Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Lidocaine inhibition of cloned hyperpolarization-activated cucline nucleotide-modulated (HCN) channels Yip, Raymond Hon Leung 2009

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2009_fall_yip_raymond.pdf [ 3.17MB ]
Metadata
JSON: 24-1.0069938.json
JSON-LD: 24-1.0069938-ld.json
RDF/XML (Pretty): 24-1.0069938-rdf.xml
RDF/JSON: 24-1.0069938-rdf.json
Turtle: 24-1.0069938-turtle.txt
N-Triples: 24-1.0069938-rdf-ntriples.txt
Original Record: 24-1.0069938-source.json
Full Text
24-1.0069938-fulltext.txt
Citation
24-1.0069938.ris

Full Text

LIDOCAINE INHIBITION OF CLONED HYPERPOLARIZATION-ACTIVATED CYCLIC NUCLEOTIDE-MODULATED (HCN) CHANNELS  by RAYMOND HON LEUNG YIP B.M.L.Sc, The University of British Columbia, 2007  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Physiology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  August 2009 © Raymond Hon Leung Yip, 2009  Abstract Hyperpolarization-activated Cyclic Nucleotide-modulated (HCN) channels underlie the hyperpolarization-activated or "funny" current (If or Ih), which is important in regulating excitability in the neurons of the central and peripheral nervous Systems and in the conduction tissue of the heart. There are four mammalian isoforms of HCN channels (HCN 1-4), each exhibiting différent activation and deactivation kinetics. Lidocaine, a local anesthetic and antiarrhythmic drug, has been shown to inhibit HCN-mediated currents in the rabbit sinoatrial (SA) node, which expresses various HCN isoforms. It also inhibits Ih in rat small DRG neurons. Lidocaine has been shown to block Na+ channels by interacting with the aromatic side chain of a phenylalanine residue in the région lining the channel pore. With the exception of a récent study on rabbit HCN4, this is the first study examining lidocaine inhibition on cloned HCN channels. I hypothesize that lidocaine, in micromolar concentrations, inhibits HCN1 channels from the intracellular side by interacting with amino acid residues lining the channel pore. Lidocaine, at concentrations ranging from 50 to 3200 uJVl, inhibited mouse HCN1 channels expressed in Chinese hamster ovary (CHO) cells with an IC50 of 669 ± 320 uM. The onset of lidocaine action is fast (~6 seconds) and concentration-dependent with maximum inhibition occurring at approximately 20 seconds, and inhibition is partially or completely réversible. The fact that currents does not return completely may be attributed to a phenomenon called current rundown, or to incomplète washout of lidocaine. Extracellular application of the permanently charged derivative of lidocaine, QX-314, did not inhibit the channel, which suggests that lidocaine inhibition, in its charged form, occurs from the intracellular side. A point mutation at the phenylalanine residue does not remove lidocaine inhibition, indicating that this residue is not responsible for interaction with lidocaine. Thèse results support the hypothesis that lidocaine ii  inhibits HCNl from the intracellular side, with molecular mechanisms of inhibition yet to be determined.  Table of Contents  Abstract  ii  Table of Contents  iv  List of Figures  »  List of Symbols and Abbreviations  vi viii  Acknowledgements  x  Chapter 1 : Introduction  1  1.1  Hyperpolarization-activated cyclic nucleotide-modulated (HCN) channels  1  1.2  Pharmacology of If  4  1.2.1  ZD7288  5  1.2.2  Ivabradine  6  1.3  Local anesthetics  1.3.1  Lidocaine  8 10  1.4  Summary  14  1.5  Hypothesis and spécifie aims  16  Chapter 2: Materials and Methods  17  2.1  Solutions and chemicals  17  2.2  Site-directed mutagenesis  18  2.3  Cell culture and transfection  19  2.4  Electrophysiological recordings  20  2.5  Protocols and data acquisition  20  2.6  Dataanalysis  24 iv  Chapter 3: Results  28  3.1  Introduction  28  3.2  Lidocaine inhibits the HÇN1 channel in micromolar concentrations  28  3.3  Lidocaine does not affect the voltage dependence of channel activation  34  3.4 Lidocaine does not affect the rates of activation and deactivation, but shortens the delay prior to deactivation 40 3.5 Extracellular application of the permanently charged derivative of lidocaine, QX-314, does not inhibit HCNl 48 3.6  Properties of HCNl mutants with single amino acid residue substitution  Chapter 4: Discussion  50  61  4.1  Summary of the study  61  4.2  Discussion of the  4.3  Significance  71  4.4  Directions of future study  74  findings  62  Références  76  APPENDICES  82  Appendix 1  82  Appendix 2  83  v  List of Figures Figure 1.1. Predicted topology of aHCN channel subunit  2  Figure 1.2. ZD7288 inhibits the HCN2 channel by interacting with hydrophobic side chains of amino acid residues A425 and 1432 in the pore-lining région 6 Figure 1.3. Chemical structure of ivabradine  8  Figure 1.4. The aromatic ring of Phel579 in the S6 région of Domain IV in Nav 1.4 is responsible for cation-7i interaction with the charged aminé group of lidocaine  13  Figure 1.5. Chemical structure of lidocaine  13  Figure 1.6. Phenylalanine is conserved in the S6 pore-lining région across the four mammalian HCN isoforms 13 Figure 2.1. If calculation from current traces elicitedby -130 mV and+30 mV  23  Figure 3.1. Lidocaine, at 200 uM, inhibits HCNl channels to a greater extent at less hyperpolarized potentials  31  Figure 3.2. Linear fit of the modified Woodhull équation shows that lidocaine inhibition is weakly voltage-dependent  33  Figure 3.3. Lidocaine inhibition of If elicited by the +30 mV deactivation step is concentrationdependent 36 Figure 3.4. Lidocaine does not affect the voltage-dependence of HCNl channel activation  38  Figure 3.5. HCNl activation currents are well described by a double exponential function  41  Figure 3.6. Lidocaine does not affect the rate of activation of HCNl channels  42  Figure 3.7. Lidocaine shortens the delay prior to the start of single exponential fit on deactivation currents  44  Figure 3.8. Lidocaine does not affect the rate of deactivation of HCNl channels, but affects the delay prior to the start of a single exponential fit 46 Figure 3.9. Extracellular application of 500 uM QX-314 reduces If elicited by the -110 mV activation step and the +30 mV deactivation step, without current recovery  49  Figure 3.10. HCNl F378A has a midpoint of activation (Vi/j) similar to that for the wild-type channel, but has a significantly différent slope factor (k) 51 vi  Figure 3.11. Activation and deactivation time constants are significantly différent between wild-type HCNl and HCNl F378A  52  Figure 3.12. HCNl F378A deactivates more slowly compared to wild-type  53  Figure 3.13. Lidocaine inhibition of If at +30 mV is fast and réversible in HCNl F378A  55  Figure 3.14. Lidocaine inhibition of If at +30 mV increases with lidocaine concentration in HCNl F378A  57  Figure 3.15. Lidocaine does not affect the rate of activation of HCNl F378A at concentrations below400uM 58 Figure 3.16. Lidocaine affects the rate of deactivation of HCNl F378A  60  Figure 4.1. Four-state cyclic allosteric model for régulation of HCN2 opening and closing by voltage 69 Figure 4.2. Réduction in Ih can affect depolarization after the initial phase of afterhyperpolarization in action potentials in DRG neurons  73  Figure Al.l. 200 uM lidocaine inhibits If at -110 mV with complète current return  82  Figure A2.1. Représentative current trace of HCNl A372G  83  Figure A2.2. Représentative current trace of HCNl V379A  84  vu  List of Symbols and Abbreviations cAMP  Adenosine 3 ' , 5'-cyclic monophosphate  cGMP  Guanosine 3', 5'-cyclic monophosphate  If  Hyperpolarization-activated cation current (in cardiac cells)  Ih  Hyperpolarization-activated cation current (in neurons)  DRG  Dorsal root ganglion  HCN  Hyperpolarization-activated cyclic nucleotide-modulated channel  mV  Millivolt  raM  Millimolar  ms  Millisecond  uM  Micromolar  ul  Microlitre  M  Molar  ZD7288  4-(N-ethyl-N-phenylamino)-l,2-dimethyl-6-(methylamino)-pyrimidinium chloride  QX-314  Lidocaine rc-ethyl bromide  DMSO  Dimethyl sulfoxide  TEA +  Tetraethylammonium ion  mRNA  Messenger ribonucleic acid  HEK293  Human embryonic kidney 293 cells  CHO  Chinese hamster ovary cells  WT  Wild-type  °C  Degrees Celsius  Ag  Silver  CO2  Carbon dioxide  vin  KC1  Potassium chloride  NaCl  Sodium chloride  MgCl2  Magnésium chloride  CaCl2  Calcium chloride  AgCl  Silver chloride  HEPES  4-(2-hydroxyethyl)-1 -piperazineethanesulfonic acid  NaOH  Sodium hydroxide  KOH  Potassium hydroxide  ml/min  Millilitres per minute  K-aspartate  L-aspartic acid potassium sait  EGTA  Ethylene glycol-bis(2-aminoethylether)-N,N,N',N'-tetraacetic acid  Mg-ATP  Adenosine 5'-triphosphate magnésium sait  PIP2  Phosphatidylinositol 4,5-bisphosphate  MQ  Megaohm  kHz  Kilohertz  SA  Sinoatrial  IX  Acknowledgements First, I want to sincerely thank my supervisor, Dr. Eric Accili, for his mentorship, support, and patience over the past two years as well as during the two summers I was working in the laboratory as an undergraduate summer student. I am very grateful to him for providing opportunities for me to develop critical and independent thinking, and to learn from my mistakes. His continuous optimism and enthusiasm provided me with confidence and support. I want to thank Dr. Christopher Ahern, member of my supervisory committee, for his help and contribution to my project. His input was invaluable in data analysis and interprétation. I want to thank my other supervisory committee members, Dr. Steven Kehl and Dr. David Mathers, for their time, advice, and feedback. I want to also thank Dr. Stephan Schwarz for taking time out of his busy schedule to serve as an examiner for my defence. I am grateful to Dr. Damiano Angoli and Vincenzo Macri for teaching me electrophysiology through sharing their expertise and providing technical assistance. Their help allowed me to troubleshoot problems and hâve a greater understanding on the techniques. I want to extend spécial thanks to Dr. Damiano Angoli for helping with expérimental design and data acquisition. I want to also thank Dr. Andrew Hegle and Christian Peters for taking time to offer advice and feedback for my project as well as proofreading my thesis. I am deeply appréciative towards my fellow lab colleagues Hamed Nazzari, Sarah Chow, Gina Whitaker, and Heather Jackson for their time and patience teaching me molecular biology, cell culture, and other laboratory techniques. Finally, I want to thank my dearest family and friends for their love and support.  x  Chapter 1: Introduction  Many cellular processes require the transduction of electrical signais to initiate a spontaneous event such as neuron activation and muscle contraction. This signal transduction involves the movement of charges between the inside and the outside of cells. In biological Systems, charges are not mobilized in the form of électrons, but rather as inorganic ions such as sodium (Na+), potassium (K+), and chloride (Cl") ions. However, thèse ions are unable to permeate the membrane of cells, and require spécifie proteins on the membrane to assist movement into and out of the cell by lowering the energy barrier posed by the lipid membrane. An interesting membrane protein is the ion channel, which is a transmembrane protein spanning the lipid membrane with access to both the extracellular and intracellular environment. Some ion channels are sélective for particular ions, while others are non-selective. Ion channels can be modulated by various factors such as ligand binding and changes in voltage. The extent to which each type of ion channel is affected by thèse factors provides ion channels with unique physiological functions and differentiates one type of channel from another.  1.1  Hyperpolarization-activated cyclic nucleotidc-modulatcd (HCN) channels Hyperpolarization-activated cyclic nucleotide-modulated (HCN) channels are unique  channels, with opening and closing controlled by both voltage and ligand binding. Their secondary, tertiary, and quaternary structure is based on similarity to voltage-gated potassium or "Kv" channels (Gauss et al, 1998; Ludwig et al., 1998; Santoro et al., 1998). Thus, the channel subunits likely tetramerize to form channels that underlie the hyperpolarization-activated cation current (If in cardiac cells or Ih in neurons), which is important in regulating excitability in the  neurons of the central and peripheral nervous Systems and in the conduction tissue of the heart. Like Kv channels, each channel subunit consists of six transmembrane domains or segments (SlS6) with a pore-forming région between the 5* and 6th transmembrane domains that contains a selectivity filter for Na+ and K+ ions (Figure 1.1). The first four transmembrane domains (S1-S4) sensé changes in voltage, and control the opening and closing of the channel. The opening and closing of the channel, also known as channel gating, can be modulated by cyclic nucleotides such as cAMP and cGMP, which interact with the cyclic nucleotide binding domain (CNBD) located in the intracellular C-terminus.  Figure 1.1, Predicted topology of a HCN channel subunit. (Biel et al, 2002), used with permission. YSYALFKAMSHMLCIGYG  The hyperpolarization-activated cation current is also known as the "funny" current (If) because the functional properties of HCN channels are unusual compared to most voltage-gated 2  Na or K channels. HCN channels activate in response to hyperpolarizing voltages, with the midpoint of activation ranging from -73 mV to -92 mV (Accili et al., 2002). If is carried by a mixture of K+ and Na+ ions with a ratio of approximately 4:1, respectively. In response to hyperpolarized voltages, thèse channels open to yield an inward current. As positive ions enter the cell, the membrane potential becomes depolarized, and gradually reaches threshold for the next action potential firing. Ih has been described in central and peripheral nervous tissue (Pape, 1996), in which it controls the extent of hyperpolarizations caused by potassium currents and inhibitory inputs (Luthi and McCormick, 1998). Ih is also known to modulate the resting potential and regulate the firing frequency of action potentials (Chan et al., 2004; Ludwig et al, 2003; Nolan et al., 2003; Pape, 1996; Pape and McCormick, 1989). In the heart, If contributes to the diastolic depolarization phase of the action potential, which brings the membrane potential from the "undershoot" back to threshold level for the next action potential firing (DiFrancesco, 1993). There are four mammalian isoforms of HCN channels (HCN 1-4), each with unique properties and patterns of distribution in différent tissues (Robinson and Siegelbaum, 2003). Ail four isoforms are expressed in the brain (Ludwig et al., 1998; Santoro et al., 1998). In the sinoatrial node of the heart, HCN4 mRNA is présent in the greatest amount, with mRNA of HCNl and HCN2 also présent (Moosmang et al., 2001; Shi et al., 1999). HCNl activâtes the fastest among the four isoforms, whereas HCN4 has the slowest time of activation (Altomare et . al., 2001). Intracellular cyclic nucleotides, most notably cAMP, can alter the voltage range in which thèse channels open, with HCNl being the least responsive (Baruscotti et al., 2005). Thus, within the heart and the brain, the amounts of différent HCN isoforms expressed can détermine  3  the kinetics, voltage dependence, and the amount of inward, time-dependent current If (Cheng et al., 2007).  1.2  Pharmacology of If Because If is important in regulating heart rate as well as action potential firing frequency  in neurons, drugs that affect this current can potentially be used to treat patients with irregular heart rates, ischémie heart disease, and neurological diseases. The main goal of HCN channel pharmacological inhibition is to reduce heart rate or firing rate in neurons without compromising other features of the action potential, leading to unwanted side effects. (3-blockers and some calcium channel antagonists are examples of agents that can reduce heart rates, but hâve side effects on the respiratory System and decrease the strength of heart muscle contraction (DiFrancesco and Camm, 2004). In the heart, bradycardic agents are particularly useful in patients with decreased oxygen availability to cardiac tissue and patients with irregular or higher than normal heart rates. Similarly, thèse agents may be useful to reduce pain in patients with conditions such as neuropathic pain, which may involve HCN channels in the dorsal root ganglion (DRG) neurons contributing to neuronal excitability (Tu et al., 2004). By studying how thèse drugs affect the channel, important information such as the locations specifically targeted by the drugs can be elucidated. Prior to the cloning of four HCN isoforms from mouse, human, and the sea urchin sperm in the late 1990s (Gauss et al., 1998; Ludwig et al., 1998; Santoro et al., 1998), pharmacological studies could only détermine whether spécifie agents had an effect on If in a particular tissue, likely containing a mixture of the isoforms expressed in différent amounts. The ability for each HCN isoform to be heterologously expressed in mammalian culture cells  4  allows for experiments examining the effect of drugs on the spécifie isoforms. The following section describes the history and properties of two known HCN channel blockers.  1.2.1  ZD7288 ZD7288, a substituted aminopyrimidinium compound, has bradycardic and antianginal  effects (Figure 1.2). It was initially found to block If in dissociated guinea-pig sinoatrial node cells with an IC50 of approximately 0.3 uM (BoSmith et al., 1993). Ca2+ currents and K+ currents were also shown to be affected in addition to If, albeit to a much lesser extent (BoSmith et al., 1993). Following the initial characterization of the effects on If, ZD7288 was also shown to inhibit Ih in substantia nigra, the hippocampal CA1 région, thalamocortical neurons and photoreceptors (Gasparini and DiFrancesco, 1997; Harris and Constanti, 1995). This inhibition was not found to be frequency- or use-dependent, indicating that inhibition does not accumulate with each subséquent opening and closing of the channel. The exact mechanism by which ZD7288 inhibits If was not known until studies on cloned HCN channels were performed. In mammalian HEK293 cells andXenopus'oocytes expressing HCNl cloned from mouse (mHCNl), ZD7288 enters the channel when it opens, and then becomes trapped in the intracellular side of the channel when the channel closes (Shin et al., 2001). The inhibition of ZD7288 on the mHCNl channel is réversible and is less at more hyperpolarized voltages; however, in another clone from sea urchin sperm (spHCN), the effect of ZD7288 was not réversible (Shin et al., 2001). Shin et al. suggest that the différences in amino acid residues in the S6 pore-lining région between mHCN 1 and spHCN may hâve attributed to the reversibility of inhibition, or lack thereof. In a récent study, the side chains of two amino acid residues lining the pore région of HCN2 were suggested to interact with ZD7288 (Cheng et al., 2007) (Figure 1.2). Thèse side 5  chains were speculated to form a hydrophobic cavity for interaction with the aromatic ring of ZD7288. Together, thèse fmdings suggest that the pore région of the channel is critical for ZD7288 inhibition.  Figure 1.2. ZD7288 inhibits the HCN2 channel by interacting with hydrophobic side chains of amino acid residues A425 and 1432 in the pore-lining région. (Left) Chemical structure of ZD7288. (Right) Aromatic ring of ZD7288 interacts with the hydrophobic pocket formed by the side chains of the amino acid residues predicted to face the central cavity of the channel pore. (Cheng et al., 2007), used with permission.  1.2.2  Ivabradine Ivabradine (S 16257) (Figure 1.3) has been shown to slow down heart rate by inhibiting If  specifïcally (DiFrancesco and Camm, 2004). However, side effects such as mild blurred vision and dizziness were observed in patients administered a single 0.2 mg/kg intravenous dose of ivabradine (Camm and Lau, 2003). Blurred vision might hâve been caused by the effect of ivabradine on Ih in photoreceptors, which express predominantly HCN1 (Barrow and Wu, 2009; Demontis et al., 2002; Moosmang et al., 2001). Ivabradine was initially found to block If in 6  rabbit sinoatrial node cells (Thollon et al., 1994). Ivabradine caused minimal prolongation of the action potential duration compared to another bradycardic agent zatebradine (Thollon et al, 1994). Because the réduction in action potential firing frequency was mainly from the réduction in diastolic depolarization, Thollon et al. suggested that ivabradine inhibition was specifically on If. Ivabradine block of If was use-dependent, with an IC50 of 2.8 uM (Bois et al., 1996). Using inside-out and cell-attached patch-clamp configurations, the effect of ivabradine on If was found to be similar to that in the whole-cell configuration, suggesting that ivabradine inhibits the channel from the intracellular side (Bois et al., 1996). The authors also showed that ivabradine had minimal effects on the K+ current responsible for repolarization of the action potential, thus further confirming its bradycardic effect is solely on the pacemaker current If. Similar to the studies performed with ZD7288, intensive investigation has been performed with cloned HCN isoforms in trying to decipher the mechanism of inhibition as well as the site(s) of inhibition. Analysis of the effects of ivabradine on cloned mouse HCN1 and human HCN4 channels in HEK293 cells revealed that ivabradine requires the HCN4 channel to open for inhibition, whereas HCN1 channels do not need to be open for inhibition to take place (Bucchi et al., 2006). In contrast to ZD7288, inhibition by ivabradine increases and accumulâtes with the number of openings and closings of the channel until a certain point when the current ceases to reduce, which may take from 13 seconds with 30 uM up to 90 seconds with 0.3 \xM (Bucchi et al., 2002). This property makes ivabradine a good bradycardic agent, as greater inhibition occurs at higher heart rates and reduced effect at lower heart rates. Currently, ivabradine is used clinically as a drug to treat angina pectoris (DiFrancesco and Borer, 2007).  7  Figure 1.3. Chemical structure of ivabradine (pKa = 8.6).  1.3  Local anesthetics Local anesthetics hâve been used for more than 150 years to relieve pain. Thèse agents  are particularly useful in patients undergoing opérations or surgery, and to reduce chronic pain. Early local anesthetics such as cocaine can be found naturally, but can only provide short-duration anesthesia. In addition, the use of cocaine is associated with high toxicity (Arthur and Strichartz, 1987). Local anesthetics with modified chemical structures were synthesized in the early to mid-1900s, and provided longer duration of anesthetic action as well as fewer side effects. Local anesthetics are characterized by particular chemical structures. Ail hâve three chemical moieties that play important functional rôles, as well as separating one from another: a lipophilic (aromatic) end, a hydrophilic (amine) end, and a linker between the two ends (Heavner, 2007). The linker can be either an aminoester group, or an aminoamide group. Early local anesthetics such as cocaine and procaine hâve an ester group as the linker, and are called ester-type local anesthetics. Local anesthetics such as lidocaine that were later synthesized in the early to mid-1900s hâve an amide group as the linker, and are called amide-type local anesthetics. The amine end is usually a tertiary amine that can be protonated (charged) or remain neutral (uncharged), depending on the pKa of the compound. Local anesthetics, in the uncharged form, move through the lipid membrane and into the cell, then block sodium channels from the  8  intracellular side (Hille, 1977a). Thus, the relative proportion of charged versus uncharged form of the compound can control the speed of block onset. Local anesthetics act in part by blocking nociceptive stimuli and transmission to the brain from various régions of the body. Local anesthetics act on the cell membrane, and are well known for their réversible, inhibitory effect on voltage-gated Na+ channels resulting in block of action potential propagation responsible for nerve conduction in both central and peripheral nerve pathways (Butterworth and Strichartz, 1990). Local anesthetics can also target cardiac cells by affecting permeability of cardiac Na+ channels (Bean et al., 1983). However, local anesthetics are not necessarily spécifie Na+ channel blockers, but may hâve effects on voltage-gated K+ channels, voltage-insensitive K+ channels, two-pore domain K+ sélective channels, and many other membrane-associated proteins (Kindler and Yost, 2005; Komai and McDowell, 2001; Olschewski et al., 1998). The non-specificity in local anesthetic action and its interaction with various channels known to contribute to the action potential lead to the possibility that HCN channels are also targeted. Because Ih has been shown in peripheral nerves and DRG neurons (Grafe et al., 1997; Mayer and Westbrook, 1983; Takigawa et al., 1998), it is possible that block of action potential génération and propagation in peripheral nerves by local anesthetics may be due to block of Ih. A récent study has demonstrated that HCN1 knockout mice expérience less cold allodynia in a neuropathic pain model (Momin et al., 2008). Also, Ih is composed of mainly HCNl-mediated currents in large-diameter and fast-conducting medium-diameter neurons (Momin et al., 2008). Thus, Ih may play a significant rôle in the génération of action potentials and transmission of pain signais. One study examined the effect of three local anesthetics on Ih in rat small dorsal root ganglion (DRG) neurons. Bupivacaine, lidocaine, and mepivacaine ail hâve significant and 9  réversible effects on Ih when applied extracellularly, with IC50 of 55, 99, 190 uM respectively (Bischoff et al., 2003). It is likely that mechanism of drug action on peripheral nerves by local anesthetics can be partly attributed to block in Ih.  1.3.1  Lidocaine Lidocaine is a common local anesthetic and a class Ib antiarrhythmic drug that is known  to inhibit conductance of Na+ channels with possible effects on K+ channels (Hille, 2001 ; Olschewski et al., 1996). Lidocaine can reduce heart rates (Goodman et al., 2001), and is contraindicated in patients with heart failure and bradycardia when administered intravenously (Harrison and Collinsworth, 1974). Lidocaine's effects both as an antiarrhythmic and a local anesthetic mostly involve lowering sodium permeability by inhibiting voltage-gated Na+ channels, in order to slow heart rates and reduce pain. The concentration needed to hâve significant effects on cardiac Na+ channels is approximately 15 to 30 uM (Bean et al., 1983; Grant et al., 1980; Hondeghem and Katzung, 1977). Therapeutic concentrations used to treat heart conditions such as arrhythmias range from approximately 5 to 20 uM (Burashnikov et al., 2007). In nerves, however, the concentration needed to hâve significant effects on Na+ channels is approximately 100 to 250 uM (Hille, 1977b). Clinically, lidocaine patches applied to the skin hâve a concentration of 5% or approximately 200 raM (Galer et al., 1999). Similar to action on Na+ channels, lidocaine has been shown to inhibit Ih in nerves and If in the heart. As described above, lidocaine inhibits Ih in rat small DRG neurons, with an IC50 of 99 uM (Rocchetti et al., 1999). In rabbit sinoatrial myocytes, lidocaine has been shown to inhibit If in a concentration-dependent manner, with an IC50 of approximately 38 uM and up to  10  70% réduction in current by 100 uM of lidocaine (Rocchetti et al., 1999). Thus, the antiarrhythmic effect of lidocaine can be attributed in part to its action on If. A récent study showed effects of various antiarrhythmic drugs on rabbit HCN4 channels expressed in HEK293 cells. In low K+ (Tyrode) extracellular solution, lidocaine inhibited HCN4-mediated currents at -70 mV with an IC50 of 276 uM (Tamura et al., 2009). Although lidocaine, at 30 uM, appeared to inhibit If to a greater extent at more hyperpolarized potentials, this was not significantly différent from control. Unlike the study examining lidocaine inhibition of If in rabbit SA myocytes (Rocchetti et al., 1999), Tamura et al. evaluated lidocaine inhibition only at a potential of-70 mV where channels were not fully-activated. Finally, it was not clear to what extent current rundown contributed to the inhibition observed by Tamura et al. Therefore, a number of issues remain to be resolved with respect to the mechanism of lidocaine block of HCN channels as well as whether inhibition by this drug is similar among HCN isoforms. Lidocaine has been shown to block Na+ channels by binding to aromatic side chains of amino acid residues lining the pore, specifically phenylalanine (Ahern et al., 2008). This interaction involves the cation from the ionized aminé group of lidocaine and the rc-electrons from the aromatic ring of phenylalanine (Ahern et al., 2008) (Figure 1.4). At a pH lower than the pKa of 7.9, most lidocaine molécules acquire a proton and become charged (Figure 1.5). At a physiological pH of 7.4, approximately 25 % of lidocaine molécules remain uncharged. In extracellular solution, uncharged lidocaine molécules diffuse through the cell membrane to the inside of the cell, and become the active form once protonated (Hille, 1977a). Thus, most of the intracellular lidocaine can readily bind to the suspected binding site in the channel pore. This claim can be supported by experiments showing intracellular effects on Ih by the quaternary, 11  permanently charged derivative of lidocaine, QX-314, in guinea pig hippocampal CA1 pyramidal cells and Cajal-Retzius cells in the neocortex of rat neonates (Kilb and Luhmann, 2000; Perkins and Wong, 1995). In ail four mouse HCN isoforms, phenylalanine residues are found in the pore-lining sixth transmembrane domain (S6) région: amino acid 378 in mHCNl, amino acid 431 in mHCN2, amino acid 341 in mHCN3, and amino acid 509 in mHCN4 (Figure 1.6). With amino acids having aromatic side chains lining the pore of HCN channels, it is reasonable to propose that lidocaine inhibits HCN channels via a cation-7t interaction.  12  Figure 1.4. The aromatic ring of Phel579 in the S6 région of Domain IV in Nav 1.4 is responsible for cation-7t interaction with the charged aminé group of lidocaine. (Ahern et al., 2008), used with permission.  Figure 1.5. Chemical structure of lidocaine (pKa = 7.9).  Figure 1.6. Phenylalanine is conserved in the S6 pore-lining région across the four mammalian HCN isoforms. S6 mHCN1 mHCN2 mHCN3 rnHCN4  (343) (396) (306) (474)  SHMLCIGYGAQAPVSMSDLWITMLSMIVGATCYAM0VGHATALIQSMDSSSRQY SHHLCIGYGP.QAPESMTDIWLTMLSMIVGATCYAMF IGHATALIQSMDSSPSOY SHMLCIGYGQQAPVGMPDVWLTMLSMIVGATCYAMFIGHATALIQSHDSSSP.QY SHMLCIGYGPQAPVGMSDVWLTMLSMIVGATCYAMFIGHATALIQSMDSSSRQY  (396) (449) (359) (527)  13  It is important to know whether lidocaine, which is an important therapeutic agent and is used extensively in clinical situations, inhibits individual HCN isoforms differently. Individual tissues express différent amounts and combinations of HCN isoforms. By examining the effects of lidocaine on each cloned HCN isoform, the relative sensitivity of each body tissue to lidocaine can be estimated, based on the prédominant isoforms expressed by each tissue. Moreover, détails about possible binding sites on the each channel isoform to lidocaine can be elucidated by comparing the actions on each isoform. This information can be bénéficiai in the development of more spécifie therapeutic stratégies for treating patients with conditions such as arrhythmia, and for reducing unwanted side effects such as bradycardia.  1.4  Summary HCN channels underlie the hyperpolarization-activated cation current, which is important  in controlling the "pace" of many physiological functions. Abnormalities in channel function can resuit in neuronal and heart diseases. Thèse can be remedied by the use of agents that control or inhibit this current specifically. Many studies hâve been performed on other types of ion channels that contribute to pacemaking such as voltage-gated Na+ and K+ channels, yielding valuable information in treating diseases associated with malfunction of thèse channels. However, the rôle of HCN channels and the pacemaker current should not be underestimated. Even though use of many clinical drugs, such as local anesthetics, may be deemed safe, thèse drugs may hâve potential side effects by targeting other molécules such as the pacemaker channels. It has been shown that lidocaine, a local anesthetic, has an inhibitory effect on the pacemaker current under clinically relevant concentrations. Inhibiting Ih may also contribute to the wanted effects such as reducing pain in achieving anesthesia. Thus, it is important to 14  understand more about the blocking mechanism so that safer drugs can be designed, and also to understand how the structure of the channel contributes to the inhibitory action. Based on the background presented, it seems reasonable to test whether lidocaine inhibits HCNl channels, and if the mechanism of block may involve the binding of the drug to an intracellular site along the S6 segment. The objectives of this thesis are to address the effects of lidocaine on It- and channel kinetics of cloned HCNl channels, and to compare our findings with studies examining effects of lidocaine on native If as well as on Na+ and K+ channels. I will first présent the hypothesis with spécifie aims used to support or disprove the hypothesis. Détails of each experiment are described in the following Materials and Methods chapter. Findings in this study are then presented, in the order of the spécifie aims, followed by discussion and interprétation of the results.  15  1.5  Hypothesis and spécifie aims  Hypothesis Lidocaine, in micromolar concentrations, inhibits HCNl channelsfromthe intracellular side by interacting with amino acid residues lining the channel pore.  Spécifie aims 1) To détermine whether lidocaine inhibits the HCNl channel in micromolar concentrations and to construct a concentration-response relationship. 2) To détermine whether lidocaine affects the steady-state activation and rates of activation and deactivation of the channel. 3) To détermine whether the permanently charged derivative of lidocaine (QX-314), applied in the extracellular solution, inhibits the HCNl channel. 4) To identify residues in the S6 pore-lining région of the channel responsible for lidocaine interaction and inhibition.  16  Chapter 2: Materials and Methods  2.1  Solutions and chemicals Low K+ extracellular solution consisted of 5.4 mM KC1, 135 mM NaCl, 0.5 mM MgCb,  1.8 mM CaCl2, 5 mM HEPES, pH 7.4 (NaOH). High K+ solution or control solution, used for recordings to maximize current amplitude, consisted of 135 mM KC1, 5.4 mM NaCl, 0.5 mM MgCl2, 1.8 mM CaCl2, 5 mM HEPES, pH 7.4 (KOH). Intracellular-like solution filled into the patch pipettes contained 130 mM K-aspartate, 10 mM NaCl, 0.5 mM MgCl2, 5 mM HEPES, 1 mM EGTA, 2 mM Mg-ATP, 2 mM creatine phosphate, 1 mM cAMP, 20 uM PIP2, pH 7.4 (KOH). Certain compounds such as Mg-ATP, cAMP, and PIP2 were included in the intracellular solution to minimize the effects of current rundown, which is known to shift the voltagedependence of activation in the hyperpolarized direction (Pian et al., 2006). Current rundown is described in détail in the discussion section on page 63. 2 mM Mg-ATP shifts the midpoint of activation (V1/2) of HCN2 in the depolarized direction by 9.5 mV (Pian et al., 2006). HCN1 is less responsive to cAMP compared to HCN2 and HCN4 (Accili et al., 2002), and 1 uM cAMP only shifts V1/2 by 2 mV in the depolarized direction (Santoro et al., 1998). Using the expérimental system and procédures in our laboratory, 1 mM cAMP was found to shift V| /2 in the depolarized direction by approximately 10 mV in HCN2 (Proenza et al., 2002). For cAMP to hâve an effect against current rundown, a concentration of 1 mM was used. Moreover, a depolarizing shift in Vi/2 of approximately 15 mV was observed with 25 uM PIP2 in HCN2 (Pian et al., 2006). This intracellular-like solution was first made in 1 mL aliquots stored at -20°C. On the day of the experiment, one tube was thawed and dispensed into a syringe. The syringe with the intracellular solution was placed in room température for at least 15 minutes before use. 17  Lidocaine (Sigma-Aldrich) was dissolved in DMSO and made into 1 mL of 100 mM and 500 mM stocks stored at 4°C. Lidocaine was diluted to the desired concentrations with high K+ solution on the day of the experiment, and was placed in room température for at least 15 minutes prior to use. The 100 mM stock was used to make final concentrations up to 100 uM, and the 500 mM stock was used to make final concentrations higher than 100 uM. Stock solutions of higher concentrations were not used because lidocaine did not completely dissolve in stock solutions higher than 500 mM. This procédure ensured that the % DMSO in the recording solution was kept to a minimum. Use of lidocaine hydrochloride may hâve provided a better indication of lidocaine inhibition, as this form is used clinically and is soluble in water. It seems possible that the structure of lidocaine might differ when dissolved in DMSO or water. QX-314 (Sigma-Aldrich) was dissolved in water, and made into 1 mL of 20 mM stock stored at 4°C. QX-314 was diluted to the desired concentration with high K+ solution on the day of the experiment, and was placed in room température for at least 15 minutes prior to use.  2.2  Site-directed mutagenesis Three mutants, each with single amino acid residue substitutions in the pore-lining S6  région, were made to détermine whether lidocaine inhibition involved interaction with thèse residues. Substitution of phenyialanine at position 378 with alanine was based on homology to a phenyialanine located in the S6 pore-lining région of the fourth homologous domain (D4) in Nav1.4, at which lidocaine, in its charged form, interacted with % électrons of the aromatic ring (Ahern et al., 2008). Na+ channels were chosen for comparison, as lidocaine inhibition on Na+ channels hâve been studied more extensively compared to K+ channels. Also, Na+ channels possess similar structure compared to HCN channels in that the S6 région lines the channel pore. 18  The A372G and V379A mutations were made based on the fact that the side chains from thèse two amino acid residues are known to form a hydrophobic cavity for the phenyl ring of ZD7288, a known HCN channel blocker (Cheng et al., 2007). There is a phenyl ring présent in lidocaine which could occupy the hydrophobic cavity. Site-directed mutagenesis was performed to produce point mutations at thèse three sites. Mutant channels were generated by polymerase chain reaction using oligonucleotide primers (synthesized by Integrated DNA Technologies) encoding the mutation of interest using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA). Ail mutations were verified by DNA sequencing (Nucleic Acid Protein Service Unit, University of British Columbia).  2.3  Cell culture and transfection The mammalian culture cells used were CHO-K1 cells (American Type Culture  Collection, Manassas, VA), maintained in Ham's F-12 médium (Invitrogen, Ontario, Canada) supplemented with 50 ug/ml penicillin/streptomycin (Invitrogen) and 10% fetal bovine sérum (Invitrogen). Cells were grown on glass coverslips (22 x 22 mm) in 35 mm dishes at 37°C with 5% CO2. Transfection of mammalian expression vectors encoding mouse HCNl (a kind gift from Dr. Martin Biel, University of Munich) or HCNl mutant channels generated (1 ug per dish) was performed using the FuGeneô transfection reagent (Roche Diagnostics, Québec, Canada) according to manufacturer's protocols. The DNA plasmid encoding Green Fluorescent Protein (GFP) was co-transfected for visualization of transfected cells.  19  2.4  Electrophysiological recordings Approximately 24 to 48 hours post-transfection, a small pièce of coverslip was placed  into the recording chamber (180 ul volume) on the stage of an inverted microscope. The recording chamber was perfused with low K+ extracellular solution at a constant rate of 0.5 - 1.0 ml/min. The voltage and current were recorded using a borosilicate glass électrode (Sutter Instruments Co., CA) filled with an intracellular-like solution. The patch pipettes had a résistance of 2.0 - 5.0 MQ. An Ag-AgCl ground wire was connected to the bath solution by a 3M KC1 agar bridge. Ail currents were measured in the whole-cell configuration. A gigaohm seal between the tip and the cell membrane and rupturing of the membrane patch were obtained via manual suction. The recording chamber (extracellular) solution was changed to high K+ solution after rupturing of the membrane patch. After one minute of complète perfusion of the new solution, voltage protocols were applied and currents were recorded using an Axopatch 200B amplifier and Clampex software (Axon Instruments Inc., Foster City, CA). Data were sampled at 1 kHz, filtered at 2 kHz, and analyzed using Clampfit (Molecular Devices, CA) and Origin (Microcal, MA) software. Data should hâve been sampled at least two times the filter -3 dB frequency, as suggested by the Nyquist Sampling Theorem. Despite sampling at a lower frequency, data in this study were still reliable and currents could be accurately quantified. Currents were not leak subtracted, and capacitance compensation was not used. AU experiments were performed at room température.  2.5  Protocols and data acquisition The whole-cell patch-clamp method was used to measure currents, which were elicited  by applying a séries of hyperpolarizing puises from a holding potential of-35 mV. Micromolar  concentrations of lidocaine were added to cells to détermine whether HCN-mediated currents (If) were reduced, similar to studies examining effects on Ih in rat small DRG neurons (Bischoff et al., 2003) and on If in rabbit SA myocytes (Rocchetti et al., 1999). Currents were recorded before, during, and after lidocaine application to the extracellular side of the membrane. Cells in which current return was partial or complète were chosen for analysis. To détermine whether the mechanism of lidocaine block was by gradually blocking ions passing through the channel pore, a protocol consisting of activation and deactivation steps was applied every 6 seconds (1/6 Hz) until reaching steady-state inhibition, followed by washout of the drug. The duration of the activation step varied from 800 to 1000 ms in order for activation currents to reach steady-state, and the duration of the deactivation step was 450 ms. This protocol was used to test whether lidocaine, when delivered directly onto the cell under study via a fast-step perfusion System (Warner Instruments, CT), caused réduction in current amplitude, which could accumulate until steady-state inhibition was reached. This protocol was used to generate a time course showing onset of inhibition and current return. If amplitude was calculated as the différence between the steady-state current and the instantaneous current (Figure 2.1). Lidocaine inhibition was expressed as percent of If réduction observed during steady-state inhibition relative to control or prior to lidocaine addition. Percent If réduction caused by 200 uM lidocaine was examined at hyperpolarizing potentials of-85 mV, -100 mV, and-HOmV. Inhibition by lidocaine at concentrations of 50, 100, 200, 400, 800, 1600, and 3200 uM was quantified by the percent If réduction elicited by the +30 mV deactivation step. If amplitude was calculated as the différence between the peak current and the steady-state current (Figure 2.1). The percent If réduction was determined by the différence between If at steady-state inhibition 21  and If prior to lidocaine addition. Similar protocols were used for experiments on the HCN1 F378A mutant, with the exception that the duration of the activation step varied from 10Q0 to 1500 ms and the duration of the deactivation step was 5000 ms for currents to reach steady-state. Because of the prolonged duration in deactivation, trains of activation/deactivation steps were applied every 8 seconds (1/8 Hz). Controls were performed using high K solution containing % DMSO équivalent to that in the lidocaine solution.  22  Figure 2.1. If calculation from current traces elicited by -130 mV and +30 mV. (Top) At -130 mV, If amplitude was calculated as the différence between the steady-state current and the instantaneous current (Ijnst)- (Bottom) At +30 mV, If amplitude was calculated as the différence between the peak current and the steady-state current.  lOOOpA  200 ms Peak current  100 pA  Steady-st )te current  200 ms  The effects of QX-314 were examined using the same expérimental protocol when delivered extracellularly at a concentration of 500 pM. This concentration was at least 2 to 5-fold lower compared to intracellular QX-314 concentrations used to inhibit Ih in guinea pig hippocampal CAl pyramidal cells and Cajal-Retzius cells in the neocortex of rat neonates (Kilb and Luhmann, 2000; Perkins and Wong, 1995). Because the objective of this experiment was to 23  détermine whether lidocaine, in its charged form, interacting with extracellular site(s), the QX-314 concentration chosen was within the range of lidocaine concentrations known to hâve substantial inhibition on If. Voltage-dependence of activation was determined from tail currents at +30 mV following test puises ranging from -150 mV to -50 mV in +20 mV steps, with 1800 ms duration for the initial step and an additional 1800 ms for each subséquent step to ensure currents reached steadystate. A 100 ms puise to +5 mV after the +30 mV step was included to ensure complète channel deactivation. For the F378A mutant, voltage-dependence of activation was determined from tail currents at -35 mV following the same set of test puises, with 1500 ms duration for the initial step and an additional 500 ms for each subséquent step.  2.6  Data analysis Ail data are presented as mean ± standard error of the mean (SEM) (n = number of cells),  except parameters generated by fitting the data points using a nonlinear least-squares procédure such as those in the concentration-response curve, which are presented as mean ± standard error (SE). Paired student's t-test was used for statistical comparisons between values from control and a spécifie lidocaine concentration tested on the same cell, and was not used to assess multi-group différences. Multi-group différences comparing lidocaine inhibition at various concentrations were assessed by one-way analysis of variance (ANOVA), with post hoc analysis of individual means using Tukey multiple comparison test to détermine which of the means were significantly différent. A p-value of less than 0.05 was deemed statistically significant. Curve fitting and génération of bar graphs were performed using Origin software (Microcal, MA), and  24  statistical analysis such as student's t-test and analysis of variance were performed using GraphPad Prism version 4.0 software (San Diego, CA).  Voltage-dependence of inhibition K1/2 values, or binding constants of lidocaine, at -110, -100, and -85 mV were calculated as the fraction of unblocked channels (Funb) divided by the fraction of blocked channels (1 - Funb). The fraction of unblocked channels was the same as the proportion of If unblocked, and the fraction of blocked channels was the proportion of If blocked. The natural logarithms of K1/2 values were plotted against membrane voltage, and were fitted with a linear function. According to the modified Woodhull équation (Woodhull, 1973) derived by Heginbotham and Kutluay, 2004, the effective valence value, zô, was obtained from the négative slope value of the linear relationship: In K1/2 = In K1/2 (0 mV) - zÔFVm/RT where F is Faraday's constant, Vm is the membrane potential, R is the gas constant, T is the absolute température, and In K1/2 (0 mV) is the voltage-independent binding constant at 0 mV.  Voltage-dependence of activation Normalized tail current amplitudes were plotted as a function of test potential and values were fitted with the Boltzmann function: XV) = I max /(r+exp((V 1 /2-V)/k)) where V is the test voltage, k is the slope factor, and V1/2 is the midpoint of activation. The Boltzmann function assumes that channels only undergo from closed to open transition, and does not account for any intermediate closed or open states that may exist. After obtaining 25  activation traces for control, a différent cell was used to evaluate the effects of lidocaine to ensure that any différence in midpoint of activation was not due to current rundown. Unpaired student's t-test was used to détermine whether V1/2 and k values were significantly différent between control and after lidocaine addition.  Concentration-response relationship The proportion of If unblocked (y) was plotted against lidocaine concentration on a logarithmic scale to generate a concentration-response relationship. Data points were fitted by a nonlinear least-squares procédure using a modified version of the Hill équation assuming that lidocaine does not inhibit currents completely: y =1 - IM«. inhibition/ (1 + (IC50/[Lidocaine])An) where IC50 is the half maximal inhibitory concentration, n is the Hill coefficient, [Lidocaine] is the lidocaine concentration in the bath solution and  IM3X. inhibition  is the proportion of If blocked at  maximal inhibition.  Determining rates of activation and deactivation Time constants for activation were generated using a double exponential fitting procédure, and a single exponential fitting procédure was used to generate the time constant for deactivation. Paired student's t-test was used to analyze différences in time constants between control and a spécifie lidocaine concentration tested on the same cell. Multi-group différences caused by various lidocaine concentrations were not assessed because of the variability in time constants between control values in différent cells. Activation currents were fitted by double exponential fonction, according to the following équation: 26  I(t) = Af exp (-t/tfast) + As exp (-t/xslow) + C where Af and As are current amplitudes of fast and slow components of current activation, respectively, Tfast and xsi0w are the time constants for fast and slow components of current activation, respectively, and C is the steady-state current amplitude. Deactivation currents were fitted by a single exponential function, according to the following équation: I(t) = A exp (-t/xdeact) + C where A is the current amplitude of current deactivation and xdeact is the time constant of current deactivation. The relative amplitude of fast component of current activation was calculated: A f /(Af+A s ) where At- and As are current amplitudes of fast and slow components of current activation, respectively.  27  Chapter 3: Results  3.1  Introduction Results in this chapter aim to address the spécifie aims presented in Chapter 1, and are  described in the order in which the spécifie aims are presented. The first section reports that inhibition of If by lidocaine is fast, réversible, and weakly voltage-dependent. This observation is supported by représentative time courses and traces for différent potentials tested. A spécifie quantitative parameter is chosen to evaluate the half maximal inhibitory concentration (IC50) of lidocaine on the HCNl channel. The next section (3.2) looks at activation curves and analysis on channel activation and deactivation kinetics, which provide information as to whether or not lidocaine affects gating of the channel. The following section (3.3) examines the effects of the permanently charged derivative of lidocaine, QX-314, on the channel when applied extracellularly. Lastly, properties of three mutants with single amino acid substitution in the pore-lining S6 région of HCNl are presented in section 3.4. Ail three mutants display kinetic properties that differ from the wild-type. The properties of one mutant and its effects by lidocaine are explored in détail. Initial experiments performed to optimize expérimental conditions are described in the discussion chapter, with the corresponding figure attached in the Appendices.  3.2  Lidocaine inhibits the HCNl channel in micromolar concentrations HCNl channel steady-state activation currents were examined in the présence of  lidocaine in micromolar concentrations. Whole-cell currents were elicited by trains of activating/deactivating voltage steps applied every 6 seconds (Figure 3.1). This protocol was  applied at least 2-3 times prior to perfusion of 200 uM lidocaine. The current generated from the activating step in each puise was plotted over time. Thèse current recordings prior to lidocaine addition were denoted as control indicated by squares in the time course in Figure 3.1. Lidocaine was then applied for approximately 3-4 sweeps or approximately 20 seconds until steady-state inhibition was reached. The onset of lidocaine inhibition was fast, which was compatible with lidocaine entry into the cell through the cell membrane in the uncharged form. Lidocaine inhibition did not appear to be use-dependent, as inhibition did not accumulate with repeated steps to open and close the channels. The control solution was re-introduced, and the protocol was run for approximately 20-30 seconds to observe whether or not currents returned to the level prior to lidocaine addition. The percent If réduction caused by 200 uM lidocaine was 6.6 ± 1.7 % (n = 3 cells), 8.3 ± 1.0 % (n = 7 cells), 17.1 ± 3.3 % (n = 7 cells) at -110 mV, -100 mV, and -85 mV respectively (Figure 3.1), suggesting greater inhibition at less hyperpolarized potentials. To assess the voltage-dependence in lidocaine inhibition, Woodhull analysis was performed. A modified version of the Woodhull équation (Heginbotham and Kutluay, 2004) was used in determining the delta (ô) value, which represents the percent pénétration of lidocaine into the transmembrane electric field. First, the fraction of unblocked channels (Funb) was calculated to be the fraction of If remaining, and the fraction of blocked channels (1-Funb) was derived from the fraction of If inhibited. The natural logarithm of the proportion of unblocked channels versus blocked channels was then plotted versus the membrane potential, and the plot was fitted with a linear function (Figure 3.2). The négative of the slope value yielded the z-delta (zô) value. Given that lidocaine can be either uncharged or possess a single positive charge, its valence (z) value is +1 assuming that only the charged form participated in channel inhibition. Figure 3.2 shows that lidocaine only sensed 4.5% of the transmembrane electric field from a ô 29  value of 0.045. Compared to a study that used a similar approach to assess voltage-dependence of inhibition (Heginbotham and MacKinnon, 1992), this z-delta value indicated weakly voltagedependent inhibition by lidocaine on the HCN1 channel at hyperpolarized potentials.  Figure 3.1. Lidocaine, at 200 uM, inhibits HCNl channels to a greater extent at less hyperpolarized potentials. Trains of activating/deactivating steps were applied every 6 seconds to Chinese hamster ovary (CHO) cells expressing HCNl channels. (Left) Sample current traces were recorded just before lidocaine was added to the recording bath (a), at steady-state inhibition (b), and following washout (c). Voltage protocol is shown in the inset. (Right) Time course of HCNl If currents recorded every 6 seconds. % If inhibition was determined by If réduction at steady-state inhibition relative to control, or the différence in If betweèn (a) and (b) in the time course. % If réduction (mean ± SEM): 6.6 ± 1.7 % (n = 3) at -110 mV; 8.3 ± 1.0 % (n = 7) at -100 mV; 17.1 ± 3.3 % (n = 7) at -85 mV.  31  A.-110mV 200 p/VI Lidocaine -Control - 200 uM Lidocaine Wash  500 pA <  -1750  >  +30 mV 450ms  E o  30  40  50  Time (sec)  500 ms  B.-100mV Control 200 uM Lidocaine  200 JJM Lidocaine  Wash  -2950  500 pA  -3000 "j£ -3050 g  -3100.  O ° -3150-  4 J5--3200 -3250-33000  r~—j  .  10  t  i  20  1  30  r"—T  40  r—)  50  1  1  .  60  [—  70  Time (sec) 500 ms  C. -85 mV 200 jiM Lidocaine Control 200 uM Lidocaine Wash 500 pA  10  500 ms  20  30  40  50  60  Time (sec)  32  Figure 3.2. Linear fit of the modified Woodhull équation shows that lidocaine inhibition is weakly voltage-dependent. Natural logarithms of the binding constant (K1/2) at -110, -100, and -85 mV were plotted against membrane potential, and fitted with a linear function. Error bars represent the SEM. The z-delta value (zô) derived from the négative slope value of the linear relationship suggests that lidocaine inhibition may be weakly voltage-dependent.  3.5-,  3.0  2.5-  2.0-1 SSÉT  _£  zô = 0.045 +/- 0.011  1.5-1 1.0-  0.5-  0.0  —I— -110  -105  -100  -95  —T— -90  -85  1 -80  mV  From current traces elicited by -110 mV for activation and +30 mV for deactivation, a greater proportion of current in the deactivation step appeared to be inhibited compared to the activation step. This current at +30 mV also appeared to be more stable during repeated activation/deactivation steps, and complète washout occurred in nearly every cell examined. Figure 3.3 shows représentative current traces and time courses demonstrating the ability for current to return completely at +30 mV but not at ail at -130 mV. The stability of the currents 33  and the reproducibility of thèse results at +30 mV led me to pursue experiments examining the effects of lidocaine at various micromolar concentrations. The percent If inhibition at +30 mV caused by 50, 100, 200, 400, 800, 1600, and 3200 uM lidocaine is shown in the concentrationresponse curve in Figure 3.3. The time for lidocaine inhibition to reach steady-state was fast (~6 seconds), and was dépendent on concentration. To evaluate the half maximal inhibitory concentration (IC50), the proportion of If unblocked was plotted versus lidocaine concentration on a logarithmic scale (Figure 3.3). The data points with the associated error bars were fitted with the Hill équation, yielding an IC50 value of 669 ± 320 uM (n = 3-7 cells), a Hill coefficient of 0.86 ±0.15, and a maximum proportion of current inhibited of 0.88 ±0.13 (mean ± SE). At the lowest lidocaine concentration tested of 50 uM, the percent If inhibited was 6.1 ± 2.0 % (mean ± SEM).  3.3  Lidocaine does not affect the voltage dependence of channel activation It is possible that the observed lidocaine inhibition was due to a négative shift of the  activation curve, leading to decreased number of channels activating at a spécifie potential. Thus, steady-state activation of HCNl channels in the présence and absence of lidocaine was examined. If activation curves fitted with a Boltzmann function (Figure 3.4) revealed that the midpoint of activation (Vi/2) was -85.7 ± 1.7 mV (mean ± SEM; n = 6 cells) in the absence of lidocaine and -88.9 ± 1.2 mV (n = 6 cells) in the présence of 200 uM lidocaine. Thèse values were not significantly différent. There was also no significant différence in the slope factor (k), 13.3 ± 0.9 mV and 12.3 ± 1.0 mV (mean ± SEM) for control and with 200 uM lidocaine, respectively. At -110 mV, both activation curves (control and with 200 uM lidocaine) showed that approximately 90% of the maximum number of channels were activated. Because there was no 34  change in the voltage range in which the channels open, lidocaine appeared to hâve no effect on the steady-state activation of HCN1.  35  Figure 3.3. Lidocaine inhibition of If elicited by the +30 mV deactivation step is concentrationdependent. (A) Trains of activating/deactivating steps were applied every 6 seconds to CHO cells expressing HCNl channels. Sample current traces were recorded just before lidocaine was added to the recording bath (a), at steady-state inhibition (b), and following washout (c). Voltage protocol is shown in the inset. (B) Current traces generated by the +30 mV step (enclosed in the red rectangle in panel A). (C) HCNl If current recordings shown in the time course were elicited by the +30 mV deactivating step. % If inhibition between (a) and (b): 23.7 ± 4.5 % (mean ± SEM; n = 7 cells) at +30 mV. (D) Proportion of unblocked If (I /10) plotted against lidocaine concentration fitted with the Hill équation. Data were obtained from % If inhibition at +30 mV deactivating step as described in the top panel. Error bars represent the SEM. The number of cells examined at each concentration is shown in parenthèses. Values generated from a nonlinear least-squares fit include IC50 (the half maximal inhibitory concentration of lidocaine), n (Hill coefficient), and proportion of If blocked at maximal inhibition. Ail three values are reported as mean ± SE.  36  100 pA  500 pA  +30 mV 450 ms  200 ms  -35 mV  -Control - 200 pM Lidocaine Wash  200 |iM Lidocaine  -130 mV 1000 ms g - 250.  E  500 ms  —, 20  .  , 40  .  1 60  .  1 80  1  , 100  1 "  t • 120  Time (sec)  D. 1.0-  (5) (5) " ^ • ^  0.8-  oo.e-  0.4-  (7)  "îs.^ *V At +30 mV: ICj,, = 669 +/- 320 MM n = 0.86 +/- 0.15  0) (4) Nnr -fc\ (3)  Max. inhibition = 0.88+/- 0.13 0.2100  1000  10000  Lidocaine oonœntration (MM)  37  Figure 3.4. Lidocaine does not affect the voltage-dependence of HCN1 channel activation. (Top) Activation curves fitted with the Boltzmann function from 6 cells not treated with lidocaine and 6 cells treated with 200 uM lidocaine. Error bars represent the SEM. The midpoints of activation {Vi/i) for control and lidocaine were -85.7 ± 1.7 mV and -88.9 ± 1.2 mV (mean ± SEM), respectively. There was no significant différence between the midpoints of activation of control and 200 uM lidocaine. (Unpaired Student's t-test; p = 0.1444) The slope factor (k) was 13.3 ± 0.9 and 12.3 ± 1.0 mV (mean ± SEM) for control and lidocaine, which were not significantly différent. (Unpaired Student's t-test; p = 0.4665) (Bottom) Représentative current traces used to generate the activation curves.  38  Control 200 p M Udocaine  1.0-  0.8-  0.6-  0.4-  0.2-  0.0—I— -150  0 -aP  = === =  -i— -130  ~l  •  -110  1  -90  -70  "i " » -50 -30  -10  Pre-pulse (mV)  ad  w — . .  55*>m»» «<•<•  1000 pA  gaa  ««*• mum mum HWHJMWIMI  '*•*«  H**w  V.  2000 ms  Control  2000 ms  200 uM Udocaine  39  3.4  Lidocaine does not affect the rates of activation and deactivation, but shortens the  delay prior to deactivation To examine how lidocaine affected the rates at which channels opened and closed, a double exponential function was used to describe activation currents and a single exponential function was used for deactivation currents. Figure 3.5 shows the double exponential function fitted curves for control and lidocaine activation currents. In both the control and lidocaine traces, an initial delay was observed prior to both activation and deactivation, which did not fit well with either a single or double exponential function and was excluded from the fitting procédure. A double exponential function fit described almost the entire activation process, whereas a single exponential function only fitted well with the initial -200 ms of the activation process (data not shown). Comparison of the slow and fast components of the double exponential fits on activation (Figure 3.6) indicated that there was no significant différence in the activation kinetics after the addition of lidocaine ranging from 50 uM to 3200 uM, with the exception of the slow component at 400 uM. The percent of current amplitude that fitted with the fast component was 69.9 ± 2.2 % and 68.4 ± 2.3 % (mean ± SEM) for control and various lidocaine concentrations combined, respectively. The delay during deactivation in the control was considerably longer compared to activation. However, with lidocaine treatment, the delay prior to start of fitting deactivation currents was shortened, as can be appreciated in Figure 3.7. Figure 3.8 shows that the addition of lidocaine ranging from 50 to 3200 uM did not significantly affect the deactivation time constants. The delay prior to the start of fitting shown in Figure 3.8 was also important to indicate that the delay was significantly reduced at ail concentrations with the exception of 200 uM. The discussion chapter addresses possible reasons for this delay and its significance in channel gating. 40  Figure 3.5. HCNl activation currents are well described by a double exponentiai fonction. (Top) Activation current trace for control. (Bottom) Activation current trace in the présence of 200 uM lidocaine. The black trace is the raw trace, the red trace is the fîtted curve, and the green trace represents the residuals of the fit (the différence between the raw trace and thefîttedcurve). With an initial delay, activation currents were fîtted well with a double exponentiai fonction. A single exponentiaifonctioncould only describe the initial -200 ms of the activation process (data not shown). The goodness of fit was determined by the residuals of the fit. o—*=1000pA  200 ms  lOOOpA  200 ms  41  Figure 3.6. Lidocaine does notaffect the rate ofactivation ofHCNl channels. Activation time constants were derived from double-exponential fits. Error bars and asterisks represent the SEM and p < 0.05, respectively. (Top) Comparison of the slow component of the fit between control and lidocaine treatment on the same cell showed no significant différence except at 400 uM. Student's paired t-test (two-tailed): 50 uM: p = 0.4768, n = 5; 100 uM: p = 0.9846, n = 5; 200 uM: p = 0.7880, n = 7; 400 uM: p = 0.0290, n = 3; 800 uM: p - 0.2533, n = 3; 1600 uM: p = 0.1060, n = 4; 3200 uM: p = 0.3696, n = 3. (Bottom) Comparison of the fast component of the fit between control and lidocaine treatment on the same cell showed no significant différence at ail concentrations. Student's paired t-test (two-tailed): 50 uM: p = 0.5944, n = 5; 100 uM: p = 0.5665, n = 5; 200 uM: p = 0.1843, n = 7; 400 uM: p = 0.0832, n = 3; 800 uM: p = 0.2592, n = 3; 1600 uM: p = 0.0986, n = 3; 3200 uM: p = 0.0801, n = 3.  Ctrl 50 (JM Ctrl 100 M M Ctrl 200 (JM Ctrl 400 (iM Ctrl 800 M M Ctrl 1600 \xM Ctrl 3200 M M I—'—!—'—i—•—i—'—i—'—i—•—(—'—i—'—i—'—i—'—i—'—i—'—r  0  200  400  600  800  Tau, slow  1000  1200  (ms) v '  Ctrl 50 \iM Ctrl 100MM  Ctrl 200 |aM Ctrl 400 \M Ctrl 800 M M Ctrl 1600 M M Ctrl 3200 \iM  H| NS " < — i — • — i — • — i — ' — r  20  40  60  80  100  Taufast (ms)  1 — r 140 120  160  43  Figure 3.7. Lidocaine shortens the delay prior to the start of single exponential fit on deactivation currents. Représentative current traces elicited by the +30 mV deactivation step for control (top) and following treatment with 3200 uM lidocaine (bottom). The black trace is the raw trace, the red trace is the fitted curve, and the green trace represents the residuals of the fit (the différence between the raw trace and the fitted curve). Fit of the respective traces with a single exponential function can only describe the deactivation currents well up to a certain point, as the initial 20 ms of the control and the initial 7 ms of the lidocaine trace cannot be described by a single exponential function. The différence in delay between control and lidocaine deactivation traces is quantified in Figure 3.8.  44  Delay prïor to start fit = 20 ms  100pA  ^'\tw\^M^^'Afi^  —  100 ms  Delay prior to start fit = 7 ms  50 pA — - _ _ i  100 ms  45  Figure 3.8. Lidocaine does not affect the rate of deactivation of HCNl channels, but affects the delay prior to the start of a single exponential fit. Error bars and asterisks represent the SEM and p < 0.05, respectively. (Top) Deactivation time constants derived from single exponential fits. Comparisons were made between values from control and lidocaine treatment on the same cell. Student's paired t-test (two-tailed): 50 uM: p = 0.0646, n = 5; 100 uM: p = 0.1158, n = 5; 200 uM: p = 0.5485, n = 7; 400 uM: p = 0.4468, n = 3; 800 uM: p = 0.8379, n = 3; 1600 uM: p = 0.6884, n = 4; 3200 uM: p - 0.2614, n = 3. (Bottom) Delay prior to the start of single exponential fit as described in Figure 3.7. The delay appeared to be removed upon addition of lidocaine at ail concentrations with the exception of 200 uM. Student's paired t-test (one-tailed): 50 uM: p = 0.0025, n = 5; 100 uM: p = 0.0201, n = 5; 200 uM: p = 0.0932, n = 7; 400 uM: p = 0.0023, n = 3; 800 uM: p = 0.0017, n = 3; 1600 uM: p = 0.0110, n = 4; 3200 uM: p = 0.0327, n =3.  46  Ctrl 50 uM  Ctrl 100 [iM Ctrl 200 M M Ctrl 400 |iM  ^] NS  Ctrl 800 |iM Ctrl 1600MM  Ctrl 3200 \iM  TaU  deact(mS)  Ctrl 50 )JM Ctrl 100 |iM  Ctrl 200 (JM Ctrl 400 jiM  Ctri 800 yxM Ctrl 1600 \iM Ctrl 3200 [xM 10  20  30  Delay prior to start fit (ms)  47  3.5  Extracellular application of the permanently charged derivative of lidocaine,  QX-314, does not inhibit HCN1 To détermine whether lidocaine inhibition, in its charged form, involved interaction with extracellular site(s), a protocol consisting of repeated activation/deactivation steps described above was performed with 500 uM QX-314 perfused into the recording chamber bath. The permanently charged derivative of lidocaine, QX-314, at a concentration of 500 uM, decreased If amplitude elicited by both the activation and deactivation steps (Figure 3.9). Currents recorded at +30 mV gradually reduced with each subséquent sweep following removal of QX-314, and did not show any current return. It is possible that this réduction in current was due to current rundown.  48  Figure 3.9. Extracellular application of 500 uM QX-314 reduces If elicited by the -110 mV activation step and the +30 mV deactivation step, without current recovery. Trains of activating/deactivating steps were applied every 6 seconds. If réduction was observed, but could not be distinguished from current rundown. (A) Sample current traces were recorded just before QX-314 was added to the recording bath (a), at steady-state inhibition (b), and foliowing washout (c). Voltage protocol is shown in the inset. (B) Current traces generated by the +30 mV step (enclosed in the red rectangle in panel A) (C) Chemical structure of QX-314. (D) Time course showing HCN1 If currents elicited by the +30 mV deactivating step.  L  A.  B.  500 pA 200 pA  +30 mV 450 ms -35 mV  -110mV  100 ms  800 ms  500 ms  D.  500 uM QX-314  310-  -Control -500 uM QX-314 Wash  300.  > o  + 270-  260-  —1— 10  - 1 —  20  -1— 40  —i— 50  70  Time (sec)  49  3.6  Properties of HCN1 mutants with single amino acid residue substitution Three HCN1 mutants were made each with a single amino acid residue substituted in the  pore-lining S6 région of the channel to détermine whether lidocaine inhibition involved interaction with thèse residues. The A372G and V379A mutants were made to test whether the aromatic group of lidocaine interacts with the suspected hydrophobic cavity formed by the side chains of thèse amino acid residues (Cheng et al., 2007). The F378A mutant was produced to examine whether lidocaine, in its charged form, interacts with n électrons from the aromatic ring in phenylalanine. Ail three mutants were functional, with currents recorded from activation and deactivation steps. F378A was chosen for detailed analysis with lidocaine, as it provided relatively stable recordings compared to A372G and V379A. Représentative current recordings of A372G and V379A can be found in the Appendix (Figures A2.1 and A2.2). Steady-state activation and rates of activation and deactivation were compared between wild-type and HCN1 F378A. Steady-state activation of F378A (Figure 3.10) showed that the midpoint of activation was -80.2 ± 2.7 mV (mean ± SEM; n = 4 cells), which was not significantly différent from that of wild-type (-85.7 ± 1.7 mV). However, the slope factor was 19.2 ± 1.1 mV, which was significantly différent from the value obtained from wild-type (13.3 ± 0.9 mV). At -130 mV, approximately 90% of the maximum number of channels was activated. The rates of activation and deactivation were significantly différent between wild-type and F378A (Figure 3.11). The rate of deactivation was significantly prolonged, as shown in représentative traces from wildtype and F378A (Figure 3.12). Activation currents in F378A appeared to reach steady-state within 1000 to 1500 ms, but deactivation currents required a minimum of 2000 ms to reach steady-state compared to 100 to 200 ms in wild-type. It is also interesting to note that the delay observed during the deactivation process in wild-type channels did not appear in F378A. 50  Figure 3.10. HCNl F378A has a midpoint of activation {Vi/i) similar to that for the wild-type channel, but has a significantly différent slope factor (k). (Top) Activation curve fitted with the Boltzmann functionfrom 4 cells expressing HCNl F378A. (Bottom) Représentative current traces used to generate the activation curve.  1.0-  0.8-  0.6V1/2 = -80.2 +/- 2.7 mV k = 19.2+/-1.1 mV  0.4-  0.2.  0.0.  -1—  -150  -1 -130  «  1— -110  -90  -70  -,  p_ -50  Pre-pulse (mV)  1000 pA  2000 ms  51  Figure 3.11. Activation and deactivation time constants are significantly différent between wild-type HCNl and HCNl F378A. Error bars and asterisks represent the SEM and p < 0.0001 (unpaired Student's t-test), respectively. (A, B) Comparison of slow and fast components of double exponential fit on activation between wild-type and F378A. (C) Comparison of deactivation time constants between wild-type and F378A. Number of cells is indicated in parenthèses.  F378A  52  Figure 3.12. HCNl F378A deactivates more slowly compared to wild-type. (Top) Deactivation currents elicited by +30 mV deactivation step reached steady-state within approximately 100 ms in wild-type. (Bottom) Deactivation currents elicited by +30 mV deactivation step did not reach steady-state within the same time scale as above. Full deactivation current trace is shown in the inset, with the portion enclosed in the red rectangle enlarged for comparison with wild-type. Note that the delay or "plateau" is présent in wild-type but absent in F378A.  HCN1 wild-type lOOpA  100 ms  200 pA  * > • * • " • <««>HiM.»l»'il t »il > «  100 ms  53  The proportion of If inhibited was then examined using -130 mV to activate the channel in trains of activation and deactivation steps repeated every 8 seconds. Similar to studies on the wild-type channel, recordings of If elicited by the +30 mV deactivation step were stable and the effect of lidocaine was réversible (Figure 3.13). The duration of the deactivation step was extended to 5000 ms to ensure that the steady-state was reached. Despite holding the cell at +30 mV for a prolonged period of time allowing up to 12 repeated puises, currents in most cells were stable enough for If to be accurately calculated. Similar to the wild-type channel, If recorded at +30 mV was plotted over time to show the effects of lidocaine and washout (Figure 3.13). Various lidocaine concentrations were tested on F378A, and the proportion of If inhibited increased with increasing lidocaine concentration (Figure 3.14). Compared to data for wild-type, a greater proportion of If was inhibited at lower concentrations tested (< 200 uM) in F378A. The data were not well described by the Hill équation, and were not presented in a concentrationresponse curve. Data from the range of concentrations tested might hâve formed the steepest portion of the Hill plot without the top and bottom plateaus. There was a significant différence in the rates of activation by 800 uM lidocaine for both the slow and fast components of the double exponential function fit, with 400 uM lidocaine also affecting the fast component (Figure 3.15). The deactivation time constant was significantly différent with lidocaine application at concentrations ranging from 10 to 800 uM (Figure 3.16). It is important to note that both decrease in If amplitude and différence in deactivation time constant were attained with 10 uM and 20 uM lidocaine.  54  Figure 3.13. Lidocaine inhibition of If at +30 mV is fast and réversible in HCNl F378A. Trains of activating/deactivating steps were applied every 8 seconds to CHO cells expressing HCNl F378A. (A) Sample current traces were recorded just before lidocaine was added to the recording bath (a), at steady-state inhibition (b), and following washout (c). Voltage protocol is shown in the inset. (B) Current traces generated by the +30 mV deactivation step (enclosed in the red rectangle in panel A). (C) Time course of HCNl If currents recorded every 8 seconds. % h- inhibition between (a) and (b) (mean ± SEM): 26.6 ± 1.4 % (n - 4) at +30 mV by 200 uM lidocaine.  55  B,  A.  2000 pA  200  pA  +30 mV 5000ms  2000 ms  1000 ms  c. IIOOT  1050-  _  > E o  1000-  950  -Contro! 200 |JM Lidkxaine Wash  rri +  900  850-  800  -T— 10  —I— 20  -T— 30  40  —1— 50  -1— 60  —I— 70  —T— 80  '90  Time (sec)  56  Figure 3.14. Lidocaine inhibition of If at +30 mV increases with lidocaine concentration in HCNl F378A. Data reported as mean ± SEM (n = 4 - 8 cells). One-way ANOVA p < 0.0001. * p < 0 . 0 5 . * * p < 0 . 0 1 . ***p<0.001.  * * * ** *  50-  40>  o m  +  30-  <3>  c O V;  20-  JE c *—  ^ o^  10-  0-  10 |jM  20 |JIV\  50 M M  200 ^M 400 pM 800 [iM  57  Figure 3.15. Lidocaine does not affect the rate of activation of HCN1 F378A at concentrations below 400 uM. Activation time constants were derived from double-exponential fits. Error bars and asterisks represent the SEM and p < 0.05, respectively. (Top) Comparison of the slow component of the fit between control and lidocaine treatment on the same cell showed no significant différence except at 800 uM. Student's paired t-test (two-tailed): 10 uM: p = 0.0567, n = 4; 20 u.M: p = 0.7530, n = 8; 50 uM: p = 0.1716, n = 6; 200 uM: p = 0.3640, n = 4; 400 uM: p = 0.1055, n = 8; 800 uM: p = 0.0022, n = 7. (Bottom) Comparison of the fast component of the fit between control and lidocaine treatment on the same cell showed no significant différence except at 400 uM and 800 uM. Student's paired t-test (two-tailed): 10 uM: p = 0.0619, n = 4; 20 uM: p = 0.0520, n = 8; 50 uM: p = 0.5786, n = 6; 200 uM: p = 0.0514, n = 4; 400 uM: p = 0.0326, n = 8; 800 uM: p = 0.0021, n = 7.  58  Ctrl 200 |iM  3 NS  Ctrl 400 [iM Ctrl 800 MM 100  200  300  400  500  Tau. slow  600  700  800  900  (ms) v  '  Ctrl 200 uM Ctrl 400 |iM Ctr 800 (JM  Tauf fast  fms)  x  '  59  Figure 3.16. Lidocaine affects the rate ofdeactivation of HCN1 F378A. Deactivation time constants derived from single exponential fits. Comparisons were made between the respective control and lidocaine treatment on the same cell. Data reported as mean ± SEM. Student's paired t-test (two-tailed): 10 ^M: p = 0.0202, n = 4; 20 ^M: p = 0.0230, n = 8; 50 u.M: p = 0.0495, n = 6; 200 uM: p = 0.0476, n = 4; 400 uM: p = 0.0427, n = 8; 800 jiM: p = 0.0010, n = 7.  Ctrl 10 uM Ctr 20 pM Ctrl 50 \xM  Ctrl 200 uM Ctr 400 uM Ctr 800 pM , 800 TaU  r  1000  dSac.(mS'  60  Chapter 4: Discussion  This chapter first provides a summary of my project, followed by a discussion of the findings in the previous chapter and work not presented, but used to optimize expérimental conditions. Also, the significance of my project and future directions are discussed.  4.1  Summary of the study The goal of my thesis is to détermine whether lidocaine, in micromolar concentrations,  inhibits the cloned HCN1 channel from the intracellular side. Spécifie aims were posed to examine various aspects of lidocaine inhibition. The first was to détermine whether lidocaine inhibition was voltage-dependent and/or coricentration-dependent. Increasing lidocaine concentrations exhibited greater percent of If inhibited at +30 mV, as represented by the concentration-response curve (Figure 3.3). Lidocaine inhibition appeared to be weakly voltagedependent at hyperpolarized voltages in which more than half of the maximal number of channels is known to activate (Figure 3.2). The next objective sought whether activation and deactivation kinetics were altered in the présence of lidocaine. Parameters used for évaluation included voltage-dependence of activation and time constants of activation and deactivation. Activation curves comparing midpoints of activation and slope factors between control cells and cells treated with lidocaine revealed no significant différence in either value (Figure 3.4). The activation curves also showed that at least 90% of the maximal number of channels was activated at -110 mV (Figure 3.4). Activation currents were well described by a double exponential function compared to a single exponential function, whereas deactivation currents were fitted well by a single exponential function. Addition of lidocaine in concentrations used in the 61  concentration-response curve did not significantly affect the time constants for both activation and deactivation. A prolonged delay or "plateau" prior to the start of single exponential function fit of deactivation currents was observed, and was shortened or eliminated in current traces recorded with lidocaine présent. Extracellular application of the permanently charged derivative of lidocaine, QX-314, at a concentration of 500 uM, decreased If amplitude, but could not be distinguished from current rundown (Figure 3.9). The last objective examined whether amino acid residues in the pore-lining S6 région were responsible for lidocaine interaction and inhibition. Three HCN1 mutants were generated, and were functional channels. F378A was selected for further analysis, as it provided relatively stable recordings during repeated activation/deactivation steps compared to A372G and V379A. Increasing lidocaine concentrations showed greater inhibition of If in F378A at +30 mV (Figure 3.14). Lidocaine, at concentrations ranging from 10 to 800 uM, increased the time constant of deactivation compared to control (Figure 3.16). The next section attempts to interpret findings described above, to address strengths and weaknesses with the study, and to relate findings in the literature.  4.2  Discussion of the findings To test my hypothesis, I first performed experiments to examine whether lidocaine, at  micromolar concentrations, inhibited currents from cloned HCN1 channels. Using trains of activation and deactivation steps, If elicited by the activation step was reduced upon addition of lidocaine. However, If started to decrease even before lidocaine was applied to the recording chamber, and steady-state inhibition could not be confidently determined. Also, in most cells, If continued to decrease after lidocaine was removed from the recording chamber, and showed no signs of current stabilizing or washout. The inability of currents returning to the control level  may be attributed to a phenomenon called current rundown, and/or the inability for lidocaine to washout completely. Current rundown might hâve interfered with data interprétation and masked the effect of lidocaine. During prolonged dialysis in whole-cell recordings with HCN channels, rundown may occur (DiFrancesco et al., 1986; Pian et al., 2006). During current rundown, the voltage range in which the channels open may gradually shift to more hyperpolarized potentials and the maximum conductance at a certain potential may decrease over time. Rundown of If was also observed in single, dissociated sinoatrial node cells in whole-cell patch-clamp configuration (BoSmith et al., 1993). The exact mechanism for rundown is not known, but may be related to the loss of intracellular contents (Pian et al., 2006). Thus, once the cell membrane enclosed by the patch pipette is ruptured during a whole-cell recording, currents start to decrease over time. In our study, we only chose cells that yielded partial or complète return in current amplitude following the removal of lidocaine, which would not occur if the initial decrease was due only to current rundown. Initial experiments were performed with lidocaine perfusion from the side of the recording chamber, and not directly on the cell under study. The intracellular solution in the patch pipette did not include 2 mM creatine phosphate, 2 mM Mg-ATP, 1 mM cAMP, and 20 u.M PIP2. Figure Al .1 in the Appendix shows représentative current traces and time course from thèse initial experiments. Inhibition of lidocaine could not be distinguished from the linear decay in current shown in the time course. Also, each experiment required approximately 2.5 minutes and as many as 13 sweeps of activation and deactivation steps for steady-state inhibition to become apparent. Although current rundown may hâve attributed to the réduction in current amplitude, it was apparent that lidocaine inhibition took place, with the observation of current 63  return to the level prior to lidocaine addition. If current réduction was solely due to current rundown, currents recorded after lidocaine removal from the recording chamber would not hâve returned partially or completely to the level prior to lidocaine addition. Possible solutions to ensure accurate quantification of lidocaine inhibition included performing perforated patch-clamp experiments, altering the constituents of the intracellular solution in the patch pipette, and choosing only cells in which washout was complète or close to complète. The latter two options were incorporated into the modifications to the expérimental conditions, in order to minimize the effects of current rundown leading to stronger, more supportive data addressing the spécifie aims and hypothesis. In addition, a fast-step perfusion setup was installed to deliver lidocaine directly onto the cell under study. Using the fast-step perfusion setup, the cell was exposed to the desired concentration of lidocaine in approximately 20 ms, so that steady-state inhibition of lidocaine could be observed in less time. The duration for each experiment was shortened, which prevented prolonged dialysis of the cell and minimized the effects of current rundown. Constituents that are known to produce a depolarizing shift in channel activation, as described in the Materials and Methods section, were added to the intracellular solution in the patch pipette, which included creatine phosphate, Mg-ATP, cAMP, and PIP2. Also, only cells in which currents returned at least 20 % to the level prior to lidocaine addition were used for analysis. Thèse methodological adjustments vastly improved the quality of data, providing more accurate quantification of lidocaine inhibition. Voltage-dependence of activation was examined to observe whether there was a shift in the midpoint of activation caused by lidocaine. To ensure that any shift in the midpoint of activation was due to lidocaine and not current rundown, one group of cells was designated as control and another was designated for lidocaine treatment. A lidocaine concentration of 200 64  uM was used, as initial experiments prior to the use of the fast-step perfusion setup showed the strongest évidence of lidocaine inhibition at this concentration. Figure 3.4 shows that there was no significant différence in either the midpoint of activation or the slope factor, which suggests that lidocaine did not affect the function of the voltage sensing régions (S1-S4) of the channel. Because there was no significant shift in the midpoint of activation, If réduction was most likely due to a decrease in maximal current amplitude. Thèse results led to experiments testing If réduction at various hyperpolarizing potentials at which the channels open. To test whether lidocaine inhibition was voltage-dependent, trains of activation and deactivation steps were applied in the présence of 200 ixM lidocaine, with If determined from activation current traces elicited by -110 mV, -100 mV, and -85 mV. Steady-state inhibition was reached within one to three sweeps (6 to 18 seconds), which suggests that lidocaine bound to open channels very quickly and dissociated slowly from closed channels according to observations in Na+ channels (Arthur and Strichartz, 1987). Even though lidocaine inhibition was clearly observed, current return was incomplète in most cells because of current rundown or incomplète washout. Also, If réduction was less than 20 % over the range of potentials tested. However, greater lidocaine inhibition at less hyperpolarized potentials may indicate that lidocaine, in its charged form, blocked from the inside of the channel. At less hyperpolarized potentials, lidocaine, with a positive charge, was more likely to be driven to the pore of the channel, as the intracellular environment was less negatively charged. Figure 3.2 shows a zô value of 0.045. In a study examining the block of TEA+ on potassium channels, the authors concluded that a zô value of 0.04 was nearly voltage-independent (Heginbotham and MacKinnon, 1992).  65  Moreover, it also appeared from the current traces that there was greater If réduction at the +30 mV deactivation step. If recorded at +30 mV was more stable, which led to better reproducibility of results and more accurate quantification of lidocaine inhibition at lower concentrations. Complète washout also occurred in more cells. The concentration-response relationship in Figure 3.3 showed the half maximal inhibitory concentration to be approximately 669 uM. This value indicated that lidocaine had a weaker effect on HCN1 compared to cloned rabbit HCN4, Ih in rat small DRG neurons, and If in rabbit sinoatrial myocytes, which had reported IC5o of 276 uM, 99 uM, and 38 uM respectively (Bischoff et al., 2003; Rocchetti et al., 1999; Tamura et al., 2009). Différences in IC50 may be due to différent isoform distribution from one tissue to another, différent response to lidocaine between isoforms, différent voltage protocols applied, and possible current rundown that was not accounted for in thèse studies. Also, in studies on HCN4 and rabbit sinoatrial myocytes, experiments were conducted at 36°C, which may hâve altered the extent of lidocaine inhibition as well as the onset of lidocaine action possibly by modifying the pKa of lidocaine (Butterworth and Strichartz, 1990). Recordings in both studies were performed in Tyrode solution or low K+ solution, in contrast to high K+ extracellular solution used in this study. This différence in recording solution composition may explain the greater effect of lidocaine in thèse studies, as decreased current amplitude or inward ion flux may provide less résistance for lidocaine molécules to reach the binding site in the channel pore. Such dependence on current has been recently described for inhibition by ivabradine which is also thought to bind on the inner portion of HCN channels (Bucchi et al., 2002; Bucchi et al., 2006). Furthermore, the relatively low IC50 value obtained in the rabbit sinoatrial myocytes study may be due to native If channels consisting of différent isoforms forming heteromers with or without accessory P-subunit(s). Comparing to results in 66  this study, it is likely that lidocaine has a greater affinity for heteromers or homomers consisting of other HCN isoforms. Other parameters generated from the concentration-response relationship include the Hill coefficient and the maximal proportion of If inhibited. The Hill coefficient was approximately one, which indicates that lidocaine has a single binding site or receptor. This is in accordance to the single receptor for local anesthetic binding in Na+ channels (Hille, 1977a). The maximal proportion of If inhibited was also approximately one, but it remains unclear whether lidocaine can inhibit If completely. To improve data generated from the concentration-response relationship, in particular to reduce the standard error associated with the IC50 value, data can be collected at lower concentrations and at a high concentration where actual maximum inhibition is observed. In wild-type HCN1, lidocaine did not affect the rates of channel activation and deactivation. The time constants of activation and deactivation for the control cells, however, varied considerably (Figures 3.6 and 3.8). For example, the activation time constant describing the fast and slow components of the double exponential fonction fit ranged from approximately 30 to 60 ms and 150 to 300 ms, respectively. The deactivation time constant ranged from approximately 25 to 60 ms. This variability could arise from cell-to-cell variability as well as errors in the fitting procédure. The standard errors of the mean were quite large in some test groups. This problem might hâve resulted in falsely determining that 400 uM lidocaine significantly affected the slow component of the activation time constant. However, the activation and deactivation time constants shown in figures 3.6 and 3.8 were comparable to previous studies describing the kinetics of HCN1 channels (Ishii et al., 2007; Moosmang et al., 2001 ; Santoro et al., 1998; Ulens and Tytgat, 2001). The fast component of the double exponential fonction fit on activation accounted for approximately 70 % of current amplitude,  which was in accordance with one of the initial studies characterizing HCN1 (Moosmang et al., 2001). From thèse findings, lidocaine appeared not to hâve disrupted the process of channel opening and closing, but affected the delay prior to the start of single exponential function fit of channel deactivation. Similar to the problem with determining activation and deactivation time constants, there was variability in the control data ranging from approximately 22 to 37 ms for the delay. Thèse findings provide information on how HCN1 channels activate and deactivate. A four-state allosteric model (Figure 4.1 ) has been described for HCN2 activation and deactivation kinetics involving voltage-dependent (CR to CA, OA to OR) and voltage-independent (CA to 0 A , OR to CR) transitions (Chen et al., 2007). Although this gating scheme appears to be quite simplified, it was shown to fit the opening and closing kinetics of HCN2 relatively well. In HCN2, both activation and deactivation could be fitted well with a single exponential function. Also, the delay or "plateau" was présent upon steps to depolarized potentials, as it was more energetically favourable for deactivation to proceed through two sequential open states (OA to OR) to reach the closed state. A similar model can be used to describe HCN1 channel gating, with the exception that activation may require an additional closed state to represent the double exponential function used to fit activation currents. The deactivation process appeared to be similar to HCN2, as the delay or "plateau" was also observed in deactivation of HCN1. However, with the addition of lidocaine, it appears that the delay was reduced or, in some cases, abolished by lidocaine. The absence of this delay suggests that the channel did not undergo the transition from OA to OR. From thèse observations, lidocaine might hâve blocked the OR state, preventing deactivation from proceeding through sequential open states. Assuming that the proposed model for HCN1 channel gating is accurate, channel deactivation would hâve to 68  undergo the reverse direction of activation. In response to the deactivation step, channels would proceed from the OR state to an intermediate closed state, which might hâve resulted in the absence of the delay (Figure 3.7).  Figure 4.1. Four-state cyclic allosteric model for régulation of HCN2 opening and closing by voltage. CR and CA represent closed states with voltage sensors in resting and activated conformations, respectively. OR and 0 A represent open states with voltage sensors in resting and activated conformations, respectively. a and J3 represent voltage-dependent rate constants for activation and deactivation for the closed state, respectively. a' and p' represent voltagedependent rate constants for activation and deactivation for the open state, respectively. g and h represent voltage-independent opening and closing rates respectively for the activated state. g' and h ' represent voltage-independent opening and closing rates respectively for the resting state. (Chen et al, 2007), used with permission.  a(V)  69  To détermine whether lidocaine inhibition, in its charged form, involved extracellular site(s), the permanently charged derivative of lidocaine, QX-314, was applied extracellularly at a concentration of 500 uM. An immédiate réduction in If was observed with the addition of QX-314 in panel D in Figure 3.9, but this current réduction could not be determined to be inhibition by QX-314 without any clear indications of current return. Currents did not return either because QX-314 did not washout, or that the observed effect was solely due to current rundown. Current rundown appeared to be greater at -110 mV than at +30 mV comparing panels A and B in Figure 3.9. However, when considering current réduction as a proportion of total current for both inward current at -110 mV and outward current at +30 mV, the proportion of current reduced due to current rundown in both cases are similar. In addition, current return probably should not be expected in the first place, as studies examining effects of intracellular QX-314 showed >90 % If inhibition at approximately 150 and 40 seconds after establishing whole-cell configuration without indications of current return (Kilb and Luhmann, 2000; Perkins and Wong, 1995). QX-314 was présent in the bath solution for only three sweeps or approximately 20 seconds, to ensure consistency between this experiment and those with lidocaine. Further experiments using millimolar concentrations of QX-314 placed in the recording chamber bath for durations longer than 150 seconds might be needed to provide better indication as to whether or not extracellular QX-314 inhibits If. Three mutants with single amino acid residue substitutions in the pore-lining S6 région were produced to détermine whether lidocaine inhibition involved interaction with thèse residues. F378A was chosen for detailed analysis with lidocaine, as it provided relatively stable recordings during repeated activation/deactivation steps compared to A372G and V379A. Voltage dependence of activation in F378A was similar to wild-type, but activation and deactivation were 70  considerably slower (Figure 3.11). This différence in activation and deactivation kinetics may indicate that the side chain of phenylalanine could contribute to channel gating. Moreover, the delay prior to deactivation was not observed in F378A (Figure 3.12), suggesting that deactivation did not undergo the OA to OR transition in the proposed gating model. There was a similar trend in If réduction at +30 mV (Figure 3.14) compared to wild-type, with a greater proportion of If inhibited at concentrations less than 200 uM. If the phenylalanine residue were the sole déterminant of lidocaine inhibition or interaction, substituting this residue with another residue without an aromatic ring in the side chain would theoretically remove lidocaine inhibition. With the proportion of If inhibited similar to wild-type at various micromolar concentrations and greater at lower concentrations, the phenylalanine residue is not responsible for lidocaine inhibition, and substitution to an alanine residue might hâve modified or enhanced lidocaine inhibition. Furthermore, with lidocaine inhibition observed in F378A, lidocaine could still inhibit the channel in the absence of the OR state. Thus, the OR state would not be responsible for lidocaine inhibition, contrary to spéculations derived from wild-type data. Because lidocaine inhibition required channels in open state, lidocaine would most likely hâve greater affinity for the OA state. This is supported by the fact that greater inhibition was observed at 20 uM lidocaine in F378A compared to 50 uM in wild-type. The increase in block at lower lidocaine concentrations in F378A compared to wild-type can be due to the OA state being available for longer duration due to prolonged deactivation.  4.3  Signifîcance Results in this study suggest that the actions of lidocaine on If may be important for its  effects as an antiarrhythmic drug and a local anesthetic. With the therapeutic concentration of 71  approximately 20 uM when administered intravenously as an antiarrhythmic drug (Burashnikov et al., 2007), it is obvious that lidocaine's antiarrhythmic effect does not require a concentration close to the IC50 value of 669 uM obtained in this study. Although the IC50 value is much higher than in the study examining inhibition of If in rabbit sinoatrial myocytes (38 uM) (Rocchetti et al., 1999), it is possible that a small percent If réduction, such as a 6.1% decrease in If with 50 uM lidocaine in this study, is sufficient to produce significant effects. However, the high IC50 value may indicate that the functional contribution to If by HCN1 may be less compared to other HCN isoforms such as HCN2 and HCN4. As a local anesthetic, lidocaine has been shown to block propagation of action potentials in peripheral nerves by inhibiting Na+ channels (Butterworth and Strichartz, 1990). HCN 1-mediated currents hâve been shown to be responsible for Ih in rat large and médium fast-conducting sensory neurons, and élimination of HCN1mediated currents can lead to réduction in nociception (Momin et al., 2008). It is likely that lidocaine, at a concentration similar to the IC50 obtained in this study, can lead to réduction in nociception by inhibiting Ih. Lidocaine can potentially affect depolarization following the initial phase of afterhyperpolarization, leading to prolongation of the interval between action potentials in DRG neurons similar to observations on clonidine inhibition of Ih (Yagi and Sumino, 1998) (Figure 4.2). Although inhibiting Ih does not prevent génération of action potentials, slowing action potential propagation can be important physiologically in reducing pain.  72  Figure 4.2. Réduction in Ih can affect depolarization after the initial phase of afterhyperpolarization in action potentials in DRG neurons. Clonidine, an agent known to produce antinociceptive effects, inhibits Ih in rat DRG neurons and prolongs the interval between action potentials. (Modified from Yagi and Sumino, 1998, used with permission.)  control . clonidine /  20 mV 2 ms Comparison of lidocaine inhibition between Na+ and HCN channels shows both similarities and différences. In our study, lidocaine inhibits If from HCN1 at +30 mV with IC50 of 669 uM, which is greater than 15 to 30 uM lidocaine used to produce significant inhibition on cardiac Na+ channels (Bean et al., 1983; Grant et al., 1980; Hondeghem and Katzung, 1977) and 100 to 250 uM lidocaine used to inhibit Na+ channels in nerves (Hille, 1977b). At a holding potential of-120 mV, approximately 10 % of Na+ current is inhibited with 50 uM lidocaine (Bean et al., 1983), which corresponds to a similar réduction in If with 50 uM lidocaine. Based on findings in this study, it appears that lidocaine inhibition on HCN1, in its charged form, occurs from the intracellular side, which has been shown to be the mechanism in Na+ channel inhibition. The phenylalanine residue in the pore-lining région is responsible for lidocaine 73  inhibition in Na channels, but not in HCN1. The similarities in lidocaine inhibition between Na+ and HCN channels suggest that the drug binding site could be located in the channel pore. The différences in inhibition between the two channels may be due to the fact that différent gating mechanisms can greatly affect the mechanism and the rate by which lidocaine inhibits the channel. The speed of drug onset can be affected by the fact that HCN channels require more time to activate and do not undergo inactivation. Higher concentrations of local anesthetics are needed to block K+ channels compared to Na+ channels (Arthur and Strichartz, 1987). With HCN channels being more structurally similar to K+ channels compared to Na+ channels, higher concentrations of local anesthetics may also be needed to block HCN channels.  4.4  Directions of future study This study has confirmed lidocaine inhibition on HCN1 channels, with molecular  mechanism of inhibition yet to be determined. An alternate approach to examine whether % électrons of the phenylalanine at position 378 interact with lidocaine is to substitute the aromatic ring of phenylalanine at position 378 to a cyclohexane, thereby removing the 71 électrons without altering the side chain. To further elucidate the molecular mechanism by which lidocaine inhibited the channel, site-directed mutagenesis could be performed on other residues in the S6 région speculated to be involved in lidocaine inhibition. Tryptophan at position 362 and tyrosine at position 375 are possible candidates, as both amino acid residues possess an aromatic ring in which K électrons may participate in lidocaine interaction. Moreover, extracellular QX-314 experiments can be improved by placing QX-314 in the recording bath for a longer duration. To confirm whether lidocaine inhibition on If is from the intracellular side, QX-314 and lidocaine can be applied to inside-out patches. The reasons for the shortened delay in deactivation after 74  lidocaine addition, as well as the prolonged deactivation in F378A, are still unclear. Using results from this study, computations can be performed to construct an allosteric model for HCN1 channel gating. To examine the relative functional contribution of each isoform to native If, protocols used in this study can be applied to examine the potencies of lidocaine on other cloned HCN isoforms such as HCN2 and HCN4, which are predominantly expressed in the heart as well as in rat small DRG neurons. Lidocaine has been shown to inhibit Ih in rat small DRG neurons, but effects on large DRG neurons hâve not been examined. Because HCNl-mediated currents hâve been shown to contribute to Ih in rat large and médium fast-conducting DRG neurons, lidocaine can be applied to rat large DRG neurons to confirm whether HCN1 currents is the sole déterminant of Ih by comparing IC5o to this study. To test the effects of lidocaine at various membrane potentials, voltage protocols can be applied to simulate an action potential. This can détermine inhibition by lidocaine on a spécifie isoform in physiological conditions. Lastly, inhibition of more potent amide-type local anesthetics such as bupivacaine and mepivacaine on HCN isoforms can be explored, ultimately determining whether thèse agents relieve pain in part by inhibiting Ih.  75  Références Accili, E.A., Proenza, C , Baruscotti, M., and DiFrancesco, D. (2002). From funny current to HCN channels: 20 years of excitation. News Physiol Sci 17, 32-37. Ahern, C.A., Eastwood, A.L., Dougherty, D.A., and Horn, R. (2008). Electrostatic contributions of aromatic residues in the local anesthetic receptor of voltage-gated sodium channels. Circulation Research 102, 86-94. Altomare, C , Bucchi, A., Camatini, E., Baruscotti, M., Viscomi, C , Moroni, A., and DiFrancesco, D. (2001). Integrated allosteric model of voltage gating of HCN channels. The Journal of General Physiology 117, 519-532. Arthur, G.R., and Strichartz, G.R. (1987). Local anesthetics (Berlin; New York, Springer-Verlag). Barrow, A.J., and Wu, S.M. (2009). Low-conductance HCN1 ion channels augment the frequency response of rod and cône photoreceptors. J Neurosci 29, 5841-5853. Baruscotti, M., Bucchi, A., and Difrancesco, D. (2005). Physiology and pharmacology of the cardiac pacemaker ("funny") current. Pharmacology & Therapeutics 107, 59-79. Bean, B.P., Cohen, C.J., and Tsien, R.W. (1983). Lidocaine block of cardiac sodium channels. The Journal of General Physiology 81, 613-642. Biel, M., Schneider, A., and Wahl, C. (2002). Cardiac HCN channels: structure, function, and modulation. Trends in Cardiovascular Medicine 12, 206-212. Bischoff, U., Brau, M.E., Vogel, W., Hempelmann, G., and Olschewski, A. (2003). Local anaesthetics block hyperpolarization-activated inward current in rat small dorsal root ganglion neurones. British Journal of Pharmacology 139, 1273-1280. Bois, P., Bescond, J., Renaudon, B., and Lenfant, J. (1996). Mode of action of bradycardic agent, S 16257, on ionic currents of rabbit sinoatrial node cells. British Journal of Pharmacology 118, 1051-1057. BoSmith, R.E., Briggs, I., and Sturgess, N.C. (1993). Inhibitory actions of ZENECA ZD7288 on whole-cell hyperpolarization activated inward current (If) in guinea-pig dissociated sinoatrial node cells. British Journal of Pharmacology 110, 343-349. Bucchi, A., Baruscotti, M., and DiFrancesco, D. (2002). Current-dependent block of rabbit sinoatrial node If channels by ivabradine. The Journal of General Physiology 120, 1-13. Bucchi, A., Tognati, A., Milanesi, R., Baruscotti, M., and DiFrancesco, D. (2006). Properties of ivabradine-induced block of HCN1 and HCN4 pacemaker channels. The Journal of Physiology 572, 335-346. 76  Burashnikov, A., Di Diego, J.M., Zygmunt, A.C., Belardinelli, L., and Antzelevitch, C. (2007). Atrium-selective sodium channel block as a strategy for suppression of atrial fibrillation: différences in sodium channel inactivation between atria and ventricles and the rôle of ranolazine. Circulation 116, 1449-1457. Butterworth, J.F.t, and Strichartz, G.R. (1990). Molecular mechanisms of local anesthesia: a review. Anesthesiology 72, 711-734. Camm, A.J., and Lau, C.P. (2003). Electrophysiological effects of a single intravenous administration of ivabradine (S 16257) in adult patients with normal baseline electrophysiology. Drugs in R&D 4, 83-89. Chan, C.S., Shigemoto, R., Mercer, J.N., and Surmeier, D.J. (2004). HCN2 and HCN1 channels govern the regularity of autonomous pacemaking and synaptic resetting in globus pallidus neurons. J Neurosci 24, 9921-9932. Chen, S., Wang, J., Zhou, L., George, M.S., and Siegelbaum, S.A. (2007). Voltage sensor movement and cAMP binding allosterically regulate an inherently voltage-independent closedopen transition in HCN channels. The Journal of General Physiology 729, 175-188. Cheng, L., Kinard, K., Rajamani, R., and Sanguinetti, M.C. (2007). Molecular mapping of the binding site for a blocker of hyperpolarization-activated, cyclic nucleotide-modulated pacemaker channels. The Journal of Pharmacology and Expérimental Therapeutics 322, 931-939. Demontis, G.C., Moroni, A., Gravante, B., Altomare, C., Longoni, B., Cervetto, L., and DiFrancesco, D. (2002). Functional characterisation and subcellular localisation of HCN1 channels in rabbit retinal rod photoreceptors. The Journal of Physiology 542, 89-97. DiFrancesco, D. (1993). Pacemaker mechanisms in cardiac tissue. Annual Review of Physiology 55, 455-472. DiFrancesco, D., and Borer, J.S. (2007). The funny current: cellular basis for the control of heart rate. Drugs 67 Suppl 2, 15-24. DiFrancesco, D., and Camm, J.A. (2004). Heart rate lowering by spécifie and sélective If current inhibition with ivabradine: a new therapeutic perspective in cardiovascular disease. Drugs 64, 1757-1765. DiFrancesco, D., Ferroni, A., Mazzanti, M., and Tromba, C. (1986). Properties of the hyperpolarizing-activated current If in cells isolated from the rabbit sino-atrial node. The Journal of Physiology 377, 61-88.  Galer, B.S., Rowbotham, M.C., Perander, J., and Friedman, E. (1999). Topical lidocaine patch relieves postherpetic neuralgia more effectively than a vehicle topical patch: results of an enriched enrollment study. Pain 80, 533-538. Gasparini, S., and DiFrancesco, D. (1997). Action of the hyperpolarization-activated current (Ih) blocker ZD 7288 in hippocampal CA1 neurons. Pflugers Arch 435, 99-106. Gauss, R., Seifert, R., and Kaupp, U.B. (1998). Molecular identification of a hyperpolarizationactivated channel in sea urchin sperm. Nature 393, 583-587. Goodman, L.S., Hardman, J.G., Limbird, L.E., and Gilman, A.G. (2001). Goodman and Gilman's the pharmacological basis of therapeutics, lOth edn (New York, McGraw-Hill, Médical Pub. Division). Grafe, P., Quasthoff, S., Grosskreutz, J., and Alzheimer, C. (1997). Function of the hyperpolarization-activated inward rectification in nonmyelinated peripheral rat and human axons. Journal of Neurophysiology 77, 421-426. Grant, A.O., Strauss, L.J., Wallace, A.G., and Strauss, H.C. (1980). The influence of pH on the electrophysiological effects of lidocaine in guinea pig ventricular myocardium. Circulation Research 47, 542-550. Harris, N.C., and Constanti, A. (1995). Mechanism of block by ZD 7288 of the hyperpolarization-activated inward rectifying current in guinea pig substantia nigra neurons in vitro. Journal of Neurophysiology 74, 2366-2378. Harrison, D.C., and Collinsworth, K.A. (1974). Antiarrhythmic actions of lidocaine. Annual Review of Medicine 25, 143-148. Heavner, J.E. (2007). Local anesthetics. Current Opinion in Anaesthesiology 20, 336-342. Heginbotham, L., and Kutluay, E. (2004). Revisiting voltage-dependent relief of block in ion channels: a mechanism independent of punchthrough. Biophysical Journal 86, 3663-3670. Heginbotham, L., and MacKinnon, R. (1992). The aromatic binding site for tetraethylammonium ion on potassium channels. Neuron 8, 483-491. Hille, B. (1977a). Local anesthetics: hydrophilic and hydrophobic pathways for the drug-receptor reaction. The Journal of General Physiology 69, 497-515. Hille, B. (1977b). The pH-dependent rate of action of local anesthetics on the node of Ranvier. The Journal of General Physiology 69, 475-496. Hille, B. (2001). Ion channels of excitable membranes, 3rd edn (Sunderland, Mass., Sinauer). 78  Hondeghem, L.M., and Katzung, B.G. (1977). Time- and voltage-dependent interactions of antiarrhythmic drugs with cardiac sodium channels. Biochimica et Biophysica Acta 472, 373398. Ishii, T.M., Nakashima, N., Takatsuka, K., and Ohmori, H. (2007). Peripheral N- and C-terminal domains détermine deactivation kinetics of HCN channels. Biochemical and Biophysical Research Communications 359, 592-598. Kilb, W., and Luhmann, H.J. (2000). Characterization of a hyperpolarization-activated inward current in Cajal-Retzius cells in rat neonatal neocortex. Journal of Neurophysiology 84, 16811691. Kindler, C.H., and Yost, C.S. (2005). Two-pore domain potassium channels: new sites of local anesthetic action and toxicity. Régional Anesthesia and Pain Medicine 30, 260-274. Komai, H., and McDowell, T.S. (2001). Local anesthetic inhibition of voltage-activated potassium currents in rat dorsal root ganglion neurons. Anesthesiology 94, 1089-1095. Ludwig, A., Budde, T., Stieber, J., Moosmang, S., Wahl, C , Holthoff, K., Langebartels, A., Wotjak, C , Munsch, T., Zong, X., et al. (2003). Absence epilepsy and sinus dysrhythmia in mice lacking the pacemaker channel HCN2. The EMBO Journal 22, 216-224. Ludwig, A., Zong, X., Jeglitsch, M., Hofmann, F., and Biel, M. (1998). A family of hyperpolarization-activated mammalian cation channels. Nature 393, 587-591. Luthi, A., and McCormick, D.A. (1998). H-current: properties of a neuronal and network pacemaker. Neuron 21, 9-12. Mayer, M.L., and Westbrook, G.L. (1983). A voltage-clamp analysis of inward (anomalous) rectification in mouse spinal sensory ganglion neurones. The Journal of Physiology 340, 19-45. Momin, A., Cadiou, H., Mason, A., and McNaughton, P.A. (2008). Rôle of the hyperpolarization-activated current Ih in somatosensory neurons. The Journal of Physiology 586, 5911-5929. Moosmang, S., Stieber, J., Zong, X., Biel, M., Hofmann, F., and Ludwig, A. (2001). Cellular expression and functional characterization of four hyperpolarization-activated pacemaker channels in cardiac and neuronal tissues. European Journal of Biochemistry / FEBS 268, 16461652. Nolan, M.F., Malleret, G., Lee, K.H., Gibbs, E., Dudman, J.T., Santoro, B., Yin, D., Thompson, R.F., Siegelbaum, S.A., Kandel, E.R., étal. (2003). The hyperpolarization-activated HCN1 channel is important for motor learning and neuronal intégration by cerebellar Purkinje cells. Cell 775,551-564. 79  Olschewski, A., Brau, M.E., Olschewski, H., Hempelmann, G., and Vogel, W. (1996). ATPdependent potassium channel in rat cardiomyocytes is blocked by lidocaine. Possible impact on the antiarrhythmic action of lidocaine. Circulation 93, 656-659. Olschewski, A., Hempelmann, G., Vogel, W., and Safronov, B.V. (1998). Blockade of Na+ and K+ currents by local anesthetics in the dorsal horn neurons of the spinal cord. Anesthesiology 88, 172-179. Pape, H.C. (1996). Queer current and pacemaker: the hyperpolarization-activated cation current in neurons. Annual Review of Physiology 58, 299-327. Pape, H.C, and McCormick, D.A. (1989). Noradrenaline and serotonin selectively modulate thalamic burst firing by enhancing a hyperpolarization-activated cation current. Nature 340, 715718. Perkins, K.L., and Wong, R.K. (1995). Intracellular QX-314 blocks the hyperpolarizationactivated inward current Iq in hippocampal CA1 pyramidal cells. Journal of Neurophysiology 73, 911-915. Pian, P., Bucchi, A., Robinson, R.B., and Siegelbaum, S.A. (2006). Régulation of gating and rundown of HCN hyperpolarization-activated channels by exogenous and endogenous PIP2. The Journal of General Physiology 128, 593-604. Proenza, C., Angoli, D., Agranovich, E., Macri, V., and Accili, E.A. (2002). Pacemaker channels produce an instantaneous current. The Journal of Biological Chemistry 277, 5101-5109. Robinson, R.B., and Siegelbaum, S.A. (2003). Hyperpolarization-activated cation currents: from molécules to physiological function. Annual Review of Physiology 65, 453-480. Rocchetti, M., Armato, A., Cavalieri, B., Micheletti, M., and Zaza, A. (1999). Lidocaine inhibition of the hyperpolarization-activated current (If) in sinoatrial myocytes. Journal of Cardiovascular Pharmacology 34, 434-439. Santoro, B., Liu, D.T., Yao, H., Bartsch, D., Kandel, E.R., Siegelbaum, S.A., and Tibbs, G.R. (1998). Identification of a gène encoding a hyperpolarization-activated pacemaker channel of brain. Cell 95,717-729. Shi, W., Wymore, R., Yu, H., Wu, J., Wymore, R.T., Pan, Z., Robinson, R.B., Dixon, J.E., McKinnon, D., and Cohen, I.S. (1999). Distribution and prevalence of hyperpolarizationactivated cation channel (HCN) mRNA expression in cardiac tissues. Circulation Research 85, el-6. Shin, K.S., Rothberg, B.S., and Yellen, G. (2001). Blocker state dependence and trapping in hyperpolarization-activated cation channels: évidence for an intracellular activation gâte. The Journal of General Physiology 777, 91-101. 80  Takigawa, T., Alzheimer, C , Quasthoff, S., and Grafe, P. (1998). A spécial blocker reveals the présence and function of the hyperpolarization-activated cation current Ih in peripheral mammalian nerve fibres. Neuroscience 82, 631 -634. Tamura, A., Ogura, T., Uemura, H., Reien, Y., Kishimoto, T., Nagai, T., Komuro, I., Miyazaki, M., and Nakaya, H. (2009). Effects of Antiarrhythmic Drugs on the Hyperpolarization-Activated Cyclic Nucleotide-Gated Channel Current. Journal of Pharmacological Sciences. 110, 150-159. Thollon, C , Cambarrat, C , Vian, J., Prost, J.F., Peglion, J.L., and Vilaine, J.P. (1994). Electrophysiological effects of S 16257, a novel sino-atrial node modulator, on rabbit and guinea-pig cardiac préparations: comparison with UL-FS 49. British Journal of Pharmacology 112,37-42. Tu, H., Deng, L., Sun, Q., Yao, L., Han, J.S., and Wan, Y. (2004). Hyperpolarization-activated, cyclic nucleotide-gated cation channels: rôles in the differential electrophysiological properties of rat primary afférent neurons. Journal of Neuroscience Research 76, 713-722. Ulens, C , and Tytgat, J. (2001). Functional heteromerization of HCN1 and HCN2 pacemaker channels. The Journal of Biological Chemistry 276, 6069-6072. Woodhull, A.M. (1973). Ionic blockage of sodium channels in nerve. The Journal of General Physiology 61, 687-708. Yagi, J., and Sumino, R. (1998). Inhibition of a hyperpolarization-activated current by clonidine in rat dorsal root ganglion neurons. Journal of Neurophysiology 80, 1094-1104.  81  APPENDICES Appendix 1. Figure Al.l. 200 uM lidocaine inhibits If at -110 mV with complète current return. (Top) Current traces showing the effects of control (a), perfusion of 200 uM lidocaine for 2.5 minutes (b), and 2.5 minutes following drug removal from bath (c). Protocol is shown in the inset to the right. (Bottom) Time-dependent current (If) plotted over time. Note that current traces and recordings of If are obtained every 12 seconds. Complète washout is observed, as currents return to the level prior to lidocaine addition. 1 sec  i  +30mV 450ms) -35 mV  -IlOmV (1800ms)  -600-  -•—Contrat -•— Lidocaine applied -*— Wash  -800-  200 uM lidocaine -1000-  VV -1200-  /\J^  " iAr  ,*-•  -1400 -  -1600  —I 50  '  1 100  '  A A-A^ ^A-^-A 1 150  •  1 200  '  1 250  '  1 300  •——I 350  '  1 400  Time (sec)  82  Appendix 2. Figure A2.1. Représentative current trace of HCNl A372G.  W - i ^ W v ^ ' " " H W * <i l*^VH»N^ "» H *i>***M»fi«wu. <  <  l  200 pA  +30 mV 1000ms -35 mV  -130 mV 1000ms  500 ms  83  Figure A2.2. Représentative current trace of HCN 1 V379A.  2000 pA  +30 mV 1000ms -35 mV  L  -130 mV 1000ms  500 ms  84  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0069938/manifest

Comment

Related Items