 4 serine residues are changed to alanine (Ana2-4A) fails to rescue or drive centriole duplication. To understand the function of Ana2 phosphorylation, Dzhindzhev et al. analyzed the interaction of Ana2 with other centriolar components. In vitro binding assay showed that wild-type Ana2 (Ana2-WT) strongly associates with SAS-6 when active PLK4 is present, whereas no interaction between SAS-6 and Ana2-4A could be detected under the same condition [6]. Such kinase activity-dependent interactions were further confirmed in cells in vivo, with wild-type Ana2,
8. Emery, N.J., and Clayton, N.S. (2004). The mentality of crows: convergent evolution of intelligence in corvids and apes. Science 306, 1903–1907. 9. Homberg, U. (2008). Evolution of the central complex in the arthropod brain with respect to the visual system. Arthropod Struct. Dev. 37, 347–362.
10. Pfeiffer, K., and Homberg, U. (2014). Organization and functional roles of the central complex in the insect brain. Annu. Rev. Entomol. 59, 165–184. 11. Whitney, D. (2009). Neuroscience: toward unbinding the binding problem. Curr. Biol. 19, 251–253.
Neuronal Plasticity: How Do Neurons Know What To Do? A recent study confirms activity-dependent co-regulation of membrane conductances as a mechanism underlying homeostatic regulation of neuronal properties. How multiple cellular and synaptic homeostatic mechanisms interact in a neuronal circuit is best studied with a combination of experimentation and modeling. Astrid A. Prinz Nervous systems face two challenges: to be plastic and able to change, adapt, and learn, while at the same time functioning reliably to ensure an animal’s survival in an ever-changing environment. A growing body of experimental and computational work indicates that to do so brains rely on multiple plasticity and homeostasis mechanisms that act on both synaptic and cell-intrinsic parameters. While the triggers and molecular mechanisms of some prominent forms of synaptic plasticity are increasingly well understood [1], what factors govern — and what mechanisms underlie — the plasticity and stability of neuronal properties is less clear. Recent work reported in Current Biology by Schulz and colleagues [2] shows that electrical activity plays a role in regulating correlated expression levels of ionic membrane channels that had previously been found to also depend on the presence of neuromodulators [3,4]. This warrants a brief review of various triggers and regulators of neuronal plasticity and stability. An important finding in the area of neuronal plasticity and homeostasis is that electrophysiologically relevant parameters of neurons — for example, the magnitudes of different ionic conductances in a neuron’s membrane — can vary widely between different neurons of the same type [5]. Such variability of parameters that support similar and physiological network output has also been confirmed in computational models
[6]. This introduces the notion of a ‘solution space’, i.e. the idea that instead of being limited to a particular and narrowly defined combination of cellular and synaptic parameters, networks can achieve functional output in an often extensive subset of their high-dimensional space of cell and synapse parameters [7]. Despite this variability in individual parameters, the electrical activity produced by neurons of the same type can be highly stereotyped. One mechanism through which neurons appear to achieve reliable and functional activity is the imposition of constraints on how cellular and synaptic parameters can vary. Such constraints often take the form of linear relationships between pairs or higher numbers of parameters [3,8], thus reducing the dimensionality of the solution space occupied by the biological system. Such linear correlations between cellular and synaptic parameters have been independently demonstrated at the level of electrophysiological properties such as ionic membrane conductances [3,9] and synapse strengths [10], and at the level of mRNA copy numbers for ion channels, like in the recent paper by Schulz and Colleagues [2]. Why might correlations between neuronal parameters be important for the ability of neurons and networks to generate and maintain proper biological output? Computational models of neurons and networks show that imposing pairwise correlations on neuronal or synaptic parameters can increase the likelihood
Department of Biology, Lund University, So¨lvegatan 35, 22362 Lund, Sweden. E-mail: [email protected]
http://dx.doi.org/10.1016/j.cub.2014.09.041
that a given neuron or network generates functional activity despite variability in individual parameters [11]. In some cases, sets of cellular parameters that are found to be correlated in biological neurons appear to be functionally tied into ‘modules’ [12]. For example, in bursting neurons, the conductances of the slow inward current IB and the potassium current IA constitute a burst generation module, while conductances for the transient calcium current ICaT and the delayed rectifier current IKd, if co-regulated, determine the peak and duration of the slow voltage oscillations underlying bursts [12]. Imposing correlations between the conductances within a given module will therefore help ensure proper neuronal behavior. Intriguingly, the recent Schulz et al. paper [2] and previous reports by this and other groups [3,4] show that correlations between cellular parameters that appear to support functional network output in an intact and unperturbed circuit are sometimes abandoned when the circuit is exposed to — and has to overcome — massive perturbation or injury. For example, in the stomatogastric nervous system of crabs, rhythmically active central pattern-generating neurons exhibit several pairwise linear correlations between mRNAs coding for various ion channels when the circuit is intact, under the influence of neuromodulators, and generating appropriate motor patterns [3,8]. In contrast, most of these pairwise correlations disappear when the circuit is challenged to produce its motor pattern in the absence of neuromodulation and after a period of quiescence [3,4]. It is as if the circuit ‘knows’ that it needs to abandon its previously implemented correlation rules in order to explore a larger swath of its parameter space and find a new solution to generating functional activity under perturbed conditions. What tells a neuron whether — and how — to adjust its properties in order to maintain proper function? Figure 1
Dispatch R1045
illustrates some of the characteristics of neuronal circuits that have been implicated as potentially involved in sensing whether a network functions properly, and in adjusting cellular and/or synaptic parameters if it doesn’t. The well-studied phenomenon of synaptic scaling [1] is thought to be triggered and regulated primarily by the overall level of activity in a neuronal network, and is thus often hypothesized to be a mechanism that supports firing rate homeostasis [13]. In contrast, some forms of synaptic plasticity, most notably spike-timing-dependent plasticity (STDP), are thought to be sensitive to the precise relative timing of action potentials in pre- and postsynaptic neurons, and thus appear to depend on the details of network activity, beyond average firing rates [14]. In some systems, synaptic plasticity and homeostasis have also been shown to be triggered and regulated by the level of neurotransmission present in the system [15], or by tissue-wide concentrations of neuromodulators or growth factors, whose release can itself be dependent on neuronal activity levels [13]. What does the recent study by Schulz and colleagues [2] add to this already complex — albeit likely far from complete — picture of the factors that can trigger and regulate plasticity and homeostasis in neuronal systems? By carefully decoupling electrical activity, neurotransmission, and modulatory state, Schulz and colleagues show that in stomatogastric neurons, the maintenance of correlations between mRNA levels for different ion channel types is activity-dependent. This is in keeping with previous predictions from computational work by various groups that indicated that homeostatic regulation of cellular properties could in principle rely on the sensing of electrical activity [16,17], most likely through calcium-based activity sensors and intracellular regulatory cascades [18]. This notion is furthermore supported by experiments that demonstrated that stomatogastric neurons, when isolated from their network, can cell-autonomously regulate their properties in an activity- and calcium-dependent manner [19]. What initially makes Schulz and colleagues finding surprising is that the
Modulation Activity
Timing Transmission Current Biology
Figure 1. Potential triggers and regulators of cellular and synaptic plasticity. Neuronal network characteristics potentially involved in triggering and regulating cellular and synaptic plasticity include the overall level of electrical activity, its precise timing within and between neurons, the presence of neuromodulatory substances, and the presence of neurotransmitters.
same and other groups have argued in earlier work that the maintenance of correlation rules between ionic conductances and mRNA levels in stomatogastric neurons does in fact depend on the presence of neuromodulators, and not the electrical activity generated by the network. How do these seemingly discrepant findings go together? Schulz and colleagues point out that their recent work examines channel mRNAs that are distinct from those whose correlations were previously found to be modulator-dependent. They further argue that they were able to tease apart the effects of neuromodulation from those of electrical activity more thoroughly in the newer work than in previous experiments. It is tempting to speculate that a further explanation of the seemingly different results lies in the fact that activity- and modulator-dependent homeostasis mechanisms need not be mutually exclusive. More likely, neuronal circuits, especially those that have evolved to reliably produce stereotyped activity patterns under various conditions, employ multiple, parallel homeostatic processes that are triggered by different neuronal network characteristics (such as those illustrated in Figure 1), target multiple network parameters, and serve to maintain various aspects of network activity [20]. Clearly, much further work, ideally in a continuing back-and-forth between experiment
and modeling, will be necessary to fully understand the complement of plasticity and homeostasis mechanisms used by neurons and neuronal networks, and how multiple mechanisms might interact to achieve and maintain proper network function. References 1. Turrigiano, G.G. (2008). The self-tuning neuron: synaptic scaling of excitatory synapses. Cell 135, 422–435. 2. Temporal, S., Lett, K.M., and Schulz, D.J. (2014). Activity-dependent feedback regulates correlated ion channel mRNA levels in single identified motor neurons. Curr. Biol. 24, 1899–1904. 3. Khorkova, O., and Golowasch, J. (2007). Neuromodulators, not activity, control coordinated expression of ionic currents. J. Neurosci. 27, 8709–8718. 4. Temporal, S., Desai, M., Khorkova, O., Varghese, G., Dai, A., Schulz, D.J., and Golowasch, J. (2012). Neuromodulation independently determines correlated channel expression and conductance levels in motor neurons of the stomatogastric ganglion. J. Neurophysiol. 107, 718–727. 5. Swensen, A.M., and Bean, B.P. (2005). Robustness of burst firing in dissociated Purkinje neurons with acute or long-term reductions in sodium conductance. J. Neurosci. 25, 3509–3520. 6. Prinz, A.A., Bucher, D., and Marder, E. (2004). Similar network activity from disparate circuit parameters. Nat. Neurosci. 7, 1345–1352. 7. Prinz, A.A. (2010). Computational approaches to neuronal network analysis. Philos. Trans. R. Soc. Lond. B Biol. Sci. 365, 2397–2405. 8. Schulz, D.J., Goaillard, J.-M., and Marder, E.E. (2007). Quantitative expression profiling of identified neurons reveals cell-specific constraints on highly variable levels of gene expression. Proc. Natl. Acad. Sci. USA 104, 13187–13191. 9. MacLean, J.N., Zhang, Y., Johnson, B.R., and Harris-Warrick, R.M. (2003). Activityindependent homeostasis in rhythmically active neurons. Neuron 37, 109–120.
Current Biology Vol 24 No 21 R1046
10. Goaillard, J.-M., Taylor, A.L., Schulz, D.J., and Marder, E. (2009). Functional consequences of animal-to-animal variation in circuit parameters. Nat. Neurosci. 12, 1424–1430. 11. Hudson, A.E., and Prinz, A.A. (2010). Conductance ratios and cellular identity. PLoS Comput. Biol. 6, e1000838. 12. Nair, S.S., Ball, J.M., Ransdell, J., and Schulz, D.J. (2011). Potassium current co-regulation preserves bursting properties in a computational model of the crustacean cardiac ganglion. Society for Neuroscience Abstract Viewer and Itinerary Planner 41. 13. Turrigiano, G.G., and Nelson, S.B. (2004). Homeostatic plasticity in the developing nervous system. Nat. Rev. Neurosci. 5, 97–107. 14. Feldman, D.E. (2012). The spike-timing dependence of plasticity. Neuron 75, 556–571.
15. Lindsly, C., Gonzalez-Islas, C., and Wenner, P. (2014). Activity blockade and GABA(A) receptor blockade produce synaptic scaling through chloride accumulation in embryonic spinal motoneurons and interneurons. PLoS One 9, e94559. 16. Liu, Z., Golowasch, J., Marder, E., and Abbott, L.F. (1998). A model neuron with activity-dependent conductances regulated by multiple calcium sensors. J. Neurosci. 18, 2309–2320. 17. Olypher, A.V., and Prinz, A.A. (2010). Geometry and dynamics of activity-dependent homeostatic regulation in neurons. J. Comput. Neurosci. 28, 361–374. 18. Gunay, C., and Prinz, A.A. (2010). Model calcium sensors for network homeostasis: sensor and readout parameter analysis from a database of model neuronal networks. J. Neurosci. 30, 1686–1698.
Centriole Duplication: When PLK4 Meets Ana2/STIL Polo-like kinase 4 is known to drive centriole duplication, but the relevant substrate remains elusive. A new study shows that PLK4 phosphorylates a key centriolar component, Ana2/STIL, to initiate centriole assembly. Minhee Kim1,2, Chii Shyang Fong1, and Meng-Fu Bryan Tsou1,2,* Animal centrioles serve as the basal body for cilia assembly, and form the core of the centrosome for microtubule organization. To faithfully execute these functions, centriole numbers are strictly regulated in cycling cells, primarily through the precise control of duplication and segregation [1,2]. Duplication involves the doubling of centrioles in S phase in which exactly one new (or daughter) centriole forms in close proximity to the pre-existing (or mother) centriole. For more than a decade, the serine/threonine-protein kinase PLK4 (ZYG-1 in worms; SAK in flies) has been demonstrated as the master kinase driving centriole biogenesis [3–5], as PLK4 associates with mother centrioles where its level profoundly affects the number of daughter centrioles assembled [3]. Conversely, loss of PLK4 abolishes all signs of centriole duplication, revealing PLK4 as one of the most upstream regulators in the pathway. Thus, there has been a strong interest in understanding how PLK4 catalyzes centriole assembly, and in identifying the relevant substrates. The search has finally led to an exciting discovery by the group of David Glover,
published in this issue of Current Biology [6]. Centriole integrity is established through a stepwise assembly of a series of structural components, many of which serve as the building block, scaffold, or stabilizing factor. While self-assembly is a key feature of centriole duplication, the initiation step is critically guarded by enzymatic regulators to allow quantity/quality control. Centriolar proteins SAS-6 and STIL (human Ana2) are the first two components recruited to the normally single assembly site specified at the periphery of each mother centriole [7,8]. SAS-6 and likely together with STIL form the primary backbone of the cartwheel [9–11], which is the geometric scaffold promoting 9-fold symmetrical assembly of the centriole. How the assembly site is chosen or limited to one per mother centriole is not fully understood, but in vertebrate cells, the initial loading of SAS-6 or STIL is strictly dependent on the kinase PLK4 [8,12]. Moreover, when PLK4 is overexpressed, additional assembly sites can form around the mother centriole [3], where SAS-6 and STIL are loaded to promote extra daughter centriole formation. A few PLK4 substrates have been reported [13–16], but a direct link to the
19. Golowasch, J., Abbott, L.F., and Marder, E. (1999). Activity-dependent regulation of potassium currents in an identified neuron of the stomatogastric ganglion of the crab Cancer borealis. J. Neurosci. 19, RC33. 20. Marder, E., and Tang, L.S. (2010). Coordinating different homeostatic processes. Neuron 66, 161–163.
Department of Biology, Emory University, O. Wayne Rollins Research Center, Room 2105, 1510 Clifton Road, Atlanta, GA 30322, USA. E-mail: [email protected]
http://dx.doi.org/10.1016/j.cub.2014.09.064
initiation step of centriole assembly remains obscure. Finding the relevant substrates of PLK4, and knowing how the phosphorylation triggers centriole assembly are no doubt two of the most urgent goals in centrosome biology. Using in vitro kinase assay with purified Drosophila components, Dzhindzhev et al. found that Ana2 (fly STIL), but not SAS-6, can be strongly phosphorylated by PLK4, consistent with the physical interaction between STIL and PLK4 reported recently [6,17]. Four phosphorylation sites (S318, S365, S370 and S373) were identified, and all of them are clustered within the STAN motif [18], a conserved domain found in all Ana2/STIL orthologues essential for centriole duplication. Importantly, the same phosphorylation sites can also be detected in vivo in cells overexpressing Ana2 and PLK4. Moreover, the authors showed that these phosphorylation sites are functionally relevant, as the non-phosphorylatable Ana2 in which the 4 serine residues are changed to alanine (Ana2-4A) fails to rescue or drive centriole duplication. To understand the function of Ana2 phosphorylation, Dzhindzhev et al. analyzed the interaction of Ana2 with other centriolar components. In vitro binding assay showed that wild-type Ana2 (Ana2-WT) strongly associates with SAS-6 when active PLK4 is present, whereas no interaction between SAS-6 and Ana2-4A could be detected under the same condition [6]. Such kinase activity-dependent interactions were further confirmed in cells in vivo, with wild-type Ana2,
