Excitation of Rat Sympathetic Neurons via M1

Muscarinic Receptors Independently of Kv7 channels

DOCTORAL THESIS AT THE MEDICAL UNIVERSITY OF VIENNA for obtaining the academic degree

Doctor of Philosophy

Submitted by Isabella Salzer, MD

Supervisor Prof. Dr. Stefan Böhm

ZENTRUM FÜR PHYSIOLOGIE UND PHARMAKOLOGIE, ABTEILUNG FÜR NEUROPHYSIOLOGIE UND -PHARMAKOLOGIE,

Schwarzspanierstraße 17, 1090 Wien

Vienna, 31. 1. 2014 Contents

List of Figures ii

List of Tables iii

List of Abbreviations iv

Acknowledgements vii

Abstract viii

Zusammenfassung ix

1 Introduction 1 1.1 Basics of nerve cell communication ...... 1 1.1.1 TheRestingMembranePotential...... 1 1.1.2 Theactionpotential...... 2 1.1.3 Synaptic transmission ...... 2 1.2Thesympatheticnervoussystem...... 4 1.3 Acetylcholine and its receptors ...... 4 1.3.1 Acetylcholine ...... 4 1.3.2 Nicotinic acetylcholine receptors ...... 5 1.3.3 Muscarinic acetylcholine receptors ...... 5 1.4Typesofionchannels...... 8 1.4.1 Ligand-gatedionchannels...... 10 1.4.2 Voltage-gatedionchannels...... 10 1.4.3 Ionchannelsneithergatedbyligandsnorvoltage...... 13 1.5 Modulation of ion channels by G-protein coupled receptors ...... 19 1.5.1 Modulation of ion channels by depletion of phosphatidylinositol 4,5 bis- phosphate ...... 21 1.5.2 Modulation of ion channels via release of calcium from intracellular stores 22 1.5.3 Modulation of ion channels via activation of protein kinase C ...... 23 1.6 Aim of the project ...... 23

2 Results 25

3 Discussion 65

Bibliography 79

i List of Figures

1.1 Nicotinic acteylcholine receptors ...... 6 1.2 Muscarinic acetylcholine receptors ...... 9 1.3 Kv7voltage-gatedpotassiumchannels...... 17 1.4 Ca2+-activated Cl− channels...... 20 1.5 Modulation of ion channels by Gαq-coupled GPCRs ...... 22

ii List of Tables

1.1 Muscarinic Acetylcholine Receptors ...... 8 1.2Ligand-gatedionchannels...... 11 1.3Voltage-gatedionchannels...... 12 1.4Ionchannelsnotprimarilygatedbyaligandorvoltage...... 13 1.5 Calcium-activated chloride channels ...... 21 1.6ProteinkinaseCisoforms...... 24

iii List of Abbreviations

5-HT serotonin ACh acetylcholine ADP adenosine diphosphate AMP adenosine monophosphate AMPA α-amino-3-hydroxy-5-methyl-4-isoxazoleproprionic acid ANO anoctamin ASIC acid sensing ion channels ATP adenosine triphosphate

B2 bradykinin B2 receptor

B2R bradykinin B2 receptor BK big-conductance K+ channel 2+ Cav voltage-gated Ca channel CaCC Ca2+-activated Cl− channel CaM calmodulin cAMP cyclic adenosine monophosphate CFTR cystic fibrosis transmembrane conductance regulator cGMP cyclic guanosine monophosphate ClC chloride channels CLCA Ca2+-activated Cl− channel family CNG cyclic nucleotide gated channels CNS central nervous system DAG diacylglycerol DIDS 5-isothiocyanato-2-[2-(4-isothiocyanato-2-sulfophenyl)ethenyl]benzene-1-sulfonic acid DRG dorsal root ganglion e. p. p. end plate potential ELKS active zone protein ENaC epithelial sodium channel EPSP excitatory postsynaptic potential ER endoplasmatic reticulum

iv List of Tables

GABA γ-aminobutyric acid GDP guanosine diphosphate GPCR G-protein coupled receptor GRK2 G-protein coupled receptor kinase 2 GTP guanosine triphosphate IK intermediate-conductance K+ channel + IKR rapid delayed rectifying K current + IKS slow delayed rectifying K current

IP3 inositol 1,4,5 trisphosphate

IP3R inositol 1,4,5 trisphosphate receptor IV curve current-voltage relation + K2P two-pore K channels 2+ + KCa Ca -activated K channels + Kir inward rectifier K channels + Kv voltage-gated K channels

KCNQ gene name of Kv7 channels

M1-M5 muscarinic M1 -M5 receptors

M1R muscarinic M1 receptor M1-4 membrane spanning helices of nAChRs mAChR muscarinic acetylcholine receptor MT3 mamba toxin 3 MT7 mamba toxin 7 NA noradrenaline + Navi voltage-insensitive Na leak channel + Nav voltage-gated Na channel nAChR nicotinic acetylcholine receptor NCS1 neuronal calcium sensor 1 NKCC1 Na+-K+-2 Cl− co-transporter NMDA N-methyl-D-aspartate NSF N-ethylmaleimide-sensitive factor OxoM oxotremorine methiodite + + − pNa/pK /pCl Na /K /Cl permeability PE phorbolester

PIP2 phosphatitylinositol 4,5 bisphosphate

v List of Tables

PKC protein kinase C PLCβ phospholipase C β PMA phorbol-12-myristate-13-acetate PNS peripheral nervous system PTX pertussis toxin Rab3 Ras GTPase superfamily RCBM regulatory calmodulin-binding motif RCK regulating conductance of K+ RIM Rab3-interacting molecule/ Unc-10 RIM-BP RIM binding protein RyR ryanodine receptor

S1-6 membrane spanning domains of Kv SCG superior cervical ganglion SITS 2-[2-(4-acetamido-2-sulfophenyl)ethenyl]-4-isothiocyanatobenzene-1-sulfonic acid SK small-conductance K+ channel SM-protein family Sec1/Munc18 protein family SNAP-25 synaptosome associated protein of 25 kDa SNARE Soluble NSF-attachment protein receptor SUR1 sulfonyl urea receptor 1 TM transmembrane domain TMD1-8 transmembrane domain 1-8 of CaCCs TMEM16 transmembrane proteins with unknown function 16 TRP transient receptor potential cation channels TTX tetrodotoxin VRAC volume regulated anion channel ZAC zinc activated channel

vi Acknowledgements

I want to thank Stefan Böhm for believing in me and letting me continue to work in his lab - even though, I tend to mess up projects. I think he is an exceptionally good supervisor, who always tries to find the best solution for all PhD students - not just his students. I want to thank him for making a PhD students life easier at the MUW.

I want to thank Klaus Schicker: for teaching me how to use electrophysiology, for discussing theoretical aspects of the technique with me, for discussing all kinds of things with me and thereby teaching me how to be critical, for teaching me how to use a computer properly, for showing me how to maintain a functional electrophysiological setup, for drinking coffee with me and finally for sharing my affinity for Songs from the "Erste Allgemeine Verunsicherung".

I want to thank Xaver König for being very patient with everyone in the lab (including me, of course) and repairing or fixing everything that needs to be fixed. You are responsible for transforming "Schrotti" into "Flotti".

I want to thank Sylvia Pagler for reducing the bureaucracy load to an absolute minimum and always taking care of us.

I want to thank Gabi Gaupmann for taking over primary cell culture from me. This saved me a lot of time, which I could spend on experimenting and writing.

I want to thank Hend, Manu, Marco, Jae-Won, Verena, Dagmar and everyone else in the lab for creating such a nice work environment - Thank you for making my job cool!

I want to thank the FWF for funding the PhD-Program CCHD. Without proper funding, none of us would be able to work.

Finally, I want to thank my family for accepting me the way I am, for accepting my moods and for accepting when I have no time - once again. I want to thank my parents for letting me choose my career. I want to thank them for supporting me, no matter what.

vii Abstract

Slow ganglionic transmission relies on activation of muscarinic M1 receptors and the subsequent inhibition of Kv7 channels. However, the activation threshold for Kv7 channels lies at more pos- itive potentials than the actually measured membrane voltage. Hence, inhibiting a non-active channel is unlikely to underlie M1 receptor mediated slow EPSPs. Additionally, the muscarinic agonist oxotremorine M (OxoM) triggered depolarizations in sympathetic neurons of the rat, but not in all investigated neurons. On the other hand, OxoM mediated Kv7 channel inhibition in all tested neurons. Changing intra- and extracellular chloride concentrations revealed an influence of chloride on both in OxoM-triggered depolarizations as well as noradrenaline release. Two non-selective chloride channel inhibitors niflumic acid and 2-[2-(4-acetamido-2-sulfophenyl)ethenyl]-4-isothiocyanato- benzene-1-sulfonic acid (SITS) reduced depolarizations and noradrenalin release evoked by OxoM, but left OxoM-mediated Kv7 channel inhibition unaltered. A slow rising inward current at -65 mV induced by OxoM was inhibited by selective antagonists of TMEM16A (anoctamin-1) and TMEM16B (anoctamin-2), who is responsible for Ca2+-activated Cl−currents. In addition, these selective inhibitors strongly reduced OxoM-evoked noradrenalin release from neuronal cultures of rat sympathetic ganglia. Activation of M1 receptors triggers formation of diacylglycerol, which activates protein kinase C (PKC). By using different kinds of PKC inhibitors, a dependence of OxoM-induced depolariza- tions and noradrenaline release from rat sympathetic neurons on classical PKCs was revealed. However, neither M-currents nor their inhibition via M1 receptors was influenced by PKC inhi- bition. Taken together, this study shows that slow cholinergic transmission is dependent on activation 2+ − of Ca -activated Cl channels in addition to being mediated by inhibition of Kv7 channels. Additionally, this pathway involves activity of PKC.

viii Zusammenfassung

Die langsame Komponente der synaptischen Übertragung in sympathischen Ganglien wird durch Aktivierung von muskarinischen M1 Rezeptoren und die dadurch vermittelte Hemmung von Kv7 Kanälen bestimmt. Allerdings liegt der Schwellenwert für eine Aktivierung von Kv7 Kanälen bei einem positiveren Potential als -60 mV und das tatsächliche Ruhemembranpotential eher zwis- chen -64 bis -67 mV. Außerdem war die ausgelöste Depolarisation durch den muskarinischen Agonisten Oxotremorin M (OxoM) in den untersuchten Zellen äußerst variabel im Gegensatz zur gleichförmigen OxoM- mediierten Hemmung von Kv7 Kanälen. Die Veränderung von intra- und extrazellulärem Chlorid und die daraus resultierenden Änderun- gen im Zellverhalten lieferten einen ersten Hinweis auf den Einfluss eines Chloridkanals. Die unspezifischen Chloridkanal-Blocker Niflumsäure und 2-[2-(4-Acetamido-2-Sulfophenyl)ethenyl]- 4-Isothiocyanatobenzen-1-Sulfonsäure (SITS) reduzierten sowohl OxoM-induzierte Depolarisa- tionen als auch Noradrenalinfreisetzung. Auch hier wurde kein Effekt auf Kv7 Kanäle und die durch OxoM induzierte Hemmung gefunden. Bei einem Haltepotential von -65 mV, löste die Applikation von OxoM einen langsam größer werdenden Einwärtsstrom aus. Dieser war durch selektive Inhibitoren von Anoctamin-1 (TMEM16A) und Anoctamin-2 (TMEM16B), welche ve- rantwortlich für einen Ca2+-aktivierten Cl−-Strom sind, zu hemmen. Ebenso wurde die durch OxoM ausgelöste Noradrenalinfreisezung durch diese beiden Antagonisten reduziert. Aktivierung von M1 Rezeptoren führt zur Bildung von Diacylglycerol (DAG), welches wiederum Proteinkinase C (PKC) aktiviert. Im Gegensatz zur Hemmung der Kv7 Kanäle durch OxoM, waren sowohl die von OxoM ausgelösten Depolarisationen, als auch die Noradrenalinfreisetzung durch Hemmung der Proteinkinase C (PKC) reduziert. Zusammenfassend zeigt diese Studie, dass Ca2+-aktivierte Cl− Kanäle über einen PKC-abhängigen Mechanismus zur langsamen cholinergen Übertragung beitragen.

ix 1 Introduction

1.1 Basics of nerve cell communication

1.1.1 The Resting Membrane Potential

In 1902 Julius Bernstein proposed a general principle for electric currents observed not only in animal tissues like muscles, neurons or exocrine cells, but also in plant tissues. He suggested the so-called "membrane theory", where the membrane of a cell was only permeable to K+ and Cl− ions at rest, but impermeable to other ions (Bernstein 1902). In order to calculate the potential difference between the two sides of the membrane caused by different concentrations of the respective ions, he used an equation introduced by Walther Nernst (Nernst 1888). The potential difference Bernstein calculated is now widely accepted to resemble the resting membrane potential of a cell of any kind. It arises from different ion concentrations outside and inside the cell. External concentrations of ions are similar to ion concentrations in seawater, consisting of high concentrations of Na+ and Cl−, but only a low concentration of K+.Incontrast,internalCl− and Na+ concentrations are fairly low compared to the concentration of K+ (70 mM (Woodward et al. 1969), 20 mM, and 204 mM, (Brown and Scholfield 1974), respectively, in superior cervical ganglion neurons). Of course, the cell membrane is impermeable to ions. Only opening of selective ion channels causes the membrane to be semipermeable, as suggested above, and let ions flow according to their concentration gradients. Thus, opening for example of a Na+ channel would cause flow of positively charged sodium ions into the cell. This flow can only be maintained as long as the interior of the cell, and with this the membrane potential, is negative enough to let Na+ pass. The point, when the concentration gradient and the electrical power cancel each other out is called the equilibrium potential. Like Bernstein, we can use Walther Nernst’s equation to calculate this potential:

RT Co E = · ln (1.1) zF Ci Nernst equation: E (Equilibrium potential), R (gas constant [8.3145 J mol−1 K−1]), T (absolute tem- −9 perature [K]), z (valence of the ion), F (Faraday’s constant [9.6485 · 10 C]), Co (concentration outside),

Ci (concentration inside)

Typical values for superior cervical ganglion neurons are ENa=+49 mV, EK=-88 mV (Brown and Scholfield 1974), ECl=-41 mV (Woodward et al. 1969). Unlike Bernstein, we used this equation to calculate the equilibrium potential, but not the mem- brane potential. Firstly, the Nernst equation only takes one kind of ion into account. Secondly, for calculating the membrane potential also permeabilities for the respective ions need to be con- sidered. For this purpose, the Goldman-Hodgkin-Katz equation is much more suitable (Goldman

1 1 Introduction

1943; Hodgkin and Katz 1949):

RT p · [Na+] + p · [K+] + p · [Cl−] E = · ln Na o K o Cl i + + − (1.2) F pNa · [Na ]i + pK · [K ]i + pCl · [Cl ]o

+ + − Goldman-Hodgkin-Katz equation: pNa, pK , pCl (permeabilities for Na ,K ,Cl ), [Ion]o (Ion concen- tration outside), [Ion]i (Ion concentration inside)

For superior cervical ganglion neurons, following permeabilities have been determined: pK :pNa:pCl = 1:0.04:0.1 (Brown and Scholfield 1974) resulting in a membrane potential of -68 mV (Woodward et al. 1969).

1.1.2 The action potential

A local rise in the membrane potential rapidly changes the permeabilities to a ratio of K:Na:Cl = 1:20:0.45 in the squid giant axon (Hodgkin and Katz 1949). Thus, sodium will follow its electrochemical gradient into the cell and will cause the membrane potential to rise towards the equilibrium potential of sodium. The depolarization during an action potential will finally also lead to an increase in potassium permeability (Keynes 1951), which repolarizes the membrane back towards its original resting potential. The changes in membrane potential are triggered by two distinct currents. The first is a fast rising sodium current, that, after reaching a peak, will also decay fairly rapidly to low values. The second is a slow rising potassium current, that reaches a plateau, and will remain unchanged until the resting membrane potential is restored (Hodgkin and Huxley 1952b; Hodgkin and Huxley 1952a). A refractory period follows the action potential depending on both the inactivation of the sodium conductance and the delayed rise in potassium conductance (Hodgkin and Huxley 1952c; Hodgkin and Huxley 1952a). It appears logical, that a mechanism is required to reverse these ion fluxes in order to maintain the resting membrane potential and enable the formation of an action potential. In 1957, a protein, being responsible for this mechanism, was discovered (Skou 1957). The Na-K-ATPase pumps three sodium and two potassium ions against their electrochemical gradients utilizing energy in form of ATP degradation (Clausen 1998).

1.1.3 Synaptic transmission

The first description of synaptic transmission was provided by Otto Loewi in 1921. He discovered that two distinct chemical transmitters are responsible for modulatory actions of sympathetic and parasympathetic neurons on heart action (Loewi 1921). A more detailed view was given later: the nerve impulse causes release of the neurotransmitter acetylcholine, which causes the formation of an end plate potential (e. p. p.) at the nerve-muscle junction. It was known that transmitter release is Ca2+-dependent and can be blocked by botulinum toxin, whereas the e. p. p. causes a rather unspecific rise of a sodium and potassium conductance, which can be inhibited by application of curare (Katz 1962). Synaptic transmitter release only takes 100 μs from the arrival of an action potential (Rizo and Rosenmund 2008). To achieve this speed, it takes place at disc-like structures of 0.2-0.5 μm diam- eter called active zones (Südhof 2012). Active zones serve three main functions: (1) They provide

2 1 Introduction a physical link between vesicles, voltage-gated Ca2+ channels and other presynaptic components to enable fast synaptic transmission, (2) Presynaptic receptors are organized for modulation of synaptic transmission, and (3) several forms of short- and long-term synaptic plasticity are me- diated (Südhof and Rizo 2011). Fast synaptic transmission requires four steps: Docking: a fraction of vesicles is in close contact with the plasma membrane of the postsynaptic site, Priming: makes the vesicles fusion com- petent, Ca2+-triggering: transduces the Ca2+-signal to the fusion complex, and Vesicle fusion: finally releases the neurotransmitter into the synaptic cleft (Wojcik and Brose 2007). One of the key components of the fusion machinery is the so-called SNARE (Soluble NSF- attachment protein receptor) complex. The SNARE complex consists of three protein members: Syntaxin-1 and SNAP-25 (synaptosome associated protein of 25 kDa) on the presynaptic mem- brane and Synaptobrevin on the vesicular membrane. Four SNARE motifs of the three SNARE proteins form a 4-helical coiled-coil (Wojcik and Brose 2007) to bring the vesicle and the plasma membrane in close proximity and provide the energy for fusion (Rizo and Rosenmund 2008). However, the SNARE complex is not specialized to active zones of synapses, but occurs in all secretory pathways (Wojcik and Brose 2007). Fast synaptic transmission requires the presence of active zone proteins: RIM (Rab3-interacting molecule/ Unc-10), Munc13, RIM-BP (RIM binding Protein), α-liprin, ELKS, Rab3 (of Ras GT- Pase superfamily). While the specific functions of some members of the complex are well defined, others remain enigmatic. Overall, active zone proteins form single large complexes, docking and priming vesicles as well as recruiting N- and P/Q-type voltage-gated Ca2+-channels (Südhof 2012). Interaction of RIM with the Ca2+-sensor Synaptotagmin-1 and N- and P/Q-type voltage-gated Ca2+-channels allows Ca2+-sensing and fast signal transduction (Wojcik and Brose 2007). Synap- tic proteins Piccolo and bassoon guide vesicles from the backfield of the synapse to active zones (Südhof 2012). Munc18, a member of the SM-protein family (from Sec1/Munc18) ensures cor- rect assembly of the SNARE complex and organizes its spatial arrangement to keep the proteins from sliding between the fusing membranes (Südhof and Rizo 2011). Finally, the SNARE com- plex coiled-coils are dissociated by NSF (N-ethylmaleimide sensitive factor) and SNAPs (soluble NSF-attachment protein - no relation to SNAP-25) to allow cycles of vesicle fusion (Wojcik and Brose 2007). When the neurotransmitter reaches the postsynaptic site, it needs a receptor to exert its func- tion. These receptors can be divided into two groups: (1) ionotropic receptors or ligand-gated ion channels: ligand binding causes opening of an ion channel within a millisecond range. This process is therefore referred to as neurotransmission. (2) metabotropic receptors or G-protein coupled receptors (GPCR): binding of a ligand will initiate a complex signaling cascade follow- ing activation of a GTP-binding protein and can for example influence functions of voltage-gated ion channels. The latter mechanism employs its function within seconds or minutes and is there- fore termed neuromodulation (Kaczmarek and Levitan 1987). After information between two neurons is transferred, the synaptic signal needs to be termi- nated. Two termination mechanisms were identified: (1) removal of the neurotransmitter from the synaptic cleft by active reuptake via neurotransmitter transporters, (2) removal of the neu-

3 1 Introduction rotransmitter by enzymatic hydrolysis (Zimmerman and Soreq 2006).

1.2 The sympathetic nervous system

The nervous system is divided into two parts: (1) the central nervous system (CNS), consisting of the brain and the spinal cord, and (2) the peripheral nervous system (PNS). The PNS can be further separated into an afferent part, responsible for transferring sensory information to the CNS, and an efferent part. In the efferent part of the PNS one can further differentiate between the somatic and the autonomic nervous system (Langley 1921). The former, being voluntary motor neurons, release acetylcholine (ACh) to excite muscles (Dale et al. 1936). The latter is sub-classified into (1) the sympathetic nervous system, (2) the parasympathetic nervous system, and (3) the enteric nervous system (Langley 1921). Sympathetic and parasympathetic fibers often regulate body functions of the same organ in an opposing manner, like heart rate or respiratory function. Generally, both systems form extraver- tebral ganglia, thus they consist of a preganglionic part, connected with the spinal cord and the brain, and a postganglionic part innervating the target organ. However, the origin and anatomy of parasympathetic and sympathetic neurons is highly diverging. Preganglionic parasympathetic neurons originate in the brainstem and the sacral spinal cord innervating parasympathetic gan- glia lying in close proximity to target organs. Contrarily, preganglionic sympathetic neurons originate from the thoracolumbar spinal cord changing over to postganglionic neurons in par- avertebral sympathetic ganglia (Langley 1921). The major neurotransmitter for all preganglionic neurons, no matter whether sympathetic or parasympathetic is acetylcholine, whereas most post- ganglionic sympathetic neurons change to noradrenaline (NA) as a transmitter. However, post- ganglionic sympathetic neurons innervating sweat glands and postganglionic parasympathetic neurons release Ach (Perry and Talesnik 1953).

1.3 Acetylcholine and its receptors

1.3.1 Acetylcholine

Acetylcholine (ACh) was the first neurotransmitter to be discovered (Loewi 1921) and to be iden- tified both in the autonomic nervous system (Loewi and Ernst 1926) and in the neuromuscular junction (Dale et al. 1936; Brown et al. 1936). The effects of ACh on both types of junctions were found to be similar: ACh caused an increase in postsynaptic conductance in muscles (Katz 1962) and in the junction between the preganglionic neuron and the sympathetic ganglion (Blackman et al. 1963). Both effects can be blocked by δ-tubocurarine, but not by atropine, and could be prolonged by the application of physostigmin (Brown 2006). Synaptic transmission via ACh is terminated by the enzyme acetylcholine esterase which hydrolyses ACh to acetate and choline (Zimmerman and Soreq 2006). Actions of ACh can be divided into a nicotinic and a muscarinic action (Dale 1914).

4 1 Introduction

1.3.2 Nicotinic acetylcholine receptors

The ligand-gated ion channel responsive to ACh is termed nicotinic acetylcholine receptor (nAChR). Together with zinc-activated channels, GABAA, glycine, and 5-HT3 receptors, they form the Cys- loop superfamily of ionotropic receptors (Alexander et al. 2011). A common feature of receptors of this class is, that five single subunits are required to form functional channels. Vertebrates ex- press seventeen different nAChR subunits (α1-10, β1-4, γ, δ, ε). Nicotinic receptors can roughly be discriminated into muscle-type and neuronal nAChRs. Muscle-type receptors are the main excitatory neurotransmitter receptor in the neuromuscular junction. They are composed of α1, β1, γ, δ, and ε subunits forming functional (α1)2β1γδ heteromers in embryonic muscles and (α1)2β1δε heteromers in adult muscles, respectively. The remaining subunits (α2-10, β2-4) are expressed in neurons where they form receptors with a huge variety of subunit compositions. The most abundant subunit compositions in central neurons are α4β2 heteromers and α7 homomers, whereas autonomic neurons most frequently express α3β4 heteromers and α7 homomers (Millar and Gotti 2009). The large extracellular N-terminus of a subunit contains a loop formed by two conserved cysteines thirteen amino acids apart. This "cys-loop" is characteristic for the whole family of Cys-loop re- ceptors (Tsetlin et al. 2011). Additionally, the N-terminus forms one part of the ligand-binding domain. The ACh binding domain needs a principal binding site, in neuronal nAChRs provided by α-subunits (except α5), and a complementary binding site provided by β2 and β4 subunits. Hence, a functional heteromeric receptor forms two binding sites, whereas a homomeric receptor can theoretically form 5 binding sites (see 1.1C). It is not known, whether more than two ACh molecules need to be bound to open the receptor (Gotti and Clementi 2004). Furthermore, one subunit consists of four membrane-spanning α-helices (M1-M4) (see 1.1 A). M2 of all five subunits provide the inner ring of the ion permeation pore, whereas the remain- ing transmembrane domains of a subunit coil around each other to shield the inner ring from the lipids of the membrane (see 1.1 B). In closed confirmation, M2 of α-subunits are rotated forming a constricted girdle, which is the barrier for permeating ions. This distorted form of α-subunits, compared to non-α subunits, is stabilized by subunit-subunit interactions. Binding of an agonist delivers the energy needed to overcome this distortion. Now, α-subunits rearrange in a shape closer related to non-α subunits, which finally opens the channel gate (Unwin 2005). Heteromeric channels are rather unselective and permeable to Na+,Ca2+ and K+ (Schicker et al. 2008). However, homomeric α7 receptors have been shown to be 20 times more permeable to Ca2+ (Fucile 2004). Central nAChRs occur preferentially at presynaptic terminals controlling the release of ACh as autoreceptors or of noradrenaline (NA), dopamine, glutamate, or GABA as heteroreceptors (Schicker et al. 2008). In autonomic ganglia, nAChRs occur both on the presynaptic as well as the postsynaptic site where they are also responsible for fast ganglionic transmission (Gotti and Clementi 2004).

1.3.3 Muscarinic acetylcholine receptors

Muscarinic acetylcholine receptors (mAChR) are members of the large family of G-protein cou- pled receptors (GPCR), in particular of the A rhodopsin-like family (Brown 2010). GPCRs bind

5 1 Introduction

Figure 1.1: Nicotinic acteylcholine receptors

The nicotinic acetylcholine receptor is part of the pentameric cys-loop superfamily of ionotropic receptors. (A) A subunit consists of 4 hydrophobic trans- membrane domains (M1 to M4), of which M2 lines the ion-conducting pore. The large extracellular N-terminus includes the cys-loop and the agonist-binding site. Two intra- cellular loops link M1 to M2 and M2 to M3, respectively, whereas a large intracellular loop links M3 to M4. The C-terminal end is rather short. (B) demonstrates the quarternary structure of a nicotinic acetylcholine receptor. (C) shows a typical α3β4 conformation of nAChRs, occurring in autonomic ganglia and an α4β2 conformation, typically occuring in the brain. In these conformations, two binding sites are created. In the α7 conformation, there are five putative binding sites created. (adapted from Absalom et al. 2004; Gotti and Clementi 2004; Gotti et al. 2006)

6 1 Introduction to the name giving heterotrimeric GTP-binding protein (G-protein). The G-protein provides the link between the receptor and its effectors. G-proteins are formed by three subunits α, β, and γ. ?nactive G-proteins have GDP bound to α-subunits, which is exchanged to GTP upon activation of the receptor. When GDP is exchanged for GTP in α subunits, the βγ-dimer is released into the cytosol where it can interact with effector molecules. The α-subunit exerts a slow GTPase ac- tivity hydrolyzing GTP to GDP that leads to reassembly of α with βγ and terminates G-protein activity. Additionally, α-subunits also facilitate signal transduction via effector molecules (Stros- berg and Nahmias 2007). Heterotrimers are classified into four families according to the structure of Gα subunits: Gαs,Gαi,Gαq/11, and Gα12/13 (Oldham and Hamm 2008). Gαs-subunits ac- tivate the adenylylcyclase that forms cyclic AMP (cAMP), whereas Gαi-subunits inhibit this enzyme, which leads to reduction of cAMP (Cabrera-Vera et al. 2003). Gαq-subunits activate phospholipase Cβ (PLCβ), which hydrolyses phosphatitylinositol 4,5-bisphosphate (PIP2)into diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP3) (Suh and Hille 2007) (see 1.2 C). Gα12/13-subunits regulate rho-GTPase activity (Suzuki et al. 2009). Additionally, the membrane- bound βγ dimer can directly interact with effector molecules, while diffusible second messengers are not necessary. Therefore, this pathway is termed membrane-delimited and usually involves βγ subunits associated with Gαi/o-coupled receptors (Zamponi and Currie 2013). The muscarinic acetylcholine family consists of five members: M1,M2,M3,M4, and M5.The odd-numbered receptors signal via the Gαq/11 pathway, while the even-numbered receptors bind Gαi/o-subunits sensitive to pertussis-toxin (PTX). Like other GPCRs, mAChRs have seven α- helices, which form the transmembrane domains (TM1-TM7) (see 1.2 A). They are linked via three extracellular and three intracellular loops, respectively. Intracellular loops i2 and i3 are different in the odd- and even-numbered groups, but are fairly homologous within the groups, which is thought to underlie the preferential coupling to G-proteins (Caulfield and Birdsall 1998). Acetylcholine binds to the top part of TM7, a binding site consistent with the composition of those of nAChRs and acetylcholine esterases (Hulme et al. 2003) (see 1.2 B). In the inactive re- ceptor, transmembrane domain 6 resides in the so-called TM6 clamp that prevents the outward movement of this transmembrane domain. Binding of an agonist causes rupture of this clamp via an inward movement of TM7, a displacement of TM5, and a lateral movement of TM3, which weakens the TM3 component of the TM6 clamp (Hulme 2013). Prolonged activation of mAChRs causes agonist-induced desensitization, which involves receptor phosphorylation. Muscarinic re- ceptors can be phosphorylated by PKC in an agonist-independent manner and by GPCR kinases and casein kinase 1α in an agonist-dependent manner. Phosphorylated receptors can bind to β-arrestin-1 and 2 and interact with clathrin leading to internalization via clathrin-coated vesi- cles. Internalization can amount up to 30% in heterologously expressed M1 receptors, which are phorphorylated by GRK2. Muscarinic M3 receptors have been shown to form dimers in native tissues (Koppen and Kaiser 2003). Muscarinic receptors serve many functions in various kinds of tissues: M1 receptors were shown to inhibit voltage-gated potassium channels of the Kv7 family (see 1.4.2 and 1.5.1). M2 receptor me- diate reduction of contractility in guinea-pig heart, which relies on an inhibition of voltage-gated 2+ Ca -channels (Cav2) and activation of inwardly rectifying potassium channels (Kir3) (Brown 2010). Receptor-mediated relaxation of vascular smooth muscles in the guinea-pig ileum is me-

7 1 Introduction

diated by M3 receptors. Relaxation of the anococcygeus muscle is mediated by release of nitric oxide via presynaptic M4 receptors (Caulfield and Birdsall 1998). Furthermore, different kinds of mAChRs are expressed in neuronal tissues where they are involved in a variety of diseases. Unfortunately, pharmacological tools for a selective characterization are limited (see table 1.1).

Table 1.1: Muscarinic Acetylcholine Receptors

M1 M2 M3 M4 M5 G-protein Gq/11 Gi/o Gq/11 Gi/o Gq/11 Second PLC - PIP2 inhibition of PLC - PIP2 inhibition of PLC - PIP2 Messenger -IP3/DAG/ AC -IP3/DAG/ AC -IP3/DAG/ PKC PKC PKC Functional M-current inhibition of inhibition of Response inhibition Ca2+ chan- Ca2+ chan- nels nels Agonist acetylcholine, acetylcholine, acetylcholine, acetylcholine, acetylcholine, muscarine, muscarine, muscarine, muscarine, muscarine, OxoM, OxoM, OxoM, OxoM, OxoM, carbachol carbachol carbachol carbachol carbachol Antagonist atropine, atropine, atropine, atropine, atropine, pirenzepine, pirenzepine pirenzepine pirenzepine, pirenzepine mamba mamba toxin MT7 toxin MT3 Expression sympathetic thalamus, cerebral striatum, substantia profile ganglia, hind brain, cortex, hip- cerebral nigra cerebral hippocam- pocampus, cortex, hip- cortex, hip- pus, stria- smooth pocampus, pocampus tum, sym- muscle, sympathetic pathetic glandular ganglia ganglia tissues Disease Alzheimer’s, Alzheimer’s, chronic ob- Parkinson’s, drug de- relevance cognitive cognitive structive schizophre- pendence, dysfunction, dysfunction, pulmonary nia, neu- Parkinson’s, schizophre- pain disease, irri- ropathic schizophre- nia table bowel pain nia syndrome

PLC - phospholipase C, PIP2 - phosphatitylinositol 4,5 bisphosphate, IP3 - inositol, 1,4,5 trispho- sphate, DAG - diacylglycerol, PKC - protein kinase C, OxoM - oxotremorine methiodite Adapted from Caulfield and Birdsall 1998; Langmead et al. 2008; Alexander et al. 2011; Mitchel- son 2012

1.4 Types of ion channels

The semipermeable membrane predicted by Julius Bernstein (Bernstein 1902) needed "holes of almost molecular dimensions" (Goldman 1943) in order to be able to maintain the membrane potential and form the action potential. These ?holes? are now called ion channels, membrane-

8 1 Introduction

Figure 1.2: Muscarinic acetylcholine receptors

The muscarinic acetylcholine receptor is part of the A rhodopsin-like family of GPCRs. (A) Like other GPCRs, mAChRs consist of seven transmembrane domains (TM1-TM7). (B) shows the quarternary structure of a muscarinic receptor, with the heterotrimeric G-protein bound to in- tracellular loops i2 and i3. The acetylcholine binding-site lies on the top part of TM7. (C) depicts the pathway of a typical Gαq-coupled receptor like the muscarinic M1 receptor. Phospholipase Cβ (PLCβ) is activated upon agonist binding to the receptor. PLCβ cleaves phosphatitylinos- itol 4,5 bisphosphate (PIP2) into diacylglycerol (DAG) and inositol 1,4,5 trisphosphate (IP3). While DAG can subsequently activate protein kinase C (PKC), IP3 can bind to its receptor on the membrane of the endoplasmatic reticulum (ER) and release Ca2+ from intracellular stores. (adapted from Zeng and Wess 2000; Oldham and Hamm 2008; Langmead et al. 2008; Haga 2013; Delmas and Brown 2005)

9 1 Introduction proteins forming pores to allow ion flux (Hille 2001). Ion channels can be classified by the origin of their gating: (1) ligand-gated ion channels, (2) voltage-gated ion channels, and (3) ion channels not gated by ligands or voltage (Alexander et al. 2011).

1.4.1 Ligand-gated ion channels

A list of ligand-gated ion channels is presented in table 1.2. However, a member of the Cys-loop superfamily of ligand-gated ion channels, nicotinic acetylcholine receptors has been described in detail in the previous chapter (1.3.2).

1.4.2 Voltage-gated ion channels

A detailed list of voltage-gated ion channels can be taken from table 1.3. Generally, voltage-gated ion channels can be distinguished according to their ion selectivity: the family of voltage-gated sodium channels (Nav1) (Catterall et al. 2003b), the family of voltage-gated calcium channels (Cav1-Cav3) (Catterall et al. 2003a), and the biggest family of voltage-gated potassium channels (Kv1-12) (Gutman et al. 2003). Additionally, one could add a family of voltage-gated chloride channels (ClC). While subunits of this family have a rather complex structure of 18 transmem- brane segments forming dimers (Verkman and Galietta 2009), all other voltage-gated channels share a somewhat similar structure (Alexander et al. 2011). Canonical voltage-gated ion chan- nels need four domains, each having six transmembrane segments. In voltage-gated calcium and sodium channels, these four domains are fused to form a big pore forming α-subunit. Both, voltage-gated calcium and sodium channels need auxiliary subunits for physiologic function, respectively (Catterall et al. 2003a; Catterall et al. 2003b). For the purpose of this thesis, I will describe voltage-gated potassium channels in more detail, exemplified by the family of Kv7 channels. Like other voltage-gated potassium channels, four Kv7 subunits are required to form a functional channel. Each subunit consists of six membrane-spanning domains (S1-S6), where S1 to S4 are described as the voltage sensor and S5 and S6 as the pore forming unit. Every third position in the S4 segment is positively charged. These residues are the primary gating charges (Chanda and Bezanilla 2008). A change in membrane potential moves S4 vertically, rotates and tilts it by 180◦ and 30◦, respectively (Bezanilla 2008b). However, additional movements opening the pore region remain poorly understood (Chanda and Bezanilla 2008). The channel tetramer arranges symmetrically around the permeation pore, leaving four voltage sensors around it. These volt- age sensors control the channel open probability, which in turn is determined by the membrane potential (Bezanilla 2008a). The long intracellular C-terminal tail provides several regulatory domains. In a screening for calmodulin (CaM) binding-sites, four α-helical potential binding partners were identified termed Helix A-D. In absence of Ca2+, CaM was found to cooperatively bind to Helix A and B and stabilize them in close proximity to each other. Upon increase of intracellular Ca2+, this inter- action is disrupted and the channel closes (Yus-Najera et al. 2002). Helix C, also known as the A-domain, is important in membrane trafficking of the channel (Schwake et al. 2000). Helix D is the domain responsible for tetramerization (Smith et al. 2001). Furthermore, there are two PIP2 binding sites described for Kv7.2 subunits. One is localized immediately adjacent to the

10 1 Introduction 2+ may alsometabotropic serve functions alsoglycine, blocked by Mg binds genome − 2+ 2+ 2+ 2+ − − cations Permeant ion Characteristics selectivity for Ca selectivity for Ca high Ca permeability; P2X5: also Cl cations,selectivity high for Ca cations absent in rodent Cl Cl cations 2, het- 2 γ ;CNS: 1 β α 3 β 1 δε -amino-3-hydroxy-5-methyl-4-isoxazoleproprionic acid, 7 1 β α α α , 4, 3 β 1 γδ α Table 1.2: Ligand-gated ion channels 1 β α , most common: 7; PNS: α 2 .1 γ 3, 3 γ .2 β - homopentamers or 2 α 4 β 3 β tetramer, di-heteromerGluN1 of & GluN2 at least homopentamer eropentamers α muscle type: α heteropentamer2 α composition , θ 1-3, 1-4, -aminobutyric acid, AMPA - ε , β β γ , A-E homo- & heteropentamers δ βα 3 ε , ρ 1-3 δ 1-4, 1-6, 1-10, , 1-3, , GluN2A-D, GluN3A-B α 5-HT α γ α γ π 3 3+ Ligand Subunits Subunit composition ATP » ADP P2X1-7 homotrimers & heterotrimers cations, P2X1: GABA acetylcholine, nicotine glutamate GluK1-5 homotetramers & heterotetramers cations, high Zn recep- receptors 5-HT A 3 NMDA glutamate GluN1, GABA 5-HT Glycine receptors glycine Kainate AMPA glutamate GluA1-4 homotetramers & heterotetramers cations, high tors Nicotinic acetyl- choline receptors Zincchannel (ZAC) activated FamilyCys-loop receptors, Member 4TM Ionotropic glutamate receptors, iGluR, 3TM P2X, 2TM NMDA - N-methyl-D-aspartate, ATP -Adapted adenosine from: triphosphate, Olsen ADP and - Sieghart adenosine 2008; diphosphate Collingridge et al. 2009; Millar and Gotti 2009; Alexander et al. 2011; Lummis 2012 5-HT - serotonin, TM - transmembrane domain, GABA -

11 1 Introduction

Table 1.3: Voltage-gated ion channels

Family Alternative name Member Characteristics Voltage-gated L-type Cav1.1-1.4 high voltage activated calcium channels P/Q-type Cav2.1 high voltage activated N-type Cav2.2 high voltage activated R-type Cav2.3 high voltage activated T-type cav3.1-3.3 low voltage activated Chloride chan- ClC ClC1-7, no canonical voltage- nels ClCKa, sensor; ClC3-7: Cl−/H+ ClCKb antiporter Voltage-gated Shaker-related Kv1.1-1.8 associated subunits: potassium chan- Kvβ1, Kvβ2 nels Shab-related Kv2.1-2.2 associated subunits: Kv5.1, Kv6.1-6.4, Kv8.1-8.2, Kv9.1-9.3 Shal-related Kv3.1-3.4 Shaw-related Kv4.1-4.3 KCNQ, M-channel, Kv7.1-7.5 Kv7.1: cardiac IKS, KVLQT (Kv7.1) Kv7.2, 7.3, 7.5: neu- ronal M-channel EAG Kv10.1-10.2 hERG Kv11.1-11.3 cardiac IKR ELK Kv12.1-12.3 Voltage-gated HVCN1 Hv1 additional pH & tem- proton channels perature sensitivity Voltage-gated Nav1.1-1.9 TTX-insensitive: sodium channels Nav1.5, 1.8, 1.9

+ IKS - slow delayed rectifying K current, IKR - rapid delayed rectifying current Adapted from: Catterall et al. 2003b; Catterall et al. 2003a; Gutman et al. 2005; Alexander et al. 2011

C-terminal end of the S6 segment, homologous to Kv7.1 (Zhang et al. 2003; Telezhkin et al. 2013), and the second binding site is thought to lie in the linker region between Helix A and Helix B (Hernandez et al. 2008) (see Fig. 1.3 A). Binding of PIP2 increases the channel open probability fivefold to physiological values (Li et al. 2005). Hence, binding of PIP2 is required for full channel function. Five different Kv7 subunits are known: Kv7.1 to Kv7.5. Only Kv7.2, Kv7.3, and Kv7.5 are ex- pressed in central neurons and autonomic ganglia. Kv7.1 is expressed in the heart and Kv7.4 in the inner ear (Brown and Passmore 2009). In sympathetic neurons, heterotetramers of Kv7.2 and Kv7.3 are the most abundant channels, but also homomers of Kv7.2 and heteromers with Kv7.5 may occur in vivo (Brown et al. 2007) (see. Fig. 1.3 B). Kv7 channels are encoded by the KCNQ gene and have therefore initially been termed KCNQ channels (Gutman et al. 2003). These channels have first been described in sympathetic neurons,

12 1 Introduction where they carry a non-inactivating potassium current at sub-threshold levels that was inhib- ited by application of muscarine. The current produced was termed M-current and the channel carrying it was termed M-channel, since the molecular correlates were not known (Brown and Adams 1980).

1.4.3 Ion channels neither gated by ligands nor voltage

Table 1.4 provides a detailed list of ion channels neither gated by a ligand nor gated by voltage. I will focus on ion channels located on the plasma membrane that are gated by rising concentrations of intracellular Ca2+. One can further distinguish these ion channels according to their ion selectivity: calcium-activated potassium channels and calcium-activated chloride channels. I will give a brief description of Ca2+-activated K+ channels and a more detailed one of Ca2+-activated Cl− channels, especially TMEM16A.

Table 1.4: Ion channels not primarily gated by a ligand or voltage

Family Abbrev. Member Permeant ion Gated by Characteristics Acid sensing ion ASIC ASIC1-3 Na+ H+ voltage- channels insensitive Chloride channels CFTR Cystic fi- Br−>Cl− cAMP 12TM, ABC brosis trans- >I−>F− transporter-type membrane protein conductance regulator 2+ − − 2+ CaCC Ca - SCN >NO3 Ca voltage-sensitive activated >I−>Br− Cl− channels >Cl−>F− Maxi Cl− I−>Br− GPCRs γ = 280-430 pS >Cl−>F− >gluconate VRAC Volume reg- SCN−>I− cell volume volume regula- − − ulated anion >NO3 >Br tors channel >Cl−>F− >gluconate Cyclic nucleotide CNG CNGA1-A3 Ca2+ >Na+ cGMP, voltage- gated channels cAMP independent Epithelial sodium ENaC Na+ regulated by: channels aldosterone, angiotensin II, vasopressin, insulin, glucocor- tocoids

13 1 Introduction

Family Abbrev. Member Permeant ion Gated by Characteristics Hyperpolarization- HCN HCN1-4 K+ >Na+ cAMP, active more neg. activated cGMP than -50 mV, nucleotide-gated cAMP shifts IV channels curve to more pos. potentials 2+ 2+ Inositol 1,4,5- IP3RIP3R1-3 Ca IP3,Ca located on in- trisphosphate tracellular Ca2+ receptors store sites (like ER) + Potassium chan- Kir1-7 K Kir3.1-3.4: Kir6 also known nels, 2TM fam- PIP2,Gβγ; as KAT P , asso- ily, inward recti- Kir6.1-6.2: ciate with SUR1 fier K+ channels ATP & SUR2a/b + + Potassium chan- TWIK K2P 1.1, K K leakcurrent nels, 4TM fam- K2P 6.1, + ily, two-pore K K2P 7.1 channels + + TREK K2P 2.1, K K leak current K2P 10.1, K2P 4.1 + + TASK K2P 3.1, K K leak current K2P 9.1, K2P 15.1 + + TALK K2P 16.1, K K leak current K2P 5.1, K2P 17.1 + + THIK K2P 13.1, K K leak current K2P 12.1 + + TRESK K2P 18.1 K K leak current + Potassium chan- BK KCa1.1 K cytosolic big-conductance nels, 6TM family Ca2+ Ca2+-activated K+ channel + SK KCa2.1-2.3 K cytosolic small- Ca2+ conductance Ca2+-activated K+ channel + IK KCa3.1 K cytosolic intermediate- Ca2+ conductance Ca2+-activated K+ channel

14 1 Introduction

Family Abbrev. Member Permeant ion Gated by Characteristics + Slack KCa4.1, K cytosolic gating inhibitors: + − 2+ (Slo2.2), KCa4.2 Na ,Cl internal Ca , Slick ATP (Slo2.1) + + Slo3 KCa5.1 K ,Na voltage, pH>7.5 Ryanodine recep- RyR RyR1-3 Ca2+ >K+ cytosolic located on in- tors Ca2+,ATP tracellular Ca2+ stores + + Sodium leak Navi Navi2.1 Na constitutively Na leak cur- channels, non- active rent, voltage- selective insensitive Transient re- TRP TRPC1,3-7 mono- & di- DAG (in- activated down- ceptor potential valent cations dependent stream of Gαq- cation channels of PKC) coupled receptors TRPM1-8 mono- & heat: divalent TRPM2-5; cations: depolar- TRPM1-3, ization: 8; monova- TRPM4, lent cations: 5, 8; con- TRPM4, stitutively 5, 7; diva- active: lent cations: TRPM6; TRPM6 cAMP, PKA: TRPM7 TRPA1 mono- & di- cooling, valent cations Ca2+ TRPV1-6 mono- & di- noxious valent cations heat: TRPV1, 2; heat: TRPV3, 4; mechanical stimuli: TRPV4; consti- tutively active: TRPV5, 6

15 1 Introduction

Family Abbrev. Member Permeant ion Gated by Characteristics TRPML1-3 TRPML1: constitutively located on intra- Fe2+, active cellular vesicles monova- lent cations; TRPML2: Na+; TRPML3: Na+>K+>Cs+ cAMP - cyclic adenosine monophosphate, cGMP - cyclic guanosine monophosphate, ATP - adenosine triphosphate, TM - transmembrane domain, GPCR - G-protein coupled receptor, ER - endoplasmatic reticulum, IV curve - current - voltage relation, IP3 - inositol 1,4,5 trisphosphate, PIP2 - phosphatitylinositol 4,5 bisphosphate, SUR1 - sulfonyl urea receptor 1 Adapted from: Wei et al. 2005; Alexander et al. 2011

2+ + The family of Ca -activated K channels consists of five members (KCa1.1-KCa5.1) that can be further divided into two structurally related groups. One group involves KCa2.1-2.3 2+ and KCa3.1, also known as small-conductance (SK) and intermediate-conductance (IK) Ca - activated K+ channels, respectively. Conductances range from 2.3 pS for SK channels to 11 pS for IK channels. Both, SK and IK channels are voltage-insensitive and are solely gated by rising 2+ concentrations of intracellular Ca with a Kd value of 0.6 μM for SK and 0.1-0.3 μMforIK channels, respectively. Both channels have six membrane spanning domains and a C-terminal 2+ + binding site for calmodulin as Ca sensor. KCa2 channels produce a selective K current, re- sponsible for the afterhyperpolarization in many neuronal and muscular tissues. On the other + + + + hand, KCa3 channels are permeable for K ,Rb ,NH4 and Cs and are expressed in a variety of non-neuronal tissues like the placenta, erythrocytes, liver but also smooth muscles (Wei et al. 2005). The second, structurally distinct, group includes KCa1.1, also known as big-conductance (BK) 2+ + Ca -activated K channels, and KCa4.1-4.2 and KCa5.1 channels. The latter channels have been included in this group due to their structural relation. However, neither KCa4 nor KCa5 2+ + − channels are activated by Ca .KCa4 channels are gated synergistically by Na and Cl and KCa5 channels are gated by intracellular alkalization. As opposed to the first group, the second group of Ca2+-activated K+ channels has seven transmembrane domains with an extracellular N-terminus and a C-terminus that consists of four additional cytosolic α-helices. The C-terminus + of KCa1.1 furthermore carries two RCK (regulating conductance of K ) domains, and a series of negatively charged amino acid residues known as the "Ca2+ bowl". Hence, BK channels do 2+ 2+ not require calmodulin as a Ca sensor but directly bind Ca with a Kd of ∼10 μM. Addi- tionally, BK channels are voltage-sensitive and need both, a depolarization and the Ca2+ signal simultaneously for gating (Berkefeld et al. 2010). The single channel contuctance of BK chan- nels amounts to 260 pS. BK channels are expressed in a variety of tissues, including skeletal and smooth muscle as well as neuronal tissues like hippocampal pyramidal cells, striatum, and the neocortex. Usually, BK channels are coupled with N-type voltage-gated Ca2+ channels and mediate the fast after hyperpolarization and regulate presynaptic transmitter release (Alexander

16 1 Introduction

Figure 1.3: Kv7 voltage-gated potassium channels

Kv7 channels are members of the superfamily of voltage-gated potassium channels. (A) represents a subunit of a neuronal Kv7 channel. In consists of 6 transmembrane segments (S1-S6). S1-S4 contain the voltage-sensor, S5 and S6 form the ion conduction pathway. The long C-terminal tail comprises of two calmodulin (CaM) binding sites, two binding sites for PIP2 as was as an additional α-helix (Helix D), that is required for tetramerization. (B) shows the quarternary structure of a typical neuronal Kv7 heteromer, representing the molecular correlate of the so- called M-current. (adapted from Delmas and Brown 2005; Telezhkin et al. 2013)

17 1 Introduction et al. 2011).

For a long time, the molecular correlate for Ca2+-activated Cl− channels (CaCC) remained enigmatic. Among the suggested candidates was the Ca2+ activated Cl− channel family (CLCA), but heterologously expressed CLCAs display a different Ca2+- and voltage-sensitivity compared to native CaCCs, additional to a different pharmacological profile. A human gene, homologous to the Drosophila gene tweety, was proposed to be the background of CaCCs, but this gene is not expressed in the same tissues as CaCCs. The best candidates for a long time were bestrophins, but failed because of a lack of voltage-sensitivity and a higher affinity for Ca2+ (Hartzell et al. 2009). Finally, TMEM16A, a member of the protein family of "Transmembrane proteins with unknown function 16" (TMEM16) was identified as the channel responsible for Ca2+-activated chloride currents. The family consists of 10 members, TMEM16A-H, J, and K (Huang et al. 2012). Mem- bers TMEM16B-K share a higher rate of homology ( 60%) with TMEM16A, whereas members G, H, J, and K match only about 20% with TMEM16A (Galietta 2009). However, TMEM16s form a unique family of ion channels and are not related to any other ion channel family, includ- ing Ca2+-activated K+ channels (Huang et al. 2012). TMEM16s are predicted to have eight transmembrane domains and some members have been as- sociated with anion channels. Hence, they were renamed into the family of anoctamins (ANO). So far, TMEM16A and TMEM16B (ANO-1 and ANO2) have been associated with classical Ca2+- activated chloride channels (Galietta 2009). TMEM16F (ANO-6) carries either Ca2+-activated Cl− currents (Szteyn et al. 2012; Shimizu et al. 2013), or Ca2+-activated cation currents (Tian et al. 2012; Kunzelmann et al. 2013). Recently, also ANO3 (TMEM16C) was found to be a func- tional ion channel but mediates a Na+-dependent K+ current (Huang et al. 2013). It is generally accepted that anoctamins comprise eight membrane-spanning regions (TMD1-8) (see Fig. 1.4 A) (Hartzell et al. 2009). It is thought, that TMD5 and TMD6 form the ion permeation pathway and mutations in the linker between the two membrane-spanning domains are described to inter- fere with anion selectivity (Galietta 2009). CaCCs are fairly unselective for Cl−, the sequence of − − − − − − selectivity is: SCN >NO3 >I >Br >Cl >F (Hartzell et al. 2005). Anion permeability and conductance are distinct features in CaCCs. Shells of water molecules stabilize ions in bulk water, which can be characterized by their hydration energy. For larger ions with lower hydration energy it is relatively easy to enter the channel, i.e. to permeate the channel, than for smaller ions with higher hydration energy. However, anions with lower hydration energy get caught in the pore and can therefore not pass through the channel (i.e. conductance). Thus, small ions with large hydration energies display a poor conductance, because they do not enter the channel, whereas large ions enter the channel, but get caught in the pore and are therefore also poorly conductive. Furthermore, large ions that get stuck in the pore also block the pore for other ions like Cl−. Additionally, CaCCs are not exclusively permeable for anions, but display a rather high pNa/pCl ratio of 0.1 (Hartzell et al. 2005). It is still a matter of debate how CaCCs sense rising intracellular Ca2+ concentrations, since there are no typical motifs for Ca2+ sensing or classical calmodulin binding-sites (Galietta 2009). There is a stretch of 5 glutamic acids in the first intracellular loop, which is thought to create a

18 1 Introduction

"calcium bowl" as in Ca2+-activated K+ channels. In favor of this theory, there is a splice vari- ant removing the last glutamic acid residue, which increases Ca2+ affinity dramatically. However, deleting the first four residues hardly changes the affinity for Ca2+ (Yu et al. 2012). Another possibility for Ca2+-sensing would be via binding of calmodulin (CaM). Three proposed CaM binding sites were found not to be significant for channel function. Nevertheless, an N-terminal regulatory calmodulin-binding motif (RCBM) seems to be a good candidate for a CaM binding. It was found to bind CaM and needs to be intact for correct Ca2+-dependent activation of CaCCs (Vocke et al. 2013). Yet, there was another possibility for direct Ca2+-sensing proposed: In the revised topology of ANO1, the new large third intracellular loop harbors two glutamic acids im- mediately adjacent to TMD7. Mutation of these residues strongly influences Ca2+ binding, even though an effect of the mutants on gating could not be excluded (Yu et al. 2012). It remains elusive how CaCCs sense voltage, since they lack the classical voltage-sensor of voltage-gated ion channels (Huang et al. 2012). Binding of Ca2+ and voltage sensing work in a coordinated manner. In the absence of intracellular Ca2+, the channel remains inactive at all voltages (Fer- rera et al. 2009). Rising intracellular Ca2+ concentrations shift the current-voltage relation to more negative values. Hence, at sub-maximal Ca2+ concentrations the current-voltage relation is outwardly rectifying. However, at maximal Ca2+ concentrations, the channel is fully active at all voltages. But changes in membrane voltage also influence the channels affinity for Ca2+.Ata membrane voltage of -60 mV the half-maximal concentration for Ca2+ is 2.6 μM, whereas at +60 mV the half-maximal concentration decreases to 0.4 μM (Galietta 2009). An interesting feature of TMEM16A is, that alternative splicing creates various functional isoforms. Currently, four alter- native exons are known termed a-d (see Fig. 1.4. A). A minimal splice-variant (TMEM16A(0)), devoid of all alternative exons, forms a functional anion channel but is voltage-independent. Hence, alternative splicing might display a way of regulating channel properties (Galietta 2009). Interestingly, exons seem to be skipped in a tissue-specific manner (Ferrera et al. 2010). This might explain the highly variable conductances between 1 and 15 pS found in different tissues (Hartzell et al. 2009). Furthermore, ANO-1 has been found to homodimerize, while it is not known whether they are also able to form heteromers or whether auxiliary subunits are required for channel function (see Fig. 1.4 B) (Kunzelmann et al. 2011). The Ca2+ necessary for activation of the channel was thought to be provided both by release from intracellular stores and by Ca2+ influx (Hartzell et al. 2005). Conversely, it was recently shown in dorsal root ganglion neurons that CaCCs only react to rising Ca2+ concentrations via release from intracellular stores rather than reacting to global changes of intracellular Ca2+ caused by influx via voltage-gated calcium channels (Jin et al. 2013). CaCCs are expressed in various tissues, linked to a broad variety of functions and diseases, which can be extracted from table 1.5 as can be pharmacological tools to characterize CaCCs.

1.5 Modulation of ion channels by G-protein coupled receptors

Ion channel function needs to be fine-tuned; therefore they are subject to modulation for example by GPCRs. Activation of a G protein-coupled receptor leads dissociation of the Gα and the βγ subunit. In case of a Gαq-coupled receptor, like muscarinic M1 receptors, the active Gαq activates phospholipase Cβ (PLCβ). The substrate for this enzyme is membrane bound phosphatitylinos-

19 1 Introduction

Figure 1.4: Ca2+-activated Cl− channels

TMEM16A (anoctamin-1) forms Ca2+-activated Cl−-channels. (A) depicts the putative structure of TMEM16A. It consists of 8 transmembrane domains (TMD1-8), where TMD5 and 6 are thought to form the ion permeation pore. Alternative splicing of four alternative exons (|empha- d) form Ca2+-activated Cl−-channels with different properties. Ca2+ sensing is thought to either take place on an N-terminal calmodulin recognition domain (RCBM), a "Ca2+-bowl" analogous to that of Ca2+-activated K+ channels, or a direct Ca2+ binding site on the C-terminal end of the large third intracellular loop. (B) TMEM16A was found to form homodimers, but it is not known, whether the two participating subunits form one or two pores. Adapted from Hartzell et al. 2009; Kunzelmann et al. 2011; Yu et al. 2012; Vocke et al. 2013

20 1 Introduction

Table 1.5: Calcium-activated chloride channels

TMEM16 anoctamin Disease correlation/ function antagonist TMEM16A ANO-1 tracheomalacia, pain sensation, gastroin- niflumic acid, flufe- testinal mobility, salivary gland secretion, namic acid, SITS, airway epithelial secretion DIDS, fluoxetine, CaCCinh-A01, T16Ainh-A01 TMEM16B ANO-2 photoreceptor signal in the retina, olfac- niflumic acid, flufe- tory transduction in sensory neurons namic acid, SITS, DIDS, fluoxetine, CaCCinh-A01 TMEM16E ANO-5 gnathodiaphyseal dysplasia, proximal niflumic acid, flufe- limb-girdle muscular dystrophy, distal namic acid, SITS, non-dysferlin Miyoshi myopathy DIDS, fluoxetine TMEM16F ANO-6 Ca2+-dependent phospholipid scram- niflumic acid, flufe- blase for phosphatitylserine exposure in namic acid, SITS, platelets, Scott’s syndrome DIDS, fluoxetine, CaCCinh-A01 TMEM16G ANO-7 expression in healthy prostate and niflumic acid, flufe- prostate cancer namic acid, SITS, DIDS, fluoxetine TMEM16K ANO-10 autosomal-recessive cerebellar ataxia niflumic acid, flufe- namic acid, SITS, DIDS, fluoxetine

SITS - 2-[2-(4-acetamido-2-sulfophenyl)ethenyl]-4-isothiocyanatobenzene-1-sulfonic acid, DIDS - 5-isothiocyanato-2-[2-(4-isothiocyanato-2-sulfophenyl)ethenyl]benzene-1-sulfonic acid Adapted from: De La Fuente et al. 2008; Namkung et al. 2011; Huang et al. 2012

itol 4,5 bisphosphate (PIP2). Hydrolysis of PIP2 results in formation of diacylglycerol (DAG), which is still membrane bound, and inositol 1,4,5 trisphosphate (IP3), which is diffusible. Thus there are at least three possible pathways to modulate channel protein functions (see. Fig. 1.5):

1.5.1 Modulation of ion channels by depletion of phosphatidylinositol 4,5 bisphosphate

For a long time it was thought, that only the products of PIP2-cleavage can have effects on other proteins. However, in the early 2000s evidence was found that already the reduction of PIP2, as a result of PIP2 hydrolysis, has effects by itself (Zhang et al. 2003). One of the best-known examples is the modulation of Kv7 channels by M1 receptors. This interaction itself was revealed when Kv7 channels were discovered in 1980 (Brown and Adams 1980). Due to the inhibition of an unknown potassium current via activation of M1 receptors, these channels were initially called M-channels and the current produced was termed M-current. Later on, the molecular correlate of neuronal M-channels was determined to be carried by heterotetramers of Kv7.2 and Kv7.3 subunits (Wang et al. 1998). Recovery of M-current inhibition was found to be dependent on PIP2 resynthesis (Suh and Hille 2002). Thus, M-current inhibition via M1 receptor activation

21 1 Introduction

Figure 1.5: Modulation of ion channels by Gαq-coupled GPCRs

Gαq-coupled GPCRs like muscarinic M1, or bradykinin B2 receptors modulate ion channel func- tions via different pathways. Activated PLCβ hydrolyses membrane-bound PIP2. This leads to a depletion of PIP2, which in turn inhibits Kv7, since they need presence of PIP2 for appropriate channel function. One cleavage product of PIP2 is IP3, which binds to ER-bound IP3 receptors and releases Ca2+ from intracellular stores. The rise of intracellular Ca2+ inhibits via binding of CaM Kv7 channels. The second cleavage product of PIP2, DAG, activates PKC, which increases N-type voltage-gated Ca2+ channel activity directly or disrupts their inhibition caused by the βγ subunit of PTX-sensitive GPCRs. seems to rely only on PIP2 depletion and not the presence of cleavage products (Zhang et al. 2003). This effect can be easily explained by the fact that channel open probability dramatically decreases in the absence of PIP2 (Li et al. 2005).

1.5.2 Modulation of ion channels via release of calcium from intracellular stores

Additionally, M-channels are also modulated by another GPCR preferentially signaling via Gαq, bradykinin B2 receptors (Higashida and Brown 1986). However, this interaction was found to depend on the formation of IP3. In contrast to M1 receptors, B2 receptors are closely linked to the endoplasmatic reticulum (ER). Hence, IP3 resulting from B2 receptor activation can diffuse 2+ to the ER and bind to ER-bound IP3 receptors, which release Ca from intracellular stores (Delmas and Brown 2005). The channel detects the resulting Ca2+ rise via interaction with CaM (Gamper and Shapiro 2003). 2+ Two questions remain: Why do B2 receptors inhibit Kv7 channels via Ca release and not via 2+ PIP2 depletion? And why doesn’t Ca elevation play a role in M-channel modulation via M1

22 1 Introduction receptors? It is proposed that a rise of intracellular Ca2+ also activates the neuronal Ca2+ sensor protein (NCS1), which in turn activates PI4-kinase, responsible for PIP2 resynthesis. Therefore, activa- tion of B2 receptors does not cause enough PIP2 depletion to inhibit the channel. On the other hand, M1 receptors are not located in close proximity to the ER and therefore do not cause a 2+ Ca elevation to an extend that is needed for Kv7 channel inhibition (Delmas and Brown 2005). However, a recent report suggested that only depletion of PIP2 is required for inhibition of Kv7 channels via Gq-coupled receptors. Neither substantial PKC activation nor rise of intracellular calcium is sufficient to influence Kv7 channel function. The detected differences of Kv7 channel modulation via two distinct Gq-coupled receptors (P2Y2 and M1 receptors) could entirely be explained by different expression levels of the two receptors (Dickson et al. 2013; Falkenburger et al. 2013). One has to keep in mind that this study was conducted in an artificial system comparing heterologously expressed M1 and endogenous P2Y2 receptors. It remains to be es- tablished, whether these findings hold true in more native systems, like neuronal cultures and whether a similar effect might explain the proposed differences in B2 and M1 receptor mediated Kv7 channel modulation.

1.5.3 Modulation of ion channels via activation of protein kinase C

The second cleavage product of PIP2 is DAG, which is the natural activator of protein kinase C (PKC) (Steinberg 2008). PKC occurs in several isoforms, classified by their mode of activation (see table 1.6). A classic example is PKC-mediated modulation of N-type voltage-gated Ca2+ channels (Zamponi and Currie 2013). Phosphorylation of the α-subunit causes two effects. First, it directly enhances currents through Cav2.2 channels (Stea et al. 1995). Second, it antagonizes the inhibition of Cav2.2 mediated by Gαi coupled receptors acting via the so-called membrane- delimited pathway (Hamid et al. 1999).

1.6 Aim of the project

In the sympathetic nervous system, stimulation of preganglionic cholinergic neurons produce two types of excitation: (1) a fast component, termed fast ganglionic transmission, that is carried by nAChRs and (2) a slow component, termed slow ganglionic transmission, that is mediated by mAChRs (Trendelenburg 1966; Brown and Selyanko 1985). The ionic correlate to the slow ganglionic transmission relies on M1 receptor mediated inhi- bition of Kv7 channels in bullfrog sympathetic neurons (Adams and Brown 1982) and also in mammalian sympathetic neurons (Brown and Selyanko 1985). Correspondingly, application of a muscarinic agonist triggers release of noradrenaline from rat sympathetic nerve endings via inhibition of Kv7 channels residing on neuronal somata. The resulting depolarization triggers ac- tion potentials, propagated to nerve endings where, in case of sympathetic neurons from superior cervical ganglia, noradrenaline is released (Lechner et al. 2003). However, neuronal Kv7 channels activate at voltages more positive to -60 mV (Owen et al. 1990; Lechner et al. 2003; Maljevic et al. 2008; Brown and Passmore 2009; Linley et al. 2012; Telezhkin et al. 2012). The Goldman-Hodgkin-Katz equation predicts resting membrane potentials be-

23 1 Introduction

Table 1.6: Protein kinase C isoforms

Isoform Members Characteristics Antagonists Expression profile in rat SCG classic α, βI, βII, Ca2+ dependent, staurosporine, α, βI, βII PKC γ DAG activated, Gö-6983, PE sensitive GF-109203X (more potent at α), Gö-6976 novel δ, ε, η, θ Ca2+ independent, staurosporine, δ, ε PKC DAG activated, Gö-6983, PE sensitive GF-109203X (more potent at α) atypical ι/λ, ζ not sensitive to staurosporine, ζ PKC Ca2+, DAG or PE Gö-6983, GF-109203X (>1μM)

DAG - diacylglycerol, PE - phorbolester Adapted from: Mellor and Parker 1998; Scholze et al. 2002; Wu-Zhang and Newton 2013

tween -64 to -67 mV for rat sympathetic neurons. Hence it appears unlikely that inhibition of an otherwise inactive channel will cause substantial depolarization and subsequent transmitter release.

• The first aim of the present study is to determine, whether Kv7 channels are involved in OxoM induced depolarizations and noradrenaline release. If not, what other conductances might be responsible? It was suggested that the slowly occurring depolarization via activation of mAChRs also involves a Cl−-component in sympathetic ganglia (Brown and Selyanko 1985). Indeed, an acetylcholine-evoked Cl− current was found in rat superior cervical ganglion cells, which involved presence of high intracellular Ca2+ as well as DAG and it was reduced when protein kinase C (PKC) was inhibited (Marsh et al. 1995).

• Thus, the second aim is to determine, if a Cl− conductance is involved in the formation − of M1 receptor mediated slow EPSPs. If so, it needs to be determined which Cl channel family is responsible.

• If the first two aims hold true, the third aim is to further determine the second messenger pathway involved after M1 receptor activation. As suggested above (Marsh et al. 1995), PKC activation could also be needed to trigger slow EPSPs followed by noradrenaline release from rat sympathetic ganglia.

24 2 Results

Resubmitted manuscript entitled "Excitation of rat sympathetic neurons via M1 muscarinic re- ceptors independently of Kv7 channels" to Pflügers Archiv - European Journal of Physiology.

25 Submission to Pflügers Archiv - European Journal of Physiology 2014/01/28

Excitation of rat sympathetic neurons via M1 muscarinic receptors independently of Kv7 channels

Isabella Salzer, Hend Gafar, Viola Gindl, Peter Mahlknecht, Helmut Drobny, and Stefan Boehm

Department of Neurophysiology and Neuropharmacology, Center for Physiology and Pharmacology, Medical University of Vienna, A-1090 Vienna, Austria

Corresponding author: Stefan Boehm

Department of Neurophysiology and Neuropharmacology

Center for Physiology and Pharmacology, Medical Univ. Vienna

Waehringerstrasse 13a, A-1090 Vienna, Austria

Phone: +43 1 40160 31200; Fax: +43 1 40160 931201 Email: [email protected]

1

Abstract The slow cholinergic transmission in autonomic ganglia is known to be mediated by an inhibition of Kv7 channels via M1 muscarinic acetylcholine receptors. However, in the present experiments using primary cultures of rat superior cervical ganglion neurons, the extent of depolarization caused by the M1 receptor agonist oxotremorine M did not correlate with the extent of Kv7 channel inhibition in the very same neuron. As activation of M1 receptors leads to a boost in protein kinase C activity in sympathetic neurons, various PKC enzymes were inhibited by different means. Interference with classical PKC isoforms led to reductions in depolarizations and in noradrenaline release elicited by oxotremorine M, but left the Kv7 channel inhibition by the muscarinic agonist unchanged. M1 receptor-induced depolarizations were also altered when extra- or intracellular Cl- concentrations were changed, as were depolarizing responses to γ-aminobutyric acid. Depolarizations and noradrenaline release triggered by oxotremorine M were reduced by the non-selective Cl- channel blockers 4-acetamido-4'-isothiocyanato-stilbene-2,2'-disulfonic acid (SITS) and niflumic acid. Oxotremorine M induced slowly rising inward currents at negative membrane potentials that were blocked by inhibitors of Ca2+- activated Cl- and TMEM16A channels and attenuated by PKC inhibitors. These channel blockers also reduced oxotremorine M-evoked noradrenaline release. Together, these results reveal that slow cholinergic excitation of sympathetic neurons involves mechanisms other than the inhibition of Kv7 channels which include an activation of classical PKCs and of Ca2+- activated Cl- channels.

Key words

2+ M1 muscarinic receptors, noradrenaline release, Kv7 channels, protein kinase C; Ca - activated Cl- channels

2

Introduction

Acetylcholine is the prime transmitter in the ganglia of the entire autonomic nervous system; it excites postganglionic neurons simultaneously via two different types of receptors: nicotinic

(nAChRs) and muscarinic (mAChRs) acetylcholine receptors. Ganglionic transmission via these two receptors can occur independently of each other [8,36]. However, the excitation of postganglionic neurons via mAChRs is much slower than that via nAChRs and involves a G protein-mediated inhibition of Kv7 channels, also known as M type K+ channels [9]. There are at least 5 different subtypes of mAChRs named M1 through M5 [38], and the inhibition of

Kv7 channels in postganglionic sympathetic neurons was found to involve M1 receptors

[5,27]. The underlying signalling mechanism is a phospholipase C-mediated depletion of phosphatidylinositol-4,5-bisphosphate (PIP2) in the neuronal plasma membrane [10].

In primary cultures of postganglionic sympathetic neurons, activation of M1 mAChRs causes depolarization and action potential firing which ultimately leads to exocytotic noradrenaline release from the axon terminals [22,26]. The following results indicated that an inhibition of

Kv7 channels contributed to this sequence of events: (i) retigabine, an activator of Kv7 channels, abolished noradrenaline release evoked by the mAChR agonist oxotremorine M, but not that triggered by electrical field stimulation [22]; (ii) direct inhibition of Kv7 channels by

Ba2+ and/or linopirdine also elicited action potential- and Ca2+-dependent noradrenaline release from sympathetic neurons [7,19]; (iii) activation of B2 bradykinin receptors on sympathetic neurons caused an inhibition of Kv7 channels [17], on one hand, and led to noradrenaline release, on the other hand, again in an action potential- and Ca2+-dependent manner [7]. However, in the case of B2 bradykinin receptors, an additional mechanism was found to mediate sympathoexcitation caused by this peptide: an activation of protein kinase C

[34]. In sympathetic neurons, protein kinase C (PKC) can also be activated via mAChRs [25], and PKC may also contribute to the muscarinic inhibition of Kv7 channels [14], most probably by regulating the PIP2 sensitivity of the channel [23]. However, it is still unknown

3 whether PKC might also be involved in the excitation of postganglionic sympathetic neurons via M1 receptors, and if so, whether a PKC-dependent excitation of sympathetic neurons also relies on an inhibition of Kv7 channels.

Noradrenaline release from sympathetic neurons is not only triggered by an activation of M1 receptors, as described above, but is also modulated (i.e. enhanced or decreased) by several muscarinic receptors located at the axon terminals where exocytosis occurs. These presynaptic receptors generally mediate a reduction of action potential-evoked noradrenaline release, which is in most instances based on an inhibition of voltage-activated Ca2+ channels via pertussis toxin-sensitive G proteins [20]. Therefore, experiments regarding the release of noradrenaline were carried out on neurons treated with pertussis toxin in order to largely eliminate confounding effects of inhibitory presynaptic muscarinic receptors. The results demonstrate that activation of M1 receptors can depolarize sympathetic neurons and induce noradrenaline release independently of Kv7 channels; the alternative signalling mechanisms include classical PKC enzymes and Ca2+-activated Cl- channels.

4

Materials and Methods

Primary cultures of rat superior cervical ganglion neurons.

Primary cultures of dissociated superior cervical ganglion (SCG) neurons from neonatal rats were prepared as described before [21]. Newborn Sprague-Dawley rats were kept and sacrificed three to ten days after birth by decapitation in full accordance with all rules of the

Austrian animal protection law (see http://ris1.bka.gv.at/Appl/findbgbl.aspx?name=entwurf&format=pdf&docid=COO_2026_100

_2_72288) and the Austrian animal experiment law (see http://www.ris.bka.gv.at/Dokumente/BgblAuth/BGBLA_2012_I_114/BGBLA_2012_I_114.p df). Ganglia were removed immediately after decapitation of the animals, cut into 3 to 4 pieces, and incubated in collagenase (1.5 mg ml-1; Sigma, Vienna, Austria) and dispase (3.0 mg ml-1; Boehringer Mannheim, Vienna, Austria) for 30 min at 36 °C. Subsequently, they were further incubated in trypsin (0.25 % trypsin; Worthington, Lakewood, NJ, USA) for 15 min at 36 °C, dissociated by trituration, and resuspended in Dulbeccos modified Eagle's

Medium (InVitrogen, Lofer, Austria) containing 2.2 g l-1 glucose, 10 mg l-1 insulin, 25000 IU l-1 penicillin and 25 mg l-1 streptomycin (InVitrogen), 50 µg l-1 nerve growth factor (R&D

Systems Inc., Minneapolis, MN, USA), and 5% fetal calf serum (InVitrogen). Finally, all cells were seeded onto 5 mm plastic discs for radiotracer release experiments and onto 35 mm culture dishes for electrophysiological experiments. All tissue culture plastic was coated with rat tail collagen (Biomedical Technologies Inc., Stoughton, MA, USA). The cultures were stored for 4 to 8 days in a humidified 5% CO2 atmosphere at 36 ºC. On days one and 4 after dissociation, the medium was exchanged entirely.

5

Electrophysiology

Recordings were carried out at room temperature (20-24°C) on the somata of single SCG neurons using the perforated-patch version of the patch-clamp technique which prevents rundown of currents through Kv7 channels [6]. Patch pipettes were pulled (Flaming-Brown puller, Sutter Instruments, Novato, CA, USA) from borosilicate glass capillaries (Science

Products, Frankfurt/Main, Germany) and front-filled with a solution consisting of (mM)

K2SO4 (75), KCl (55), MgCl2 (8), and HEPES (10), adjusted to pH 7.3 with KOH. Then, electrodes were back-filled with the same solution containing 200 μg/ml amphotericin B or 50

μg/ml gramicidin D (in 0.8 % DMSO) which yielded tip resistances of 1 to 3 MΩ. In some current clamp recordings, KCl in the pipette was replaced by identical concentrations of either

CsCl or K-gluconate. For the measurement of oxotremorine M-induced currents, the pipette solution contained (mM) KCl (140), CaCl2 (1.0), MgCl2 (0.7), EGTA (10) and HEPES (10), adjusted to pH 7.3 with KOH. The bathing solution consisted of (mM) NaCl (140), KCl (3.0),

CaCl2 (2.5), MgCl2 (2.0), glucose (20), HEPES (10), adjusted to pH 7.4 with NaOH. With this bath solution, liquid junction potentials ranged between –8 mV for pipette solutions containing K2SO4 plus KCl or CsCl and -12.4 mV for the pipette solution containing 140 mM K-gluconate. These values were corrected for during experimentation. Tetrodotoxin (0.5

μM) was included to suppress action potential firing which interferes with the precise determination of depolarizations. All other drugs were applied via a DAD-12 drug application device (Adams & List, Westbury, NY,USA) which permits a complete exchange of solutions surrounding the cells under investigation within less than 100 ms. To investigate currents through Kv7 channels, cells were held at a potential of -30 mV, and four times per minute 1 s hyperpolarisations to -55 mV were applied to deactivate the channels; the difference between current amplitudes 20 ms after the onset of hyperpolarisations and 20 ms prior to re- depolarisation was taken as a measure for currents through Kv7 channels. Amplitudes

6 obtained during the application of test drugs (b) were compared with those measured before

(a) and after (c) application of these drugs by calculating 200b / (a+c) = % of control or 100 -

(200b/[a+c]) = % inhibition [6].

Determination of [3H]noradrenaline release

[3H]noradrenaline uptake and superfusion were performed as described [22]. Briefly, plastic discs with dissociated neurons were incubated at 36 °C for 1 h in 0.05 µM [3H]noradrenaline

(specific activity 42.6 Ci/mmol) in culture medium containing 1 mM ascorbic acid.

Thereafter, these discs were introduced into small chambers and superfused with a solution consisting of (mM) NaCl (120), KCl (6.0), CaCl2 (2.0), MgCl2 (2.0), glucose (20), HEPES

(10), fumaric acid (0.5), Na-pyruvate (5.0), ascorbic acid (0.57), adjusted to pH 7.4 with

NaOH. Superfusion was performed at 25 °C at a rate of about 1.0 ml/min. Collection of 4 min superfusate fractions was started after a 60 min washout period to remove excess radioactivity.

To investigate noradrenaline release evoked by oxotremorine M, the muscarinic agonist was included in the superfusion buffer for 2 minutes, unless indicated otherwise. For comparison, tritium overflow was also elicited by the application of 60 monophasic rectangular electrical pulses (0.5 ms, 60 mA, 50 V/cm) delivered at a frequency of 1.0 Hz. Modulatory agents, i.e.

PKC inhibitors, ion channel blockers, furosemide, or bumetanide were present from min 50 of superfusion (i.e. 10 min prior to the start of sample collection) onward. The radioactivity remaining in the cells after finishing the experiments was extracted by immersion of the discs in 2 % (v/v) perchloric acid. Radioactivity in extracts and collected fractions was determined by liquid scintillation counting (Perkin Elmer Tri-Carb 2800 TR). Radioactivity released in response to electrical field stimulation from rat sympathetic neurons after labelling with tritiated noradrenaline under conditions similar to those of this study had been shown to

7 consist mainly of the authentic transmitter and to contain only small amounts (<15%) of metabolites [35]. Therefore, the outflow of tritium as determined here was assumed to reflect the release of noradrenaline and not that of metabolites.

The spontaneous (unstimulated) rate of [3H] outflow was obtained by expressing the radioactivity of a collected fraction as percentage of the total radioactivity in the cultures at the beginning of the corresponding collection period. Stimulation-evoked tritium overflow was calculated as the difference between the total [3H] outflow during and after stimulation and the estimated basal outflow which was assumed to decline linearly throughout experiments. Therefore, basal outflow during periods of stimulation was assumed to equate to the arithmetic mean of the samples preceding and those following stimulation, respectively.

The difference between the total and the estimated basal outflow was expressed as a percentage of the total radioactivity in the cultures at the beginning of the respective stimulation (% of total radioactivity). The amount of electrically or oxotremorine M-evoked tritium release may vary considerably between different SCG preparations [22]. Therefore, tritium overflow in the presence of release altering agents, such as PKC, transporter or channel inhibitors, was always compared with that obtained within the same SCG preparation in the presence of solvent. To directly compare effects of different modulatory agents upon electrically and oxotremorine M-evoked overflow, respectively, the values obtained in the presence of these modulators were expressed as percentage of the corresponding values in the presence of solvent within the same preparation.

Statistics

All values are arithmetic means + standard error of the mean. n values reflect single cells in electrophysiological experiments and numbers of cultures in radiotracer release experiments.

Statistical significance of differences between two groups was determined by the Mann-

Whitney test. Statistical significance of differences between multiple groups was performed

8 by Kruskal-Wallis tests followed by Dunn’s multiple comparison tests. P values < 0.05 were considered as indicating statistical significance.

Materials

(-)-[Ring-2,5,6-3H]noradrenaline was obtained from PerkinElmer (Vienna, Austria); amphotericin B, gramicidin D, oxotremorine M, staurosoporine, 12-(2-Cyanoethyl)-6,7,12,13- tetrahydro-13-methyl-5-oxo-5H-indolo[2,3-a]pyrrolo[3,4-c]carbazole (Gö6976), 3-[1-[3-

(Dimethylamino)propyl]-5-methoxy-1H-indol-3-yl]-4-(1H-indol-3-yl)-1H-pyrrole-2,5-dione

(Gö 6983), bisindolylmaleimide I (GF 109203 X), phorbol 12-myristate 13-acetate, 10,10- bis(4-pyridinylmethyl)-9(10H)-anthracenone (XE 991), disodium 4-acetamido-4'- isothiocyanato-stilbene-2,2'-disulfonic acid (SITS), niflumic acid and pertussis toxin from

Sigma (Vienna, Austria); 6-(1,1-Dimethylethyl)-2-[(2-furanylcarbonyl)amino] 4,5,6,7- tetrahydrobenzo[b]thiophene-3-carboxylic acid (CaCCinh-A01) and 2-[(5-Ethyl-1,6-dihydro-

4-methyl-6-oxo-2-pyrimidinyl)thio]-N-[4-(4-methoxyphenyl)-2-thiazolyl]acetamide

(T16Ainh-A01) from Tocris (Bristol, UK); tetrodotoxin from Latoxan (Rosans, France).

Water-insoluble drugs were first dissolved in DMSO and then diluted into buffer to yield final

DMSO concentrations of up to 0.3% which were also included in control solutions.

Results

Depolarization of SCG neurons via M1 receptors is not matched with the inhibition of Kv7 channels.

In order to evaluate the relation between the depolarization of SCG neurons and the inhibition of Kv7 channels, both through the activation of M1 receptors by oxotremorine M, an initial set of 42 neurons was investigated. The values of resting membrane potentials in these neurons

9 ranged between -55 and -75 mV. Changes in membrane potential caused by 10 µM oxotremorine M varied between – 1 mV and +13 mV. There was no correlation between these values of resting membrane potential and the changes induced by the muscarinic agonist

(Figure 1A to D). In a subset of neurons with oxotremorine M-induced depolarizations of either less (n = 8) or more (n = 8) than 5 mV, currents through Kv7 channels were determined subsequently to the current clamp measurements (Figure 1E and F). The densities of Kv7 deactivation currents (triggered by hyperpolarisations from -30 to -55 mV) were comparable in these two groups of neurons (Figure 1G). Likewise, extent as well as time course of current inhibition by 10 µM oxotremorine M for these two sets of neurons was indiscernible (Figure

1H). We therefore concluded that mechanisms other than the inhibition of Kv7 channels also contribute to the depolarization caused by oxotremorine M. As the extent of oxotremorine M- induced depolarization varies considerably between single neurons (Figure 1B), the underlying signalling cascade was investigated only in neurons that displayed depolarisations of at least 5 mV.

Activation of PKC contributes to the depolarization of SCG neurons by oxotremorine M, but not to the inhibition of Kv7 channels.

Activation of PKC contributes to the depolarization of SCG neurons through B2 bradykinin receptors [34]. Therefore, various kinase inhibitors were tested for their effect on depolarisations triggered by 10 µM oxotremorine M (which was applied repeatedly once every four minutes; Figure 2A and B). Staurosporine (1 µM), a broad spectrum kinase inhibitor, increasingly reduced oxotremorine M-evoked depolarisations over a time period of

20 minutes (Figure 2A and B): initial depolarisations amounted to 8.4 + 1.1 mV, after 20 min of staurosporine exposure this value had decreased to 4.0 + 0.75 mV (n = 6; p < 0.01;

Kruskal-Wallis test). An analogous effect was observed when the PKC inhibitor GF 109203

X (1 µM) was used instead (Figure 2A and B). After 20 min of its presence the extent of

10 depolarization caused by oxotremorine M had decreased from 6.4 + 0.7 mV to 3.0 + 0.6 mV

(n = 6; p < 0.01; Kruskal-Wallis test). However, the solvent (0.1 % DMSO) did not cause significant changes when present as long as the kinase inhibitors, and the depolarisations amounted to 7.0 + 0.9 mV in the beginning and to 6.9 + 1.1 mV (n = 6; p > 0.1; Kruskal-

Wallis test) 20 min later. For a comparison with PKC inhibitors, effects of the Kv7 channel blocker XE 991 (3 µM) were investigated in an analogous manner: in its presence, the oxotremorine M-induced depolarizations decreased from 8.5 + 1.0 mV to 3.8 + 0.9 mV (n =

6; p < 0.05; Kruskal-Wallis test). When directly comparing the effects of staurosporine, GF

109203 X, XE 991, and DMSO by normalizing the oxotremorine M-induced depolarisations to the respective first value, the values after 20 min exposure to staurosporine, GF 109203 X, or XE 991 were smaller than those after exposure to the solvent (Figure 2B). Thus, the PKC inhibitors staurosporine and GF 109203 X significantly attenuated the depolarizing action of the muscarinic agonist, as did the Kv7 channel blocker XE 991.

Activation of PKC may also contribute to the inhibition of Kv7 channels via M1 muscarinic receptors [14]. Hence, the parallel effects described above might occur through actions converging at the level of Kv7 channels. To clarify whether the PKC inhibitors interfered with the muscarinic inhibition of Kv7 channels, deactivation currents through Kv7 channels were determined, and 10 µM oxotremorine M was applied once every four minutes again. In the presence of solvent (DMSO; Figure 2C), the inhibition of deactivation amplitudes by the agonist declined from 77.9 + 5.7 % to 71.1 + 4.9 % (n = 9). Likewise, in the presence of 1 µM staurosporine and 1 µM GF 109203 X, this inhibition decreased from 81.9 + 5.2 % to 72.2 +

3.0 % (n = 6) and from 68.1 + 11.1 % to 56.1 + 8.1 % (n = 7), respectively (Figure 2C and

D). None of these changes in Kv7 inhibition were statistically significant (p > 0.05; Kruskal-

Wallis test). A direct comparison of normalized inhibition values did not reveal any differences between staurosporine, GF 109203 X, or solvent (Figure 2D). Hence, in these

11 experiments with repeated Kv7 inhibition by 10 µM oxotremorine M, the employed PKC inhibitors did not cause any alteration.

Inhibition or downregulation of PKC attenuates oxotermorine M-induced noradrenaline release.

To investigate the signalling cascade of oxotremorine M-induced depolarizations not only in single cells which are quite heterogeneous in this response, a large population of neurons was investigated simultaneously. This was achieved by loading the neurons with [3H] noradrenaline and by subsequently stimulating the overflow of radioactivity by this muscarinic agonist. In such experiments, oxotremorine M, at concentrations <100 µM, triggers overflow of radioactivity through the selective activation of M1 receptors [22]. To control for effects of PKC inhibitors unrelated to the signalling cascade of M1 receptors, tritium overflow was also elicited by electrical field stimulation. Staurosporine (1 µM) did not alter tritium overflow triggered by electrical fields, but reduced that evoked by oxotremorine

M by more than 50 % (Figure 3A and B). Exposure of SCG cultures to phorbol 12-myristate

13-acetate for 24 h downregulates all but atypical PKC isoforms [34]. In cultures treated in that way, electrically evoked release was the same as in untreated sister cultures. However, oxotremorine M-induced tritium overflow in phorbol 12-myristate 13-acetate-treated cultures amounted to only 10 % of that in untreated cultures (Figure 3C and D).

Together, the above results indicate that some PKC isoforms, with the exception of atypical ones, are involved in the excitation of SCG neurons via M1 receptors. To further elaborate which PKC subtypes may contribute, GF 109203 X and related PKC inhibitors (GÖ 6976 and

GÖ 6983) with divergent subtype preferences [37] were employed. None of these drugs caused significant alterations in electrically induced tritium overflow (Figure 4B). In contrast, at 0.01 µM GÖ 6976 and GÖ 6983, but not GF 109203 X, significantly diminished oxotremorine M-evoked overflow, and at higher concentrations all the PKC inhibitors shared

12 this effect (Figure 4C). Thus, with respect to the inhibition of noradrenaline release caused by oxotremorine M, GÖ 6976 and GÖ 6983 were more potent than GF 109203 X.

Cl- conductances contribute to the depolarization of SCG neurons by oxotremorine M.

To elucidate the ionic basis of oxotremorine M-induced depolarisations, the composition of the pipette solution was changed by replacing 55 mM KCl by equimolar concentrations of either CsCl or potassium gluconate. The resting membrane potentials determined with these three different pipette solutions were -65.2 + 2.4 mV (KCl; n = 7), -77.7 + 3.3 mV (K- gluconate; n = 5), and -62.8 + 1.9 mV (CsCl; n = 5). While changes in membrane voltage caused by 10 µM oxotremorine M were not affected by alterations in K+, the reduction of intracellular Cl- clearly reduced the depolarizing response (Figure 5A).

Since this result suggested that Cl- was the relevant ion, effects of oxotremorine M were compared with those of GABA. As in the case of oxotremorine M, 10 µM GABA depolarized the neurons. The depolarizations caused by GABA developed instantaneously and then decayed during the presence of the transmitter, whereas the oxotremorine M-induced depolarisations developed slowly during the one minute exposure towards the agonist (Figure

5B). In perforated patch recordings with amphotericin B, intracellular anion concentrations depend on those of the pipette solution (see above). This disruption of the intracellular anion homeostasis can be prevented by using gramicidin D instead of amphotericin B [2]. With our standard pipette solution containing 55 mM KCl, depolarizations caused by 10 µM oxotremorine M were the same, whether amphotericin B or gramicidin D was used as ionophore, and the same was true for GABA-evoked changes in membrane voltage (Figure

5B and C). Thus, depolarisations caused by M1 receptor activation appear to rely on high intracellular Cl- concentrations and hence on a Cl- conductance.

13

Alterations in extracellular Cl- affect noradrenaline release induced by either oxotremorine M or GABA.

To explore the role of Cl- conductances in the stimulatory action of oxotremorine M on noradrenaline release, [3H] overflow was triggered by this agonist either in quasi physiological solution (containing 134 mM Cl-), or in a solution in which 60 mM NaCl had been replaced by 60 mM sodium gluconate. In these two different solutions, tritium overflow triggered by electrical field stimulation was essentially the same (and if anything, reduced in the presence of sodium gluconate rather than enhanced; figure 6D). Oxotremorine M-evoked overflow, however, was significantly enhanced when extracellular Cl- had been reduced

(Figure 6A and B). For comparison, cultures were exposed to 10 µM GABA instead of the same concentration of the muscarinic agonist. As expected, GABA-induced overflow was also enhanced by lowering extracellular Cl- (Figure 6 C and D). Thus, the stimulation of noradrenaline release from SCG neurons through activation of M1 receptors depends on the extracellular Cl- concentration.

Cl- channel blockers diminish the depolarization of SCG neurons by oxotremorine M.

The above results hint to a role of Cl- conductances in the excitatory action of oxotremorine

M. There is a large number of different voltage- and Ca2+-gated Cl- channels, but only a comparably low number of relatively unselective blockers [11,31]. Two frequently used representatives of these blockers are SITS and niflumic acid which were tested for their effects on depolarisations triggered by 10 µM oxotremorine M (which was applied repeatedly as in Figure 2). As the effects of Cl- channel blockers on the channels are complex (with voltage-dependent enhancing and decreasing activities) and develop slowly [32], these agents were applied for prolonged periods of time. In the presence of 300 µM niflumic acid or SITS

(Figure 7 A), oxotremorine M-induced depolarisations decreased from 7.4 + 0.8 mV to 4.4 +

0.6 mV (n = 7; p < 0.05; Kruskal-Wallis test). An equivalent decline was observed with 300

14

µM SITS (Figure 7A): the extent of depolarization caused by oxotremorine M fell from 6.6 +

0.4 mV to 4.2 + 0.5 mV (n = 7; p < 0.001; Kruskal-Wallis test). However, the solvent did not cause significant changes, and the depolarisations amounted to 8.2 + 0.8 mV in the beginning and to 7.2 + 0.9 mV (n = 7; p > 0.1; Kruskal-Wallis test) at the end of experiments. When directly comparing these changes by normalizing the oxotremorine M-induced depolarisations, the values after exposure to either SITS or niflumic acid were significantly smaller than those obtained in solvent (Figure 7B). Thus, the two Cl- channels blockers significantly attenuated the depolarizing action of the muscarinic agonist.

To reveal whether these channel blockers might also affect Kv7 channels or their inhibition via muscarinic receptors, currents through these latter channels were determined again. In the presence of 300 µM niflumic acid (Figure 2C) and 300 µM SITS, the inhibition of Kv7 deactivation currents decreased from 89.7 + 4.5 % to 68.2 + 12.4 % (n = 7) and from 87.0 +

2.5 % to 78.1 + 3.5 % (n = 7), respectively. In solvent, a similar trend was observed and the oxotremorine M-induced inhibition was 90.8 + 4.2 % in the beginning and 74.8 + 12.5 % (n =

7) at the end of recordings. All these changes in Kv7 inhibition were statistically non- significant. Furthermore, a direct comparison of normalized inhibition values did not reveal any differences between niflumic acid, SITS, and solvent (Figure 7D).

Inward currents induced by oxotremorine M are attenuated by blockers of Ca2+-activated Cl- channels and by PKC inhibitors.

To find out whether oxotremorine M can induce depolarizing currents, neurons were clamped to -65 mV and the M1 receptor agonist was applied for periods of two minutes. The potential of -65 mV was chosen as this was the median value of membrane potentials as determined in current clamp experiments (Figure 1B). Moreover, at this voltage Kv7 channels of SCG neurons are not activated [22]. In the presence of 10 µM oxotremorine M, inward currents

15 developed slowly and reached a maximum after 30 s to two minutes (Figure 7E). Maximal current amplitudes ranged between 20 and 120 pA.

As these currents were triggered by the activation of M1 receptors, but not by changes in membrane voltage, it appeared straightforward to assume that they were carried by Ca2+- activated rather than voltage-gated Cl- channels. Hence, currents were induced again by a second application of oxotremorine M to the very same cells in the presence of CaCCinh-A01 or T16Ainh-A01, two selective blockers of different Ca2+-activated Cl- channels [30]. The muscarinic agonist triggered currents of similar amplitudes again when re-applied in the presence of solvent (0.1% DMSO). In contrast, in the presence of 3 µM CaCCinh-A01 or 3

µM T16Ainh-A01, current amplitudes caused by the second oxotremorine M application were reduced significantly (Wilcoxon matched pair signed rank test; Figure 7E and F).

To reveal whether the triggering of these currents by oxotremorine M does also involve PKC, the agonist was applied repeatedly in the presence of staurosporine, GF 109203 X, or solvent

(as shown for depolarizations in figure 2A and B). Staurosporine (1 µM) increasingly reduced oxotremorine M-evoked currents: initial amplitudes amounted to 64.8 + 20.1 pA, and these were reduced to 36.5 + 19.6 pA (n = 7; p < 0.05; Kruskal-Wallis test) 20 min later. Likewise, when the PKC inhibitor GF 109203 X (1 µM) was used, amplitudes decreased from 77.8 +

14.4 pA to 33.8 + 10.6 pA (n = 7; p < 0.05; Kruskal-Wallis test). However, the solvent (0.1 %

DMSO) did not cause significant changes, and the amplitudes amounted to 95.6 + 37.8 pA and to 71.3 + 26.4 pA (n = 7; p > 0.1; Kruskal-Wallis test) in the beginning and at the end, respectively (Figure 7G and H).

Blockers of Cl- channels and Cl- transporters reduce noradrenaline release induced by oxotremorine M.

To corroborate the data shown above by an independent approach, SITS and niflumic acid were the first blockers to be tested for their effects on noradrenaline release. These two agents

16 did not affect electrically evoked tritium overflow, but significantly reduced overflow induced by oxotremorine M (Figure 8 A to C).

The release enhancing effect of reductions in extracellular Cl- concentrations (Figure 6) points to a role of high intracellular Cl- in the depolarizing action oxotremorine M. Neurons in the peripheral nervous system express Na+/K+/Cl- cotransporters which mediate Cl- uptake and intracellular Cl- accumulation. These transporters can be blocked by diuretics such as furosemide and bumetanide [18]. In the presence of 300 µM of either of these two drugs, overflow of radioactivity triggered by electrical fields remained unaltered, whereas oxotremorine M-induced overflow was significantly reduced (Figure 8 D to F). Thus, hindrance of Cl- uptake into the neurons selectively diminished the secretagogue action of oxotremorine M.

The results obtained with CaCCinh-A01 or T16Ainh-A01 in electrophysiological experiments were also confirmed with respect to [3H] noradrenaline release: both blockers at 1 to 10 µM did reduce tritium overflow triggered by the muscarinic agonist (Figure 8G and I). Electrically evoked overflow, however, was not reduced, but rather enhanced by CaCCinh-A01 as well as

T16Ainh-A01 in a concentration-dependent manner (Figure 8G and H).

Discussion

Transmission in autonomic ganglia involves acetylcholine as prime transmitter which triggers fast and slow EPSPs mediated by nicotinic and muscarinic receptors, respectively. The slow component of ganglionic transmission has been known to be mediated by an inhibition of Kv7 channels via M1 receptors for more than three decades [9]. In primary cultures of rat SCG neurons, activation of M1 receptors causes depolarization and ensuing noradrenaline release, and evidence has been presented that suggests these effects to rely on an inhibition of Kv7 channels as well [22]. However, the present results reveal an additional and novel mechanism

17 of M1 receptor-dependent excitation of sympathetic neurons that is independent of Kv7 channels.

Depolarizations caused by the mAChR agonist oxotremorine M were remarkably variable between single neurons and not related to the levels of resting membrane potential. However, there was no correlation between the extent of depolarization and either Kv7 channel current densities or the degree of Kv7 channel inhibition by oxotremorine M. Furthermore, the Kv7 channel blocker XE 991 reduced depolarizations caused by this muscarinic agonist, but did not abolish them. Hence, it appeared obvious to search for additional mechanisms involved in the depolarization of SCG neurons via M1 mAChRs.

The activation of M1 receptors in sympathetic neurons turns on the entire Gq- and phospholipase C-linked signalling cascade which includes a boost of PKC [25]. Activated

PKC may contribute to the muscarinic inhibition of Kv7 channels [14], but this effect is generally believed to be mainly mediated by the depletion of membrane phosphatidylinositol

4,5-bisphosphate [13]. In accordance with this latter concept, the present experiments did not reveal any effect of PKC inhibitors on the oxotremorine M-induced inhibition of currents though Kv7 channels. In contrast, staurosporine and GF 109203 X both reduced depolarisations caused by the muscarinic agonist. Moreover, noradrenaline release evoked by oxotremorine M (but not that induced by electrical field stimulation) was also reduced by various measures employed to prevent PKC activity, as detailed below.

In primary cultures of rat SCG, the expression of PKC α, βI, βII, δ, ε, and ζ has been documented by immunoblots, whereas PKC γ or μ were absent. All of the former isoforms with one exception, PKC ζ, are downregulated by a long lasting phorbol ester treatment [34].

Since exposure of the neurons to phorbol 12-myristate 13-acetate for 24 h did reduce oxotremorine M-induced noradrenaline release, a contribution of PKC ζ can be excluded. GF

109203 X is more potent at PKC α than at PKC β, and at least tenfold less potent at PKC δ

18 and PKC ε than at PKC β; atypical PKCs are not affected by GF 109203 X concentrations up to 1 µM. GÖ 6983, in contrast, is equipotent at virtually all PKC isoforms including atypical ones, whereas Gö 6976 does not inhibit Ca2+-independent (δ and ε) and atypical PKC enzymes at concentrations up to 3 µM [29, 37]. In the present experiments (using concentrations of 0.01 to 1 µM), GÖ 6983 and GÖ 6976 turned out to be more potent in reducing oxotremorine M-induced noradrenaline release than GF 109203 X. Thus, the PKC enzymes involved can only be classical (Ca2+-sensitive) ones and include β rather than α subtypes. Hence, M1 receptors appear to engage other PKC enzymes than B2 bradykinin receptors to excite SCG neurons, as the latter receptors were found to be linked to PKC δ and

ε [34].

To elucidate the ionic mechanisms underlying the excitation of SCG neurons via M1 receptors, various types of experiments were performed. First, the intracellular concentrations of K+ and Cl- were changed in perforated patch experiments, and only the changes in Cl- caused alterations in oxotremorine M-induced depolarizations. Second, changes in membrane potentials caused by the mAChR agonist were compared with those caused by GABA, and both turned out to be depolarizing. Moreover, both agonists triggered noradrenaline release in

SCG cultures. Third, extracellular Cl- was partially substituted by gluconate, and this enhanced noradrenaline release elicited by oxotremorine M and GABA, respectively, but not that induced by electrical field stimulation. Together, these results hint to an induction of a Cl- conductance as one mechanism involved in the excitatory actions of M1 receptor activation.

This conclusion leads to at least two additional questions: (i) why is the activation of a Cl- conductance depolarizing? (ii) Which are the channels mediating this Cl- conductance? As far as question (i) is concerned, GABA has been reported previously to depolarize rat SCG neurons [1]. This depolarizing action of GABA is based on the fact that these neurons accumulate high intracellular Cl- concentrations of about 30 mM [3]. High intracellular Cl-

19 levels in neurons involve Cl- uptake via Na+/K+/Cl- cotransporters which can be blocked by bumetanide and related drugs [18]. The relevance of this mechanism for the excitatory actions of M1 receptors in SCG neurons was documented by the inhibition of oxotremorine M-evoked noradrenaline release by bumetanide and furosemide.

- With respect to the Cl channels involved in the action of M1 receptors, several blockers have been employed in the present experiments. The non-selective Cl- channel blockers niflumic acid and SITS reduced depolarisations as well as noradrenaline release triggered by oxotremorine M. This effect was specific for the depolarising action of the muscarinic agonist, as neither the oxotremorine M-dependent inhibition of Kv7 channels nor electrically evoked noradrenaline release were altered by these agents. While these results confirmed the

- contribution of some Cl channels to the excitatory action of M1 receptor activation, it remained unclear which of the numerous Cl- channel subtypes might be involved [11,31].

Previously, oxotremorine M had been reported to enhance a depolarization-evoked Ca2+- dependent Cl- current in rat SCG neurons [28]. More recently, mAChR agonists have been found to increase Ca2+-dependent Cl- currents in interstitial cells of Cajal [39]. There is quite a variety of different Ca2+-activated Cl- channels expressed in various cell types including neurons. Recently, TMEM16 proteins, in particular TMEM16A, also known as anoctamin 1

(ANO1), were found to contribute to the formation of Ca2+-activated Cl- channels [4, 15].

Potent blockers that are selective for Ca2+- activated Cl- channels in general and for

TMEM16A in particular have been synthesized recently [30]. In the present experiments, these blockers (CaCCinh-A01 and T16Ainh-A01) reduced inward currents evoked by oxotremorine M and largely attenuated noradrenaline release triggered by the muscarinic agonist. Thus, TMEM16A is the most likely candidate to mediate the excitatory action of M1 receptor activation in SCG neurons. For comparison, sensory neurons were among the first cells that were revealed to exhibit Ca2+-dependent Cl- currents and to express

TMEM16A/ANO1 [15], and the latter channels have been demonstrated recently to be

20 involved in the nociceptive activity of bradykinin [24]. TMEM16A/ANO1 is activated by

Ca2+ concentrations in the low micromolar range [15], but oxotremorine M-induced increases in intracellular Ca2+ in SCG neurons as quantified by fura-2 microfluorometry do not exceed

1 µM and are only observed at depolarized membrane potentials [33]. However, activation of

TMEM16A/ANO1 in sensory neurons via G protein coupled receptors relies on spatially restricted Ca2+ signals that hardly correlate with global cellular Ca2+ as determined by Ca2+ indicator microfluorometry [16]. Hence, the specific features of the Ca2+ signals that may link

M1 receptors to TMEM16A/ANO1 remain to be determined.

Considering that the M1 receptor agonist led to an activation of PKC, on one hand, and to the gating of TMEM16A/ANO1 channels, on the other hand, the causal relation between these two events remained to be determined. The TMEM16A/ANO1 amino acid sequences in mammals contain putative phosphorylation sites for PKC [15], and currents through

TMEM16A/ANO1 in biliary epithelial cells were found to be enhanced through an activation of classical PKC enzymes via P2Y receptors [12]. In the present experiments, PKC inhibitors attenuated the oxotremorine M-evoked currents that were otherwise reduced by blockers of

Ca2+-dependent Cl- currents and TMEM16A/ANO1 channels. Previously, the facilitation of depolarization-evoked Ca2+-dependent Cl- currents in SCG neurons by oxotremorine M was also reported to involve PKC [28]. Thus, PKC is involved in the activation of Ca2+- activated

- Cl channels via M1 receptors.

In summary, this report demonstrates that slow cholinergic excitation of sympathetic neurons involves mechanisms other than the inhibition of Kv7 channels which include an activation of classical PKCs and of Ca2+- activated Cl- channels.

21

Conflict of Interest

The authors declare that they have no conflict of interest.

Acknowledgements

This study was supported by the Austrian Science Fund (FWF; P23658, P23670 and W1205);

IS and HG are members of the doctoral programme “Cell Communication in Health and

Disease” (CCHD; co-financed by FWF and the Medical University of Vienna). The excellent technical assistance of G. Gaupmann is gratefully acknowledged.

Ethical standards

The present experiments comply with the current law in Austria.

Abbreviations

CaCCinh-A01, 6-(1,1-Dimethylethyl)-2-[(2-furanylcarbonyl)amino] 4,5,6,7- tetrahydrobenzo[b]thiophene-3-carboxylic acid; GÖ 6976, 12-(2-Cyanoethyl)-6,7,12,13- tetrahydro-13-methyl-5-oxo-5H-indolo[2,3-a]pyrrolo[3,4-c]carbazole; GÖ 6983, 3-[1-[3-

(Dimethylamino)propyl]-5-methoxy-1H-indol-3-yl]-4-(1H-indol-3-yl)-1H-pyrrole-2,5-dione; mAChRs, muscarinic acetylcholine receptors; nAChRs, nicotinic acetylcholine receptors;

PKC, protein kinase C; SCG, superior cervical ganglion; SITS, disodium 4-acetamido-4'- isothiocyanato-stilbene-2,2'-disulfonic acid; T16Ainh-A01, 2-[(5-Ethyl-1,6-dihydro-4-methyl- 6-oxo-2-pyrimidinyl)thio]-N-[4-(4-methoxyphenyl)-2-thiazolyl]acetamide; XE 991, 10,10- bis(4-pyridinylmethyl)-9(10H)-anthracenone.

22

References

[1] Adams PR, Brown DA (1975) Actions of gamma-aminobutyric acid on sympathetic ganglion cells. J. Physiol. 250:85–120.

[2] Akaike N (1996) Gramicidin perforated patch recording and intracellular chloride activity in excitable cells. Prog. Biophys. Mol. Biol. 65:251–64

[3] Ballanyi K, Grafe P (1985) An intracellular analysis of gamma-aminobutyric-acid- associated ion movements in rat sympathetic neurones. J. Physiol. 365:41–58.

[4] Berg J, Yang H, Jan LY (2012) Ca2+-activated Cl- channels at a glance. J. Cell Sci. 125:1367–71.

[5] Bernheim L, Mathie A, Hille B (1992) Characterization of muscarinic receptor subtypes inhibiting Ca2+ current and M current in rat sympathetic neurons. Proc Natl Acad Sci U S A 89:9544–9548.

[6] Boehm S (1998) Selective inhibition of M-type potassium channels in rat sympathetic neurons by uridine nucleotide preferring receptors. Br. J. Pharmacol. 124:1261–1269.

[7] Boehm S, Huck S (1997) Noradrenaline release from rat sympathetic neurones triggered by activation of B2 bradykinin receptors. Br. J. Pharmacol. 122:455–462.

[8] Brown AM (1967) Cardiac sympathetic adrenergic pathways in which synaptic transmission is blocked by atropine sulfate. J. Physiol. 191:271–88.

[9] Brown DA (1983) Slow cholinergic excitation — a mechanism for increasing neuronal excitability. Trends Neurosci. 6:302–307.

[10] Delmas P, Brown DA (2005) Pathways modulating neural KCNQ/M (Kv7) potassium channels. Nat Rev Neurosci 6:850–862.

[11] Duran C, Thompson CH, Xiao Q, Hartzell HC (2010) Chloride channels: often enigmatic, rarely predictable. Annu. Rev. Physiol. 72:95–121.

[12] Dutta AK, Woo K, Khimji A, Kresge C, Feranchak AP (2013) Mechanosensitive Cl- secretion in biliary epithelium mediated through TMEM16A. Am. J. Physiol. Gastrointest. Liver Physiol. 304:G87–98.

[13] Hernandez CC, Zaika O, Tolstykh GP, Shapiro MS (2008) Regulation of neural KCNQ channels: signalling pathways, structural motifs and functional implications. J Physiol 586:1811–1821.

[14] Hoshi N, Zhang J-S, Omaki M, Takeuchi T, Yokoyama S, Wanaverbecq N, Langeberg LK, Yoneda Y, Scott JD, Brown DA, Higashida H (2003) AKAP150 signaling complex promotes suppression of the M-current by muscarinic agonists. Nat. Neurosci. 6:564–71.

23

[15] Huang F, Wong X, Jan LY (2012) International Union of Basic and Clinical Pharmacology. LXXXV: calcium-activated chloride channels. Pharmacol. Rev. 64:1– 15.

[16] Jin X, Shah S, Liu Y, Zhang H, Lees M, Fu Z, Lippiat JD, Beech DJ, Sivaprasadarao A, Baldwin SA, Zhang H, Gamper N (2013) Activation of the Cl- channel ANO1 by localized calcium signals in nociceptive sensory neurons requires coupling with the IP3 receptor. Sci. Signal. 6:ra73.

[17] Jones S, Brown DA, Milligan G, Willer E, Buckley NJ, Caulfield MP (1995) Bradykinin excites rat sympathetic neurons by inhibition of M current through a mechanism involving B2 receptors and G alpha q/11. Neuron 14:399–405.

[18] Kahle KT, Staley KJ, Nahed B V, Gamba G, Hebert SC, Lifton RP, Mount DB (2008) Roles of the cation-chloride cotransporters in neurological disease. Nat. Clin. Pract. Neurol. 4:490–503.

[19] Kristufek D, Koth G, Motejlek A, Schwarz K, Huck S, Boehm S (1999) Modulation of spontaneous and stimulation-evoked transmitter release from rat sympathetic neurons by the cognition enhancer linopirdine: insights into its mechanisms of action. J Neurochem 72:2083–2091.

[20] Kubista H, Kosenburger K, Mahlknecht P, Drobny H, Boehm S (2009) Inhibition of transmitter release from rat sympathetic neurons via presynaptic M(1) muscarinic acetylcholine receptors. Br J Pharmacol 2009;156:1342–1352.

[21] Lechner SG, Hussl S, Schicker KW, Drobny H, Boehm S (2005) Presynaptic inhibition via a phospholipase C- and phosphatidylinositol bisphosphate-dependent regulation of neuronal Ca2+ channels. Mol. Pharmacol. 68:1387–96.

[22] Lechner SG, Mayer M, Boehm S (2003) Activation of M1 muscarinic receptors triggers transmitter release from rat sympathetic neurons through an inhibition of M- type K+ channels. J. Physiol. 553:789–802.

[23] Lee S-Y, Choi H-K, Kim S-T, Chung S, Park MK, Cho J-H, Ho W-K, Cho H (2010) Cholesterol inhibits M-type K+ channels via protein kinase C-dependent phosphorylation in sympathetic neurons. J. Biol. Chem. 285:10939–50.

[24] Liu B, Linley JE, Du X, Zhang X, Ooi L, Zhang H, Gamper N (2010) The acute nociceptive signals induced by bradykinin in rat sensory neurons are mediated by inhibition of M-type K+ channels and activation of Ca2+-activated Cl- channels. J Clin Invest 120:1240–1252.

[25] Malhotra RK, Bhave S V, Wakade TD, Bhave AS, Wakade AR (1990) Effects of neurotransmitters and peptides on phospholipid hydrolysis in sympathetic and sensory neurons. FASEB J. 4:2492–8.

[26] Marrion N V (1997) Control of M-current. Annu. Rev. Physiol. 59:483–504.

24

[27] Marrion N V, Smart TG, Marsh SJ, Brown DA (1989) Muscarinic suppression of the M-current in the rat sympathetic ganglion is mediated by receptors of the M1-subtype. Br. J. Pharmacol. 98:557–73.

[28] Marsh SJ, Trouslard J, Leaney JL, Brown DA (1995) Synergistic regulation of a neuronal chloride current by intracellular calcium and muscarinic receptor activation: a role for protein kinase C. Neuron 15:729–37.

[29] Martiny-Baron G, Kazanietz MG, Mischak H, Blumberg PM, Kochs G, Hug H, Marme D, Schachtele C (1993) Selective inhibition of protein kinase C isozymes by the indolocarbazole Go 6976. J Biol Chem 268:9194–9197.

[30] Namkung W, Phuan P-W, Verkman AS (2011) TMEM16A inhibitors reveal TMEM16A as a minor component of calcium-activated chloride channel conductance in airway and intestinal epithelial cells. J. Biol. Chem. 286:2365–74.

[31] Nilius B, Droogmans G (2003) Amazing chloride channels: an overview. Acta Physiol. Scand. 177:119–47.

[32] Piper AS, Greenwood IA, Large WA (2002) Dual effect of blocking agents on Ca2+- activated Cl(-) currents in rabbit pulmonary artery smooth muscle cells. J. Physiol. 539:119–31.

[33] Del Río E, Bevilacqua JA, Marsh SJ, Halley P, Caulfield MP (1999) Muscarinic M1 receptors activate phosphoinositide turnover and Ca2+ mobilisation in rat sympathetic neurones, but this signalling pathway does not mediate M-current inhibition. J. Physiol. 520 Pt 1:101–11.

[34] Scholze T, Moskvina E, Mayer M, Just H, Kubista H, Boehm S (2002) Sympathoexcitation by bradykinin involves Ca2+-independent protein kinase C. J Neurosci 22:5823–5832.

[35] Schwartz DD, Malik KU (1993) Cyclic AMP modulates but does not mediate the inhibition of [3H]norepinephrine release by activation of alpha-2 adrenergic receptors in cultured rat ganglion cells. Neuroscience 52:107–113.

[36] Trendelenburg U (1966) Observations on the ganglion-stimulating action of angiotensin and bradykinin. J Pharmacol Exp Ther 154:418–425.

[37] Way KJ, Chou E, King GL (2000) Identification of PKC-isoform-specific biological actions using pharmacological approaches. Trends Pharmacol. Sci. 21:181–187.

[38] Wess J, Eglen RM, Gautam D (2007) Muscarinic acetylcholine receptors: mutant mice provide new insights for drug development. Nat Rev Drug Discov 6:721–733.

[39] Zhu MH, Sung IK, Zheng H, Sung TS, Britton FC, O’Driscoll K, Koh SD, Sanders KM (2011) Muscarinic activation of Ca2+-activated Cl- current in interstitial cells of Cajal. J. Physiol. 589:4565–82.

25

Fig. 1 Comparison of depolarization and Kv7 channel inhibition by oxotremorine M.

Membrane potential and currents through Kv7 channels in SCG neurons were recorded in current-clamp and voltage-clamp mode, respectively, using the amphotericin B-perforated patch technique. A shows the time course of membrane voltage in two different SCG neurons; oxotremorine M (OxoM, 10 μM) was present as indicated by the bars. B displays the frequency distribution of resting membrane potentials (voltage in bins of 2 mV) as determined in 42 SCG neurons. C displays the frequency distribution of changes in membrane potentials

(Δ mV in bins of 2 mV) caused by 10 µM oxotremorine M as determined in these 42 SCG neurons. D displays a correlation between resting membrane potential (voltage) and membrane potential changes (Δ mV) caused by 10 µM oxotremorine M in the same 42 SCG neurons; experiments were carried out as shown in A. The Spearman coefficient for this correlation is -0.053 (95% confidence interval: -0.3594 to 0.2638). E A subset of 16 neurons was categorized according to the extent of depolarization caused by 10 µM oxotremorine M

(<5 mV, n = 8; >5 mV, n = 8); the means of the depolarization observed in these two groups are shown. F Subsequently, currents through Kv7 channels were recorded by holding these 16 cells at a voltage of -30 mV and by applying hyperpolarizations to -55 mV once every 15 s.

The traces show current responses of two neurons, one out of each of these two categories. G displays the mean values of densities of deactivation currents caused by the steps from -30 to

-55 mV in the neurons from both categories; n = 8. H displays the time course of deactivation current amplitudes caused by the steps from -30 to -55 mV in the neurons from both categories; oxotremorine M (OxoM, 10 μM) was present as indicated by the bar; n = 8.

Fig. 2 Effects of PKC inhibitors on depolarization and Kv7 channel inhibition by oxotremorine M. Membrane potential and currents through Kv7 channels in SCG neurons were recorded in current-clamp and voltage-clamp mode, respectively, using the amphotericin

B-perforated patch technique. Oxotremorine M (OxoM, 10 μM) was present for 6 periods of 26

60 s each; these periods of oxotremorine M application were separated by 3 min intervals.

From minute two after the first oxotremorine M application onward, PKC inhibitors, XE 991, or solvent (0.1% DMSO) were present throughout the remaining measurements. A shows the time course of membrane voltage in two different SCG neurons during the first and the 6th exposure to oxotremorine M (OxoM, 10 μM); the agonist was present as indicated by the bars. After the first oxotremorine M exposure, either 1 µM staurosporine or 1 µM GF 109203

X were present. B displays the changes in membrane voltage (Δ mV) caused by these 6 oxotremorine M applications (O1 to O6) in the presence of DMSO, staurosporine, GF 109203

X, or XE 991 (3 µM); the values of these 6 depolarizations were normalized to the value of the first one; n = 6; * indicates a significant difference between the four values at O6 (p <

0.05; Kruskal-Wallis test). C shows current responses of one neuron that was clamped at a voltage of -30 mV and hyperpolarized to -55 mV once every 15 s and that has been exposed to 1 µM staurosporine. The traces were obtained before (solvent) and during (OxoM) the first application (O1) of 10 μM oxotremorine M as well as before (stauro) and during (stauro +

OxoM) the sixth application (O6) of oxotremorine M. D displays the changes in Kv7 inhibition (quantified by deactivation current amplitudes) caused by these 6 oxotremorine M applications (O1 to O6) in the presence of either DMSO, staurosporine, or GF 109203 X; these 6 values of Kv7 inhibition were normalized to the value of the first one; n = 6 to 9.

Fig. 3 Effect of PKC inhibition on noradrenaline release evoked by electrical field stimulation or oxotremorine M. Cultures of SCG were labelled with [3H]noradrenaline, superfused, and subsequent to a 60 min washout period, 4 min fractions of superfusate were collected. 60 monophasic rectangular pulses (0.5 ms, 60 mA, 50 V/cm) were applied in minute 73, and oxotremorine M (10 μM) was present in minutes 93 and 94. From minute 50 of superfusion onward, the buffer contained either solvent (0.1% DMSO) or 1 μM staurosporine.

Alternatively, cultures had been treated with either 0.1% DMSO (untreated) or 1 μM phorbol 27

12-myristate 13-acetate (PMA-treated) for 24 h. A shows the time course of [³H] outflow as percentage of radioactivity in the cells in the presence of either solvent (○) or staurosporine

(●); n = 3. B summarizes the effect of 1 μM staurosporine on [³H] overflow evoked by electrical field stimulation (EFS) or oxotremorine M (OxoM); n = 8 to 9. C shows the time course of [³H] outflow as percentage of radioactivity in the cells which had either been treated with phorbol 12-myristate 13-acetate (●) or had remained untreated (○); n = 3. D summarizes the effect of phorbol 12-myristate 13-acetate treatment on [³H] overflow evoked by electrical field stimulation (EFS) or oxotremorine M (OxoM); n = 12. ** indicates a significant difference vs solvent and untreated, respectively, at p < 0.01, n.s. indicates no significance.

Fig. 4 Effects of subtype preferring PKC inhibitors on noradrenaline release evoked by electrical field stimulation or oxotremorine M. Cultures of SCG were labelled with

[3H]noradrenaline, superfused, and subsequent to a 60 min washout period, 4 min fractions of superfusate were collected. 60 monophasic rectangular pulses (0.5 ms, 60 mA, 50 V/cm) were applied in minute 73, and oxotremorine M (10 μM) was present in minutes 93 and 94. From minute 50 of superfusion onward, the buffer contained either solvent (0.1% DMSO) or 0.01 to

1 μM of GF 109203 X (GF), Gö 6976, or Gö 6983. A shows the time course of [³H] outflow as percentage of radioactivity in the cells in the presence of either solvent (○) or 1 μM of GF

109203 X (●); n = 3. B summarizes the effects of the indicated concentrations of PKC inhibitors on [³H] overflow evoked by electrical field stimulation (EFS). Overflow in the presence of the inhibitors is depicted as percentage of the overflow in the presence of solvent

(% of control). C summarizes the effects of the indicated concentrations of PKC inhibitors on

[³H] overflow evoked by oxotremorine M (OxoM). Overflow in the presence of the inhibitors is depicted as percentage of the overflow in the presence of solvent (% of control). In B and C n = 6 to 9. *, **, *** indicate significant differences vs. solvent at p < 0.05, p < 0.01, and p <

0.001, respectively. 28

Fig. 5 Effects of Cl- and ionophore substitution on depolarizations by oxotremorine M.

Membrane potential in SCG neurons was recorded in current-clamp mode using the amphotericin B- or gramicidin D-perforated patch technique. A displays the extent of depolarizations elicited by 10 μM oxotremorine M in amphotericin B-perforated patch recordings with pipette solutions containing 75 mM K2SO4 plus 55 mM KCl, 55 mM potassium gluconate (KGluc), or 55 mM CsCl; n = 5 to 7; *, ** indicate significant differences vs. potassium gluconate at p < 0.05 and p < 0.01, respectively (Kruskal-Wallis test). B shows original current clamp traces using either amphotericin B (upper traces) or gramicidin D (lower traces). 10 μM oxotremorine M (OxoM) or 10 μM GABA were present as indicated by the bars. C depicts the extent of depolarizations elicited by 10 μM oxotremorine M or 10 μM GABA in either amphotericin B- or gramicidin D-perforated patch recordings; n = 5 to 7.

Fig. 6 Effects of Cl- substitution on noradrenaline release evoked by electrical field stimulation, oxotremorine M, or GABA. Cultures of SCG were labelled with

[3H]noradrenaline, superfused, and subsequent to a 60 min washout period, 4 min fractions of superfusate were collected. 60 monophasic rectangular pulses (0.5 ms, 60 mA, 50 V/cm) were applied in minute 73, and oxotremorine M (10 μM; in A and B) or GABA (10 μM; in C and

D) were present in minutes 93 and 94. From minute 50 of superfusion onward, the buffer contained 134 or 74 mM Cl- (the lacking Cl- was replaced by gluconate). A and C shows the time course of [³H] outflow as percentage of radioactivity in the cells; n = 3. B summarizes the amount of [³H] overflow evoked by electrical field stimulation (EFS) and oxotremorine M

(OxoM), respectively; n = 12. D summarizes the amount of [³H] overflow evoked by electrical field stimulation (EFS) and GABA, respectively; n = 10 to 12. ** indicates a significant difference vs. 134 mM Cl- at p < 0.01, n.s. indicates no significance.

29

Fig. 7 Effects of Cl- channel blockers on depolarizations, Kv7 channel inhibition, and inward currents induced by oxotremorine M. Membrane potential and currents in SCG neurons were recorded in current-clamp and voltage-clamp mode, respectively, using the amphotericin B- perforated patch technique. In A to D and G to H, oxotremorine M (OxoM, 10 μM) was present for 6 periods of 60 s each; these periods of oxotremorine M application were separated by 3 min intervals. From minute two after the first oxotremorine M application onward, Cl- channel blockers or solvent were present throughout the remaining measurement.

A shows the time course of membrane voltage in two different SCG neurons during the first and the 6th exposure to oxotremorine M (OxoM, 10 μM); the agonist was present as indicated by the bars. After the first oxotremorine M exposure, either 300 µM niflumic acid or 300 µM

SITS were present. B displays the changes in membrane voltage (Δ mV) caused by these 6 oxotremorine M applications (O1 to O6) in the presence of either solvent, niflumic acid, or

SITS; the values of these 6 depolarizations were normalized to the value of the first one; n =

7; * indicates a significant difference between the three values at O6 (p < 0.05; one way

Kruskal-Wallis test). C shows current responses of one neuron that was clamped at a voltage of -30 mV and hyperpolarized to -55 mV once every 15 s and that has been exposed to 300

µM niflumic acid. The traces were obtained before (solvent) and during (OxoM) the first application (O1) of 10 μM oxotremorine M as well as before (niflumic) and during (niflumic

+ OxoM) the sixth application (O6) of oxotremorine M. D displays the changes in Kv7 inhibition (quantified by deactivation current amplitudes) caused by these 6 oxotremorine M applications (O1 to O6) in the presence of either solvent, niflumic acid, or SITS; these 6 values of Kv7 inhibition were normalized to the value of the first one; n = 7. In E and F, currents were recorded at a holding potential of -65 mV. Oxotremorine M (10 μM) was present for two periods of 120 s each; these periods of oxotremorine M application were separated by 3 min intervals. From minute two after the first oxotremorine M application 30 onward, Cl- channel blockers or solvent were present throughout the remaining measurement.

E shows the time course of membrane currents in two different SCG neurons during the first and the second exposure to oxotremorine M (OxoM, 10 μM); the agonist was present as indicated by the bars. After the first oxotremorine M exposure, either 3 µM CaCCinh or 0.1%

DMSO (solvent) were present. F displays the current amplitudes caused by the first and the second oxotremorine M application in the presence of either solvent, 3 µM CaCCinh, or 3 µM

T16Ainh; n = 6. * indicates a significant difference at p < 0.05 (Wilcoxon matched pairs signed rank test). G shows the time course of membrane currents in two different SCG neurons during the first and the 6th exposure to oxotremorine M (OxoM, 10 μM); the agonist was present as indicated by the bars. After the first oxotremorine M exposure, either 1 µM GF

109203 X or 1 µM staurosporine was present. H displays the changes in the amplitudes of the currents caused by these 6 oxotremorine M applications (O1 to O6) in the presence of either solvent, 1 µM GF 109203 X, or 1 µM staurosporine; the values of these 6 current amplitudes were normalized to the value of the first one; n = 7; * indicates a significant difference between the three values at O6 (p < 0.05; Kruskal-Wallis test).

Fig. 8 Effects of Cl- channel blockers and inhibitors of Cl- transporters on noradrenaline release evoked by electrical field stimulation or oxotremorine M. Cultures of SCG were labelled with [3H]noradrenaline, superfused, and subsequent to a 60 min washout period, 4 min fractions of superfusate were collected. 60 monophasic rectangular pulses (0.5 ms, 60 mA, 50 V/cm) were applied in minute 73, and oxotremorine M (10 μM) was present in minutes 93 and 94. From minute 50 of superfusion onward, the buffer contained solvent, 300

μM SITS, 300 μM niflumic acid, 300 μM bumetanide, 300 μM furosemide, CaCCinh, or

T16Ainh, the latter two at concentrations of 1, 3, or 10 µM. A shows the time course of [³H] outflow as percentage of radioactivity in the cells in the presence of either solvent (○) or SITS

(●); n = 3. B summarizes the effect of 300 μM SITS on [³H] overflow evoked by electrical 31 field stimulation (EFS) or oxotremorine M (OxoM); n = 18. C summarizes the effect of 300

μM niflumic acid on [³H] overflow evoked by electrical field stimulation (EFS) or oxotremorine M (OxoM); n = 9. D shows the time course of [³H] outflow as percentage of radioactivity in the cells in the presence of either solvent (○) or bumetanide (●); n = 3. E summarizes the effect of 300 μM bumetanide on [³H] overflow evoked by electrical field stimulation (EFS) or oxotremorine M (OxoM); n = 12. F summarizes the effect of 300 μM furosemide on [³H] overflow evoked by electrical field stimulation (EFS) or oxotremorine M

(OxoM); n = 12. *, ** indicate significant differences versus the respective value obtained in solvent at p < 0.05 and p < 0.01, respectively. G shows the time course of [³H] outflow as percentage of radioactivity in the cells in the presence of either solvent (○) or 3 µM CaCCinh

(●); n = 3. H summarizes the effects of the indicated concentrations of CaCCinh or T16Ainh on [³H] overflow evoked by electrical field stimulation (EFS). Overflow in the presence of the inhibitors is depicted as percentage of the overflow in the presence of solvent (% of control). I summarizes the effects of the indicated concentrations of CaCCinh or T16Ainh on [³H] overflow evoked by oxotremorine M (OxoM). Overflow in the presence of the inhibitors is depicted as percentage of the overflow in the presence of solvent (% of control). In H and I, n

= 6 to 9. *, **, *** indicate significant differences vs. solvent at p < 0.05, p < 0.01, and p <

0.001, respectively (Kruskal-Wallis test).

32

2 Results

58 2 Results

59 2 Results

60 2 Results

61 2 Results

62 2 Results

63 2 Results

64 3 Discussion

Inputs from preganglionic neurons cause two forms of excitatory postsynaptic potentials (EPSP) in postganglionic sympathetic neurons that both contribute to ganglionic transmission. Acetyl- choline, released from preganglionic nerve terminals, causes opening of nicotinic acetylcholine receptors, members of the superfamily of ligand-gated ion channels. As these ion channels can be activated and inactivated within a time range of milliseconds, this process is referred to as fast ganglionic transmission. On the other hand, the released acetylcholine also activates muscarinic acetylcholine receptors. These receptors are G-protein coupled and mediate their effects within a time range of seconds or even minutes. Hence, this form is termed slow ganglionic transmission (Kobayashi and Libet 1968; Kobayashi and Libet 1970; Brown and Selyanko 1985). In sympa- thetic neurons, muscarinic M1, M2, and M4 receptors are expressed, but only Gq-coupled M1 receptors mediate the slow EPSP (Marrion et al. 1989). Whether the EPSPs are fast or slow, they lead to the generation of action potentials that are propagated alongside the postganglionic axons to invade the varicosities located within target organs. There, noradrenaline is released when action potentials arrive. In primary cultures of dis- sociated sympathetic ganglia all these mechanisms occur within one dish (Kubista and Boehm 2006). Following M1 receptor activation, phospholipase Cβ hydrolyzes phosphatitylinositol 4,5 bispho- sphate (PIP2) and thus depletes it (Suh and Hille 2002). Accordingly, reduction of PIP2 leads to M1 receptor mediated inhibition of so-called M-currents (Brown and Adams 1980), carried by heterotetramers of Kv7.2 and Kv7.3 (Wang et al. 1998). This process is thought to underlie the slow EPSP in sympathetic neurons (Constanti and Brown 1981; Adams and Brown 1982; Brown and Selyanko 1985). Likewise, Application of a muscarinic agonist, like oxotremorine M (OxoM) triggers noradrenaline release from rat superior cervical ganglion (SCG) neurons via M1 recep- tors. Activated M1 receptors cause a slow EPSP via inhibition of Kv7 channels located on the neuronal somata. The depolarization results in formation of action potentials, which will be car- ried to nerve terminals via voltage-gated sodium channels resulting in noradrenaline release from sympathetic neurons. Hence, this effect is sensitive towards tetrodotoxin (TTX). Noradrenaline release is triggered by activation of M1 receptors, but inhibited by activation of Gi-coupled M2 and M4 receptors located at the nerve terminals. Therefore, preincubation of neurons in pertussis toxin (PTX) will actually enhance OxoM-mediated transmitter release (Lechner et al. 2003). Kv7 channels are active at subthreshold voltage levels that means they activate at voltage levels of -50 to -60 mV (Owen et al. 1990; Lechner et al. 2003; Maljevic et al. 2008; Brown and Passmore 2009; Linley et al. 2012; Telezhkin et al. 2012). With the experimental settings of the present study, the Goldman-Hodgkin-Katz equation predicts a resting membrane potential of -64 mV for release experiments and -67 mV for patch clamp experiments, respectively. Indeed, the measured membrane potential in current clamp recordings amounted to -65 mV and thus peaks in the

65 3 Discussion range of the predicted value. The logic question arises: How should inhibition of an already inactive channel be responsible for depolarization of a neuron and subsequent transmitter release? On top of this, only a fraction of sympathetic neurons developed a depolarization upon stimula- tion with OxoM. One could argue that only the fraction of cells having membrane potentials more positive than -60 mV would respond to OxoM with depolarization, but the measured membrane voltage did not correlate with the ability to form a depolarization. Furthermore, all investigated cells, no matter whether they depolarized or not, steadily expressed a measurable M-current carried by Kv7 channels. Additionally, this current was inhibited by application of OxoM both in depolarizing and non-depolarizing cells. If Kv7 channels are not involved in generation of the slow EPSP, is another family of potassium channels responsible? Replacing a portion of K+ in the internal solution with Cs+ reduces the driving force for K+. Hence, if K+ channels were involved, OxoM-induced depolarizations should be strongly reduced. However, reduction of K+ driving force did not change OxoM-mediated depolarizations. What other ion conductance can trigger depolarizations in rat sympathetic neurons? Early on, when Kv7 channel inhibition was introduced as mediator of the slow EPSP, an in- fluence of a chloride conductance could not be excluded (Brown and Selyanko 1985). But why should activation of a Cl− conductance lead to depolarization, when in adult central neurons activation of a chloride conductance, for example by activating the ligand-gated Cl−-channel GABAA, leads to hyperpolarization of the neuron (Kahle et al. 2008)? Well, γ-aminobutyric acid (GABA) was found to depolarize cat sympathetic neurons in vivo (Groat 1970) and rat sympathetic neurons in vitro (Bowery and Brown 1972). This phenomenon is caused by a rather high intracellular concentration of Cl−, which amounts up to 70 mM in rabbit SCG neurons, (Woodward et al. 1969) and up to 30 mM in rat SCG neurons (Ballanyi and Grafe 1985). The resulting Cl− equilibrium potential lies at -42 mV in rat SCG neurons, leading to an outward directed driving force (Adams and Brown 1973). First, it needed to be clarified, whether OxoM would also trigger depolarizations at native inter- nal Cl− concentrations. In the perforated patch-clamp technique, presence of amphothericin B in the patch pipette creates pores in the plasma membrane that are permeable for Na+,K+ and Cl−, respectively. Hence, after allowing for equilibration, the respective intracellular ion concen- trations will be equal to the pipette solution. However, using gramicidin D creates pores only permeable for Na+ and K+, thus leaving intracellular Cl− levels unaltered (Kyrozis and Reichling 1995). Nevertheless, application of both OxoM and GABA triggered comparable depolarizations, no matter if gramicidin D or amphothericin B was used. In order to further test, whether chloride channels are involved in OxoM-mediated depolar- izations, one can use a similar approach as before. Substitution of a major part of Cl− with equimolar amounts of gluconate in the pipette solution will reduce the Cl− driving force. Hence an OxoM-evoked depolarization should be reduced compared to measurements with the regular internal solution or the previously introduced Cs+-based solution. Indeed, substitution for Cl− diminished OxoM-evoked depolarizations indicating a role for a Cl− but not a K+ conductance. In radiotracer release experiments, the cytosol is left unaltered. However, one can use the op-

66 3 Discussion posite approach and replace external Cl− for gluconate and thus increase the Cl− driving force. Indeed, this strongly increased noradrenaline release evoked both by OxoM and GABA. Peripheral GABA receptors were found to have similar properties, like chemical specificity or ionic permeability, as central GABA receptors. However, the difference that remained was the direction of the potential difference created by application of GABA. It was suggested to rely on different intracellular Cl− concentrations, requiring opposing directions of an active Cl− transport mechanism (Adams and Brown 1975). As opposed to central neurons, peripheral neurons express the sodium - potassium - 2-chloride co-transporter (NKCC1), responsible for pumping chloride into the cell. NKCC1 can be inhibited by sulfamoylbenzoic acid derivatives, like furosemide and bumetanide (Isenring and Forbush 2001). Inhibition of NKCC1 reduces the intracellular Cl− concentration and thus the electrochemical driving force. The present result show, that OxoM- induced transmitter release was reduced in the presence of either NKCC1 antagonist. Taken together, these results point towards a chloride conductance as mediator of the slow EPSP and subsequent transmitter release. What type of chloride channel might be involved in the formation of the slow EPSP? A broad variety of Cl− channels is expressed in mammalian tissues, including the ClC family, CFTR, Ca2+ activated Cl− channels (CaCC), Maxi Cl− channels, and volume regulated anion channels (VRAC) (Duran et al. 2010). A Ca2+-activated Cl− current was described in mouse SCG neurons. However, this current was only found in 50% of the studied neurons (De Castro et al. 1997). In a similar study, 75% of rat SCG neurons displayed a similar current but only after axotomy. In non-axotomized neurons, the Ca2+-activated Cl− current was absent (Sánchez- Vives and Gallego 1994). Neurobiotin labeling of mouse SCG neurons revealed, that neurons with shorter dendrites show a more pronounced Ca2+-activated Cl− current than those with longer dendrites. Hence it was concluded, that channels responsible for Ca2+-activated Cl− currents are most likely expressed in distal dendrites. As mouse SCG neurons have shorter dendrites, Ca2+-activated Cl− currents can be detected more easily than in rat SCG neurons with longer dendrites. Only dendritic retraction produced by axotomy also reveals Ca2+-activated Cl− cur- rents in rat SCG neurons (De Castro et al. 1997). OxoM-induced depolarizations in this study were only found in a portion of analyzed rat SCG neurons, hence it appears likely that CaCCs play a role in OxoM-mediated depolarizations. In order to test this hypothesis, one can use a combination of otherwise rather unspecific drugs. Niflumic acid is frequently used as inhibitor of CaCCs, but it also blocks members of the ClC-family of Cl− channels. On the other hand, 2-[2-(4-acetamido-2-sulfophenyl)ethenyl]-4- isothiocyanatobenzene-1-sulfonic acid (SITS) can also be used to block CaCCs but Maxi Cl− channels as well (Alexander et al. 2011). Indeed, both inhibitors were able to reduce OxoM- induced depolarizations. Likewise, OxoM-triggered noradrenaline release was reduced in pres- ence of either of the two inhibitors. On the other hand, neither niflumic acid nor SITS affected M-currents or their inhibition by OxoM. CaCCs were reported to produce inward currents in rodent SCG neurons (De Castro et al. 1997; Martínez-Pinna et al. 2000). Hence, it was tested, whether OxoM was able to trigger a similar in- ward current. These voltage-clamp measurements were undertaken at a holding voltage of -65 mV. This value was chosen for two reasons: first, this was the mean value of measured membrane po-

67 3 Discussion

tentials in current-clamp experiments; and second, at this holding voltage, a contribution of Kv7 channels is unlikely. Recently, a member of the TMEM16 (= anoctamin) family, TMEM16A (= anoctamin-1, ANO-1) was identified as the molecular correlate of CaCCs (Schroeder et al. 2008; Caputo et al. 2008; Yang et al. 2008). Additionally, a second member of the family, TMEM16B (ANO-2), was shown to mediate a Ca2+ activated Cl− current (Galietta 2009). The discovery of the molecular correlate of CaCCs led to the synthesis of specific small-molecule inhibitors. The substance CaCCinh-A01, is an unspecific blocker of all CaCCs, whereas the inhibitor T16Ainh- A01 is more selective for TMEM16A at low concentrations (Namkung et al. 2011). Hence, these two inhibitors were tested in voltage-clamp experiments. Indeed, both inhibitors markedly di- minished OxoM-induced inward currents and additionally reduced OxoM-triggered noradrenaline release. In summary, these results suggest that M1 receptor mediated slow EPSPs in sympathetic neurons 2+ − involve Ca activated Cl channels of the anoctamin family in addition to Kv7 channels.

A while ago, a Cl− current was identified in rat sympathetic neurons. This current could be triggered by acetylcholine and was dependent on rise of intracellular Ca2+ as well as formation of diacylglycerol. Furthermore, inhibition of PKC abolished these currents. It was suggested, that rising concentrations of intracellular Ca2+ were caused by influx via nicotinic acetylcholine receptors, whereas the DAG necessary for PKC activation resulted from activation of muscarinic acetylcholine receptors (Marsh et al. 1995). The finding, that Ca2+ via nicotinic receptors is necessary, correlates with the reports, that M1 receptors cannot trigger a sufficient rise of intra- 2+ cellular Ca to inhibit Kv7 channels via interaction with calmodulin. However, it is reported that this is the prime mode of bradykinin B2 receptor mediated inhibition of M-currents (Delmas and Brown 2005). This difference occurs, because in contrast to B2 receptors, M1 receptors are not located adjacent to intracellular Ca2+ stores and can thus not provide a sufficient Ca2+ rise (Delmas and Brown 2002). However, a recent report suggests that activation of heterologously 2+ expressed M1 receptors is able to trigger substantial release of Ca from intracellular stores in tSA201 cells (Dickson et al. 2013; Falkenburger et al. 2013). Additionally, it was shown that CaCCs in dorsal root ganglion (DRG) neurons are preferentially activated by a local rise of intra- cellular Ca2+ provided by release from intracellular stores, rather than global changes mediated by influx via voltage-gated Ca2+ channels or nicotinic acetylcholine receptors (Jin et al. 2013). 2+ Furthermore, in interstitial cells of Cajal, M1 receptor-mediated Ca release from intracellular stores was sufficient to activate CaCCs (Zhu et al. 2011). Hence, it appears straightforward to assume that M1 receptors can also trigger CaCC activation in rat sympathetic neurons.

However, as Marsh and colleagues reported, the Cl− current, requiring the presence of Ca2+ was also sensitive towards an inhibition of PKC (Marsh et al. 1995). Thus, the next step was to determine, whether PKC activation following stimulation of Gq-coupled M1 receptors also influenced CaCC-mediated slow EPSPs, inward currents and noradrenaline release in rat SCG neurons. Activation of a Gq-coupled receptor, like M1 receptors, leads to activation of phospholipase Cβ.

68 3 Discussion

Subsequently, membrane-bound PIP2 is hydrolyzed into IP3 and DAG. The emerging DAG then activates PKC (Nishizuka 1995). The PKC family consists of 10 members that can be sub clas- sified into three groups according to their regulatory domains. Isoenzymes α, βI, βII and γ form the group of classical PKCs (cPKC), which are characterized by their sensitivity towards Ca2+ and DAG. Novel PKCs (nPKC) do not require the presence of Ca2+ but display a 100-fold higher affinity towards DAG. The group of nPKCs is formed by isoenzymes δ, ε, η, and θ.The third group, atypical PKCs (aPKC) is formed by ι/λ and ζ. These isoenzyems are not Ca2+ sensitive and cannot be activated by DAG but are responsive to anionic phospholipids and are activated by protein-protein interaction (Wu-Zhang and Newton 2013). Cultures of rat SCG neu- rons express isoenzymes of all groups, namely cPKCs α, βI, and βII, nPKC ε,aswellasaPKCζ (Scholze et al. 2002). These PKC groups can be pharmacologically distinguished. Phorbolesters, like phorbol-12-myristate-13-acetate (PMA), can be used to activate DAG-sensitive PKCs, but sustained treatment will downregulate and thus inhibit the respective PKCs (Fu et al. 1988). The indolocarbazole staurosporine is frequently used as an inhibitor of all PKC groups, but is also fairly unspecific and inhibits other kinases as well. More specific towards PKC inhibition is the bisindolylmaleimide GF-109203X (also known as bisindolylmaleimide I or Gö 6850). It in- hibits all groups of PKC but shows a higher potency at cPKCs, especially isoform α. The related compound Gö 6983 displays a similar affinity towards all groups of PKCs. The staurosporine- related Gö 6976 has a high affinity for all cPKC isoforms, but only poorly inhibits other isoforms (Wu-Zhang and Newton 2013). The present results revealed a sensitivity of OxoM-induced EPSPs as well as OxoM-mediated inward currents towards application of staurosporine and GF-109203X. Likewise, OxoM-induced noradrenaline release was also reduced in presence of staurosporine and GF-109203X. Corre- spondingly, OxoM-induced slow EPSPs, inward currents and noradrenaline release involve PKCs. Additionally, transmitter release was strongly reduced when cultures were pretreated with PMA, indicating a dependence on DAG-sensitive PKCs (cPKCs or nPKCs). The last step was to test whether transmitter release was also sensitive towards treatment with Gö 6976. Indeed, nora- drenaline release was strongly reduced already at low concentrations of Gö 6976, indicating that the PKC involved is a classical, Ca2+ dependent PKC. Together with the slight preference of-GF 109203X for cPKC α and the higher concentration of this compound needed for a reduction of noradrenaline release, one might speculate that the involved isoenzme is PKCβ. It has been suggested, that PKC might phosphorylate Kv7 channels and thereby influence PIP2 affinity (Hoshi et al. 2003; Nakajo and Kubo 2005). However, other groups have excluded an in- volvement of PKCs in M1 receptor mediated inhibition of Kv7 channels (Bosma and Hille 1989; Suh and Hille 2002; Zhang et al. 2003; Dickson et al. 2013; Falkenburger et al. 2013). In accor- dance with the latter reports, the presence of staurosporine or GF-109203X did neither influence M-currents, analyzed in this study, nor their inhibition via OxoM. In support of these latter results, ANO-1 was found to carry putative phosphorylation sites for a variety of kinases, including PKC (Huang et al. 2012). Thus it appears likely, that PKCs can directly phosphorylate TMEM16A.

In conclusion, this report demonstrates that M1 receptor mediated slow EPSPs and subsequent

69 3 Discussion

noradrenaline release does not involve inhibition of Kv7 channels. However, it favors M1 receptor- induced activation of Ca2+ activated Cl− channels and additionally requires activation of PKCβ.

70 Bibliography

Absalom, Nathan L, Trevor M Lewis, and Peter R Schofield (2004). “Mechanisms of channel gating of the ligand-gated ion channel superfamily inferred from protein structure.” In: Exp Physiol 89.2, pp. 145–53. doi: 10.1113/expphysiol.2003.026815. Adams, P R and D A Brown (1973). “Action of -aminobutyric acid (GABA) on rat sympathetic ganglion cells.” In: Br J Pharmacol 47.3, 639P–640P. – (1975). “Actions of gamma-aminobutyric acid on sympathetic ganglion cells.” In: J Physiol 250.1, pp. 85–120. – (1982). “Synaptic inhibition of the M-current: slow excitatory post-synaptic potential mecha- nism in bullfrog sympathetic neurones.” In: J Physiol 332, pp. 263–72. Alexander, Stephen P H, Alistair Mathie, and John A Peters (2011). “Guide to Receptors and Channels (GRAC), 5th edition.” In: Br J Pharmacol 164 Suppl 1, S1–324. doi: 10.1111/j. 1476-5381.2011.01649_1.x. Ballanyi, K and P Grafe (1985). “An intracellular analysis of gamma-aminobutyric-acid-associated ion movements in rat sympathetic neurones.” In: J Physiol 365, pp. 41–58. Berkefeld, Henrike, Bernd Fakler, and Uwe Schulte (2010). “Ca2+-activated K+ channels: from protein complexes to function.” In: Physiol Rev 90.4, pp. 1437–59. doi: 10.1152/physrev. 00049.2009. Bernstein, Julius (1902). “Untersuchungen zur Thermodynamik der bioelektrischen Ströme.” In: Pflugers Arch 92.10-12, pp. 521–562. Bezanilla, Francisco (2008a). “How membrane proteins sense voltage.” In: Nat Rev Mol Cell Biol 9.4, pp. 323–32. doi: 10.1038/nrm2376. – (2008b). “Ion channels: from conductance to structure.” In: Neuron 60.3, pp. 456–68. doi: 10.1016/j.neuron.2008.10.035. Blackman, J G, B L Ginsborg, and C Ray (1963). “Synaptic transmission in the sympathetic ganglion of the frog.” In: J Physiol 167, pp. 355–73. Bosma, M M and B Hille (1989). “Protein kinase C is not necessary for peptide-induced suppres- sion of M current or for desensitization of the peptide receptors.” In: Proc Natl Acad Sci U S A 86.8, pp. 2943–7. Bowery, N G and D A Brown (1972). “Depolarization of isolated rat ganglia by -aminobutyric acid and related compounds.” In: Br J Pharmacol 45.1, 160P–161P. Brown, D A and P R Adams (1980). “Muscarinic suppression of a novel voltage-sensitive K+ current in a vertebrate neurone.” In: Nature 283.5748, pp. 673–6. Brown, D A and C N Scholfield (1974). “Changes of intracellular sodium and potassium ion concentrations in isolated rat superior cervical ganglia induced by depolarizing agents.” In: J Physiol 242.2, pp. 307–19.

71 Bibliography

Brown, D A and A A Selyanko (1985). “Membrane currents underlying the cholinergic slow excitatory post-synaptic potential in the rat sympathetic ganglion.” In: J Physiol 365, pp. 365– 87. Brown, David A (2006). “Acetylcholine.” In: Br J Pharmacol 147 Suppl 1, S120–6. doi: 10. 1038/sj.bjp.0706474. – (2010). “Muscarinic acetylcholine receptors (mAChRs) in the nervous system: some functions and mechanisms.” In: J Mol Neurosci 41.3, pp. 340–6. doi: 10.1007/s12031-010-9377-2. Brown, David A and Gayle M Passmore (2009). “Neural KCNQ (Kv7) channels.” In: Br J Phar- macol 156.8, pp. 1185–95. doi: 10.1111/j.1476-5381.2009.00111.x. Brown, David A et al. (2007). “Regulation of M(Kv7.2/7.3) channels in neurons by PIP(2) and products of PIP(2) hydrolysis: significance for receptor-mediated inhibition.” In: J Physiol 582.Pt 3, pp. 917–25. doi: 10.1113/jphysiol.2007.132498. Brown, G L, H H Dale, and W Feldberg (1936). “Reactions of the normal mammalian muscle to acetylcholine and to eserine.” In: J Physiol 87.4, pp. 394–424. Cabrera-Vera, Theresa M et al. (2003). “Insights into G protein structure, function, and regula- tion.” In: Endocr Rev 24.6, pp. 765–81. Caputo, Antonella et al. (2008). “TMEM16A, a membrane protein associated with calcium- dependent chloride channel activity.” In: Science 322.5901, pp. 590–4. doi: 10.1126/science. 1163518. Catterall, William A et al. (2003a). “International Union of Pharmacology. XL. Compendium of voltage-gated ion channels: calcium channels.” In: Pharmacol Rev 55.4, pp. 579–81. doi: 10.1124/pr.55.4.8. Catterall, William A et al. (2003b). “International Union of Pharmacology. XXXIX. Compendium of voltage-gated ion channels: sodium channels.” In: Pharmacol Rev 55.4, pp. 575–8. doi: 10.1124/pr.55.4.7. Caulfield, M P and N J Birdsall (1998). “International Union of Pharmacology. XVII. Classifi- cation of muscarinic acetylcholine receptors.” In: Pharmacol Rev 50.2, pp. 279–90. Chanda, Baron and Francisco Bezanilla (2008). “A common pathway for charge transport through voltage-sensing domains.” In: Neuron 57.3, pp. 345–51. doi: 10.1016/j.neuron.2008.01.015. Clausen, T (1998). “Clinical and therapeutic significance of the Na+,K+ pump*.” In: Clin Sci (Lond) 95.1, pp. 3–17. Collingridge, Graham L et al. (2009). “A nomenclature for ligand-gated ion channels.” In: Neu- ropharmacology 56.1, pp. 2–5. doi: 10.1016/j.neuropharm.2008.06.063. Constanti, A and D A Brown (1981). “M-Currents in voltage-clamped mammalian sympathetic neurones.” In: Neurosci Lett 24.3, pp. 289–94. Dale, H H (1914). “The action of certain esters and ethers of choline, and their relation to muscarine.” In: J Pharmacol Exp Ther 6, pp. 147–190. Dale, H H, W Feldberg, and M Vogt (1936). “Release of acetylcholine at voluntary motor nerve endings.” In: J Physiol 86.4, pp. 353–80. De Castro, F, E Geijo-Barrientos, and R Gallego (1997). “Calcium-activated chloride current in normal mouse sympathetic ganglion cells.” In: J Physiol 498 ( Pt 2), pp. 397–408.

72 Bibliography

De La Fuente, Ricardo et al. (2008). “Small-molecule screen identifies inhibitors of a human intestinal calcium-activated chloride channel.” In: Mol Pharmacol 73.3, pp. 758–68. doi: 10. 1124/mol.107.043208. Delmas, Patrick and David A Brown (2002). “Junctional signaling microdomains: bridging the gap between the neuronal cell surface and Ca2+ stores.” In: Neuron 36.5, pp. 787–90. – (2005). “Pathways modulating neural KCNQ/M (Kv7) potassium channels.” In: Nat Rev Neu- rosci 6.11, pp. 850–62. doi: 10.1038/nrn1785. Dickson, Eamonn J, Björn H Falkenburger, and Bertil Hille (2013). “Quantitative properties and receptor reserve of the IP(3) and calcium branch of G(q)-coupled receptor signaling.” In: J Gen Physiol 141.5, pp. 521–35. doi: 10.1085/jgp.201210886. Duran, Charity et al. (2010). “Chloride channels: often enigmatic, rarely predictable.” In: Annu Rev Physiol 72, pp. 95–121. doi: 10.1146/annurev-physiol-021909-135811. Falkenburger, Björn H, Eamonn J Dickson, and Bertil Hille (2013). “Quantitative properties and receptor reserve of the DAG and PKC branch of G(q)-coupled receptor signaling.” In: J Gen Physiol 141.5, pp. 537–55. doi: 10.1085/jgp.201210887. Ferrera, Loretta et al. (2009). “Regulation of TMEM16A chloride channel properties by alterna- tive splicing.” In: JBiolChem284.48, pp. 33360–8. doi: 10.1074/jbc.M109.046607. Ferrera, Loretta, Antonella Caputo, and Luis J V Galietta (2010). “TMEM16A protein: a new identity for Ca(2+)-dependent Cl channels.” In: Physiology (Bethesda) 25.6, pp. 357–63. doi: 10.1152/physiol.00030.2010. Fu, T, Y Okano, and Y Nozawa (1988). “Bradykinin-induced generation of inositol 1,4,5-trisphosphate in fibroblasts and neuroblastoma cells: effect of pertussis toxin, extracellular calcium, and down-regulation of protein kinase C.” In: Biochem Biophys Res Commun 157.3, pp. 1429–35. Fucile, Sergio (2004). “Ca2+ permeability of nicotinic acetylcholine receptors.” In: Cell Calcium 35.1, pp. 1–8. Galietta, Luis J V (2009). “The TMEM16 protein family: a new class of chloride channels?” In: Biophys J 97.12, pp. 3047–53. doi: 10.1016/j.bpj.2009.09.024. Gamper, Nikita and Mark S Shapiro (2003). “Calmodulin mediates Ca2+-dependent modulation of M-type K+ channels.” In: J Gen Physiol 122.1, pp. 17–31. doi: 10.1085/jgp.200208783. Goldman, D E (1943). “POTENTIAL, IMPEDANCE, AND RECTIFICATION IN MEMBRANES.” In: J Gen Physiol 27.1, pp. 37–60. Gotti, C and F Clementi (2004). “Neuronal nicotinic receptors: from structure to pathology.” In: Prog Neurobiol 74.6, pp. 363–96. doi: 10.1016/j.pneurobio.2004.09.006. Gotti, Cecilia, Michele Zoli, and Francesco Clementi (2006). “Brain nicotinic acetylcholine recep- tors: native subtypes and their relevance.” In: Trends Pharmacol Sci 27.9, pp. 482–91. doi: 10.1016/j.tips.2006.07.004. Groat, W C de (1970). “The actions of gamma-aminobutyric acid and related amino acids on mammalian autonomic ganglia.” In: J Pharmacol Exp Ther 172.2, pp. 384–96. Gutman, George A et al. (2003). “International Union of Pharmacology. XLI. Compendium of voltage-gated ion channels: potassium channels.” In: Pharmacol Rev 55.4, pp. 583–6. doi: 10.1124/pr.55.4.9.

73 Bibliography

Gutman, George A et al. (2005). “International Union of Pharmacology. LIII. Nomenclature and molecular relationships of voltage-gated potassium channels.” In: Pharmacol Rev 57.4, pp. 473–508. doi: 10.1124/pr.57.4.10. Haga, Tatsuya (2013). “Molecular properties of muscarinic acetylcholine receptors.” In: Proc Jpn Acad Ser B Phys Biol Sci 89.6, pp. 226–56. Hamid, J et al. (1999). “Identification of an integration center for cross-talk between protein kinase C and G protein modulation of N-type calcium channels.” In: JBiolChem274.10, pp. 6195–202. Hartzell, Criss, Ilva Putzier, and Jorge Arreola (2005). “Calcium-activated chloride channels.” In: Annu Rev Physiol 67, pp. 719–58. doi: 10.1146/annurev.physiol.67.032003.154341. Hartzell, H Criss et al. (2009). “Anoctamin/TMEM16 family members are Ca2+-activated Cl- channels.” In: J Physiol 587.Pt 10, pp. 2127–39. doi: 10.1113/jphysiol.2008.163709. Hernandez, Ciria C, Oleg Zaika, and Mark S Shapiro (2008). “A carboxy-terminal inter-helix linker as the site of phosphatidylinositol 4,5-bisphosphate action on Kv7 (M-type) K+ chan- nels.” In: J Gen Physiol 132.3, pp. 361–81. doi: 10.1085/jgp.200810007. Higashida, H and D A Brown (1986). “Two polyphosphatidylinositide metabolites control two K+ currents in a neuronal cell.” In: Nature 323.6086, pp. 333–5. doi: 10.1038/323333a0. Hille, Bertil (2001). Ion channels of excitable membranes. 3rd ed. Sunderland, Mass.: Sinauer. isbn: 0878933212. Hodgkin, A L and A F Huxley (1952a). “A quantitative description of membrane current and its application to conduction and excitation in nerve.” In: J Physiol 117.4, pp. 500–44. – (1952b). “Currents carried by sodium and potassium ions through the membrane of the giant axon of Loligo.” In: J Physiol 116.4, pp. 449–72. – (1952c). “The dual effect of membrane potential on sodium conductance in the giant axon of Loligo.” In: J Physiol 116.4, pp. 497–506. Hodgkin, A L and B Katz (1949). “The effect of sodium ions on the electrical activity of giant axon of the squid.” In: J Physiol 108.1, pp. 37–77. Hoshi, Naoto et al. (2003). “AKAP150 signaling complex promotes suppression of the M-current by muscarinic agonists.” In: Nat Neurosci 6.6, pp. 564–71. doi: 10.1038/nn1062. Huang, Fen, Xiuming Wong, and Lily Y Jan (2012). “International Union of Basic and Clinical Pharmacology. LXXXV: calcium-activated chloride channels.” In: Pharmacol Rev 64.1, pp. 1– 15. doi: 10.1124/pr.111.005009. Huang, Fen et al. (2013). “TMEM16C facilitates Na(+)-activated K+ currents in rat sensory neurons and regulates pain processing.” In: Nat Neurosci 16.9, pp. 1284–90. doi: 10.1038/ nn.3468. Hulme, E C et al. (2003). “Structure and activation of muscarinic acetylcholine receptors.” In: Biochem Soc Trans 31.Pt 1, pp. 29–34. doi: 10.1042/. Hulme, Edward C (2013). “GPCR activation: a mutagenic spotlight on crystal structures.” In: Trends Pharmacol Sci 34.1, pp. 67–84. doi: 10.1016/j.tips.2012.11.002. Isenring, P and B Forbush (2001). “Ion transport and ligand binding by the Na-K-Cl cotrans- porter, structure-function studies.” In: Comp Biochem Physiol A Mol Integr Physiol 130.3, pp. 487–97.

74 Bibliography

Jin, Xin et al. (2013). “Activation of the Cl- channel ANO1 by localized calcium signals in nociceptive sensory neurons requires coupling with the IP3 receptor.” In: Sci Signal 6.290, ra73. doi: 10.1126/scisignal.2004184. Kaczmarek, Leonard K and Irwin B Levitan (1987). Neuromodulation: the biochemical control of neuronal excitability. New York: Oxford University Press. isbn: 019504097X (alk. paper). url: http://www.loc.gov/catdir/enhancements/fy0603/86005124-d.html. Kahle, Kristopher T et al. (2008). “Roles of the cation-chloride cotransporters in neurological disease.” In: Nat Clin Pract Neurol 4.9, pp. 490–503. doi: 10.1038/ncpneuro0883. Katz, B (1962). “The transmission of impulses from nerve to muscle and the subcellular unit of synaptic action.” In: Proc Roy Soc B 155, pp. 455–482. Keynes, R D (1951). “The ionic movements during nervous activity.” In: J Physiol 114.1-2, pp. 119–50. Kobayashi, H and B Libet (1968). “Generation of slow postsynaptic potentials without increases in ionic conductance.” In: Proc Natl Acad Sci U S A 60.4, pp. 1304–11. – (1970). “Actions of noradrenaline and acetylcholine on sympathetic ganglion cells.” In: J Phys- iol 208.2, pp. 353–72. Koppen, Chris J van and Björn Kaiser (2003). “Regulation of muscarinic acetylcholine receptor signaling.” In: Pharmacol Ther 98.2, pp. 197–220. Kubista, Helmut and Stefan Boehm (2006). “Molecular mechanisms underlying the modulation of exocytotic noradrenaline release via presynaptic receptors.” In: Pharmacol Ther 112.1, pp. 213– 42. doi: 10.1016/j.pharmthera.2006.04.005. Kunzelmann, Karl et al. (2011). “Anoctamins.” In: Pflugers Arch 462.2, pp. 195–208. doi: 10. 1007/s00424-011-0975-9. Kunzelmann, Karl et al. (2013). “Molecular functions of anoctamin 6 (TMEM16F): a chlo- ride channel, cation channel, or phospholipid scramblase?” In: Pflugers Arch. doi: 10.1007/ s00424-013-1305-1. Kyrozis, A and D B Reichling (1995). “Perforated-patch recording with gramicidin avoids arti- factual changes in intracellular chloride concentration.” In: J Neurosci Methods 57.1, pp. 27– 35. Langley, J N (1921). “The Autonomic Nervous System.” In: Cambridge. Langmead, Christopher J, Jeannette Watson, and Charlie Reavill (2008). “Muscarinic acetyl- choline receptors as CNS drug targets.” In: Pharmacol Ther 117.2, pp. 232–43. doi: 10.1016/ j.pharmthera.2007.09.009. Lechner, Stefan G, Martina Mayer, and Stefan Boehm (2003). “Activation of M1 muscarinic receptors triggers transmitter release from rat sympathetic neurons through an inhibition of M- type K+ channels.” In: J Physiol 553.Pt 3, pp. 789–802. doi: 10.1113/jphysiol.2003.052449. Li, Yang et al. (2005). “Regulation of Kv7 (KCNQ) K+ channel open probability by phos- phatidylinositol 4,5-bisphosphate.” In: J Neurosci 25.43, pp. 9825–35. doi: 10.1523/JNEUROSCI. 2597-05.2005. Linley, John E et al. (2012). “M channel enhancers and physiological M channel block.” In: J Physiol 590.Pt 4, pp. 793–807. doi: 10.1113/jphysiol.2011.223404.

75 Bibliography

Loewi, Otto (1921). “Ueber die humorale Uebertragbarkeit der Herznervenwirkung.” In: Pfluegers Arch. ges. Physiol. 189, pp. 239–242. Loewi, Otto and Navratil Ernst (1926). “Über humorale Übertragbarkeit der Herznervenwirkung.” In: Pfluegers Arch. ges. Physiol. 214, pp. 689–696. Lummis, Sarah C R (2012). “5-HT(3) receptors.” In: JBiolChem287.48, pp. 40239–45. doi: 10.1074/jbc.R112.406496. Maljevic, Snezana, Thomas V Wuttke, and Holger Lerche (2008). “Nervous system KV7 disor- ders: breakdown of a subthreshold brake.” In: J Physiol 586.7, pp. 1791–801. doi: 10.1113/ jphysiol.2008.150656. Marrion, N V et al. (1989). “Muscarinic suppression of the M-current in the rat sympathetic ganglion is mediated by receptors of the M1-subtype.” In: Br J Pharmacol 98.2, pp. 557–73. Marsh, S J et al. (1995). “Synergistic regulation of a neuronal chloride current by intracellular calcium and muscarinic receptor activation: a role for protein kinase C.” In: Neuron 15.3, pp. 729–37. Martínez-Pinna, J, E M McLachlan, and R Gallego (2000). “Distinct mechanisms for activation of Cl- and K+ currents by Ca2+ from different sources in mouse sympathetic neurones.” In: J Physiol 527 Pt 2, pp. 249–64. Mellor, H and P J Parker (1998). “The extended protein kinase C superfamily.” In: Biochem J 332 ( Pt 2), pp. 281–92. Millar, Neil S and Cecilia Gotti (2009). “Diversity of vertebrate nicotinic acetylcholine receptors.” In: Neuropharmacology 56.1, pp. 237–46. doi: 10.1016/j.neuropharm.2008.07.041. Mitchelson, Frederick J (2012). “The pharmacology of McN-A-343.” In: Pharmacol Ther 135.2, pp. 216–45. doi: 10.1016/j.pharmthera.2012.05.008. Nakajo, Koichi and Yoshihiro Kubo (2005). “Protein kinase C shifts the voltage dependence of KCNQ/M channels expressed in Xenopus oocytes.” In: J Physiol 569.Pt 1, pp. 59–74. doi: 10.1113/jphysiol.2005.094995. Namkung, Wan, Puay-Wah Phuan, and A S Verkman (2011). “TMEM16A inhibitors reveal TMEM16A as a minor component of calcium-activated chloride channel conductance in airway and intestinal epithelial cells.” In: J Biol Chem 286.3, pp. 2365–74. doi: 10.1074/jbc.M110. 175109. Nernst, Walther (1888). “Zur Kinetik der in Lösung befindlichen Körper. Erste Abhandlung. Theorie der Diffusion.” In: Z. phys. Chem. 2.9, pp. 613–637. Nishizuka, Y (1995). “Protein kinase C and lipid signaling for sustained cellular responses.” In: FASEB J 9.7, pp. 484–96. Oldham, William M and Heidi E Hamm (2008). “Heterotrimeric G protein activation by G- protein-coupled receptors.” In: Nat Rev Mol Cell Biol 9.1, pp. 60–71. doi: 10.1038/nrm2299. Olsen, Richard W and Werner Sieghart (2008). “International Union of Pharmacology. LXX. Subtypes of gamma-aminobutyric acid(A) receptors: classification on the basis of subunit composition, pharmacology, and function. Update.” In: Pharmacol Rev 60.3, pp. 243–60. doi: 10.1124/pr.108.00505. Owen, D G, S J Marsh, and D A Brown (1990). “M-current noise and putative M-channels in cultured rat sympathetic ganglion cells.” In: J Physiol 431, pp. 269–90.

76 Bibliography

Perry, W L M and J Talesnik (1953). “The role of acetylcholine in synaptic transmission at parasympathetic ganglia.” In: J Physiol 119.4, pp. 455–69. Rizo, J and C Rosenmund (2008). “Synaptic vesicle fusion.” ENG. In: Nat Struct Mol Biol 15.n7, pp. 665–674. issn: 1545-9985 (Electronic). doi: 10.1038/nsmb.1450. Sánchez-Vives, M V and R Gallego (1994). “Calcium-dependent chloride current induced by axotomy in rat sympathetic neurons.” In: J Physiol 475.3, pp. 391–400. Schicker, K W, M M Dorostkar, and S Boehm (2008). “Modulation of transmitter release via presynaptic ligand-gated ion channels.” In: Curr Mol Pharmacol 1.2, pp. 106–29. Scholze, Thomas et al. (2002). “Sympathoexcitation by bradykinin involves Ca2+-independent protein kinase C.” In: J Neurosci 22.14, pp. 5823–32. doi: 20026590. Schroeder, Björn Christian et al. (2008). “Expression cloning of TMEM16A as a calcium-activated chloride channel subunit.” In: Cell 134.6, pp. 1019–29. doi: 10.1016/j.cell.2008.09.003. Schwake, M et al. (2000). “Surface expression and single channel properties of KCNQ2/KCNQ3, M-type K+ channels involved in epilepsy.” In: JBiolChem275.18, pp. 13343–8. Shimizu, Takahiro et al. (2013). “TMEM16F is a component of a Ca2+-activated Cl- channel but not a volume-sensitive outwardly rectifying Cl- channel.” In: Am J Physiol Cell Physiol 304.8, pp. C748–59. doi: 10.1152/ajpcell.00228.2012. Skou, J C (1957). “The influence of some cations on an adenosine triphosphatase from peripheral nerves.” In: Biochim Biophys Acta 23.2, pp. 394–401. Smith, J S et al. (2001). “Differential expression of kcnq2 splice variants: implications to m current function during neuronal development.” In: J Neurosci 21.4, pp. 1096–103. Stea, A, T W Soong, and T P Snutch (1995). “Determinants of PKC-dependent modulation of a family of neuronal calcium channels.” In: Neuron 15.4, pp. 929–40. Steinberg, Susan F (2008). “Structural basis of protein kinase C isoform function.” In: Physiol Rev 88.4, pp. 1341–78. doi: 10.1152/physrev.00034.2007. Strosberg, A D and C Nahmias (2007). “G-protein-coupled receptor signalling through protein networks.” In: Biochem Soc Trans 35.Pt 1, pp. 23–7. doi: 10.1042/BST0350023. Südhof, Thomas C (2012). “The presynaptic active zone.” In: Neuron 75.1, pp. 11–25. doi: 10.1016/j.neuron.2012.06.012. Südhof, Thomas C and Josep Rizo (2011). “Synaptic vesicle exocytosis.” In: Cold Spring Harb Perspect Biol 3.12. doi: 10.1101/cshperspect.a005637. Suh, Byung-Chang and Bertil Hille (2002). “Recovery from muscarinic modulation of M current channels requires phosphatidylinositol 4,5-bisphosphate synthesis.” In: Neuron 35.3, pp. 507– 20. – (2007). “Regulation of KCNQ channels by manipulation of phosphoinositides.” In: J Physiol 582.Pt 3, pp. 911–6. doi: 10.1113/jphysiol.2007.132647. Suzuki, Nobuchika, Nicole Hajicek, and Tohru Kozasa (2009). “Regulation and physiological functions of G12/13-mediated signaling pathways.” In: Neurosignals 17.1, pp. 55–70. doi: 10.1159/000186690. Szteyn, Kalina et al. (2012). “Expression and functional significance of the Ca(2+)-activated Cl(-) channel ANO6 in dendritic cells.” In: Cell Physiol Biochem 30.5, pp. 1319–32. doi: 10.1159/000343321.

77 Bibliography

Telezhkin, Vsevolod, David A Brown, and Alasdair J Gibb (2012). “Distinct subunit contributions to the activation of M-type potassium channels by PI(4,5)P2.” In: J Gen Physiol 140.1, pp. 41– 53. doi: 10.1085/jgp.201210796. Telezhkin, Vsevolod et al. (2013). “A basic residue in the proximal C-terminus is necessary for efficient activation of the M-channel subunit Kv7.2 by PI(4,5)P.” In: Pflugers Arch 465.7, pp. 945–53. doi: 10.1007/s00424-012-1199-3. Tian, Yuemin, Rainer Schreiber, and Karl Kunzelmann (2012). “Anoctamins are a family of Ca2+-activated Cl- channels.” In: J Cell Sci 125.Pt 21, pp. 4991–8. doi: 10.1242/jcs.109553. Trendelenburg, U (1966). “Transmission of preganglionic impulses through the muscarinic recep- tors of the superior cervical ganglion of the cat.” In: J Pharmacol Exp Ther 154.3, pp. 426– 40. Tsetlin, Victor, Dmitry Kuzmin, and Igor Kasheverov (2011). “Assembly of nicotinic and other Cys-loop receptors.” In: J Neurochem 116.5, pp. 734–41. doi: 10.1111/j.1471-4159.2010. 07060.x. Unwin, Nigel (2005). “Refined structure of the nicotinic acetylcholine receptor at 4A resolution.” In: J Mol Biol 346.4, pp. 967–89. doi: 10.1016/j.jmb.2004.12.031. Verkman, Alan S and Luis J V Galietta (2009). “Chloride channels as drug targets.” In: Nat Rev Drug Discov 8.2, pp. 153–71. doi: 10.1038/nrd2780. Vocke, Kerstin et al. (2013). “Calmodulin-dependent activation and inactivation of anoctamin calcium-gated chloride channels.” In: J Gen Physiol 142.4, pp. 381–404. doi: 10.1085/jgp. 201311015. Wang, H S et al. (1998). “KCNQ2 and KCNQ3 potassium channel subunits: molecular correlates of the M-channel.” In: Science 282.5395, pp. 1890–3. Wei, Aguan D et al. (2005). “International Union of Pharmacology. LII. Nomenclature and molec- ular relationships of calcium-activated potassium channels.” In: Pharmacol Rev 57.4, pp. 463– 72. doi: 10.1124/pr.57.4.9. Wojcik, Sonja M and Nils Brose (2007). “Regulation of membrane fusion in synaptic excitation- secretion coupling: speed and accuracy matter.” eng. In: Neuron 55.1, pp. 11–24. issn: 0896- 6273 (Print). doi: 10.1016/j.neuron.2007.06.013. Woodward, J K, C P Bianchi, and S D Erulkar (1969). “Electrolyte distribution in rabbit superior cervical ganglion.” In: J Neurochem 16.3, pp. 289–99. Wu-Zhang, Alyssa X and Alexandra C Newton (2013). “Protein kinase C pharmacology: refining the toolbox.” In: Biochem J 452.2, pp. 195–209. doi: 10.1042/BJ20130220. Yang, Young Duk et al. (2008). “TMEM16A confers receptor-activated calcium-dependent chlo- ride conductance.” In: Nature 455.7217, pp. 1210–5. doi: 10.1038/nature07313. Yu, Kuai et al. (2012). “Explaining calcium-dependent gating of anoctamin-1 chloride channels requires a revised topology.” In: Circ Res 110.7, pp. 990–9. doi: 10.1161/CIRCRESAHA.112. 264440. Yus-Najera, Eva, Irene Santana-Castro, and Alvaro Villarroel (2002). “The identification and characterization of a noncontinuous calmodulin-binding site in noninactivating voltage-dependent KCNQ potassium channels.” In: JBiolChem277.32, pp. 28545–53. doi: 10 . 1074 / jbc . M204130200.

78 Bibliography

Zamponi, Gerald W and Kevin P M Currie (2013). “Regulation of Ca(V)2 calcium channels by G protein coupled receptors.” In: Biochim Biophys Acta 1828.7, pp. 1629–43. doi: 10.1016/ j.bbamem.2012.10.004. Zeng, F and J Wess (2000). “Molecular aspects of muscarinic receptor dimerization.” In: Neu- ropsychopharmacology 23.4 Suppl, S19–31. doi: 10.1016/S0893-133X(00)00146-9. Zhang, Hailin et al. (2003). “PIP(2) activates KCNQ channels, and its hydrolysis underlies receptor-mediated inhibition of M currents.” In: Neuron 37.6, pp. 963–75. Zhu, Mei Hong et al. (2011). “Muscarinic activation of Ca2+-activated Cl- current in interstitial cells of Cajal.” In: J Physiol 589.Pt 18, pp. 4565–82. doi: 10.1113/jphysiol.2011.211094. Zimmerman, Gabriel and Hermona Soreq (2006). “Termination and beyond: acetylcholinesterase as a modulator of synaptic transmission.” In: Cell Tissue Res 326.2, pp. 655–69. doi: 10. 1007/s00441-006-0239-8.

79 Curriculum Vitae

Personal Data

Name Isabella Salzer Date of Birth June 3, 1985 - Vienna, Austria Nationality Austria Gender Female

Education

2010-2014 PhD studies at the Medical University of Vienna 2003-2009 Medical Course at the Medical University of Vienna 1995-2003 High School, Hagenmüllergasse 30, 1030 Vienna

Teaching

2010-2012 Lecturer at the University of Applied Sciences - Midwifery, Pharmacology 2009-2012 Lecturer at the Medical University of Vienna - Human Medicine, Practical Course Neurophysiology (Block 4), Seminar ECG & Seminar Pharmacology (Block 11)

Publications

Chandaka G, Salzer I, Drobny H, Boehm S, Schicker KW (2012), Facilitation of transmit- ter release from rat sympathetic neurons via presynaptic P2Y(1) receptors., Br J Pharmacol. 164(5):1522-33.

Klinger F, Geier P, Dorostkar MM, Chandaka GK, Yousuf A, Salzer I, Kubista H, Boehm S (2012), Concomitant facilitation of GABAA receptors and KV7 channels by the non-opioid analgesic flupirtine., Br J Pharmacol.166(5):1631-42.

Salzer I, Gafar H, Gindl V, Mahlknecht P, Drobny H, Boehm S (2014), Excitation of rat sympathetic neurons via M1 muscarinic receptors independently of Kv7 channels. Pflugers Arch resubmitted.

80