Involvementof C-type inactivationgating in the actions of voltage-gated K +channel inhibitorsYuk-Man Leung ⁎Graduate Institute of Neural and Cognitive Sciences,College of Life Sciences,China Medical University,91Hsueh Shih Road,Taichung 40402,Taiwan,ROCa b s t r a c ta r t i c l e i n f o Keywords:Voltage-gated K +channels BlockersC-type inactivation GatingVoltage-gated K +(Kv)channels serve multiple functions.Besides the most well known function of control-ling membrane excitability,they may also play roles in cell death and differentiation.Pharmacological activa-tors and inhibitors of Kv channels therefore offer potential therapeutic treatments for a variety of diseases.of Kv channels by classical blockers such as tetraethylammonium and 4-aminopyridine,and peptides such as scorpion toxins,are believed to result from a direct intervention or occlusion of the permeation pathway.During prolonged depolarization,most Kv channels undergo a process called slow or C-type inactivation,by which the selectivity filter destabilizes and thus limits K +flux.Increasing amount of evidence shows that there are certain compounds which inhibit Kv currents not by directly obstructing the K +conduction pathway,but by accelerating or intensifying selectivity filter destabilization once the channels open.This mode of block represents an alternative mechanism of Kv channel inhibition.Indeed,some of the classical Kv channel blockers are to some extent,or in certain circumstances,involved in hastening slow inactivation.This review begins with a brief description of structure –functions of Kv chan-nels,and then discusses the multiple mechanisms of Kv channel inhibition by classical blockers and how cer-tain compounds inhibit Kv channels by accelerating C-type inactivation.The pharmacological and therapeutic potentials of these C-type inactivation-dependent Kv channel inhibitors are discussed.©2011Elsevier Inc.All rights reserved.Contents1.Introduction .............................................1512.Voltage-gated K +channel gating and C-type inactivation........................1523.Classical blockers of voltage-gated K +channels:extracellular block ....................1534.Classical blockers of voltage-gated K +channels:intracellular block pounds whose block is dependent on C-type inactivation gating ....................1546.Pharmacological and therapeutic potentials of C-type inactivation-dependent voltage-gated K +channel inhibitors .1567.Conclusion and perspectives .......................................157Acknowledgment ..............................................157References .................................................1571.IntroductionK +channels are integral membrane proteins which provide selec-tive conduction pathways for K +flux along the electrochemical gra-dient.Their general role is to stabilize the cell membraneelectrically.Voltage-gated K +(Kv)channels open upon depolariza-tion and the K +ef flux accounts for the repolarization phase of an ac-tion potential or acts by reducing the neuronal spiking frequency (Hille,2001).Blocking Kv channels pharmacologically prolongs the duration of an action ck of expression,mutation or re-duced activities (loss-of-functions)of certain Kv channel members may lead to neuronal hyperexcitability in rodents or humans,mani-fested as neurological disorders such as ataxia and epilepsy (Browne et al.,1994;Bernard et al.,2004;Maljevic et al.,2008;Miceli et al.,2008).Besides their most well known function in dampening cellular ex-citability,convincing evidence supports the involvement of KvPharmacology &Therapeutics 133(2012)151–158Abbreviations:4-AP,4-aminopyridine;AHR,6β-acetoxy-7α-hydroxyroyleanone;AVE0118,(2′-{[2-(4-methoxy-phenyl)-acetylamino]-methyl}-biphenyl-2-carboxylic acid (2-pyridin-3-yl-ethyl)-amide);KN-93,2-[N-(2-hydroxyethyl)]-N-(4methoxyben-zenesulfonyl)amino-N-(4-chlorocinnamyl)-N-methylbenzylamine;Kv channels,voltage-gated K +channels;QA,quaternary ammoniums;TEA,tetraethylammonium.⁎Tel.:+886422053366x2185;fax:+886422076853.E-mail address:ymleung@.tw .0163-7258/$–see front matter ©2011Elsevier Inc.All rights reserved.doi:10.1016/j.pharmthera.2011.10.005Contents lists available at SciVerse ScienceDirectPharmacology &Therapeuticsj o u r n a l h o m e p a ge :w ww.e l s e v i e r.c o m/l o c a t e /p ha rm t h e r achannels in cell death and differentiation of a wide variety of cells.For instance,there is an overexpression of Kv channels during apoptosis of neurons,immune cells and muscle cells(Yu et al.,1997;Yu, 2003).The massive K+efflux through over-expressed Kv channels leads to loss of intracellular K+and the reduced[K+]i results in relief of inhibition(thus activation)of pro-apoptotic caspases and nucle-ases(Hughes et al.,1997;Hughes &Cidlowski,1999).Kv channels have also been implicated in the axon development in Xenopus retinal cells,as pharmacological inhibition of Kv channels suppresses axon extension and produces a prominent effect on guidance cues (McFarlane&Pollock,2000;Pollock et al.,2005).More recently,it was found that K+efflux through Kv1.1,Kv1.4and Kv2.1was respon-sible for cAMP-stimulated neuritogenesis in mouse neuroblastoma N2A cells(Leung et al.,2011).All these evidence suggests that Kv channels are versatile molecules which have a variety of functions in addition to provision of K+efflux for repolarization.Specific Kv channel blockers are useful tools for probing the func-tional roles of Kv channels.On the other hand,Kv channel blockers may offer therapeutic opportunities by their ability to enhance cellu-lar excitability.Augmenting conductivity of demyelinated axons via inhibition of Kv channels by4-aminopyridine(4-AP)is a potential therapeutic treatment of multiple sclerosis(Judge&Bever,2006). Kv channel blockers such as4-AP and tetraethylammonium(TEA) could be employed to prevent neuronal apoptosis by suppressing ex-cessive K+efflux(Yu et al.,1997;Yu,2003;Hu et al.,2006).The po-tential of using Kv channel blockers for neuroprotection is discussed in a recent review(Leung,2010).The block of Kv channels by classical blockers(TEA and4-AP)is fast,and it is believed that these molecules directly interfere with the K+conduction pathway or even plug the channel pore.A process by which Kv channels limit K+efflux during continuous depolariza-tion is destabilization or distortion of the selectivityfilter(Hoshi et al.,1991;see Kurata&Fedida,2006for an excellent review).This has been coined slow inactivation or C-type inactivation(see next section).Increasing amount of evidence has shown that there are some compounds which do not appear to act by directly obstructing the pore;instead,they inhibit the channel by hastening or intensify-ing the destabilization of the selectivityfilter.This review begins by describing the fundamental structure-function aspects of Kv channel gatings and then recapitulates some basic understanding of the ac-tions of classical Kv channel blockers(especially TEA).The observa-tions that some classical blockers,to certain extent or under special circumstances,are involved in promoting Kv channel slow inactiva-tion are explained.The mode of actions of C-type inactivation-dependent blockers,as exemplified by several recently reported drugs,will be elaborated.Finally,the pharmacological and potential therapeutic values of C-type inactivation-dependent blockers are discussed.2.Voltage-gated K+channel gating and C-type inactivation2.1.Voltage-gated K+channel gatingBased on sequence homology,Kv channels are classified into Kv1 to Kv12subfamilies;each Kv subfamily contains multiple subtype members.Theα-subunit of Kv channel is the conducting subunit, tetramer polypeptides clustering around a central pore There are six transmembrane helices(S1–S6)The S5–S6linker is a pore-forming P-loop,orientate to constitute sensorma membrane is depolarized and S5–S6helices then undergo conformational changes,thus opening the cytoplasmic activation gate to allow K+efflux(Choe et al.,1999;Yellen,2002).On the basis of gating behaviors,Kv channels could be broadly classified into two major types,namely,A-type K+channels and delayed rectifiers.A-type K+channels,exemplified by Kv1.4,Kv3.3 and Kv4.2,are fast-inactivating Kv channels with low activation thresholds(around−60to−40mV)(Hollerer-Beitz et al.,1999; Hille,2001;Sacco et al.,2006).These channels serve to control the spiking orfiring frequency.The fast inactivation,usually with time constants around tens of milliseconds,results from the rapid occlu-sion by the cytoplasmic N-terminus at the internal vestibule of the opened channel—“ball-and-chain mechanism”(Hoshi et al.,1990; Zagotta et al.,1990;Hoshi et al.,1991;Kurata&Fedida,2006).There-fore fast inactivation is also called N-type inactivation.Delayed recti-fiers(e.g.Kv1.2,Kv2.1,Kv3.1)have higher activation thresholds (around−30to−20mV)and inactivate very slowly,usually with time constants in the range of seconds.Such slow inactivation(also termed C-type inactivation)is generally thought to involve a localized conformational change at the channel outer pore mouth,such as a distortion or destabilization of the selectivityfilter during persistent depolarization(Hoshi et al.,1991;Kurata&Fedida,2006;Cordero-Morales et al.,2007)(see Fig.1).However,slow inactivation may also involve more global conformational change of the channel(see below).2.2.C-type inactivation gatingIn Shaker channels,the role of T449,a residue at the external pore mouth,in C-type inactivation has been studied in details.Mutation of this amino acid to arginine,lysine,alanine or glutamate accelerates C-type inactivation,while mutation to valine or tyrosine retards C-type inactivation(Lopez-Barneo et al.,1993).However,how pivotal this residue determines C-type inactivation rate is still questionable,as this site does not appear to produce similar effects in mammalian Kv1C-type inactivation(Rasmusson et al.,1995;Fedida et al.,1999).A mechanistic model was proposed in which the interaction strength(via hydrogen bonds)between residues in the selectivityfil-ter and the adjacent pore helix is crucial for C-type inactivation (Cordero-Morales et al.,2007).These authors adopted the prokaryot-ic proton-activated KcsA channel as a model.Upon extracellular acid-ification,KcsA channels open and immediately undergo C-type inactivation at a decent rate.Interaction via hydrogen bonds between E71and D80(equivalent to Shaker438and447)is important in con-trolling C-type inactivation rate.Mutating glutamate at71to several amino acids including glycine,alanine and valine strongly retards in-activation.Kv1.2WT channels inactivate very slowly.Introducing equivalent mutations in Kv1.2yields expected results.Thus,in Kv1.2,mutating Val370into Glu370greatly speeds up inactivation as this manipulation renders possible a stronger interaction with Asp379;such interaction acts like a spring to pull and collapse the se-lectivityfilter(Cordero-Morales et al.,2007).Recent evidence reveals that hydrogen bonding between D80and another amino acid,W67,is also crucial in determining the rate and extent of C-type inactivation in the KcsA channel(Cordero-Morales et al.,2011).Similarly,it was shown in the same work that W434–D447interaction and W366–D379interaction are crucial in controlling C-type inactivation in Shaker and Kv1.2,respectively.It is interesting to note that the V370–D379interaction(Fig.2)in the Kv1.2channel or the E71–D80interaction in the KcsA channel does not only control C-type inactivation(pore stability),but also af-fects K+selectivity(Chao et al.,2010;Cheng et al.,2011).In these re-ports,it was shown that weakening such interactions(thus reduced C-type inactivation)results in increased Na+permeability.Therefore, mechanisms governing C-type inactivation appear to be intimately coupled to the control of cationic selectivity.152Y.-M.Leung/Pharmacology&Therapeutics133(2012)151–158A study using solution NMR spectroscopy revealed that the selec-tivity filter of C-type inactivated KcsA channel has a higher af finity for water than K +;the latter's permeation is thus obstructed by water molecules during C-type inactivation (Imai et al.,2010).This finding is supported by a recent molecular mechanics simulation study (Boiteux &Bernèche,2011).In a report by the Perozo lab (Cuello et al.,2010),five crystal structures of the KcsA channels at different degrees of open conformations were studied.Subtle changes in the 4K +binding sites (S1–S4)in the selectivity filter were observed which may represent C-type inactivation:an initial loss of the S2binding site (S2destabilization)and a subsequent loss of the S3bind-ing site (S3destabilization).Evidence suggests that C-type inactivation may not only involve localized conformational change at the selectivity filter,but may also have a global impact on the channel conformation (Jiang et al.,2003).In this report,using N-terminal-deleted mutant of Kv1.4expressed in Xenopus oocytes,the authors showed that the internal vestibule indeed shrinks in size during C-type inactivation;the vol-ume change is similar to that which happens during activation and deactivation.3.Classical blockers ofvoltage-gated K +channels:extracellular block3.1.Occlusion of the K +conduction pathway by tetraethylammonium TEA could block extracellularly and intracellularly.Early mutagen-esis studies showed that TEA possibly binds to T449(only three amino acids external to the GYG sequence)of Shaker (Heginbotham &MacKinnon,1992).Since T449has also been found to be a residue affecting C-type inactivation (see above)and often TEA block is ac-companied by a slowing of C-type inactivation,it appears reasonable to conceive that TEA inhibits C-type inactivation (i.e.,prevents selec-tivity filter constriction)in a “foot-in-the-door ”fashion.Subsequent mutagenesis work revealed that the TEA binding site(s)are rather diffuse,since other residues at the P-loop are also responsible for TEA binding,namely,T370,T373,Y376,D378,Y380and K382of Kv2.1(Pascual et al.,1995).It was thus proposed that TEA does not bind like a tight plug;rather,TEA resides in the aqueous environment immediately outside the outer lip (Fig.1).3.2.Tetraethylammonium action may involve C-type inactivation under special circumstancesAlthough TEA block of wild type Kv channels has been considered not associative with enhancement of selectivity filter constriction,a twist in our understanding of its action comes from a report suggest-ing that TEA might hasten C-type inactivation gate closing under cer-tain circumstances.Ahern et al.(2009)showed that in the Shaker T449F mutant,inactivation rate is much reduced,as the aromatic rings of the four Phe residues repel each other.Addition of TEA,in-stead of causing the “foot-in-the-door ”effect,indeed acceleratesC-Fig.2.Molecular graphics of the Kv1.2channel.The Kv1.2channel is a tetramer of 4polypeptides.Only one monomer is shown in this diagram.Each monomer has six transmembrane domains:S1and S2(yellow),S3and S4(green),S5and S6(red).The P-loop between S5and S6is shown in purple color.Val370and Asp379are repre-sented by the small yellow sticks.intensified BATEA blockNormal conductionDCC-type inactivationC-type inactivationintensifiedFig.1.Schematic diagrams of TEA block and C-type inactivation in Kv channels.(A )Normal conduction of K +through the selectivity filter of the Kv channel.(B )TEA resides at theouter pore mouth,blocking the ef flux of K +.(C )C-type inactivation happens when the selectivity filter of the Kv channel constricts or destabilizes during prolonged depolarization,thus obstructing K +out flow.(D )A drug (purple square)binds to an extracellular domain and accelerates/intensi fies C-type inactivation.153Y.-M.Leung /Pharmacology &Therapeutics 133(2012)151–158type inactivation.The authors suggested that there is an interaction between the cation(TEA)and theπelectrons of the Phe aromatics, so that the inactivation gate is pulled to close rapidly,resulting in a “spring-in-the-door”mechanism.To support their hypothesis,they showed that a V438A mutation,possibly producing an allosteric reor-ientation of T449F,restores the“foot-in-the-door”effect.The authors further showed that by serialfluorination of the Phe449residues of the single mutant T449F and hence drawing electrons away from the aromatics,the“spring-in-the-door”is replaced by the“foot-in-the-door”mechanism.Thus,under certain circumstances(T449F), the classical Kv channel blocker TEA may not simply act as a pore blocker,but may exert its inhibitory effect in part through promoting C-type inactivation(Ahern et al.,2009;Olcese,2009).3.3.ToxinsVarious toxins targeting Kv channels have been isolated from the venom of a diversity of animal species,such as scorpions,snakes,con-osnails,sea anemones and spiders(Mouhat et al.,2008).These peptide toxins are usually very potent and some have been shown to be specific for a particular Kv subfamily.For instance,dendrotoxins and stromatoxin-1are specific blockers for Kv1and Kv2subtypes,respec-tively(Shiau et al.,2003;Harvey&Robertson,2004).Most of the Kv-acting toxins bind extracellularly.While some(e.g.tityustoxin-Kα) have been known to have no effect on C-type inactivation(Rodrigues et al.,2003)or interfere with C-type inactivation(e.g.κ-Conotoxin PVIIA;Koch et al.,2004),there is hitherto no toxin known to block Kv channels mainly by accelerating the C-type inactivation gate.4.Classical blockers ofvoltage-gated K+channels:intracellular block4.1.Quaternary ammoniums and quinidineaction may in part involve C-type inactivationQuaternary ammoniums(QA),4-AP and quinidine could directly block Kv channels intracellularly,that is,they could block at the inter-nal cavity once the Kv channel opens.These drugs could affect C-type inactivation.In particular,part of the blocking mechanisms of QA and quinidine involves acceleration of C-type inactivation,although these two drugs achieve this via different modes of action.It is believed that K+occupies a selective site near the outer lip of the channel pore;such occupancy tends to maintain the open state of the Kv channel and prevents its C-type inactivation(Baukrowitz& Yellen,1995).It is found that upon N-type inactivation and blockade of the internal cavity by QA,the reduced K+efflux prevents the re-plenishment of this K+occupancy site.If the extracellular concentra-tion of K+is low,there is a reduced occupancy rate of this K+-selective site;and N-type inactivation and QA blockade could thus help accelerating C-type inactivation(Baukrowitz&Yellen,1996).The membrane-permeant4-AP blocks intracellularly,and prefer-entially blocks the open state of Kv channels.Thus4-AP block de-pends on the opening of the cytoplasmic activation gate(Russell et al.,1994).Once the drug gains access into the internal cavity,it can be trapped inside after the channel closes.Interestingly,C-type inac-tivation appears to have a significant effect on the conformation of the internal cavity,as it obscures the binding of4-AP into the internal cavity.This is consistent with the observation that C-type inactivation also constricts the internal cavity size(Jiang et al.,2003).That C-type inactivation and4-AP block at the internal cavity is mutually exclu-sive is further substantiated by thefinding that4-AP block could pre-vent C-type inactivation via an allosteric mechanism in Shaker (Claydon et al.,2007).Therefore,there are certain allosteric interac-tions between the slow inactivation gate and the internal vestibule.Quinidine,an anti-arrhymic drug,binds to the internal vestibule and therefore blocks Kv channels in the open state.In a Kv1.4mutant lacking N-type inactivation,quinidine has been shown to accelerate C-type inactivation by acting from the internal vestibule via an allo-steric manner,which is believed to involve orientation of the S6 (Wang et al.,2003).These studies suggest that drugs acting intracel-lularly are also able to affect the inactivation gate closing.The differ-ence between the effects of intracellular quinidine and that of internal 4-AP on C-type inactivation remains unknown.4.2.Direct internal block without involving C-type inactivationOf note,the aforementioned allosteric drug action has to be distin-guished from the effects of a few drugs known to speed up current decay not by promoting C-type inactivation,but by blocking at the internal cavity of Kv channels in an open-channel block and time-dependent manner.The classical example of this is block of Kv chan-nels by intracellular nonyltriethylammonium or TPeA,which are TEA analogs(Armstrong,1971).Block has to wait until the cytoplasmic activation gate opens:hence open-channel block.The block by these more hydrophobic drugs shows time-dependence.Internal TEA block does not display time-dependence,since TEA has a much faster off-rate and thus block by TEA happens as soon as the channel opens.More recent observations of open-channel and time-dependent block by intracellularly applied drugs include AVE0118 (2′-{[2-(4-methoxy-phenyl)-acetylamino]-methyl}-biphenyl-2-car-boxylic acid(2-pyridin-3-yl-ethyl)-amide),an anti-arrhythmic agent, and verapamil,a classical L-type Ca2+channel antagonist.AVE0118 was found to bind to at least six residues orienting toward the inter-nal vestibule,and by plugging at the internal cavity,AVE0118also slows channel deactivation(Decher et al.,2006).Block of Kv1.3inter-nal cavity by intracellularly applied TEA reduces the efficacy of verap-amil block while outer mouth pore mutations and extracellularly applied TEA fail to produce such an effect(Rauer&Grissmer,1996). These data suggest that verapamil binds to the internal cavity.pounds whose block isdependent on C-type inactivation gating5.1.HMJ-53A,rhynchophylline and6β-acetoxy-7α-hydroxyroyleanone(AHR)A variety of drugs reportedly accelerate Kv current decay,and this could be interpreted in multiple ways which include a time-dependent occlusion of the ion permeation pathway(e.g.intracellu-lar blockers such as nonyltriethylammonium,verapamil and AVE0118discussed in the previous section)or acceleration of C-type inactivation.We recently introduced several novel Kv channel blockers which inhibit Kv channels not by obstructing the pores but by enhancing the C-type inactivation gate closing.Channel inhibition is manifested as a hastened decay of currents.The inactivation time constants are often tens of milliseconds,close to those of N-type inac-tivation.These compounds include HMJ-53A(Chao et al.,2008), rhynchophylline(Chou et al.,2009)and6β-acetoxy-7α-hydroxyroy-leanone(AHR)(Leung et al.,2010).HMJ-53A is a synthetic compound which possesses inhibitory effect on platelet aggregation(Hour& Hung,2000).Rhynchophylline is an alkaloid isolated from the hook part of Uncaria rhynchophylla Miq.(Rubiaceae)and exhibits neuro-protective functions(Kang et al.,2004).AHR,isolated from Taiwania cryptomerioides,is a diterpenoid compound previously found to have anti-oxidative effects(Wang&Wu,2002).Several characteris-tics were investigated to consider if these drugs act on the inactiva-tion gate:(1)whether their blockade of Kv channels is extracellular and whether the degree of block is independent of intracellular K+ concentration,(2)whether they have significant effect on steady-state inactivation gating,and(3)their effects in C-type inactivation-defective mutants.These aspects are elaborated as follows.154Y.-M.Leung/Pharmacology&Therapeutics133(2012)151–158HMJ-53A,rhynchophylline and AHR all act extracellularly to hasten current decay and they exert no effect when dialyzed into the cell via the recording pipette (Chao et al.,2008;Chou et al.,2009;Leung et al.,2010;also see Fig.3).These data suggest that these few drugs do not act at the internal cavity,thus distinguishing their pharmacolog-ical properties from the intracellular actions of QA and quinidine,which are able to promote C-type inactivation (by different mechanisms,as mentioned above).Further,while the intracellularly acting QA and quinidine only mildly promote C-type inactivation,the effects of HMJ-53A,rhynchophylline and AHR on C-type inactivation are drastic,de-creasing inactivation time constants to around 30–100ms.What are the mechanisms of these blocks?Are these drugs directly occluding the permeation pathway,or intensifying the destabilization of the selectivity filter?If these drugs occlude the K +permeation path-way,the drug should reside immediately outside the outer pore to block K +ef flux directly;the drug and K +ions will encounter each other at the exit of the narrowest part of the ion conduction pathway.It could be reasoned that the lower the concentration of the intracellu-lar K +,the less the physical expulsion of the extracellular drug and thus the greater the block.Our experiments showed that the block by all these agents was not signi ficantly affected when intracellular K +con-centration was reduced from 140to 70mM.These results are incom-patible with the proposal that these drugs are direct pore blockers.All the three drugs caused a signi ficant left shift in steady-state in-activation curve,further supporting the notion that they effect on the inactivation gate.They all do not affect the activation kinetics.With the exception of rhynchophylline,which causes a left-shift in activa-tion curve and thus lowers the activation threshold (see below),they do not affect the voltage-dependence of activation.As mentioned,interaction between V370and Asp379in Kv1.2is important in pore stability and C-type inactivation (Cordero-Morales et al.,2007).We mutated V370into G370and this mutant in-activates signi ficantly more slowly than the WT does (Leung et al.,2010).While AHR causes a drastic acceleration of current decay in WT Kv1.2channels,its effect on current decay is much attenuated in Kv1.2V370G (Leung et al.,2010;Fig.4).This provides a strong sup-port that the effect of AHR on current decay depends on C-type inac-tivation.Given that V370is facing the internal cavity (Fig.2)and that AHR action is extracellular,it is unlikely that AHR directly binds to this residue.Instead,AHR may bind to certain extracellular domain and then accelerate C-type inactivation,which in part hinges on the V370residue.The observation that AHR has no effect on ATP-sensitive K +channels,which are devoid of C-type inactivation,fur-ther supports that AHR inhibition of Kv channels depends on C-type inactivation gating (Leung et al.,2010).5.2.2-[N-(2-hydroxyethyl)]-N-(4methoxybenzenesulfonyl)]amino-N-(4-chlorocinnamyl)-N-methylbenzylamine,dofetilide and other drugs 2-[N-(2-hydroxyethyl)]-N-(4methoxybenzenesulfonyl)]amino-N-(4-chlorocinnamyl)-N-methylbenzylamine (KN-93)is a potent inhibi-tor of Ca 2+/calmodulin-activated protein kinase II (CaMKII).KN-93,acting extracellularly but not intracellularly,was shown to enhance slow inactivation in a number of Kv channels including Kv1.2,Kv1.5,Kv2.1and Kv3.2(Ledoux et al.,1999;Rezazadeh et al.,2006).Such an action is independent of its action on CaMKII,since an inactive analog also accelerates slow inactivation.KN-93has a small effect on activa-tion gating:slight left-shift in activation curve (Rezazadeh et al.,CADIntracell. AHRB0.40.60.81.00.00.2I e o p /I p e a kTime (s)Fig.3.Extracellular,but not intracellular,application of HMJ-53A and AHR accelerates slow inactivation of Kv currents.(A )Representative traces showing outward Kv currents in N2A cells triggered by +30mV just after whole-cell con figuration was established,after 15min of 30μM HMJ-53A dialysis and 8min after bath application of 30μM HMJ-53A to the same cell.(B )The decay time constants before and after HMJ-53A treatment as observed in (A )is plotted against time.(C )Outward Kv currents were triggered by +30mV in a Kv1.2-expressing HEK293cell.Traces shown are currents just after whole-cell con figuration was established,after 10min of 50μM AHR dialysis and 3min after bath application of 50μM AHR to the same cell.(D )The ratio of end-of-pulse current/peak current (Ieop/Ipeak)as observed in (C )are plotted against time.Panels A and B are from Chao et al.(2008)Neuropharmacology 54:1128–1135.Panels C and D are from Leung et al.(2010)Cell Mol Life Sci 67:147–156.155Y.-M.Leung /Pharmacology &Therapeutics 133(2012)151–158。
More informationPhase Noise and Frequency Stability in OscillatorsPresenting a comprehensive account of oscillator phase noise and frequency stability,this practical text is both mathematically rigorous and accessible.An in-depth treatmentof the noise mechanism is given,describing the oscillator as a physical system,andshowing that simple general laws govern the stability of a large variety of oscillatorsdiffering in technology and frequency range.Inevitably,special attention is given to am-plifiers,resonators,delay lines,feedback,andflicker(1/f)noise.The reverse engineeringof oscillators based on phase-noise spectra is also covered,and end-of-chapter exercisesare given.Uniquely,numerous practical examples are presented,including case studiestaken from laboratory prototypes and commercial oscillators,which allow the oscillatorinternal design to be understood by analyzing its phase-noise spectrum.Based on tuto-rials given by the author at the Jet Propulsion Laboratory,international IEEE meetings,and in industry,this is a useful reference for academic researchers,industry practitioners,and graduate students in RF engineering and communications engineering.Additional materials are available via /rubiola.Enrico Rubiola is a Senior Scientist at the CNRS FEMTO-ST Institute and a Professorat the Universit´e de Franche Comt´e.With previous positions as a Professor at theUniversit´e Henri Poincar´e,Nancy,and in Italy at the University Parma and thePolitecnico di Torino,he has also consulted at the NASA/Caltech Jet PropulsionLaboratory.His research interests include low-noise oscillators,phase/frequency-noisemetrology,frequency synthesis,atomic frequency standards,radio-navigation systems,precision electronics from dc to microwaves,optics and gravitation.More informationThe Cambridge RF and Microwave Engineering SeriesSeries EditorSteve C.CrippsPeter Aaen,Jaime Pl´a and John Wood,Modeling and Characterization of RF andMicrowave Power FETsEnrico Rubiola,Phase Noise and Frequency Stability in OscillatorsDominique Schreurs,M´a irt´ın O’Droma,Anthony A.Goacher and Michael Gadringer,RF Amplifier Behavioral ModelingFan Y ang and Y ahya Rahmat-Samii,Electromagnetic Band Gap Structures in AntennaEngineeringForthcoming:Sorin V oinigescu and Timothy Dickson,High-Frequency Integrated CircuitsDebabani Choudhury,Millimeter W aves for Commercial ApplicationsJ.Stephenson Kenney,RF Power Amplifier Design and LinearizationDavid B.Leeson,Microwave Systems and EngineeringStepan Lucyszyn,Advanced RF MEMSEarl McCune,Practical Digital Wireless Communications SignalsAllen Podell and Sudipto Chakraborty,Practical Radio Design TechniquesPatrick Roblin,Nonlinear RF Circuits and the Large-Signal Network AnalyzerDominique Schreurs,Microwave Techniques for MicroelectronicsJohn L.B.Walker,Handbook of RF and Microwave Solid-State Power AmplifiersPhase Noise and Frequency Stability in OscillatorsENRICO RUBIOLAProfessor of Electronics FEMTO-ST Institute CNRS and Universit´e de Franche Comt´e Besanc ¸on,FranceMore informationMore informationCAMBRIDGE UNIVERSITY PRESSCambridge,New Y ork,Melbourne,Madrid,Cape Town,Singapore,S˜a o Paulo,DelhiCambridge University PressThe Edinburgh Building,Cambridge CB28RU,UKPublished in the United States of America by Cambridge University Press,New Y orkInformation on this title:/9780521886772C Cambridge University Press2009This publication is in copyright.Subject to statutory exceptionand to the provisions of relevant collective licensing agreements,no reproduction of any part may take place withoutthe written permission of Cambridge University Press.First published2009Printed in the United Kingdom at the University Press,CambridgeA catalog record for this publication is available from the British LibraryISBN978-0-521-88677-2hardbackCambridge University Press has no responsibility for the persistence oraccuracy of URLs for external or third-party internet websites referred toin this publication,and does not guarantee that any content on suchwebsites is,or will remain,accurate or appropriate.More informationContentsForeword by Lute Maleki page ixForeword by David Leeson xiiPreface xv How to use this book xviSupplementary material xviii Notation xix 1Phase noise and frequency stability11.1Narrow-band signals11.2Physical quantities of interest51.3Elements of statistics91.4The measurement of power spectra131.5Linear and time-invariant(LTI)systems191.6Close-in noise spectrum221.7Time-domain variances251.8Relationship between spectra and variances291.9Experimental techniques30Exercises33 2Phase noise in semiconductors and amplifiers352.1Fundamental noise phenomena352.2Noise temperature and noisefigure372.3Phase noise and amplitude noise422.4Phase noise in cascaded amplifiers492.5 Low-flicker amplifiers522.6 Detection of microwave-modulated light62Exercises65 3Heuristic approach to the Leeson effect673.1Oscillator fundamentals673.2The Leeson formula72More informationvi Contents3.3The phase-noise spectrum of real oscillators753.4Other types of oscillator824Phase noise and feedback theory884.1Resonator differential equation884.2Resonator Laplace transform924.3The oscillator964.4Resonator in phase space1014.5Proof of the Leeson formula1114.6Frequency-fluctuation spectrum and Allan variance1164.7 A different,more general,derivation of the resonatorphase response1174.8 Frequency transformations1215Noise in delay-line oscillators and lasers1255.1Basic delay-line oscillator1255.2Optical resonators1285.3Mode selection1305.4The use of a resonator as a selectionfilter1335.5Phase-noise response1385.6Phase noise in lasers1435.7Close-in noise spectra and Allan variance1455.8Examples1466Oscillator hacking1506.1General guidelines1506.2About the examples of phase-noise spectra1546.3Understanding the quartz oscillator1546.4Quartz oscillators156Oscilloquartz OCXO8600(5MHz AT-cut BV A)156Oscilloquartz OCXO8607(5MHz SC-cut BV A)159RAKON PHARAO5MHz quartz oscillator162FEMTO-ST LD-cut quartz oscillator(10MHz)164Agilent10811quartz(10MHz)166Agilent noise-degeneration oscillator(10MHz)167Wenzel501-04623(100MHz SC-cut quartz)1716.5The origin of instability in quartz oscillators1726.6Microwave oscillators175Miteq DRO mod.D-210B175Poseidon DRO-10.4-FR(10.4GHz)177Poseidon Shoebox(10GHz sapphire resonator)179UWA liquid-N whispering-gallery9GHz oscillator182More informationContents vii6.7Optoelectronic oscillators185NIST10GHz opto-electronic oscillator(OEO)185OEwaves Tidalwave(10GHz OEO)188 Exercises190Appendix A Laplace transforms192References196Index202More informationForeword by Lute MalekiGiven the ubiquity of periodic phenomena in nature,it is not surprising that oscillatorsplay such a fundamental role in sciences and technology.In physics,oscillators are thebasis for the understanding of a wide range of concepts spanningfield theory and linearand nonlinear dynamics.In technology,oscillators are the source of operation in everycommunications system,in sensors and in radar,to name a few.As man’s study ofnature’s laws and human-made phenomena expands,oscillators have found applicationsin new realms.Oscillators and their interaction with each other,usually as phase locking,and withthe environment,as manifested by a change in their operational parameters,form thebasis of our understanding of a myriad phenomena in biology,chemistry,and evensociology and climatology.It is very difficult to account for every application in whichthe oscillator plays a role,either as an element that supports understanding or insight oran entity that allows a given application.In all thesefields,what is important is to understand how the physical parametersof the oscillator,i.e.its phase,frequency,and amplitude,are affected,either by theproperties of its internal components or by interaction with the environment in whichthe oscillator resides.The study of oscillator noise is fundamental to understanding allphenomena in which the oscillator model is used in optimization of the performance ofsystems requiring an oscillator.Simply stated,noise is the unwanted part of the oscillator signal and is unavoidablein practical systems.Beyond the influence of the environment,and the non-ideality ofthe physical elements that comprise the oscillator,the fundamental quantum nature ofelectrons and photons sets the limit to what may be achieved in the spectral purity of thegenerated signal.This sets the fundamental limit to the best performance that a practicaloscillator can produce,and it is remarkable that advanced oscillators can reach it.The practitioners who strive to advance thefield of oscillators in time-and-frequencyapplications cannot be content with knowledge of physics alone or engineering alone.The reason is that oscillators and clocks,whether of the common variety or the advancedtype,are complex“systems”that interact with their environment,sometimes in waysthat are not readily obvious or that are highly nonlinear.Thus the physicist is needed toidentify the underlying phenomenon and the parameters affecting performance,and theengineer is needed to devise the most effective and practical approach to deal with them.The present monograph by Professor Enrico Rubiola is unique in the extent to which itsatisfies both the physicist and the engineer.It also serves the need to understand bothMore informationx Forewordsthe fundamentals and the practice of phase-noise metrology,a required tool in dealingwith noise in oscillators.Rubiola’s approach to the treatment of noise in this book is based on the input–output transfer functions.While other approaches lead to some of the same results,this treatment allows the introduction of a mathematical rigor that is easily tractable byanyone with an introductory knowledge of Fourier and Laplace transforms.In particular,Rubiola uses this approach to obtain a derivation,fromfirst principles,of the Leesonformula.This formula has been used in the engineering literature for the noise analysisof the RF oscillator since its introduction by Leeson in1966.Leeson evidently arrivedat it without realizing that it was known earlier in the physics literature in a differentform as the Schawlow–Townes linewidth for the laser oscillator.While a number ofother approaches based on linear and nonlinear models exist for analyzing noise inan oscillator,the Leeson formula remains particularly useful for modeling the noisein high-performance oscillators.Given its relation to the Schawlow–Townes formula,it is not surprising that the Leeson model is so useful for analyzing the noise in theoptoelectronic oscillator,a newcomer to the realm of high-performance microwave andmillimeter-wave oscillators,which are also treated in this book.Starting in the Spring of2004,Professor Rubiola began a series of limited-timetenures in the Quantum Sciences and Technologies group at the Jet Propulsion Labo-ratory.Evidently,this can be regarded as the time when the initial seed for this bookwas conceived.During these visits,Rubiola was to help architect a system for themeasurement of the noise of a high-performance microwave oscillator,with the sameexperimental care that he had previously applied and published for the RF oscillators.Characteristically,Rubiola had to know all the details about the oscillator,its principleof operation,and the sources of noise in its every component.It was only then that hecould implement the improvement needed on the existing measurement system,whichwas based on the use of a longfiber delay in a homodyne setup.Since Rubiola is an avid admirer of the Leeson model,he was interested in applyingit to the optoelectronic oscillator,as well.In doing so,he developed both an approachfor analyzing the performance of a delay-line oscillator and a scheme based on Laplacetransforms to derive the Leeson formula,advancing the original,heuristic,approach.These two treatments,together with the range of other topics covered,should makethis unique book extremely useful and attractive to both the novice and experiencedpractitioners of thefield.It is delightful to see that in writing the monograph,Enrico Rubiola has so openlybared his professional persona.He pursues the subject with a blatant passion,andhe is characteristically not satisfied with“dumbing down,”a concept at odds withmathematical rigor.Instead,he provides visuals,charts,and tables to make his treatmentaccessible.He also shows his commensurate tendencies as an engineer by providingnumerical examples and details of the principles behind instruments used for noisemetrology.He balances this with the physicist in him that looks behind the obvious forthe fundamental causation.All this is enhanced with his mathematical skill,of which healways insists,with characteristic modesty,he wished to have more.Other ingredients,missing in the book,that define Enrico Rubiola are his knowledge of ancient languagesMore informationForewords xi and history.But these could not inform further such a comprehensive and extremelyuseful book on the subject of oscillator noise.Lute MalekiNASA/Caltech Jet Propulsion Laboratoryand OEwaves,Inc.,February2008More informationForeword by David LeesonPermit me to place Enrico Rubiola’s excellent book Phase Noise and Frequency Stabilityin Oscillators in context with the history of the subject over the pastfive decades,goingback to the beginnings of my own professional interest in oscillator frequency stability.Oscillator instabilities are a fundamental concern for systems tasked with keeping anddistributing precision time or frequency.Also,oscillator phase noise limits the demod-ulated signal-to-noise ratio in communication systems that rely on phase modulation,such as microwave relay systems,including satellite and deep-space parablyimportant are the dynamic range limits in multisignal systems resulting from the mask-ing of small signals of interest by oscillator phase noise on adjacent large signals.Forexample,Doppler radar targets are masked by ground clutter noise.These infrastructure systems have been well served by what might now be termedthe classical theory and measurement of oscillator noise,of which this volume is acomprehensive and up-to-date tutorial.Rubiola also exposes a number of significantconcepts that have escaped prior widespread notice.My early interest in oscillator noise came as solid-state signal sources began to beapplied to the radars that had been under development since the days of the MIT RadiationLaboratory.I was initiated into the phase-noise requirements of airborne Doppler radarand the underlying arts of crystal oscillators,power amplifiers,and nonlinear-reactancefrequency multipliers.In1964an IEEE committee was formed to prepare a standard on frequency stability.Thanks to a supportive mentor,W.K.Saunders,I became a member of that group,whichincluded leaders such as J.A.Barnes and L.S.Cutler.It was noted that the independentuse of frequency-domain and time-domain definitions stood in the way of the develop-ment of a common standard.To promote focused interchange the group sponsored theNovember1964NASA/IEEE Conference on Short Term Frequency Stability and editedthe February1966Special Issue on Frequency Stability of the Proceedings of the IEEE.The context of that time included the appreciation that self-limiting oscillators andmany systems(FM receivers with limiters,for example)are nonlinear in that theylimit amplitude variations(AM noise);hence the focus on phase noise.The modestfrequency limits of semiconductor devices of that period dictated the common usage ofnonlinear-reactance frequency multipliers,which multiply phase noise to the point whereit dominates the output noise spectrum.These typical circuit conditions were secondnature then to the“short-term stability community”but might not come so readily tomind today.More informationForewords xiii Thefirst step of the program to craft a standard that would define frequency stabilitywas to understand and meld the frequency-and time-domain descriptions of phaseinstability to a degree that was predictive and permitted analysis and optimization.Bythe time the subcommittee edited the Proc.IEEE special issue,the wide exchange ofviewpoints and concepts made it possible to synthesize concise summaries of the workin both domains,of which my own model was one.The committee published its“Characterization of frequency stability”in IEEE Trans.Instrum.Meas.,May1971.This led to the IEEE1139Standards that have served thecommunity well,with advances and revisions continuing since their initial publication.Rubiola’s book,based on his extensive seminar notes,is a capstone tutorial on thetheoretical basis and experimental measurements of oscillators for which phase noiseand frequency stability are primary issues.In hisfirst chapter Rubiola introduces the reader to the fundamental statistical de-scriptions of oscillator instabilities and discusses their role in the standards.Then in thesecond chapter he provides an exposition of the sources of noise in devices and circuits.In an instructive analysis of cascaded stages,he shows that,for modulative or parametricflicker noise,the effect of cascaded stages is cumulative without regard to stage gain.This is in contrast with the well-known treatment of additive noise using the Friisformula to calculate an equivalent input noise power representing noise that may originateanywhere in a cascade of real amplifiers.This example highlights the concept that“themodel is not the actual thing.”He also describes concepts for the reduction offlickernoise in amplifier stages.In his third chapter Rubiola then combines the elements of thefirst two chapters toderive models and techniques useful in characterizing phase noise arising in resonatorfeedback oscillators,adding mathematical formalism to these in the fourth chapter.Inthefifth chapter he extends the reader’s view to the case of delay-line oscillators suchas lasers.In his sixth chapter,Rubiola offers guidance for the instructive“hacking”ofexisting oscillators,using their external phase spectra and other measurables to estimatetheir internal configuration.He details cases in which resonatorfluctuations mask circuitnoise,showing that separately quantifying resonator noise can be fruitful and that devicenoisefigure and resonator Q are not merely arbitraryfitting factors.It’s interesting to consider what lies ahead in thisfield.The successes of today’sconsumer wireless products,cellular telephony,WiFi,satellite TV,and GPS,arise directlyfrom the economies of scale of highly integrated circuits.But at the same time thisintroduces compromises for active-device noise and resonator quality.A measure ofthe market penetration of multi-signal consumer systems such as cellular telephonyand WiFi is that they attract enough users to become interference-limited,often fromsubscribers much nearer than a distant base station.Hence low phase noise remainsessential to preclude an unacceptable decrease of dynamic range,but it must now beachieved within narrower bounds on the available circuit elements.A search for new understanding and techniques has been spurred by this requirementfor low phase noise in oscillators and synthesizers whose primary character is integrationand its accompanying minimal cost.This body of knowledge is advancing througha speculative and developmental phase.Today,numerical nonlinear circuit analysisMore informationxiv Forewordssupports additional design variables,such as the timing of the current pulse in nonlinearoscillators,that have become feasible because of the improved capabilities of bothsemiconductor devices and computers.Thefield is alive and well,with emerging players eager tofind a role on the stage fortheir own scenarios.Professionals and students,whether senior or new to thefield so ablydescribed by Rubiola,will benefit from his theoretical rigor,experimental viewpoint,and presentation.David B.LeesonStanford UniversityFebruary2008More informationPrefaceThe importance of oscillators in science and technology can be outlined by two mile-stones.The pendulum,discovered by Galileo Galilei in the sixteenth century,persistedas“the”time-measurement instrument(in conjunction with the Earth’s rotation period)until the piezoelectric quartz resonator.Then,it was not by chance that thefirst inte-grated circuit,built in September1958by Jack Kilby at the Bell Laboratories,was aradio-frequency oscillator.Time,and equivalently frequency,is the most precisely measured physical quantity.The wrist watch,for example,is probably the only cheap artifact whose accuracy ex-ceeds10−5,while in primary laboratories frequency attains the incredible accuracy ofa few parts in10−15.It is therefore inevitable that virtually all domains of engineeringand physics rely on time-and-frequency metrology and thus need reference oscillators.Oscillators are of major importance in a number of applications such as wireless com-munications,high-speed digital electronics,radars,and space research.An oscillator’srandomfluctuations,referred to as noise,can be decomposed into amplitude noise andphase noise.The latter,far more important,is related to the precision and accuracy oftime-and-frequency measurements,and is of course a limiting factor in applications.The main fact underlying this book is that an oscillator turns the phase noise of itsinternal parts into frequency noise.This is a necessary consequence of the Barkhausencondition for stationary oscillation,which states that the loop gain of a feedback oscillatormust be unity,with zero phase.It follows that the phase noise,which is the integral ofthe frequency noise,diverges in the long run.This phenomenon is often referred to asthe“Leeson model”after a short article published in1966by David B.Leeson[63].Onmy part,I prefer the term Leeson effect in order to emphasize that the phenomenon isfar more general than a simple model.In2001,in Seattle,Leeson received the W.G.Cady award of the IEEE International Frequency Control Symposium“for clear physicalinsight and[a]model of the effects of noise on oscillators.”In spring2004I had the opportunity to give some informal seminars on noise in oscil-lators at the NASA/Caltech Jet Propulsion Laboratory.Since then I have given lecturesand seminars on noise in industrial contexts,at IEEE symposia,and in universities andgovernment laboratories.The purpose of most of these seminars was to provide a tuto-rial,as opposed to a report on advanced science,addressed to a large-variance audiencethat included technicians,engineers,Ph.D.students,and senior scientists.Of course,capturing the attention of such a varied audience was a challenging task.The stimu-lating discussions that followed the seminars convinced me I should write a workingMore informationxvi Prefacedocument1as a preliminary step and then this book.In writing,I have made a seriouseffort to address the same broad audience.This work could not have been written without the help of many people.The gratitudeI owe to my colleagues and friends who contributed to the rise of the ideas containedin this book is disproportionate to its small size:R´e mi Brendel,Giorgio Brida,G.JohnDick,Michele Elia,Patrice F´e ron,Serge Galliou,Vincent Giordano,Charles A.(Chuck)Greenhall,Jacques Groslambert,John L.Hall,Vladimir S.(Vlad)Ilchenko,LaurentLarger,Lutfallah(Lute)Maleki,Andrey B.Matsko,Mark Oxborrow,Stefania R¨o misch,Anatoliy B.Savchenkov,Franc¸ois Vernotte,Nan Yu.Among them,I owe special thanks to the following:Lute Maleki for giving me theopportunity of spending four long periods at the NASA/Caltech Jet Propulsion Labora-tory,where I worked on noise in photonic oscillators,and for numerous discussions andsuggestions;G.John Dick,for giving invaluable ideas and suggestions during numerousand stimulating discussions;R´e mi Brendel,Mark Oxborrow,and Stefania R¨o misch fortheir personal efforts in reviewing large parts of the manuscript in meticulous detail andfor a wealth of suggestions and criticism;Vincent Giordano for supporting my effortsfor more than10years and for frequent and stimulating discussions.I wish to thank some manufacturers and their local representatives for kindness andprompt help:Jean-Pierre Aubry from Oscilloquartz;Vincent Candelier from RAKON(formerly CMAC);Art Faverio and Charif Nasrallah from Miteq;Jesse H.Searles fromPoseidon Scientific Instruments;and Mark Henderson from Oewaves.Thanks to my friend Roberto Bergonzo,for the superb picture on the front cover,entitled“The amethyst stairway.”For more information about this artist,visit the website.Finally,I wish to thank Julie Lancashire and Sabine Koch,of the Cambridge editorialstaff,for their kindness and patience during the long process of writing this book.How to use this bookLet usfirst abstract this book in one paragraph.Chapter1introduces the language ofphase noise and frequency stability.Chapter2analyzes phase noise in amplifiers,includ-ingflicker and other non-white phenomena.Chapter3explains heuristically the physicalmechanism of an oscillator and of its noise.Chapter4focuses on the mathematics thatdescribe an oscillator and its phase noise.For phase noise,the oscillator turns out to bea linear system.These concepts are extended in Chapter5to the delay-line oscillatorand to the laser,which is a special case of the latter.Finally,Chapter6analyzes indepth a number of oscillators,both laboratory prototypes and commercial products.Theanalysis of an oscillator’s phase noise discloses relevant details about the oscillator.There are other books about oscillators,though not numerous.They can be divided intothree categories:books on radio-frequency and microwave oscillators,which generallyfocus on the electronics;books about lasers,which privilege atomic physics and classical1E.Rubiola,The Leeson Effect–Phase Noise in Quasilinear Oscillators,February2005,arXiv:physics/0502143,now superseded by the present text.PrefacexviideeperreadingbasictheoreticaladvancedtheoreticallegendexperimentalistlecturerdeeperreadingFigure1Asymptotic reading paths:on the left,for someone planning lectures on oscillatornoise;on the right,for someone currently involved in practical work on oscillators.optics;books focusing on the relevant mathematical physics.The present text is uniquein that we look at the oscillator as a system consisting of more or less complex interactingblocks.Most topics are innovative,and the overlap with other books about oscillatorsor time-and-frequency metrology is surprisingly small.This may require an additionaleffort on the part of readers already familiar with the subject area.The core of this book rises from my experimentalist soul,which later became con-vinced of the importance of the mathematics.The material was originally thought anddrafted in the following(dis)order(see Fig.1):3Heuristic approach,6Oscillator hack-ing,4Feedback theory,5Delay-line oscillators.Thefinal order of subjects aims at amore understandable presentation.In seminars,I have often presented the material in the3–6–4–5order.Y et,the best reading path depends on the reader.T wo paths are suggestedin Fig.1for two“asymptotic”reader types,i.e.a lecturer and experimentalist.Whenplanning to use this book as a supplementary text for a university course,the lecturer More information。
