UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Glycinergic and GABAergic inhibition in ventrobasal nuclei of rat thalamus Alavian Ghavanini, Ahmad (Amer Ghavanini) 2006

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-ubc_2006-199268.pdf [ 12.64MB ]
Metadata
JSON: 831-1.0092823.json
JSON-LD: 831-1.0092823-ld.json
RDF/XML (Pretty): 831-1.0092823-rdf.xml
RDF/JSON: 831-1.0092823-rdf.json
Turtle: 831-1.0092823-turtle.txt
N-Triples: 831-1.0092823-rdf-ntriples.txt
Original Record: 831-1.0092823-source.json
Full Text
831-1.0092823-fulltext.txt
Citation
831-1.0092823.ris

Full Text

GLYCINERGIC AND GABAERGIC INHIBITION IN VENTROBASAL NUCLEI OF RAT THALAMUS by A H M A D A L A V I A N GHAVANINI (AMER GHAVANINI) M . D . , Shiraz University of Medical Sciences, 1998 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES (Neuroscience) THE UNIVERSITY OF BRITISH C O L U M B I A July 2006 © Amer Ghavanini, 2006 Abstract This thesis examines the major inhibitory systems in ventrobasal nuclei of rat thalamus. The ventrobasal nuclei participate in relay and processing of somatosensory information and generation of thalamocortical rhythms. Inhibitory synaptic responses play an essential role in function of ventrobasal thalamus. A major objective of this thesis is to examine the hypothesis that glycine receptors mediate inhibition in ventrobasal thalamus. The thesis pharmacologically characterizes the synaptic and extrasynaptic inhibition and describes synaptic inhibition originating in nuclei surrounding ventrobasal thalamus. The thesis further characterizes the biophysical properties of glycine and y-aminobutyric acid type A ( G A B A A ) receptors in thalamus. We established that ( X i and CC2 glycine receptor subunits are expressed in the ventrobasal nuclei, using immunohistochemical staining. The functional nature of these subunits was confirmed by demonstrating the effects of glycine receptor agonists on thalamocortical neurons. Functional glycine receptors were likely limited to larger type-I thalamocortical neurons. Synaptic inhibition evoked from medial lemniscus uncovered a heterogeneous inhibitory input to ventrobasal thalamus. G A B A A and glycine receptors mediated synaptic inhibition in a majority of neurons. Inhibition in two minor groups of neurons was exclusively mediated by either G A B A A or glycine receptors. Occasionally, there was an additional G A B A B component of synaptic inhibition. The medial lemniscal mediated inhibition was likely polysynaptic, and resulted from co-transmission of G A B A and glycine. Stimulations within surrounding nuclei 11 evoked monosynaptic inhibition. The highest relative glycinergic strength was evoked from the ethmoid nucleus. Another major finding of these studies was the diverse kinetics of the glycinergic inhibition. The observations were consistent with the activation of two kinetically distinct populations of glycine receptors, segregated under separate nerve terminals. The kinetics of synaptic receptors mirrored the kinetics of extrasynaptic receptors. Synaptic channels displayed higher CI" permeability than their extrasynaptic counterparts. These studies involved examination of previous methods. The specificity of strychnine was investigated and the optimal concentration for discrimination between glycine and G A B A A receptors was established. Synaptic currents were simulated to examine the sources of error in non-stationary fluctuation analysis. This procedure resulted in an improved method for non-stationary fluctuation analysis. In summary, this thesis unveils the heterogeneous nature of synaptic inhibition in somatosensory thalamus. The findings open frontiers in the research and development of new drugs targeting glycinergic inhibition in the thalamus. iii Table of Contents Abstract 1 1 Table of Contents i v List of Tables v i i List of Figures V 1 " Acknowledgements and Dedication x i i Chapter 1. Introduction 1 1.1. Scope of thesis 1 1.2. Background 3 1.2.1. Major inhibitory systems 3 1.2.1.1. The glycinergic system 3 1.2.1.2. The GABAergic system 14 1.2.2. Ventrobasal nuclei of thalamus 19 1.2.2.1. Functional anatomy 20 1.2.2.2. Synaptic inhibition in ventrobasal thalamus 25 1.3. Rationale and Objectives : 28 Chapter 2. Methods 31 2.1. Immunohistochemistry 31 2.1.1. Tissue fixation and slice preparation 31 2.1.2. Staining procedures 31 2.1.3. Imaging 33 2.2. Electrophysiological recordings 33 2.2.1. Whole-cell recordings 33 2.2.1.1. Slice preparation 33 2.2.1.2. Electrophysiological recording 34 2.2.1.3. Electrical stimulation 36 2.2.2. Single channel recording 36 2.2.3. Computer simulation of synaptic currents 38 2.3. Drugs and application 40 iv 2.4. Data analysis 40 2.4.1. General measurements 40 2.4.2. Rheobase and chronaxie 41 2.4.3. Latency and latency fluctuation 41 2.4.4. Detection of spontaneous IPSCs (sIPSCs) 42 2.4.5. Rise and decay time constants 42 2.4.6. Non-stationary fluctuation analysis 43 2.4.7. Concentration-response analysis 45 2.4.8. Statistical analysis 46 Chapter 3. Results 47 Part I - Presence of Glycine Receptors in Ventrobasal Thalamus 47 3.1. Immunohistochemical evidence 47 3.1.1. Glycine receptor subunits 47 3.1.2. Glycine and taurine molecules 50 3.2. Electrophysiological evidence 51 3.2.1. GlyR 52 3.2.2. G A B A A receptor 60 Part II - Synaptic Glycinergic Inhibition in Ventrobasal Thalamus 64 3.3. Medial lemniscal evoked inhibitory responses 64 3.3.1. Susceptibility of inhibitory receptors to antagonists 67 3.3.2. Strychnine concentration response relationship 69 3.3.3. Slower component of synaptic inhibition 72 3.4. Comparison of inhibition evoked from other sources 73 3.4.1. Susceptibility to antagonists 73 3.4.2. Latency and latency fluctuations 78 3.4.3. Rheobase and chronaxie 80 3.4.4. IPSP shape parameters 82 Part III - Biophysical Properties of Synaptic Inhibitory Receptors 85 3.5. Kinetics of synaptic receptors 85 3.5.1. Evoked IPSCs 85 3.5.2. Spontaneous IPSCs 88 v 3.6. Permeability of synaptic receptors 93 3.6.1. Methodological sources of error in non-stationary fluctuation analysis 94 3.6.2. Evoked IPSCs 104 3.6.3. Spontaneous IPSCs 110 3.7. Biophysical properties of extrasynaptic inhibitory receptors 113 3.7.1. Kinetic properties of extrasynaptic receptors 113 3.7.2. Permeability of extrasynaptic receptors 116 3.8. Comparison of synaptic and extrasynaptic inhibitory receptors 118 3.9. Effect of A M B D on IPSCs 120 Chapter 4. Discussion 122 4.1. Functional glycine receptors in ventrobasal thalamus 122 4.2. Synaptic inhibition in ventrobasal thalamus 126 4.3. Biophysical characteristics of inhibition in ventrobasal thalamus 131 4.4. Functional implications 138 4.5. Significance 140 4.6. Conclusions 141 Abbreviations 143 Bibliography 146 Appendix A : Matlab codes for simulation of synaptic currents 163 Appendix B : Matlab codes for non-stationary fluctuation analysis 168 vi List of Tables Table 1.1. E C 5 0 s of GlyR agonists 7 Table 1.2. IC 5 0 s for GlyR antagonists 11 Table 2.1. Intracellular whole-cell patch clamp solutions 35 Table 3.1. IPSP shape parameters 83 Table 3.2. Decay kinetics of evoked and spontaneous glycinergic IPSCs 92 Table 3.3. Estimates of iJU with altering bin width and sampling rates 98 Table 3.4. Effect of pre-pulse on channel conductance 110 Table 3.5. Biophysical properties of extrasynaptic receptors 118 Table 3.6. Permeabilities of synaptic and extrasynaptic receptors 120 vi i List of Figures Figure 1.1. Schematic drawing of GlyR 4 Figure 1.2. Chemical structure of p-amino acids 6 Figure 1.3. Chemical structure of GlyR antagonists 10 Figure 1.4. Chemical structure of G A B A A agonists and antagonists 15 Figure 1.5. Chemical structure of G A B A C agonists and antagonists 17 Figure 1.6. Schematic drawing of ventrobasal thalamus in rat brain 21 Figure 1.7. Camera lucida drawings of typical TC neurons 22 Figure 2.1. Anatomical locations of electrophysiological studies 35 Figure 3.1. Bright field microscopy of rat brain 48 Figure 3.2. Antibody staining for subunits of GlyR 49 Figure 3.3. Confocal microscopic image of negative control 50 Figure 3.4. Antibody staining for glycine and taurine 51 Figure 3.5. Electrophysiological characteristics of typical ventrobasal neuron 52 Figure 3.6. Effect of glycine on action potential firing and input resistance 54 Figure 3.7. Strychnine antagonism of glycine actions 54 Figure 3.8. Effect of P-amino acids on action potential firing 55 Figure 3.9. Strychnine antagonism of P-alanine actions 56 Figure 3.10. Effect of L-serine on action potential firing 57 Figure 3.11. Time course of glycine actions 57 Figure 3.12. Effect of GlyR agonists on current-voltage relationship 58 Figure 3.13. Picrotoxinin antagonism of glycine actions 59 Figure 3.14. Comparison of bicuculline and strychnine effects on p-amino acids actions ... 59 vi i i Figure 3.15. Concentration-response relationships for p-amino acids actions 60 Figure 3.16. Effect of G A B A on action potential firing 61 Figure 3.17. Effect of G A B A on current-voltage relationship 61 Figure 3.18. Concentration-response relationship for G A B A actions 62 Figure 3.19. Strychnine antagonism of G A B A actions 63 Figure 3.20. Unmasking hyperpolarizing responses 64 Figure 3.21. Burst firings during recovery from hyperpolarizing responses 65 Figure 3.22. Action potential dependence of hyperpolarizing responses 66 Figure 3.23. Reversal of hyperpolarizing responses 66 Figure 3.24. Reversal potential of evoked postsynaptic current 67 Figure 3.25. Sensitivity of medial lemniscal IPSPs to bicuculline and strychnine 68 Figure 3.26. Concentration-response relationship for strychnine blockade IPSPs 70 Figure 3.27. Sensitivity of medial lemniscal IPSPs to gabazine and strychnine 71 Figure 3.28. Strychnine concentration-response relationship during gabazine application ... 72 Figure 3.29. Occasional slower component of synaptic inhibition 73 Figure 3.30. IPSPs evoked from the caudal zona incerta 74 Figure 3.31. IPSCs evoked from nRt 75 Figure 3.32. IPSP evoked from Eth 75 Figure 3.33. IPSPs evoked from Pfc 76 Figure 3.34. Correlation of antagonist sensitivity with the stimulation site 77 Figure 3.35. Correlation of latency and latency fluctuation with the stimulation site 79 Figure 3.36. Calculation of chronaxie and rheobase 80 Figure 3.37. Correlation of chronaxie and rheobase with the stimulation site 81 ix Figure 3.38. Stimulus duration-amplitude relationship for stimulation sites 82 Figure 3.39. Calculation of shape index 83 Figure 3.40. Isolation of glycinergic and G A B A A e r g i c components... of IPSCs 85 Figure 3.41. Decay kinetics of glycinergic and G A B A A e r g i c IPSCs 87 Figure 3.42. sIPSCs recorded in the V B thalamus 88 Figure 3.43. Kinetics of sIPSCs in the V B thalamus 89 Figure 3.44. Comparison of decay kinetics of spontaneous and evoked IPSCs 91 Figure 3.45. Effect of strychnine on decay kinetics of GABAAergic IPSCs 93 Figure 3.46. Computer simulated synaptic current using the classic algorithm 95 Figure 3.47. Effect of input parameters on unitary current estimate 96 Figure 3.48. Effect of bin width on unitary current estimate 97 Figure 3.49. Interplay between bin width and the length of stationary segments 99 Figure 3.50. Effect of added Gaussian noise on unitary current estimate 101 Figure 3.51. Applicability of simlPSC findings to G A B A A e r g i c IPSCs 103 Figure 3.52. Applicability of simlPSC findings to glycinergic IPSCs 104 Figure 3.53. Stability of quantal release 105 Figure 3.54. Estimation of unitary channel currents from glycinergic IPSCs 106 Figure 3.55. Current-voltage relationship for glycinergic and GABAAergic currents 107 Figure 3.56. Reversal potential for IPSCs recorded with Cs-QX patches 108 Figure 3.57. Effect of depolarizing pre-pulse on IPSCs 109 Figure 3.58. Estimation of unitary channel currents from glycinergic sIPSCs I l l Figure 3.59. Current-voltage relationships for z'cis estimated from sIPSCs 112 Figure 3.60. CI" permeability of GlyR estimated from evoked and spontaneous IPSCs .. . . 112 x Figure 3.61. Kinetics of single channels activated by glycine receptor agonists 115 Figure 3.62. Conductance of single channels activated by glycine receptor agonists 117 Figure 3.63. Comparison of biophysical properties of synaptic and extrasynaptic GlyR ... 119 Figure 3.64. Effect of A M B D on IPSCs 121 xi Acknowledgements and Dedication I would like to express sincere gratitude to my supervisor Dr. Ernie Pui l , who not only taught me the concepts of Neurosciences and Pharmacology, but also taught me how to think and how to learn. I am extremely grateful to my co-supervisor Dr. David Mathers for his guidance and the fruitful discussions that nourished my education and the research. I am very thankful to Drs. Bernie MacLeod and Craig Ries who provided excellent comments and suggestions as my PhD committee members. I am thankful to M i t i Isbasescu for his technical assistant in writing MATLAB codes and to Dr. Hee-Soo K i m for recording and analysis of single channel currents. I would also like to acknowledge M s . Viktoriya Dobrovinska and M r . Christian Caritey for excellent technical assistance. I thank M r . Kev in Hodgson and M r . Douglas Brown for assistance with confocal and light microscopy. M y gratitude for invaluable intellectual and moral support goes to my friends and colleagues. Israeli Ran deserves a special thank for instructing me on some of the techniques, discussing scientific issues and being a good friend in the past years. Canadian Institutes for Health Research, the Jean Templeton Hugi l l Foundation Chair for Anaesthesia and Analgesia, and University Graduate Fellowship through the U B C Neurosciences Program financially supported this work. I would like to thank my family for their moral and emotional support throughout my life. I dedicate this work to my father who has always been my role model, and my inspiration for pursuing education in Neuroscience. x i i Chapter 1 I N T R O D U C T I O N 1.1. Scope of thesis The scope of this thesis is to investigate the major inhibitory systems in ventrobasal nuclei of rat thalamus. A major objective of this investigation is to examine the hypothesis that glycine mediates inhibition in ventrobasal nuclei of rat thalamus. Glycine is a well-established neurotransmitter of synaptic inhibition in spinal cord and brainstem (Werman et al., 1968). The identity of glycine as transmitter is elusive in the rostral areas of the central nervous system despite a substantial number of fibres showing immunoreactivity for glycine in these areas (Rampon et al., 1996). Recent evidence, however, suggests that glycine receptors may participate in synaptic inhibition in the thalamus (Ran et al., 2004). Glycinergic inhibition in thalamus may play a critical role in processing somatosensory information and in the pathophysiology of neuropathic pain. Initially, immunohistochemical staining with antibodies raised against CC] and 0:2 glycine receptor subunits was used to examine the presence of the functional subunits of glycine receptor in ventrobasal thalamus. We next examined the actions of glycine and related amino acids on evoked firing and resting membrane properties in thalamocortical neurons. We also examined the ability of glycine and y-aminobutyric acid ( G A B A A ) receptor antagonists to antagonize these effects. The results of these studies relate to a synaptic role of glycine in thalamus. For example, a neuronal glycine transporter is evident in ventrobasal thalamus (Jursky and Nelson, 1995; Zeilhofer et al., 2005), consistent with glycinergic transmission. Chapter 1. Introduction - 2 -To provide initial rationale for glycinergic transmission, we addressed the issue whether glycine and G A B A A receptor antagonists affected synaptic inhibition in ventrobasal thalamocortical neurons. We compared the effects of these antagonists on the inhibition evoked from the medial lemniscus, the major somatosensory input to ventrobasal neurons (Feldman and Kruger, 1980). We also contrasted the effects on inhibitory postsynaptic potentials and currents (IPSPs and IPSCs) evoked from the nearby nucleus reticularis thalami (nRt), zona incerta, ethmoid nucleus, and nucleus parafascicularis (Pfc). The results of these studies reveal synaptic glycinergic inhibition in ventrobasal thalamus. Glycine receptor CI" channels containing ai or a 2 subunits display differences in biophysical properties (Takahashi et al. 1992, Singer et al. 1998). We compared biophysical characteristics of IPSCs evoked by medial lemniscal stimulation as well as spontaneous IPSCs to single channel, extrasynaptic currents evoked by glycine and related amino acids or G A B A in ventrobasal neurons We determined the decay rates of IPSCs, and examined the predicted relationship with mean burst durations of P-amino acid- or GABA-act ivated CI" channels. We also compared CI" permeability of synaptic and extrasynaptic receptor channels. These studies shed light on the nature of the inhibitory transmitters and possible functions of extrasynaptic glycine receptors in the thalamus. Chapter 1. Introduction - 3 -1.2. Background 1.2.1. Major inhibitory systems A n inhibitory synaptic signal can be defined as a signal that decreases the probability of an action potential occurring in the neuron receiving that signal. Inhibition may counteract an excitatory synaptic signal, control spontaneous activity of a neuron, or provide for integration of information (Kandel et al., 1991). Inhibition often results from hyperpolarization of the postsynaptic membrane, which drives the membrane potential away from the threshold for action potential generation and results from shunting of the excitatory synaptic current. Shunting is far more effective in blocking action potentials than hyperpolarization. G A B A and glycine are major inhibitory neurotransmitters. Glycinergic inhibition is more prevalent in the caudal neuroaxis. In supraspinal regions, G A B A is the dominant mediator of synaptic inhibition (Sivilotti and Nistri , 1991). However, this paradigm is changing as more evidence for the presence of glycinergic inhibition in rostral central nervous system is uncovered (Darstein et al., 2000; Dudeck et al., 2003). Moreover, G A B A and glycine often act as co-transmitters on release from separate nerve endings (Donato and Nistr i , 2000). In some areas of the nervous system, these amino acids are co-released from the same nerve terminals (Jonas et al., 1998). This may provide a means for neuronal integration or modulation of signals. 1.2.1.1. The glycinergic system Based on the seminal investigations of Werman et al. (1968), glycine has become a wel l -established neurotransmitter of synaptic inhibition in spinal cord and brainstem. The action of glycine is mediated via glycine receptor (GlyR), a ligand-gated CI" channel (reviewed by Lynch, Chapter 1. Introduction - 4 -2004). Activation of GlyR results in a change in membrane potential towards the CP equilibrium potential. The Cl~ equilibrium potential is usually more negative than the cell resting potential. Hence, activation of GlyR hyperpolarizes the neuron resulting in inhibition. Inhibition also results from an increase in membrane conductance that shunts excitatory responses (cf. Ries and Puil, 1999). In some nerve terminals and dendrites, however, CI" pump is not expressed resulting in an elevated concentration of intracellular Cl~. In these cases, the effect of GlyR activation is a depolarization, resulting in excitation (cf. El-Beheiry and Puil, 1990; Kuner and Augustine, 2000). Figure 1.1. Schematic drawing of GlyR showing the binding sites for glycine and strychnine on the a subunit and the interaction between the P subunit and gephyrin. Drawing by A. A. Ghavanini. GlyR (Fig. 1.1) is a pentamer composed of the 48 kDa (a) and the 58 kDa (p) glycoprotein subunits, the predominant stoichiometry being 3a2p (Grenningloh et al., 1987; Grenningloh et al., 1990). The a subunit bears the ligand-binding site and contributes to the formation of CI" selective channel and is required for GlyR to be functional (cf. Ruiz-Gomez et al., 1990; Lynch, Chapter 1. Introduction - 5 -2004). While the a subunit can form functional homomeric G l y R , homomeric receptors containing only p subunit are not functional. However, the p subunit mediates the clustering o f G l y R at postsynaptic sites (Kirsch and Betz, 1995). G l y R is anchored to the cell cytoskeleton via interactions between the p subunit and the cytoplasmic protein, gephyrin (Kirsch and Betz, 1995). Gephyrin is a 93-kDa protein that bridges the P subunit of G l y R with cytoskeletal structures, including tubulin (Meyer et al., 1995). Hence, the P subunit of G l y R is required for synaptic anchoring and targeting. The a subunit of G l y R exists in four different isoforms, ai.4 (Matzenbach et al., 1994). The adult receptors contain ai and (X3 isoforms, while 0x2 and cu isoforms are seen during embryonic and neonatal periods (Malosio et al., 1991b; Harvey et al., 2000). In some brain regions, however, detectable amounts of (X2 transcripts persist into adulthood (Piechotta et al., 2001). The subunit isoforms determine single channel conductance and kinetic properties as well as pharmacological characteristics of G l y R (Takahashi et al., 1992; Singer et al., 1998). In addition to isoform diversity, alternative splicing generates several splicing variants of the a subunit of glycine receptor with distinct physiological and pharmacological characteristics (Kuhse et al., 1991). A l l G l y R subtypes form ion channels with multiple conductance states (Hamill et al., 1983). These conductance states range between 20 and 90 pS for ai homomeric receptors and 20-110 pS for a 2 and a 3 homomeric receptors (Lynch, 2004). Both heteromeric G l y R containing P subunit and native G l y R have significantly lower unitary conductance (Bormann et al., 1993). This is consistent with the native G l y R existing as heteromers. Native and recombinant homomeric cci Chapter 1. Introduction and (X2 G lyRs have distinct mean channel open times (Takahashi et al., 1992). Mean open times of 0Ci homomeric channels are much shorter than those of 0C2 homomeric channels (Takahashi et al., 1992). Hence channel open times decrease as ai subunits replace 0C2 subunits during development. o o HoN. HoN. O H v SO3H H 2r\T ~ O H Glycine Taurine p-alanine Figure 1.2. Chemical structure of major inhibitory amino acids. Native and recombinant GlyRs are activated by inhibitory amino acids, such as glycine, taurine and P-alanine. Figure 1.2 shows the chemical structures of these compounds. In addition, D -and L-serine may also act as G l y R agonists (Horikoshi et a., 1988; Tokutomi et al., 1989). For spinal neurons, the relative abilities of these agonists to activate G l y R generally follow the potency rank order: glycine > P-alanine > taurine » serine (Curtis et al., 1968; also hippocampal neurons, M o r i et al., 2002). However, the efficacy and affinity of these agonists vary for the G l y R isoform subtype, species, and the tissue where it is expressed (Table 1.1). Hence, it is not clear i f these pharmacological characteristics can be used to accurately identify the receptor subtype. Chapter 1. Introduction Table 1.1. Variability of agonist EC50 for recombinant GlyRs . Expression system Subunit Agonist E C 5 0 ( H M ) Reference Glycine 70-290 Grenningloh et al., 1990; Schmieden and Betz, 1995; Masciaetal., 1998 Oil Human Taurine 1700-2200 Schmieden et al., 1992; Schmieden and Betz, 1995 P-alanine 600-730 Schmieden et al., 1992; Schmieden et al., 1999 OllZebrafish Glycine Taurine 67-116 188 David-Watine et al., 1999; Fucile et al., 1999 David-Watine et al., 1999 Xenopus oocyte 0l2Human Glycine Taurine (3-alanine 290-390 3700 2800 Grenningloh et al., 1990; Hosie et al., 1999; Schmieden et al., 1992 Schmieden et al., 1992 0l2Zebrafish Glycine Taurine 216 5500 Imboden et al., 2001 OllpHuman Glycine 187 Lewis etal., 1998 Oil HumanPRat Glycine 380 Langosch et al., 1994 Glycine 18-58 Lynch et al., 1997; Moorhouse et al., 1999; De Saint Jan et al., 2001 Oil Human Taurine 153-310 Lynch et al., 1997, Moorhouse et al., 1999; De Saint Jan etal., 2001 P-alanine 52 Lynch etal., 1997 OllZebrafish Glycine Taurine 25-33 25 David-Watine et al., 1999; Fucile et al., 1999 David-Watine et al., 1999 Ol2Human Glycine 60-97 Pribilla et al., 1992; De Saint Jan et al., 2001 Taurine 752 De Saint Janet al., 2001 Human embryonic cell lines 0l2Rat Glycine Taurine p-alanine 220 740 760 Farroni and McCool, 2004 0l3Human Glycine 60 Pribilla et al., 1992 OllpHuman Glycine 74 Handfordet al., 1996 0llHumanj3Rat Glycine 48-55 Pribilla et al., 1992; Bormann et al., 1993 0l2HumanPRat Glycine 86 Pribilla et al., 1992 OfepRat Glycine Taurine p-alanine 280 2200 570 Farroni and McCool, 2004 0l3HumanpRat Glycine 75 Pribilla et al., 1992 Alternative splicing of a subunit transcripts further increases the heterogeneity of GlyRs . A variant of a i isoform, a i ' n s , originates from alternative splicing of cii isoform in rat spinal cord (Malosio et al., 1991a). This variant does not show major differences from a i isoform in Chapter 1. Introduction - ° -channel properties when expressed in Xenopus oocyte. The variant may have a role in receptor-cytoskeleton interactions or post-translational modulation of G l y R function (Malosio et al., 1991a). Three variants for the rat 0C2 isoform are CC2A, 0C2B and 0C2* (Kuhse et al., 1990; Kuhse et al., 1991). The C62A and OC2B variants are similar to the 0C2 isoform in their channel properties and are likely involved in synaptogenesis (Kuhse et al., 1991). The 0C2* variant confers 40-fold lower affinity for glycine and a 500-fold lower affinity for the antagonist strychnine upon expression in Xenopus oocytes. Two alternative splice variants of the OC3 isoform have been described, the 0C3L and the 0C3K variants (Nikolic et al., 1998). Functional expression of these two variants results in the formation of GlyRs that differ from the (X3 isoform in channel desensitization kinetics (Nikolic et al., 1998). Post-translational modifications result in G l y R with unique physiological and pharmacological characteristics (Lynch, 2004). G l y R can undergo post-translational modifications via several mechanisms. These mechanisms are likely responsible for long-term changes in the functional properties of G l y R . Phosphorylation constitutes the major post-translational modification of G l y R . Intracellular signalling pathways determine the phosphorylation state of G l y R by coordinating the activities of kinases, which induce phosphorylation, and phosphatases, which reverse it. In addition, kinases and phosphatases may indirectly modify G l y R function by controlling the phosphorylation state of modulatory proteins (cf. Agopyan et al., 1993). Complex effects of phosphorylation on G l y R currents have been observed in neurons from various parts of the brain. Activation of cAMP-dependent protein kinase A can enhance G l y R currents (Vaello et al., 1994; G u and Huang, 1998). Inconsistent effects have been observed Chapter 1. Introduction - 9 -following activation of protein kinase C. While some investigators have observed that activation of protein kinase C potentiates G l y R currents (Nishizaki and Ikeuchi, 1995), others have observed no effects (Gu and Huang, 1998) or even a decrease in G l y R currents (Vaello et al., 1994) following activation of protein kinase C. In some experimental settings where protein kinase C did not alter G l y R currents, it inhibited the potentiating action of protein kinase A (Gu and Huang, 1998). Hence, these differences can be related to the phosphorylation of G l y R modulatory proteins, rather than to the phosphorylation of G l y R subunits. Antagonists of G l y R include strychnine, picrotoxin, and 6-aminomemyl-3-methyl,l-4H-1,2,6-benzothiadiazine-l,l-diazide hydrochloride ( A M B D , also referred to as T A G ) . Figure 1.3 shows the chemical structure of these compounds. Strychnine is a highly selective and extremely potent antagonist of glycine receptors (Legendre, 2001). It is a competitive antagonist of G l y R , whose binding site overlaps but is not identical to the glycine-binding site (Marvizon et al., 1986). Subtypes of G l y R are sensitive to nanomolar concentrations of strychnine (Table 1.2). The only exception is the G l y R 0C2* variant, which exhibits a greatly reduced strychnine sensitivity (Kuhse et al., 1990). A high sensitivity to strychnine is currently the most definitive means of discriminating glycinergic from G A B A e r g i c synaptic currents (Legendre, 2001; Lynch, 2004). Because of its high affinity and specificity, [ 3H]strychnine has been used in autoradiographic localization of G l y R (Young and Snyder, 1973; Zarbin et al 1981; Frostholm and Rotter 1985). Chapter 1. Introduction - 10 -Strychnine Picrotoxinin Picrotin AMBD Figure 1.3. Chemical structure of GlyR antagonists. Picrotoxin comprises an equimolar mixture of picrotoxinin and picrotin. A M B D , 6-aminomethyl-3-methyl, 1-4H-1,2,6-benzothiadiazine-1,1-diazide hydrochloride, is also referred to as T A G . Picrotoxin, another antagonist of GlyR, is much less potent than strychnine, exhibiting an EC50 in the micromolar range (Table 1.2). It comprises an equimolar mixture of picrotoxinin and picrotin. The former shows little discrimination between GlyR and G A B A A receptor (Lynch, 2004). Picrotoxinin inhibits GlyR in a use-dependent and non-competitive manner, consistent with its C l _ channel blocking properties (Lynch, 2004). While picrotin is as efficacious as picrotoxinin in inhibiting GlyRs, it is generally inactive at the G A B A A receptor. The effect of picrotin is not use dependent and its inhibition is dependent on agonist concentration. Besides, it acts as a potentiator at low concentrations and a non-competitive inhibitor at higher concentrations on some mutated GlyR (Lynch et al., 1995). Hence, picrotin is likely an allosteric inhibitor of GlyR. Chapter 1. Introduction Table 1.2. IC50 for antagonism of currents activated by glycine at EC50. -11 -Expression system Subunit Antagonist ICsoinM) Reference Strychnine 16-37 Schmieden et al., 1992; Hosie et al., 1999; Schmieden et al., 1999 OCl Human Picrotoxinin 1890 Hosie etal., 1999 Picrotoxin 25000 Handford et al., 1996 OCiZebrafish Strychnine 26 David-Watine et al., 1999 Xenopus oocyte 0 C 2 H u m a n Strychnine 50 Schmieden etal., 1992 Picrotoxinin 141 Hosie etal., 1999 0C2Zebrafish Strychnine 20 Imboden et al., 2001 0Cl(3Human Strychnine 28 Lewis etal., 1998 OCl Human \at Strychnine 17 Langosch et al., 1994 OClHuman Strychnine Picrotoxinin 13 9000 Pribilla et al., 1992 OCiZebrafish Strychnine 26 David-Watine et al., 1999 0 C 2 H u m a n Strychnine Picrotoxinin 18 6000 Pribilla et al., 1992 0C2Rat Strychnine 49 Farroni and McCool, 2004 Human embryonic 0C3Human Strychnine Picrotoxinin 20 5000 Pribilla et al., 1992 cell lines OCl PHuman Strychnine Picrotoxin 14 420000 Handford et al., 1996 OCl Human Prat Strychnine Picrotoxinin 11 >1 mM Pribilla et al., 1992 0C2umanPrat Strychnine Picrotoxinin 12 300,000 Pribilla et al., 1992 0C3umanPrat Strychnine 15 Pribilla et al., 1992 Picrotoxinin >1 mM A M B D is another competitive antagonist of GlyR. A M B D initially gained attention for its potential utility in pharmacological differentiation between the action of glycine, taurine, and P -alanine (Yarbrough et al., 1981; Girard et al., 1982). When examined in the amphibian spinal cord, A M B D selectively blocked the action of taurine but not glycine, p-alanine and G A B A (Padjen et al., 1989). Early iontophoretic studies of A M B D in mammalian nervous system supported the notion of A M B D selectivity for taurine (Yarbrough et al., 1981; Okamoto et al., 1983). Other investigators using iontophoretic techniques in the mammalian nervous system Chapter 1. Introduction - 12 -were not able to reproduce these results (Curtis et al., 1982). A technical limitation of iontophoretic application is that the concentration of the drug at the effector site is not known. Hence, the latter investigation may have involved application of A M B D at very high concentrations, where its selectivity was lost. In other words, A M B D may selectively block taurine action only over a narrow range of concentration depending on the system (Mathers, 1993). Glycine transporters (GlyT) mediate the uptake of glycine from the extracellular space into the cytosol. Two subtypes of G l y T have been identified, G l y T i and GlyT2, which differ in anatomical and cellular distribution as well as in pharmacology (reviewed by Eulenburg et al., 2005). G l y T i is expressed throughout the central nervous system (CNS) while G l y T 2 is expressed exclusively in C N S regions where synaptic glycinergic inhibition exists (Jursky and Nelson, 1995). G l y T i is specific to glial cells, in particular astrocytes, although some G l y T i immunoreactivity is observed in nerve terminals of excitatory glutamatergic neurons (Jursky and Nelson, 1995; Cubelos et al., 2005). GlyT2 is present only in presynaptic glycinergic terminals, where it is closely related to G l y R containing postsynaptic specializations (Jursky and Nelson, 1995). Hence, the presence of G l y T 2 is a good marker of glycinergic inhibitory terminals. A recently developed transgenic mouse that expresses enhanced green fluorescent protein under the control of G l y T 2 gene promotor provides an excellent in vitro model for identification and investigation of glycinergic inhibition in the C N S (Zeilhofer et al., 2005). Both GlyTs take up glycine by coupling its transport to the transmembrane N a + and C F gradients maintained by N a + / K + - A T P a s e (Eulenburg et al., 2005). The stoichiometry of this co-transport is Chapter 1. Introduction - 13 -3Na +/C17glycine for GlyT2, whereas G l y T i shows a stoichiometry of 2Na + /C17glycine (Roux and Supplisson, 2000). There are two implications of these differing stoichiometries. One is that the transport of glycine across the membrane is electrogenic, i.e., it depolarizes the membrane. Secondly, under physiological conditions, the driving force available for glycine transport by GlyT2 is much greater than that by G l y T i . Hence, the neuronal GlyT2 has a higher efficacy in maintaining extracellular glycine levels than G l y T i . The two subtypes of GlyTs are pharmacologically distinct. G l y T i is preferentially inhibited by sarcosine and N[3-(4'-fluorophenyl)-3-(4'-phenylphenoxy)propyl]sarcosine (Aubrey and Vandenberg, 2001). On the other hand, the antidepressant, amoxapine, is ten times more potent at inhibiting G l y T 2 (Nunez et al., 2000). Given the fact that G l y T 2 , but not G l y T i , is associated with glycinergic inhibition (cf. Jursky and Nelson, 1995), the pharmacologically distinct GlyTs are useful targets for drug development. Taurine is taken up by a transporter distinct from GlyTs. Taurine transporter (TauT) belongs to the same family of amino acid transporters as GlyTs. It co-transports a taurine molecule across the cell membrane along with two N a + ions and one CI" ion (Nelson, 1998). TauT shows specificity for (3-amino acids. In addition to taurine, it can transport hypotaurine and (3-alanine (Tappaz, 2004). TauT is competitively blocked by guanidinoethanesulfonic acid (Barakat et al., 2002). However, guanidinoethanesulfonic acid has been shown to inhibit G A B A uptake ( L i and Lombardini, 1990) and directly activate G A B A A receptors (Mori et a l , 2002) in addition to TauT blockade. This lack of selectivity limits the usefulness of guanidinoethanesulfonic acid as a pharmacological tool for identification of taurine transport. Chapter 1. Introduction - 14 -1.2.1.2. The G A B A e r g i c system G A B A is a major inhibitory neurotransmitter in the C N S , particularly the rostral parts. It mediates inhibition via three distinct receptor types, the ionotropic G A B A A and G A B A c and metabotropic G A B A B receptors (Johnston 1996; Bettler et al., 2004; Chebib, 2004). While G A B A B receptors can be found both on presynaptic and postsynaptic membranes, G A B A A receptors are involved in postsynaptic inhibition and thus far identified G A B A C receptors in C N S are presynaptic (Matthews et al., 1994; Johnston 1996; Bettler et al., 2004). The ionotropic G A B A A and G A B A c receptors are members of the ligand gated ion channel superfamily, which also includes G l y R (Chebib, 2004). Both receptors are comprised of five subunits assembling to form a ligand-gated C F channel (Chebib and Johnston, 2000), a structure closely resembling that of G lyR . G A B A A receptor is probably the most diverse of this superfamily. A t least 19 distinct subunit isoforms have been identified for G A B A A receptor (Hevers and Luddens, 1998; Simon et al., 2004). G A B A C is less diverse, with only five known subunit isoforms (Chebib and Johnston, 2000). The nineteen isoforms identified for G A B A A receptor subunits are ai_6, P i _ 3 , y i _ 3 , 8, s, n, p\.j, and 9 (Sieghart and Sperk, 2002; Simon et al., 2004; Khom et al., 2005). The most common stoichiometry is two a-subunits, two |3-subunits and one y-subunit (Sieghart and Sperk, 2002). The subunit composition is variable and greatly affects the pharmacology of G A B A A receptors (Johnston, 1996). Binding sites for G A B A A receptor modulators often exist at the interfaces between different pairs of subunits making up the heterooligomeric receptors (Wingrove et al., 1997). This phenomenon greatly increases the pharmacological diversity of G A B A A receptors, Chapter 1. Introduction -15-because not only the subunit composition, but also the arrangement of these subunits can affect the pharmacological properties of G A B A A receptor. Five isoforms of GABAc receptor subunits (pi.5) have been identified (Chebib and Johnston, 2000). These subunit isoforms can assemble in a variety of combinations. However, they do not co-assemble with G A B A A receptor subunits. Hence, the two families of GABAA and GABAc receptors have distinct isoform combinations (Hackam et al., 1997). This structural differences results in pharmacological characteristics that distinguish between G A B A A and GABAc receptors (Chebib, 2004). COOH CK . N s NH, O l-LCO COOH O Gabazine Bicuculline Figure 1.4. Chemical structure of selected GABAA agonists (GABA and muscimol) and antagonists (gabazine and bicuculline). The most significant distinction between G A B A A receptor and other ligand gated CI" channels such as GlyR and GABAc receptors lies in their pharmacological characteristics. The chemical structures of selective GABAA agonist and antagonists are depicted in Figure 1.4. While picrotoxin is able to block all ligand gated CI" channels (Dong and Werblin, 1996; Johnston, 1996; Lynch, 2004), other antagonists are selective enough to 'define' these receptors (Johnston, Chapter 1. Introduction - 16 -1996). Bicuculline was the first competitive antagonist of G A B A A used to probe the inhibitory action of G A B A in the C N S (Curtis et al., 1970). Bicuculline inhibits G A B A A receptors with an IC50 of 2-3 p M . Complete blockade of G A B A A receptors occurs at bicuculline concentrations of 10-25 p M (Shirasaki et al., 1991; Takahashi et al., 1994). In this concentration range, bicuculline does not have an action on strychnine-sensitive G l y R currents (Shirasaki et al., 1991; Siebler et al., 1993), making it suitable for identification of CI" mediated G A B A A inhibition. L ike all pharmacological agents, the selectivity of strychnine and bicuculline for G l y R and G A B A A receptors is limited to a range of concentrations. When bicuculline is applied at higher than 50 p M concentrations, it can antagonize G l y R in rat hippocampal pyramidal cells (Shirasaki et al., 1991) and acetylcholine receptors in mouse brain (Rognan et al., 1992). In spinal cord, bicuculline concentrations of greater than 20 p M can antagonize G l y R ( L i et al., 2003). On the other hand, when strychnine is applied at higher than 1-2 p M concentrations, it can antagonize G A B A A receptors in rat hippocampal pyramidal neurons and dorsal horn of spinal cord (Shirasaki et al., 1991; L i et al., 2003). Moreover, certain salts of bicuculline such as bicuculline methiodide, methobromide, and methochloride can inhibit the afterhyperpolarization current (/AHP) in thalamus, resulting in an enhancement of the low-threshold calcium spikes (Debarbieux et al., 1998; Seutin et al., 1997). While non-specific actions of strychnine and bicuculline appear at concentrations much higher than that required for a maximal specific action, it is prudent to carefully examine the specificity of these drugs when strychnine and bicuculline are used to identify G l y R and G A B A A receptors. Gabazine (Fig. 1.4), also referred to as SR-95531, is another G A B A A antagonist that largely overcomes the methodological difficulties due to non-specific actions of bicuculline (Rognan et Chapter 1. Introduction - 17 -al., 1992; Wermuth et al., 1987). Gabazine is more potent than bicuculline in blocking G A B A A receptors. It is blocks G A B A A currents activated by muscimol or G A B A at an IC50 three times lower than bicuculline (Yu and Ho, 1990; Ueno et al., 1997). A t 10 |0.M, gabazine completely blocks G A B A activated currents (Ueno et al., 1997). Initial iontophoretic findings indicated that gabazine does not antagonize mammalian G l y R (Michaud et al., 1986). This finding has been confirmed by bath application of gabazine, where 10 (IM concentration does not block glycine activated CI" currents in hippocampus (Mori et al., 2002). The improved specificity over bicuculline for G A B A A receptors makes gabazine suitable for differentiation between G A B A and glycine mediated CI" currents. COOH CACA TPMPA Figure 1.5. Chemical structure of selective G A B A C agonists, cw-4-aminocaproic acid ( C A C A ) , and antagonist, (l,2,5,6-tetrahydropyridine-4-yl)methylphosphinic acid ( T P M P A ) . G A B A c receptors have their unique pharmacological characteristics. Unlike G l y R and G A B A A , G A B A c receptors are not antagonized by bicuculline or strychnine (Chebib and Johnston, 2000). Instead, they are selectively antagonized by (l,2,5,6-tetrahydropyridine-4-yl)methylphosphinic acid ( T P M P A ; Ragozzino et al., 1996). In addition, G A B A c , but not G A B A A or G l y R receptors, are activated by cw-4-aminocaproic acid ( C A C A ; Matthews et al., 1997). While antagonist sensitivity is an excellent pharmacological tool for differentiating G l y R and G A B A A receptors from G A B A c receptors, the selective G A B A C agonist C A C A has limited utility because it can Chapter 1. Introduction -18-indirectly activate G A B A A receptors by inhibiting G A B A transporters (Chebib and Johnston, 1997). The chemical structures of selective G A B A C agonist and antagonists are depicted in Figure 1.5. T P M P A does not affect resting permeability and G A B A e r g i c actions of pentobarbital on thalamic neurons (Wand and Pui l , 2002). G A B A B receptors are distinct from G l y R and G A B A A receptors, both electrophysiologically and pharmacologically (reviewed by Bettler et al., 2004). G A B A B receptors are G protein-coupled receptors, sensitive to G T P analogs (Hi l l et al., 1984; Asano et al., 1985). The predominant G proteins activated by G A B A B receptors include inhibitory GoCj and Gcc 0 proteins (Morishita et al., 1990; Campbell et al., 1993; Greif et al., 2000). Agonist binding to G A B A B receptors results in inhibition of adenylyl cyclase and inositol triphosphate synthesis, thus activating postsynaptic inwardly rectifying K + channels (GIRKs) and inhibiting presynaptic voltage gated C a 2 + channels (Campbell et a l , 1993; Mott and Lewis, 1994; Greif et al., 2000). In addition to the ionic nature, the slow kinetics of postsynaptic G A B A s e r g i c K + currents differentiates them from glycinergic and G A B A A e r g i c currents (Malouf et al., 1990; Liischer et al., 1997). G A B A B receptors are pharmacologically distinct from other inhibitory receptors. In fact, the G A B A B receptor was first discovered by its unique pharmacology compared to other G A B A receptors (Bowery et al., 1980). A number of selective and potent antagonists have been synthesized (Brauner-Osborne and Krogsgaard-Larsen, 1999). Among these, CGP-35348 is exceptionally useful for pharmacological identification of G A B A B currents, due to its high potency and selectivity (Olpe et al., 1990). CGP-35348 completely blocks G A B A B in thalamus at nanomolar concentrations (Tennigkeit et al., 1998; Ran et al., 2004). Hence, a combination of Chapter 1. Introduction - 19 -pharmacological and electrophysiological tools can be used to identify postsynaptic G A B A R inhibition in thalamus. 1.2.2. Ventrobasal nuclei of thalamus The thalamus is a nuclear complex of the diencephalon. Thalamic neurons and nuclei can be classified into first and higher order relays (Sherman, 2005). A first order relay receives input from a subcortical site and relays it to cortex for the first time. Higher order relays receive input from one cortical area and relays that information to another cortical area (Sherman, 2005). Hence, the higher order relays are involved in a more elaborate role of thalamus in processing of sensory information and pain. This elaborate role involves awareness (Smythies, 1997), attention (Paus, 2002), sleep and wakefulness (Steriade, 2005), and memory (van Groen et al., 2002). Ventrobasal nuclei of thalamus (VB) constitute the principal thalamic relay centre for somatosensory information. These nuclei receive somatosensory information via medial lemniscus and spinothalamic tracts and project to the primary somatosensory cortex (Feldman and Kruger, 1980; M a et al., 1986). In addition, V B receives input from somatosensory cortex and interconnects with other thalamic nuclei (Zhang and Deschenes, 1997). In the cat, corticothalamic axons make excitatory synapses with ventrobasal inhibitory interneurons (Barbaresi and Manzoni, 2003). These interconnections imply an essential role for the V B thalamus in somatosensory processing, pain perception, and sleep and wakefulness (Jones, 1991; Steriade et al., 1997; Steriade, 2005). Chapter 1. Introduction - 20 -1.2.2.1. Functional anatomy The ventrobasal nuclei of thalamus are the most caudal segment of the ventral group of thalamic nuclei, located just ventral and rostral to the posterior/pulvinar group of thalamic nuclei (Fig. 1.6; Paxinos and Watson, 1986). These nuclei consist of ventral posterior lateral ( V P L ) and ventral posterior medial ( V P M ) nuclei (Jones, 1991). They receive somatosensory information on touch, pressure, and proprioception via the dorsal column - medial lemniscus pathway (Tsumoto, 1974), and on pain and temperature via the spinothalamic pathway (Jones, 1991). Neurons in ventrobasal thalamus are classified into two types based on their anatomical projections; thalamocortical relay (TC) neurons that project outside the ventrobasal thalamus (Yen and Jones, 1983) as well as a small number of local circuit interneurons (Harris and Hendrickson, 1987; Ohara and Lieberman, 1993). T C neurons comprise the majority of neurons in V B . In contrast to local circuit interneurons, T C neurons are larger with many short radiating dendrites, giving them a bushy appearance (Tombol, 1967). They can be electrophysiologically identified by a highly non-linear voltage-current relationship and burst firing mediated by low threshold calcium channels (Turner et al., 1997). L o w threshold calcium potential appears to be a characteristic of all T C neurons o f mammalian thalamus (Pape and McCormick, 1995). The expression of low threshold calcium channels provides T C neurons with the ability to exhibit two distinct firing modes. In the tonic mode, membrane potential is predominantly depolarized resulting in inactivation of low threshold calcium channels. In this mode, a depolarizing input results in low frequency tonic firing of action potentials. In the burst mode, the membrane potential is predominantly hyperpolarized resulting in de-inactivation of low threshold calcium channels. In this mode, a Chapter 1. Introduction -21 -depolarization input triggers a low-threshold calcium conductance that results in high frequency bursts of action potentials (Weyand et al., 2001). In addition to above characteristics, TC neurons in V B exhibit a delayed tonic firing upon depolarization, as compared to TC neurons in other thalamic nuclei (Turner et al., 1997). Figure 1.6. Schematic drawing of ventrobasal thalamus and associated structures in rat. B is the magnification of the boxed area in A. Abbreviations: Eth, Ethmoid nucleus; Hipp, Hippocampus; ic, Internal capsule; ml, Medial lemniscus; nRt, Nucleus reticularis thalami; Pfc, Parafascicular nucleus; Po, Posterior/Pulvinar thalamus. Morphologically, at least two types of TC neurons have been identified in ventrobasal thalamus (Yen and Jones, 1983; Hirai et al., 1988; Turner et al., 1997). Type-I TC neurons have larger polygonal soma and larger dendritic field (Fig. 1.7A), while type-II TC neurons have smaller, more spherical soma and smaller, more compact dendritic field (Fig. 1.75; Turner et al., 1997). A B Chapter 1. Introduction - 22 -The two types of T C neurons likely represent distinct functional entities as they project to separate layers of somatosensory cortex. Type-I T C neurons project to layers I V and V of somatosensory cortex, while type-II T C neurons project to layers I and II (Penny et al. 1982). The two types of T C neurons exhibit distinct electrophysiological characteristics. Type-I T C neurons have lower input resistance than the type-II T C neurons and exhibit less inward rectifying behaviour. Type-I T C neurons also exhibit a smaller after-hyperpolarization (Turner et al., 1997). The greater peak input resistance of type-II neurons leads to a greater predisposition to delta oscillation characteristics of deep sleep state in this type of neurons. This predisposition to exhibit delta oscillation could have physiological significance during slow-wave sleep (Turner et al., 1997). Hence, the two types of T C neurons are not only morphologically, but also functionally distinct. Figure 1.7. Camera lucida drawings of typical type-I T C neuron (A), type-II T C neuron (B) and local circuit interneuron ( Q in the cat V B nuclei. Adopted from Jones, 1991. Interneurons are morphologically distinguished by their small size and wavy dendrites that show smooth proximal segments and frequent dendritic spines towards the periphery (Fig. 1.7C; Tombdl, 1967). The morphologically identified local circuit interneurons in cat V B show a more Chapter 1. Introduction - 23 -depolarized membrane potential (relative to T C neurons), have a more linear voltage-current relationship and lack the high frequency (burst) firing associated with a low threshold calcium channels (Turner et al., 1997). The proportion of local circuit interneurons greatly varies among different species and depends on the method used to distinguish interneurons from T C neurons. When morphologic criteria are used to identify local circuit interneurons (Ohara and Lieberman, 1993), rat V B does not contain a significant number of interneurons. This finding is in contrast with the V B in cat and monkey, where significant numbers of local circuit interneurons are observed (Ohara and Lieberman, 1993; Barbaresi and Manzoni, 2003). When stained for glutamate decarboxylase ( G A D ) , a marker for G A B A e r g i c neurons, Harris and Hendrickson observed that local circuit interneurons comprise less than 0.4% of rat V B neurons (1987). However, their study was based on the assumption that V B interneurons use G A B A as their neurotransmitter. Major inputs to somatosensory thalamus include the medial lemniscal tract, the spinothalamic tract and the trigeminothalamic tract (Jones, 1981; Welker, 1973). The areas of termination of the spinothalamic and the medial lemniscal tracts largely overlap in the rat V B . This overlap is somatotopically organized, i.e. inputs from a particular part of the body through both pathways terminate in the same area of the V B (Ma et al., 1986). Nevertheless, there is some organization of sensory modalities in the V B . For example, neurons receiving wide dynamic range and nociceptive specific mechanical inputs tend to exist in the periphery of the V B , while neurons receiving low-threshold mechanical inputs towards the centre (Rausell and Jones, 1991). Chapter 1. Introduction -24-The V B nuclei relay the somatosensory input to Brodman areas 1, 2, 3 a and 3b of cortex, the primary somatosensory cortex. In the monkey, the V B nuclei project to the layer I, III and I V of primary somatosensory cortex (Burton and Jones, 1976). In the rat, these nuclei project to layer I, III, IV , V and V I (Herkenham, 1980). Corticothalamic neurons located in layer V I of the primary somatosensory cortex provide feedback to T C neurons, which is thought to influence the activity patterns and sensory response properties of T C neurons (Alitto and Usrey, 2003). The V B nuclei of thalamus are surrounded by, and interconnected with several thalamic and perithalamic nuclei. The major nucleus in this category is nRt, which surrounds most of the rostrolateral aspect of the V B thalamus (Fig. 1.6; Herrero et al., 2002). The main sources of inputs to nRt are collaterals from thalamocortical and corticothalamic axons (Crabtree, 1996). In return, the nRt neurons mainly project to the V B thalamus as well as other nuclei of dorsal thalamus (Ilinsky et al., 1999; Gentet and Ulr ich, 2003). Zona incerta covers the ventral aspect of the V B , appearing as a layer of grey matter continuous with nRt (Fig. 1.6). This nucleus is a heterogeneous collection of neurons with extensive interconnections with somatosensory thalamus (Mitrofanis, 2005). Zona incerta receives some input from the V B nuclei (Roger and Cadusseau, 1985) and projects extensively to the somatosensory as well as other areas of thalamus (Power et al., 1999). Caudal to the V B thalamus is the parafascicular nucleus (Pollin et al., 1997). This nucleus constitutes the caudal division of intralaminar thalamic nuclei, a group of midline thalamic nuclei with interconnections with many areas of the cortex, thus sometimes referred to as "non-specific Chapter 1. Introduction - 25 -thalamic nuclei" (Herrero et al., 2002). Reciprocal interactions between the parafascicular nucleus and the V B thalamus have been observed in rats (Pollin et al., 1997). Between the parafascicular nucleus and the rostral end of the medial lemniscus, lies a distinct nucleus that appears under low-power light microscopy to be perforated by axons seemingly coming from the medial lemniscus. Paxinos and Watson (1986) named this nucleus "ethmoid", due to its sieve-like appearance due to the perforating axons. The scarce literature on the ethmoid nucleus is rather confusing, as some authors have erroneously referred to this nucleus as the medial part of the medial geniculate nucleus (cf. Paxinos and Watson, 1986; Lund and Webster, 1967a & 1967b). However, this nucleus appears to receive somatosensory input and to interconnect with the V B thalamus (Lund and Webster, 1967a & 1967b). In addition to the above-mentioned nuclei, ventrobasal thalamus receives inputs from an exhaustive list o f nuclei in C N S (cf. Krauthamer et al., 1977). We have limited the focus of this thesis to the nuclei surrounding ventrobasal thalamus for practical reasons. 1.2.2.2. Synaptic inhibition in ventrobasal thalamus Somatosensory stimulation produces a sequence of an excitatory postsynaptic potential (EPSP) followed by an IPSP in ventrobasal neurons (Salt and Eaton 1990). Electrical stimulation usually produces EPSPs, and occasionally, IPSPs in these nuclei (Mishima 1992). Excitatory inputs to the V B thalamus are mainly monosynaptic glutamatergic spinothalamic (Blomqvist et al., 1996) and medial lemniscal pathways (Salt and Eaton, 1996). The nature of inhibitory inputs to the V B thalamus, however, is more complex. Chapter 1. Introduction - 2 6 -Inhibitory responses of somatosensory origin in TC neurons of the VB thalamus are disynaptic (Baldissera and Margnelli, 1979). A typical disynaptic inhibitory circuit involves an interneuron that can be a local circuit or a projecting interneuron. While significant number of local circuit interneurons in the VB of cat and monkey may account for these inhibitory responses, rat VB contains a negligible number of local circuit interneurons (Ohara and Lieberman, 1993; Harris and Hendrickson, 1987). Hence, it is possible that inhibition in rat VB essentially originates outside these nuclei. As discussed below, however, the absence of local circuit interneurons in the VB thalamus must be viewed with caution. There is a strong inhibitory input to the VB neurons of rat from projecting interneurons located in the nRt (Cox et al., 1997; Gentet and Ulrich, 2003). This inhibition is mostly mediated by GABA A and less frequently by GABA B receptors (von Krosigk et al., 1993; Zhang et al., 1997). The nRt is not the only source of inhibition in thalamus. The ZI terminations within somatosensory thalamic nuclei carry GABAergic inhibitory signals (Bartho et al., 2002). Moreover, electrical stimulation of parafasicular nucleus evokes an inhibitory response in the VB thalamic neurons, the nature of which is poorly understood (Pollin et al., 1997). Inhibitory synaptic responses play an essential role in function of ventrobasal thalamus. Inhibition originating from the nRt plays a critical role in the generation of thalamocortical sleep rhythms. Activation of the nRt neurons results in the generation of IPSPs in TC neurons. These IPSPs de-inactivates low threshold calcium channels in neurons (burst mode), resulting in a rebound low-threshold calcium bursts of action potentials on depolarizing return to the resting Chapter 1. Introduction - 27 -membrane potential. These bursts of action potentials activate thalamocortical neurons that send glutamatergic inputs to nRt. A s a result, EPSPs are generated in the nRt neurons, which w i l l activate these neurons. The resulting cycle of inhibition-excitation generates spindle waves and rhythmic oscillations. The spindle waves and oscillations are crucial in the generation of some sleep rhythms and in inhibiting the relay of peripheral signals through thalamocortical neurons, which leads to unconsciousness (McCormick and Ba i , 1997; Steriade, 2005). The functions of other types of inhibition in the V B thalamus are less understood. Inhibition originating in the zona incerta has been associated with attention and interactions between motor and sensory systems (Mitrofanis, 2005). Inhibition originating in the parafascicular nucleus to V B thalamus is likely associated with nociceptive system (Pollin et al., 1997). The nature of interconnections between the ethmoid nucleus and the V B thalamus are yet to be understood. Chapter 1. Introduction - 28 -1.3. Rationale and objectives A neurotransmitter role for glycine is well established in spinal cord and brainstem, but not the rostral C N S (Werman et al., 1968; Young and Snyder, 1973). In the latter region, G A B A is the dominant mediator of synaptic inhibition (Sivilotti and Nistri , 1991). The anatomical distinction between glycinergic and G A B A e r g i c inhibitions, however, is less clear in light of the observations that glycine and G A B A are often co-transmitters (Donato and Nistr i , 2000; Dumoulin et al., 2001). In some regions of C N S , these two neurotransmitters are co-released from the same synaptic terminals (Jonas et al., 1998). The presence or absence of glycinergic inhibition in the forebrain deserves re-assessment. Several lines of evidence support the hypothesis that glycine has a neurotransmitter role in somatosensory thalamus. A substantial number of nerve fibres in rat somatosensory thalamus show immunoreactivity for glycine (Rampon et al., 1996). Glycine is released in ventrobasal nuclei of thalamus of cats during the slow-wave sleep (Kekesi et al., 1997). The thalamus exhibits a strong autoradiographic signal of m R N A for the P subunit of glycine receptor (Malosio et al., 1991). There is less strong indication for the presence of a\ and a 2 subunits, which are required to form functional G l y R (Malosio et al., 1991). However, low sensitivity limits the interpretation of autoradiographic m R N A studies. Significant [ 3H]strychnine binding in the somatosensory thalamus of rat and mice is consistent with the presence of G l y R (Young and Snyder, 1973; Zarbin et al 1981; Frostholm and Rotter 1985). The neuronal form of glycine transporter is evident in ventrobasal thalamus (Jursky and Nelson, 1995; Zeilhofer et a l , 2005), consistent with glycinergic transmission. Most recently, strychnine was shown to abolish pentobarbital-induced oscillations in ventrobasal thalamus (Ran et al., 2004). Hence, it is Chapter 1. Introduction -29 -reasonable to hypothesize an inhibitory role for glycine-like amino acids in somatosensory thalamus. While the indirect evidence seems robust, glycinergic inhibition in the somatosensory thalamus has not been directly substantiated. This thesis addresses the existence of glycinergic inhibition in the major somatosensory complex of rat thalamus, the ventrobasal nuclei. To provide initial rationale for glycinergic transmission, we looked for the presence of functional G l y R in T C neurons of the rat ventrobasal thalamus. The expression of G l y R a subunits, required for the formation of functional receptors was examined by immunohistochemical staining of ventrobasal thalamus with antibodies against the a\ and 0C2 subunits. To study the presence of functional receptors, we investigated the effects of G l y R agonists on T C neurons of ventrobasal thalamus. Electrophysiological and pharmacological characteristics of these responses were analyzed to assess whether G l y R mediated the effects. The results substantiated the presence of functional GlyRs in somatosensory thalamus. Hence, we asked whether GlyRs in ventrobasal thalamocortical neurons were involved in synaptic transmission or are limited to extrasynaptic sites? To answer this question, IPSPs and IPSCs in rat ventrobasal thalamus were evoked from medial lemniscus, the major somatosensory input to thalamus, as well as the neighbouring zona incerta, nRt, the ethmoid nucleus, and Pfc. The nature of the evoked IPSPs and IPSCs were electrophysiologically and pharmacologically investigated, using G l y R and G A B A receptors antagonists. To determine the validity of the antagonists as a pharmacological tool to differentiate between inhibitory receptors, possible non-Chapter 1. Introduction - 30 -specific actions of the antagonists were examined. Mono- or polysynaptic nature of the synaptic inhibition was investigated by comparing the latency fluctuation of the evoked IPSPs and IPSCs. The physiological implication of multiple inhibitory systems in ventrobasal thalamus is an intriguing question. A hypothetical heterogeneity of the biophysical characteristics of these inhibitory systems would allow the ventrobasal T C neurons to integrate sensory information. To test this hypothesis, we compared the biophysical characteristics of glycinergic and G A B A e r g i c synaptic inhibition in ventrobasal thalamus. The kinetics and CI" permeability of the synaptic receptor channels, estimated from whole-cell IPSCs, were the endpoints. The biophysical characteristics of the synaptic channels were also compared and contrasted with those of extrasynaptic single channel currents in the ventrobasal T C neurons, obtained in our laboratory by Dr. Hee-Soo K i m . The model used for estimating CI" permeability of the synaptic receptor channels was extensively examined by computer simulation studies and pharmacological isolation of CI" currents, to ensure reliable estimates. Overall, the results shed new light on transmitter-receptor systems involved in thalamic inhibition. -31 -Chapter 2 METHODS The Animal Care Committee of University of British Columbia approved the use of animals in these experiments. 2.1. Immunohistochemistry 2.1.1. Tissue fixation and slice preparation Sprague-Dawley rats (13-15 day-old), deeply anesthetized with halothane, were transcardially perfused with cold 4% formaldehyde in Dulbecco's phosphate-buffered saline (PBS; Invitrogen Corp., Carlsbad, U .S .A. ) at p H = 7.4. After decapitation, the brain was removed and blocked by sectioning along the interhemispheric fissure. The tissue block was fixed overnight and cryoprotected in 30% sucrose for 3 days. Using a freezing microtome, we sagittally sectioned the tissue at 10 pm intervals into 10-20 pm thick slices containing forebrain regions, including dorsal thalamus. These sections included medulla as a control. 2.1.2. Staining procedures Freshly cut sections, rinsed with P B S , were incubated for 20 min in 10% normal blocking serum to block non-specific staining. On rinsing, they were incubated for 2 h at 4 °C with a primary goat antibody against the C-terminal of cii subunit, or N-terminal of a 2 subunit of the rat G l y R (Santa Cruz Biotechnology, Santa Cruz, U .S .A . ) . These polyclonal antibodies were raised against peptide sequences of 449 and 452 amino acid lengths, respectively. Information supplied by the commercial provider indicated that a B L A S T search for these sequences yielded a perfect Chapter 2. Methods - 32 -match with G l y R ai and G l y R a 2 subunits of rat, mouse, and human. N o significant homologous sequences were found in other known proteins (Santa Cruz Biotechnology, Santa Cruz, U . S . A . ) , indicating that the antibodies were likely specific. The primary antibody solution contained 1.5% normal blocking serum. Although we tested primary antibody dilutions in P B S that ranged from 1:100 to 1:1000, we obtained optimal staining on 1: 1000 dilutions. We also attempted co-staining with primary rabbit antibodies against glycine or taurine molecules and primary guinea pig antibody against the C-terminal of Vesicular Inhibitory Amino A c i d Transporter (Chemicon Inc., Temecula, U .S .A . ) . We tested the primary antibody against glycine or taurine of two different batches over a dilution range of 1:100 to 1:1000 (n = 5 rats for each dilution). Following thorough rinsing, the sections were again incubated for 2.5 h at room temperature (22 °C) with a secondary chicken anti-goat IgG antibody, conjugated with Alexafluor 488 (Molecular Probes, Eugene, U . S . A ) and diluted 1:400 in P B S . After rinsing, the sections were treated for 5 min with 4',6-diamidino-2-phenylindol (DAPI , 300 nM) to visualize nuclei. After rinsing, the sections were treated with a drop of Prolong gold antifade (Molecular Probes) and left overnight. Some sections in each experiment were incubated without either primary antibody, to control for non-specific secondary staining (negative controls). Some sections were stained with haematoxylin and eosin ( H & E ) for visualization of cellular morphology. After rinsing with water, freshly cut sections were incubated for 5 min in haematoxylin. The sections were then rinsed in tap water and placed in 1% acid alcohol (1% Chapter 2. Methods - 33 -hydrochloric acid in 70% ethanol) for <5 seconds. After washing off acid alcohol with tap water, the sections were placed in eosin (1 g/100 ml) for 5 min. Sections were then washed, dehydrated and mounted on slides for light microscopy. 2.1.3. Imaging Immunostained sections were visualized with x l O dry (N .A. 0.3) and x63 oil-immersion ( N . A . 1.4) objectives on a multi-photon laser scanning confocal microscope (Leica T C S M P , Leica Microsystems, Wetzlar, Germany). Bright field microscopy was performed on H & E stained sagittal brain sections with x2.5 dry (N.A. 0.075) and x63 oil-immersion (N .A. 1.4) objective lenses on a Zeiss Axiolmager Model Z I (Carl Zeiss, Germany). Another set of bright field microscopy was performed on unstained sagittal brain sections with a x2.5 dry (N .A. 0.075) objective lens on a Zeiss AxioVert Model 25 (Carl Zeiss, Germany). The images were processed using Adobe Photoshop software (Adobe, San Jose, U .S .A . ) . 2.2. Electrophysiological recordings 2.2.1. Whole-cell recordings 2.2.1.1. Slice preparation Sprague-Dawley rats (13-15 day-old) were decapitated while under deep halothane anesthesia. The brains were rapidly removed and submerged in oxygenated solution at 4 °C containing (in HIM): N a H C 0 3 , 26; N a H 2 P 0 4 , 1.25; KC1, 2.5; M g C l 2 , 2; C a C l 2 , 2; dextrose, 25; sucrose, 250. Two brain tissue blocks were obtained by sectioning along the interhemispheric fissure, and parallel to interhemispheric fissure at ~ 0.25 cm from the lateral aspects of each hemisphere. The medial surface of the block was glued to the Teflon stage of a vibroslicer (Campden Chapter 2. Methods - 34 -Instruments, London, U K ) . The tissue was sectioned into 250-300 [im thick slices, containing ventrobasal thalamic nuclei and medial lemniscus (Paxinos & Watson, 1986). The para-sagittal slices were incubated for 1-5 h in artificial cerebrospinal fluid (aCSF) at 23-25 °C, saturated with 95% C- 2: 5% C 0 2 . The aCSF contained (in mM): N a C l , 124; N a H C 0 3 , 26; N a H 2 P 0 4 , 1.25; KC1 , 2.5; M g C l 2 , 2; C a C l 2 , 2; and dextrose, 10. The p H was 7.3-7.4 and the average osmolality was ~ 305 mosmols measured by a freezing point osmometer (Advanced Instruments, Norwood, U S A ) . 2.2.1.2. Electrophysiological recording Slices were placed in a Perspex recording chamber with a -1.5-2 ml volume and were immobilized with a polypropylene mesh. They were perfused with oxygenated aCSF (21-23°C) at a rate of 1.5-2 ml/min. Ventrobasal neurons were identified under a differential interference contrast microscope at x400 (Axioscope, Carl Zeiss, Germany). The anatomical location of the ventrobasal nuclei of thalamus, where electrophysiological recordings were done, is shown in Figure 2.1. Recording microelectrodes were made from thin-wall borosilicate glass tubing (World Precision Instruments, Sarasota, U .S .A. ) using a Narishige puller, and filled with an intracellular solution with a comparison of composition given in Table 2.1. The p H was adjusted to 7.3-7.4. The electrode resistances ranged between 4-9 M Q with an average of 5 M Q . Whole cell patch-clamp recordings were obtained using an Axoclamp-2A (Axon Instruments, U .S .A . ) or a List EPC-7 amplifier ( H E K A , Germany) in both the current- and voltage-clamp Chapter 2. Methods - 35 -modes. Signals were filtered at 3 kHz, digitized at 10 kHz with a 16-bit data acquisition system (Axon Instruments) and stored using pClamp software (Axon Instruments). Figure 2.1. Anatomical locations of electrophysiological recordings and electrical stimulations. A, position of thalamus in the parasagittal slices. B, magnification of boxed area in (A) identifying ventrobasal thalamus and the surrounding structures. Electrophysiological recordings were done in the ventrobasal thalamus, mainly in the rostral and ventral outskirts. Eth, Ethmoid nucleus; Hipp, Hippocampus; ml, Medial lemniscus; nRt, Nucleus reticularis thalamus; Pfc, Parafascicular nucleus; Thai, thalamus; V B , ventrobasal thalamus; ZI , zona incerta. Table 2.1. Intracellular whole-cell patch clamp solutions. Normal Cs C s - Q X K-gluconate 140 mM 0 mM 0 mM Cs-gluconate 0 mM 16.5 mM 16.5 mM KC1 5 mM 0 mM 0 mM C s C l 0 mM 128.5 mM 128.5 mM N a C l 4 mM 4 mM 4 mM M g C l 2 3 mM 3 mM 3 mM C a C l 2 1 mM 1 mM 1 mM E G T A 10 mM 10 mM 10 mM H E P E S 10 mM 10 mM 10 mM M g A T P 3 mM 3 mM 3 mM N a 2 G T P 0.3 mM 0.3 mM 0.3 mM Q X - 3 1 4 f 0 0 3 mM Eci -53 m V O m V O m V EK* -84 m V N A N A Free rCa 2 +f 5 nM 5 nM 5 nM 'Calculated 'Lidocaine Af-efhyl bromide N A , not applicable. Chapter 2. Methods - 36 -2.2.1.3. Electrical stimulation IPSCs were evoked by stimulation with a bipolar tungsten electrode (~5 M Q resistance, Wor ld Precision Instruments). The distance between the two poles of the electrode was ~0.3 mm. The electrode was placed in one of the following locations: the medial lemniscus, the caudal part of ZI , nRt, Eth, and Pfc. The distance between the stimulating electrode and the recording electrode was ~ 3 mm for stimulations at medial lemniscus, the caudal part of zona incerta, the ethmoid nucleus, and Pfc. This distance was ~1.5 mm for stimulations at nRt. These anatomical locations are shown in Figure 2.1. A n isolated stimulator (Digitimer, Hertfordshire, U K ) was employed for electrical stimulation. The stimulator was connected to the bipolar electrode and controlled by pClamp software. Single pulse stimuli (duration, 0.05-1 ms) were delivered at <0.5 H z to avoid short-term depression of IPSPs/IPSCs. Typically, stimuli were adjusted to evoke maximal responses with no failures (1.1 x threshold). Long intervals separated a group of 10 stimuli. Stimulus parameters remained constant for paired analysis of receptor antagonist effects. A s a control, stimulation of subcortical white matter at ~ 3 mm from the ventrobasal thalamus in 4 slices did not yield synaptic responses. 2.2.2. Single channel recording Single channel recording and analysis were results of a collaborative study in our laboratory. Dr. Hee-Soo K i m carried out the laboratory preparation and data acquisition, the method of which is briefly described for completeness. Biophysical properties of extrasynaptic receptors were Chapter 2. Methods - 37 -obtained from the single channel data and used for comparative studies of synaptic and extrasynaptic receptors. Acutely dissociated T C neurons were prepared from slices containing the ventrobasal nuclei. The method used to prepare brain slices was the same as described for whole-cell recordings. The slices were initially incubated at room temperature (21-22 °C) for 10 min in an oxygenated, Ca 2 +-free media composed of (in mM): 120 N a C l ; 5 KC1; 1 M g C l 2 ; 5 D-glucose; 20 1,4 piperazine-bis-(2-ethanesulfonic acid) (PIPES); ethylene-glycol-bis(2-aminoethylether)-N, N , N ' , N'-tetraacetic acid ( E G T A ) , and 2 mg/ml bovine serum albumin (BSA) at p H = 7.3. The tissue was then stirred at 32°C for 45 minutes in a solution of composition (in mM): 120 N a C l ; 5 KC1 ; 1 M g C l 2 ; 1 C a C l 2 ; 20 PIPES; 2 mg/ml B S A ; and, 14 units/ml papain (Sigma Chemical Co. , St. Louis, U .S .A . ) at p H = 7.0. The tissue was rinsed and left for 15 minutes at room temperature. The cells were mechanically dispersed in 2 ml of C a 2 + - and BSA-free PIPES solution and plated on uncoated 35 mm tissue culture dishes. The cells remained in PIPES buffered solution at room temperature until needed for recording. Single channel currents were recorded at room temperature ( K i m et al., 2004). Dispersed ventrobasal neurons were bathed in a saline containing (in mM): 4 KC1; 135 N a C l ; 10 C a C l 2 ; 1 M g C l 2 ; 10 H E P E S ; and, 5 D-glucose. The p H was 7.3. Patch pipettes (10-15 M i l ) contained a solution (pH = 7.3) composed of (in mM): 135 C s C l ; 1 M g C l 2 ; 0.267 C a C l 2 ; 10 H E P E S ; 3 E G T A ; and 5 D-glucose. Ea was zero millivolts in these recordings. Chapter 2. Methods - 38 -Outside-out membrane patches were voltage-clamped with a List EPC-7 amplifier at a holding potential, V n = -60 m V . The currents were filtered at D C to 1 kHz , digitized (8 kHz) and analyzed off-line with commercial software (Instrutech Corp., New York, U . S . A . ) . Single channel openings were detected as transients exceeding 50% of the difference between the averaged baseline and open channel currents, disregarding events briefer than 180 ps. Distributions of open channel times were fitted by a triexponential function. Closed time distributions were fitted by 4 exponential terms. Exponential fitting was performed using Simplex maximization of likelihood. Groups of openings were defined as bursts, provided that the openings were separated by gaps shorter than tc, a specified critical time. t c was calculated by solving l-e c3 = e a where xC2 and xC3 were the time constants of the second and third fastest components in closed time distributions (Colquhoun and Sakmann, 1985). Using Simplex methods, 1 or 2 Gaussian terms were fitted to the amplitude distributions of single channel currents. Mean channel conductance was calculated as the weighted sum of the Gaussian fit components. 2.2.3. Computer simulation of synaptic currents Synaptic currents were simulated using MATLAB Version 7.0 (R14) (Mathworks Inc., Natick, U S A ) . Simulations were performed using two algorithms referred to as the "classic algorithm" and the "stationary-segments algorithm" (see below). Synaptic currents were first simulated in an optimal (no-noise) condition. The effect of noise was studied by adding simulated random Gaussian noise to these simulated currents (Benke et al. 1998). Chapter 2. Methods - 3 9 -The classic algorithm was based on assumptions originally outlined by Sigworth (Sigworth 1980; reviewed by Alvarez et al., 2002). Current fluctuations due to the gating of single channels were simulated, which could be closed (no current passing) or open (passing unitary current of i). Using the M A T L A B built-in binomial random generator and open probability input parameter of Po( t), an open or closed state was assigned to each channel during a synaptic current for the duration of 8 x decay time constant at time intervals of 1/sampling rate. The open probability of a single channel during a synaptic current was calculated as, z i z i P o ( t ) = ( l - e T ' ) ^ T 2 where t was time and Ti and %2 were rise and decay time constants of the synaptic current respectively. The sum of N number of simulated single channel currents produced a synaptic current. Simulating single channel currents separately has the advantage that the open or closed state of a single channel at each time point is independent from the channel activity before or after that time point or the activity of other channels. The duration 8 xdecay time constant was chosen to allow complete decay of the synaptic current. The stationary-segments algorithm simulated synaptic currents that contained segments during which the single channel current did not change (stationary segments). To obtain this stationarity, the code used for the classic algorithm simulations was modified such that when the output of the binomial random generator changed from closed to open and vice versa, the channel would maintain the new state for the duration of stationary segment. We simulated synaptic currents for sets of defined i, N , %\, i2, and sampling rate. Every ten, twenty, forty, sixty, eighty, or 100 synaptic responses generated with the same parameters were Chapter 2. Methods - 40 -grouped and treated as recordings from one "virtual" neuron. Five neurons were simulated for each condition. In total, we simulated more than 6400 currents grouped as more than 200 "virtual" neurons. The Matlab codes for simulation of synaptic currents are presented in Appendix A . 2.3. Drugs and application Drugs were applied by bath perfusion at 1-2 ml/min. The bath kinetics were established using opaque and vital dyes, showing wash-in time of ~ 3 min. The bath was cleared from the dye after ~ 30 min (bath washout time). Stock drug solutions were prepared in distilled water and diluted in aCSF just prior to use. Drug solutions were bubbled with 95% O2, 5% CO2. (3-Alanine, 2-amino-5-phosphonovalerate ( A P V ) , bicuculline methiodide, CGP-35348, 6-cyano-7-nitro-quinoxaline-2,3-dione ( C N Q X ) , 4',6-diamidino-2-phenylindol (DAPI) , G A B A , gabazine, glycine, kynurenic acid, D-serine, L-serine, strychnine, taurine, and tetrodotoxin ( T T X ) were purchased from Sigma Chemical Co. (St. Louis, U .S .A . ) . (RS)-a-methyl-4-carboxyphenylglycine ( M C P G ) was purchased from Precision Biochemicals (Vancouver, Canada). Pancuronium bromide was purchased from Baxter Corporation (Ontario, Canada). 6-aminomethyl-3-methyl,l-4H-l,2,6-benzothiadiazine-l,l-diazide hydrochloride ( A M B D ) was a gift from Merck Frosst Co. (Quebec, Canada). 2.4. Data analysis 2.4.1. General measurements Electrophysiological data were analyzed using pClamp software (Axon Instruments). Membrane Chapter 2. Methods -41-voltages were corrected for a liquid junction potential of-11 m V . Input resistance was measured from <5 m V voltage responses to hyperpolarizing current pulses. Slope resistance was measured in current-clamp from a linear portion of voltage-current relationships between -65 and -85 m V . 2.4.2. Rheobase and chronaxie We determined chronaxie and rheobase using the method of Nowak and Bull ier (1998). The strength and duration of electrical stimulation were gradually increased in a step-wise manner until IPSPs or IPSCs were evoked with minimal (<10%) failure. Min imum stimulation strength required to reliably evoke IPSPs and IPSCs was recorded for all stimulation durations. The recorded stimulation strengths were plotted against inverse stimulation durations. The relationships were fitted with a linear function using least squares fit. The linear function was, S = R.C.— + R D where S was minimum stimulation strength, D was stimulation duration, R was rheobase and C was chronaxie. 2.4.3. Latency and Latency fluctuation We calculated the mean latency for 5-10 IPSPs or IPSCs of each neuron. Latency of the response was measured as the time interval from the start of stimulus to the onset of the response, defined as the point at which the postsynaptic potential or current visibly exceeded the noise level. In some cases where the stimulus artifact masked the onset of the response, we estimated the onset by fitting a single exponential function to the rising phase of the response and determining the point where the curve crossed the baseline. The latencies measured by both procedures were consistent with each other and the data were pooled. Chapter 2. Methods -42-We calculated the latency fluctuation (also referred to as latency jitter by some authors) for IPSPs and IPSCs (Wierenga and Wadman, 2003) from the absolute differences between the latencies of 5 to 10 individual IPSPs or IPSCs and the averaged IPSP or IPSC latency. Latency fluctuation was then calculated as the coefficient of variation of these differences using the formula, where A/ , is the absolute differences between the latency of i IPSPs or IPSCs and the averaged IPSP or IPSC latency, Almean is the average of the absolute differences between the latency of individual IPSPs or IPSCs and the averaged IPSP or IPSC latency and n is the number of IPSPs/IPSCs. 2.4.4. Detection of spontaneous IPSCs (sIPSCs) Spontaneous IPSCs were recorded during intracellular application of C s + and QX-314 in neurons that were voltage clamped at -60 m V . Single sIPSCs were visually selected for averaging and creation of search templates. We used the sliding-template procedure of pClamp software, setting the template match stringency to a medium level. Given the observed variablity of sIPSC time courses, multiple template searches were required for precise detection of all sIPSCs. The events were monitored visually during the entire procedure, for rejection of sIPSPs with more than a single peak and noise. Latency fluctuation = A/. mean 2.4.5. Rise and decay time constants Exponential functions were fitted to the rise and decay phases of IPSCs, averaged from 5-10 Chapter 2. Methods -43-individual currents. Single exponential function equation was, -t Aex where A was the amplitude and x was the time constant. The double exponential function was the sum of two terms, Ax.eh +A2.e%2 where A i and A 2 were the amplitudes of the terms with time constants xi and x 2 , respectively. 2.4.6. Non-stationary fluctuation analysis Non-stationary fluctuation analysis was used to estimate unitary current from IPSCs as well as simulated postsynaptic currents (Sigworth, 1980; Traynelis et al., 1993; De Koninck and Mody, 1994). We assessed the stability of quantal release at a stimulation frequency of <0.5 H z by lack of changes in IPSC amplitude and coefficient of variation for groups of three consecutive IPSCs in the train (Scheuss and Neher, 2001). Coefficient of variation for groups of three consecutive IPSCs was calculated as, where C V was the coefficient of variation for amplitude, A , was amplitude of individual IPSCs in the group, Amea„ was average amplitude of the IPSCs in the group, and n was the number of IPSCs in the group (i.e. 3). This was not required for simulated IPSCs because the quantal release did not change during simulations. -t -t A mean C h a p t e r 2. M e t h o d s - 44 -We averaged successive IPSCs, after aligning their peaks in time. Starting at the IPSC peak, the decays were divided into bins (segments) of equal sizes, /mean(t), or the average current of each bin (t) was calculated from the relationship, i __ X*(t)j mean(t) n where I(t>j was the current amplitude for trial j for bin t, and n was the number of trials. The variance (a ) of each bin was calculated from the difference between the scaled average and individual IPSC (cf. Traynelis et al. 1993). For simulated currents, the variance was calculated directly from IPSCs rather than from the difference between the scaled average and individual IPSC. Using a least squares algorithm, the resulting plot was fitted with the quadratic function, 0 2 ( t ) = J c l . / m e a n ( t ) - — 72mean(t) + O t h 2 ( t ) where G th (t) denoted residual noise and ic\ was the elementary current through the agonist-gated channel. The parameter N (total number of channels at the synaptic site) was not considered further for IPSCs since scaling the average IPSC to individual IPSC increases the accuracy of iC\ but decreases accuracy of N (Traynelis et al. 1993). We also obtained *'ci as the slope of the initial part of the variance-to-mean current relationship, fitted by linear regression. The results of the two estimates were in good agreement. The channel conductance was calculated as y = /c i /AV, where y was the channel conductance and A V was the driving force calculated as the difference between the holding potential and Ec\. The channel CI" permeability (PCi) was calculated according to the Goldman-Hodgkin-Katz ( G H K ) constant-field relationship, Chapter 2. Methods - 45 -VF - • R T \-e™ *C1 _ lC\ • y p 2 • VF [ C l - ] i - [ C l ] 0 . e R T where R was the real gas constant (8.3144 J/°K.mol), T was the absolute temperature (°K), F was the Faraday constant (9.65 x 10 4 C/mol), V was membrane potential, and [CT]j and [CT] 0 were the intracellular and extracellular CI" concentrations, respectively. We applied a similar procedure to calculate the Pci from single channel currents. The Matlab codes for non-stationary fluctuation analysis are presented in Appendix B . 2.4.7. Concentration-response analysis Concentration-response relationships were established by cumulative application of the drugs in a step-wise manner. The responses were plotted against the logarithm of drug concentration. These relationships were fitted with one or two sigmoid curves using the least squares fit in Prism GraphPad software (San Diego, U .S .A . ) . The fitting equation for single sigmoid-relationship was, _ max y " ( l + E C 5 0 - [ d r u g ] ) n where max referred to the maximum response (plateau), EC50 (or IC50 for antagonists) was the cconcentration of the drug that produces a half-maximal effect (or inhibition for antagonists), and n was the slope o f the sigmoid curve. The fitting function for double-sigmoid relationship was, _ max,, max, Y _ (1 + E C 5 0 / i - [drug])"* + (1 + E C 5 0 , - [drug])"' where max/, and max/ referred to the maximum response (plateau) minus the minimum response (foot) for the high affinity and low affinity components, ECso/, and EC50/ referred to the Chapter 2. Methods - 46 -respective half-maximal effect concentrations and n/, and n/ referred to the respective sigmoid slopes. 2.4.8. Statistical analysis Data are expressed as mean ± S E M and n denotes number of neurons or patches. Data were statistically analyzed using N C S S software (Kaysville, U .S .A . ) . We examined the goodness of fit for all data sets to a normal distribution (Kolmogorov-Smirnov test). The equation for the sum of Gaussian functions, used to describe frequency distribution histograms was, " e -(*-I2"! f(*)=lLAle- 7 r = — + C When the data set conformed to normal distribution, Student's f-test was used for comparing two unpaired groups, paired f-test was used for comparing two paired groups, A N O V A was used for multiple comparisons of unpaired groups, repeated measures A N O V A was used for multiple comparisons of paired groups, and a Tukey post hoc test for comparing group pairs. When the data set did not conform to normal distribution, Mann-Whitney U test was used for comparing two unpaired groups, Wilcoxon test was used for comparing two paired groups, and Kruskal-Wall is test was used for multiple comparisons. Using bootstrap methods, we estimated the 95% confidence interval for parabolic fits to variance-to-mean current relationships. Significance was defined as P < 0.05. CorelDraw and Corel PhotoPaint softwares were used for preparing figures (Corel Corp., Ottawa, Canada). - 4 7 -Chapter 3 RESULTS Part I - Presence of Glycine Receptors in Ventrobasal Thalamus 3.1. Immunohistochemical evidence Light microscopy of H & E stained slices was performed to identify the cellular morphology o f neurons in ventrobasal nuclei (Fig. 3.1). The medulla, a known glycinergic region (Araki et al., 1988), was examined for comparison. Cellular morphology of neurons was preserved during tissue fixation. Typical ventrobasal neurons, identified from their anatomical location, had a round shape with 20-25 Jim diameter. In contrast, typical medullary neurons had a triangular shape with 15-20 | i m diameter (Fig. 3.IB). 3.1.1. Glycine receptor subunits Multiphoton confocal microscopy images of sections stained with antibodies for ( X i or a 2 subunits are shown in Figure 3.2. Immunoreactivity for ( X i or a 2 subunits was observed in both thalamus and medulla. Ventrobasal thalamus showed low to moderate staining with antibodies for the oti and a 2 subunits. A comparison of Figures 3.1 and 3.2 reveals preservation of cellular morphology in confocal images. A s shown in the insets of Figure 3.2, the staining was both punctate and diffuse, localized mainly to somata and large processes. There was some cytosolic staining evident for both subunits. In the medulla, there was similar but heavier staining for the c c i and oc2 subunits (Fig. 3.2). Negative controls showed negligible staining (Fig. 3.3). We quantitatively assessed the immunoreactivity for a i or oc2 subunits in ventrobasal thalamus Chapter 3. Results - 48 -and medulla by counting positively stained cells in low power field. In ventrobasal thalamus, 19 ± 4.3 % of 511 cells (6 sections) stained for the a i subunit, whereas 30 ± 2.7 % of 555 cells (10 sections) were positive for the a 2 subunit. Medullary cell counts revealed that 53 ± 3.5 % of 311 cells (5 sections) stained for the 0Ci subunit, whereas 50 ± 2.7 % of 269 cells (4 sections) were positive for the ot2 subunit. Hence, although less heavily stained than medulla, ventrobasal thalamus apparently expressed a i and oc2 subunits of glycine receptors. Figure 3.1. Bright field microscopy of parasagittal section of rat brain. A, low-power magnification (objective, x2.5) of slice (10 u,m, thick), stained with haematoxylin and eosin ( H & E ) showing the locations of ventrobasal complex and medulla. B, high-power magnification (objective, x63) shows typical neurons in ventrobasal thalamus and medulla. Chapter 3. Results - 4 9 -Figure 3.2. Antibody staining for oti and 0C2 subunits of G l y R shows low to moderate staining of both subunits (green) in ventrobasal nuclei. In the medulla, more prominent staining was observed. Diffuse (arrows) and punctate (arrowheads) staining was observed in high-power views of selected cells (rectangles) in both areas. There was some cytosolic staining evident for both subunits. Multiphoton confocal microscopy images at low power (objective, x lO) and high power (objective, x63) images also show nuclei stained with D A P I (blue). Each high power image represents a 1.2 pm thick optical section at the nuclear plane. Chapter 3. Results -50-Figure 3.3. Multiphoton confocal microscopy image of ventrobasal nuclei in a negative control slice shows negligible staining. Nuclei are stained with D A P I (blue). 3.1.2. Glycine and taurine molecules Multiphoton confocal microscopy images of sections stained with primary antibodies against glycine and taurine molecules are shown in Figure 3.4. Staining appeared to be all-or-none, i.e. either diffuse or no staining was observed over antibody dilution range of 1:100 to 1:1000 (n = 5 rats for each dilution). Occasional areas of higher density staining were not readily associated with specific histological elements. Repeating the experiments with a different batch of primary antibodies yielded similar results (dilution range of 1:100 to 1:1000; n = 3 rats for each dilution). Hence, the staining was concluded to be non-satisfactory possibly due to a poor specificity of the commercial antibodies and was not further pursued. 8 10 Mm Chapter 3. Results A 10 jim Figure 3.4. Antibody staining for glycine (A) and taurine (2?) shows diffuse staining for both molecules (green) in ventrobasal nuclei. Images also show nuclei stained with D A P I (blue). Each image represents a 1.2 pm thick optical section. 3.2. Electrophysiological evidence Given the a subunit staining, we examined whether these subunits form functional glycine receptors by determining the effects of G l y R agonist applications on the firing and resting membrane properties of in ventrobasal thalamocortical neurons. Under current clamp and using normal patch solution, these neurons typically had a mean resting potential o f -64 ± 0.7 m V , mean input resistance of 405 ± 47 Mf2 and capacitance of 200 ± 23 pF (n = 25). Although there was a general trend for neurons with higher capacitance (indicative of larger somata) to have lower input resistance, there was no clear correlation between the two parameters (R = 0.1, P > 0.05, Kolmogorov-Smirnov goodness-of-fit test). Repetitive firing of action potentials (tonic firing) was evoked by injection of 10 to 100 p A depolarizing current pulses (Fig. 3.5^4). A silent period of ~ 50 ms typically preceded the tonic firing. Hyperpolarizing current pulses of 10 to 100 p A were followed by a rebound Chapter 3. Results 52-depolarization that could initiate high frequency burst firing of action potentials (low threshold spikes, LTS; Fig. 3.5^4). The current-voltage relationship exhibited slight inward rectification at membrane potentials that were 20 mV or more negative to the resting membrane potential (Fig. 3.55). 50 msec B -100 r 4 0 100 -40 Voltage (mV) Figure 3.5. Current clamp recordings a typical ventrobasal TC neuron. A, tonic firing was evoked by injection of 100 pA depolarizing current pulse. Rebound depolarization upon 100 pA hyperpolarizing current pulse is accompanied by burst firings. B, current-voltage relationship for the same neuron exhibited slight inward rectification. Resting membrane potential was ~ -60 mV. 3.2.1. GlyR Glycine (100 (0.M) was bath applied in the presence of N M D A antagonist kynurenic acid or A P V (10 |J,M) to block N M D A receptors (cf. Parsons et al., 1998). An initial application of glycine reversibly blocked action potential firing evoked by current pulse injection and depolarized ventrobasal thalamic neurons by 2 to 10 mV. Nine out of 25 tested neurons (36%) did not respond to applications of 100 ^ i m glycine. Neurons that responded to glycine had a mean input Chapter 3. Results - 53 -resistance of 291 ± 29 MQ and capacitance of 231 ± 32 pF (n = 16). Neurons that did not respond to glycine had a mean input resistance of 607 ± 88 MQ and capacitance of 144 ± 15 pF (n = 9). The glycine-responsive neurons had significantly lower input resistance and higher capacitance than the non-responsive neurons (P < 0.05, t-test). The higher capacitance of the glycine-responsive neurons was compatible with our qualitative observation that larger neurons tended to respond to glycine application. In the later experiments, we visually selected the larger neurons. This biased our electrophysiological findings towards a higher incidence of glycinergic responses. This fact has been considered in the Discussion. Applications of 100 |1M glycine markedly decreased action potential firing, evoked by current pulse injection in 37 ventrobasal thalamocortical neurons (Figure 3.6). This action of glycine was reversible upon washout. Glycine-induced blockade of action potential firing was surmountable by injecting larger depolarizing pulses, implicating a shunt of membrane voltage by an increased conductance (Figure 3.6). Glycine (100 |iM) reversibly decreased input resistance from 372 ± 26 MQ to 230 ± 36 MQ (P < 0.05, paired Mest) and depolarized ventrobasal thalamic neurons by 2 to 10 mV. These changes were reversible upon washout. Chapter 3. Results - 5 4 -Figure 3.6. Glycine (Gly, 100 p M ) reversibly decreased action potential firing and decreased input resistance in ventrobasal thalamic neurons. The blockade of action potential firing was surmountable by injecting of larger depolarizing pulses. Strychnine application (1-2 p M , cf. Pribilla et al., 1992) abolished the effects of glycine on action potential firing, input resistance and resting membrane potential in all tested neurons (Fig. 3.7). Strychnine applied alone, did not have an effect on action potential firing, input resistance and resting membrane potential of T C neurons. Application of the G A B A A antagonist, bicuculline (25 or 50 pM) , did not block the above-mentioned effects of glycine (data not shown). t 25mv| 200 ms Control Gly Wa shout Str Str + Gly Figure 3.7. Strychnine (Str, 1 p M ) abolished the decreases in action potential firing induced by glycine (Gly, 100 p M ) . Bath applications of 500 p M taurine in 13 neurons and 750 p M P-alanine in 12 neurons markedly decreased action potential firing, evoked by current pulse injection (Figure 3.8). Taurine reversibly decreased input resistance from 248 ± 38 Mf2 to 157 ± 25 M f J (P < 0.05, paired f-test) Chapter 3. Results - 55 -and depolarized the membrane potential by 2 to 10 mV. P-alanine reversibly decreased input resistance from 259 ± 22 MQ to 92 ± 42 MQ (P < 0.05, paired Mest) and depolarized the membrane potential by 2 to 10 mV. Attenuation of glutamatergic excitation with kynurenic acid application (1 mM) did not affect these responses. The reversible blockade by taurine and (3-alanine was surmountable by injecting larger depolarizing pulses, implicating a shunt of membrane voltage by an increased conductance (Figure 3.8). Control Control •* ' i I min i II nun i -Taurine p-alanine j i — i i — Figure 3.8. Application of taurine (500 u,M) and P-alanine (750 U.M) decreased the action potential firing of ventrobasal thalamocortical neurons. This decrease was surmountable by injecting larger depolarizing pulses. Action potentials were truncated for presentation purposes. Strychnine application (1-2 U . M ) abolished the effects of taurine on action potential firing, input resistance and resting membrane potential in all tested neurons (n = 6). Application of bicuculline (25 or 50 | ! M ) , did not block the effects of taurine (n = 6, data not shown). The effects of p-alanine on action potential firing, input resistance and resting membrane potential was significantly antagonized by 1-2 u,M strychnine application but not by 25 or 50 u , M Chapter 3. Results 56-bicuculline (n = 5). The blockade of P-alanine by up to 2 u.M strychnine, however, was not complete. Application of bicuculline (25 or 50 U.M), did not further block the residual effect left following the application of strychnine (Fig. 3.9). Hence, neither G l y R nor G A B A A receptors mediated the residual effect of P-alanine. The nature of this effect was not further investigated. \MMMM 20 m V 500 ms Control p-ala p-ala + Str p-ala + Str + Bic Figure 3.9. Strychnine (Str, 2 uM) significantly, but not completely, blocked the decreases in action potential firing induced by (3-alanine (P-ala, 750 u,M). The residual effect of P-alanine was not blocked by bicuculline (Bic, 25 uM). Current pulse amplitudes are ± 50 p A for all conditions. We tested the effects of D- and L-serine on neurons where glycine had an effect in the presence of kynurenic acid (1 mM) to block N M D A receptors. Applications of D-serine (n = 4) or L-serine (n = 4) at 1 to 5 mM did not significantly alter input resistance, resting membrane potential and action potential firing evoked by current pulse injection that depolarized the neurons by 5-20 m V (cf. F ig . 3.10). Input resistance was 300 ± 69 M Q before and 294 ± 72 M Q after the application of D-serine (P > 0.05, paired Mest). Input resistance was 226 ± 27 M Q before and 233 ± 31 M Q after the application of L-serine (P > 0.05, paired /-test). Chapter 3. Results - 5 7 -mmmm 20 m V 500 ms Control L-Serine 1 mM Figure 3.10. L-serine (1 mM) did not inhibit action potential firing. Act ion potentials are truncated for presentation purposes. The maximum action of glycine on input resistance and resting membrane potentials was typically observed 3-5 min after bath application of glycine and was completely reversed 30-40 min after washout (Fig. 3.11). Glycine 200 nM 0 i | I * - 5 -50 £ 1 0 0 150 7.5 I 5 15 30 45 Time (min) Figure 3.11. Time course of changes in input resistance (ARj) and membrane potential ( A V m ) , induced by glycine application (200 pM, bar) on 3 neurons. Chapter 3. Results -58-We obtained evidence that action of glycine agonists on input resistance and membrane potential was mediated via CI" channels. The reversal potentials were obtained from the intersections of curves describing voltage-current relationships before and after agonist application. Glycine agonists decreased the slope resistance and linearized the voltage-current relationship, partly by shunting rectification (Fig. 3.12). The average reversal potentials for glycine, taurine, and |3-alanine were -54 ± 3 m V (n = 37), -53 + 4 m V (n = 13), and -55 + 4 m V (n = 12), respectively. The reversal potentials did not differ significantly from Ea (P > 0.05, one sample /-test). L (PA) L (PA) L (PA) -150 -100 -50 0 50 100 150 -150 -100 -50 0 50 100 150 -150 -100 -50 0 5 0 100 150 4 4 4 V„ = -55 mV Glycine 100 p M V\fa shout Control Taurine 500 Control •alanine 750 u M Figure 3.12. Glycine (A), taurine (5) and p-alanine (C) decreased the slope resistance and linearized the voltage-current relationship. The reversal potentials obtained from the intersections of control and agonist curves were within a few m V of Ea-Picrotoxinin (50 pM) abolished the effect of 200 p M glycine in three tested neurons (Fig. 3.13). When tested on the same neurons, strychnine (1-2 pM) abolished the decrease in input resistance produced by glycine and taurine, and significantly antagonized the change due to p-alanine. Application of bicuculline (25 or 50 pM) , did not block these agonist effects (Fig. 3.14). Strychnine, picrotoxinin and bicuculline did not have significant actions on resting membrane properties that could account for their antagonist effects (cf. F ig . 3.7 and 3.13). Chapter 3. Results -59-Control, PTX, PTX + Gly Figure 3.13. Picrotoxinin (PTX, 50 pM) abolished the effect of glycine (Gly, 200 JIM) on input resistance of ventrobasal T C neurons. Figure 3.14. Strychnine (2 pM) significantly blocked the effects of glycine (200 p M , n = 7), taurine (500 p M , n = 5), and P-alanine (750 p M , n = 5) on input resistance. Bicuculline (25-50 pM) applied to the same neurons did not block these effects. Asterisks denote significantly different effects of agonist, without and with antagonist (P < 0.05, A N O V A ) . In cumulative concentration-response studies, the active agonists produced concentration-dependent decreases in input resistance (Fig. 3.15). Glycine application decreased input resistance with an EC50 of 126 ± 19 p M and H i l l slope of 2.3 + 0.8 (« = 11). Taurine application decreased input resistance with EC50 of 590 ± 124 p M and slope of 2.0 ± 0.7 (n = 5). P-Alanine application decreased input resistance with an EC50 of 703 ± 245 p M and slope of 1.1 ± 0.5 (n = m + Bicuculline • + Strychnine Glycine Taurine p-Alanine Chapter 3. Results - 60 -12). The apparent difference between slopes did not reach statistical significance ( A N O V A , P < 0.01). The EC 5 o for glycine was significantly lower than the EC50S for taurine and P-alanine (P < 0.01, Repeated Measures A N O V A ) . The agonists displayed similar efficacies, evident from their maximal responses. 100 g 75 S3 in - 5 0 H tu or 25 a Glycine • Taurine O P-Alanine O i l r 0.01 -1 1—I I I I 111 1 1—I I I I 111 0.1 1.0 10 Agonist Concentration (mM) Figure 3.15. The decreases in input resistance induced by glycine, taurine and P-alanine were concentration-dependent. 3.2.2. GABA A receptor We examined the action of G A B A on 19 ventrobasal T C neurons. Bath application of 50-200 U.M G A B A markedly decreased action potential firing, evoked by current pulse injection in all neurons. This action of G A B A was reversible. The blockade of action potentials by G A B A was surmountable by injecting larger depolarizing pulses, implicating a shunt of membrane voltage by an increased conductance (Fig. 3.16). Chapter 3. Results -61 -Control GABA Washout 20 m v 200 pA I ' 1 11 100 m s Figure 3.16. Application of G A B A (50 u M ) decreased the action potential firing of ventrobasal thalamocortical neurons. This decrease was surmountable by injecting larger depolarizing pulses. Application of 50 U.M G A B A reversibly decreased the input resistance from 252 ± 42 M Q to 113 ± 30 M Q (n = l; P < 0.05, paired Mest) and depolarized ventrobasal thalamic neurons by up to 10 m V . The reversal potentials for these actions, obtained from the intersections of control and agonist curves in the voltage-current relationships was -55.6 ± 1 m V (n = 5; cf. F ig . 3.17). The calculated reversal potential did not significantly differ from Eci (P > 0.05, one sample /-test). 'm<(*> V m (mV) Figure 3.17. G A B A (50 U.M) decreased the slope resistance and linearized the voltage-current relationship in this neuron. The reversal potentials obtained from the intersections of control and agonist curves was within few m V of Ecu Chapter 3. Results 62 In cumulative concentration-response studies, G A B A produced concentration-dependent decreases in input resistance (Fig. 3.18). The EC50 and slope for this action of G A B A were ~ 44 ± 5 p M and ~ 3.5 ± 1.8, respectively (n = 15). 100 —1 8 75 ro 50 o ts "O 03 or 25 OH 1 1 1 1 I I I I I 1 1 1 1 I M I | 100 1000 GABA Concentration (nM) Figure 3.18. The decrease in the input resistance was dependent on the concentration of G A B A . Strychnine and bicuculline applied alone did not have significant actions on resting membrane properties that could account for their antagonist effects (cf. F ig . 3.7 and 3.13). Application of bicuculline (25-50 pM) abolished the decrease in input resistance produced by 200 p M G A B A (Fig. 3.19,4). Application of strychnine (1 pM) did not antagonize the effects of G A B A (n = 5). However, when strychnine was applied at concentrations equal to or higher than 10 p M , it slightly antagonized the action of G A B A (Fig. 3.19^4). Cumulative application of strychnine indicated that this non-specific antagonism was concentration dependent (Fig. 3.195). Application of strychnine at concentrations as high as 100 p M significantly attenuated, but did not abolish the effect of 200 p M G A B A on input resistance (n = 5). Hence, we were not able to Chapter 3. Results 63 calculate the IC50 for the non-specific inhibition of G A B A by strychnine from these data. However, the apparent estimate of IC50 calculated from our data was 93 + 14 U.M (Fig. 3.195). A B 100 75 50 o a 25 GABA GABA GABA GABA + + + Bic Str 1 nM Str 50 \xM 1 1 1—I I I I I -I r — | — r 10 100 Strychnine concentration (nM) Figure 3.19. ^4, application of G A B A (200 U.M) decreased the membrane input resistance (n = 5). This effect was abolished by bicuculline (Bic, 25-50 U.M). L o w concentration (1 U.M) of strychnine (Str) did not antagonize this effect. However, a higher concentration (50 (IM) o f strychnine significantly but not completely antagonized this effect. B, the antagonism of G A B A effect by strychnine was concentration dependent. The attempted sigmoid fit is shown. In summary, these experiments for the first time demonstrated that oti and 0C2 G l y R subunits are moderately expressed in the ventrobasal nuclei. The effects of glycine agonists along with the exclusive blockade of these effects by the specific G l y R antagonist, strychnine, showed that these subunits form functional receptors. We found that glycine, taurine, and P-alanine had similar efficacies, apparent from their maximal responses. Functional glycine receptors were likely limited to the larger type-I thalamocortical neurons. The findings raise the question whether the GlyRs participate in synaptic inhibition? Chapter 3. Results - 6 4 -Part II - Synaptic Glycinergic Inhibition in Ventrobasal Thalamus 3.3. Medial lemniscal evoked inhibitory responses We initially studied thalamocortical IPSPs or IPSCs in ventrobasal nuclei evoked by electrical stimulation of medial lemniscus (Ec\ = -53 mV) . For isolation of the inhibitory responses, we blocked glutamatergic excitation by applying A P V (50 U.M) and C N Q X (50 U.M), or kynurenic acid (1 mM). We observed a hyperpolarizing response during glutamatergic blockade in 72 out of 83 neurons (87%) at V m = -40 to -45 m V held with D C injection (Fig. 3.20). In the rest of the neurons, there was no response at the presence of glutamatergic blockade. Figure 3.20. Co-application of ionotropic glutamate receptor blockers, A P V (50 |!M) and C N Q X (50 UM), unmasked a hyperpolarizing response in current clamp and at V m = -40 m V (Ea = -53 m V ) . In two neurons, a slow depolarizing potential immediately followed the isolated hyperpolarizing response at Vh = -45 m V . These depolarizing potentials were accompanied by high frequency burst firings at their peaks (Fig. 3.21). These potentials resembled low threshold calcium spikes expected in T C neurons, were infrequent and were not further studied. -40 mV 50 ms Chapter 3. Results -65-20 mV 200 ms Figure 3.21. L o w threshold calcium potential accompanied with a high frequency burst firings following an evoked hyperpolarizing potential. We evoked response by electrical stimulation at medial lemniscus. Application of T T X (700 n M , n = 3) abolished action potentials evoked by intracellular current injection as wel l as the hyperpolarizing responses (Fig. 3.22). Application of a metabotropic glutamate antagonist, M C P G (500 pM , n = 4), nicotinic receptor antagonists, pancuronium (10-20 p M , n = 2) and mecamylamine (10 p M , n = 1), a muscarinic receptor antagonist, atropine (4 p M , n = 1), and a dopaminergic receptor antagonist, haloperidol (20 pM , n = 1), did not affect the inhibitory postsynaptic responses (data not shown). We concluded that the isolated responses were postsynaptic potentials requiring an action potential-evoked release of transmitter (cf. Flint et al., 1998). Chapter 3. Results -66-50 rns | Figure 3.22. Application of T T X (700 nM) abolished the action potential firing (A) and hyperpolarizing responses evoked by electrical stimulation of medial lemniscus (B). Each trace in B represents an average of 10 responses. The observed postsynaptic responses were likely mediated by C f . This was evident from the observation that the polarity of postsynaptic potentials reversed near Ec\ upon current injection (current clamp recordings, Fig . 3.23). 3mV 50 ms Figure 3.23. The polarity of response reversed between -50 and -60 m V (EQ\ = -53 m V ) . Chapter 3. Results - 6 7 -We calculated the reversal potential for the evoked responses in voltage clamp, where we plotted the amplitude of evoked currents against the holding potential. Postsynaptic currents demonstrated an average reversal potential of-52.3 ± 2.6 m V (n = 5, F ig . 3.24). This calculated reversal potential did not significantly differ from Ea (P > 0.05, one sample /-test). Hence, the postsynaptic responses were CI" mediated IPSPs and IPSCs. Figure 3.24. Postsynaptic currents were likely CI" mediated. A, postsynaptic currents reversed near the holding potential, Vh = -50 m V . B, the average reversal potential calculated from the amplitude versus Vh curve was -53 ± 2.6 m V (n = 5). Error bars indicate standard error. 3.3.1. Susceptibility of inhibitory responses to antagonists We examined the susceptibility of inhibitory responses to antagonists. Negligible sensitivity to an antagonist was defined as < 10% reduction in the amplitude of IPSP or IPSC. A n antagonist was said to eliminate a response when > 90% reduction in the response amplitude occurred upon bath application of the antagonist. Partial blockade was defined as between 10-90% reduction in the amplitude of IPSP or IPSC. A B Vh (mV) -25 Chapter 3. Results - 68 -We applied strychnine in 1 and 2 U M concentrations to 27 previously untreated neurons, during medial lemniscal stimulation (Fig. 3.25). Strychnine reduced IPSP amplitude by 10-90% in 13 neurons, eliminated the IPSPs in 4 neurons, and had negligible effects in the remaining 10 neurons. Thus, a first application of strychnine reversibly reduced the IPSP in 63% of the neurons. Figure 3.25D illustrates the typical time course of antagonism and recovery from strychnine at 1 U M . Here, a maximal effect was typically observed at ~ 6 min, with full recovery at ~ 35 min after discontinuing the application. We observed recovery in most neurons. Substantial recovery (>50%) from applications at concentrations >2 U M required 60 to 150 min. Figure 3.25. Sensitivities of inhibition evoked from medial lemniscus to strychnine and bicuculline in four neurons (A-D). A, strychnine (Str, 1 U M ) reversibly reduced IPSP. Co-application with bicuculline (Bic, 50 U M ) abolished the IPSP. Fu l l recovery was observed at ~1 h after discontinuing strychnine and bicuculline. B, strychnine (50 U M ) did not greatly affect the IPSP, subsequently eliminated by bicuculline (50 U M ) . C, IPSP was slightly sensitive to bicuculline (50 U M ) , but was subsequently eliminated by strychnine (10 U M ) . Substantial recovery was observed at ~90 min after discontinuing strychnine. D, time course of IPSC antagonism by strychnine application (arrows) and recovery (Vh = -80 mV) . Maximal effects were observed between 6 and 10 min of application, with full recovery at 35 min after discontinuing strychnine. Each trace (A-D) represents an average of 10 responses. Chapter 3. Results - 69 -We tested the effects of strychnine in conjunction with bicuculline in 16 neurons. Where strychnine (1-2 (IM) reduced the IPSP amplitude by 50-90%, subsequent co-application with bicuculline (50 pM) abolished the remaining response (Fig. 3.254 n = 7). Where bicuculline had small or negligible effects, strychnine eliminated the IPSP (Fig. 3.25C, n = 4). Where strychnine had negligible effects, bicuculline eliminated the IPSP (Fig. 3.255, n = 5). The picture emerging from these studies was that a majority of IPSPs had sensitivities to both antagonists, whereas a minority exhibited nearly exclusive sensitivity to either antagonist. 3.3.2. Strychnine concentration-response relationship We determined the effects of strychnine, applied in a cumulative manner over a wide concentration range (0-50 p M ) . Each of the 19 neurons had not received prior drug application, before testing with 3 to 5 strychnine concentrations in a stepwise manner. Figure 3.26 shows the concentration-response relationship for strychnine's effects on medial lemniscal IPSPs. A double, sigmoid function best fitted the relationship between concentration and suppression o f IPSP amplitude. The plateau at concentrations of 300 n M to 2 p M corresponded to - 3 0 % blockade of IPSP amplitude. The first sigmoid curve had an apparent IC 5 n o f 59 ± 15 n M , whereas the second curve had an apparent IC50 of 5.7 ± 1.2 p M . Hence, - 2 log units separated the two components of this relationship. In subsequent experiments, we considered the possibility that strychnine at >2 p M may have blocked G A B A A receptors (cf. Shirasaki et al., 1991; L i et al., 2003), whereas strychnine at <2 p M exerted only selective actions on glycine receptors. Chapter 3. Results - 7 0 -100n I 1 1—I l l l l l l 1 1—I I I I I 11 1 1—I I'l I I l | 1 1—I I I I l l | 0.01 0.1 1 10 100 Strychnine Concentration (p.M) Figure 3.26. Concentration-response relationship for strychnine blockade of medial lemniscal IPSPs. Strychnine was applied cumulatively in a step-wise manner to 19 neurons (Vh = -80 mV). The smooth curve shows a fit to a double sigmoid equation. Dotted lines show apparent IC50S at 59 n M and 5.7 p M for the two components. The decrease in IPSP amplitude caused by strychnine was concentration-dependent (P < 0.05, ANOVA). Number of neurons tested at each concentration is shown in parentheses. To test this hypothesis, we applied a maximal blocking concentration of gabazine (Mori et al., 2002), a more specific G A B A A antagonist than bicuculline (cf. Rognan et al., 1992). Initial application of 10 p M gabazine resulted in an average change in baseline membrane current of A I m = +45 ± 60 pA at holding potential, Vh = -80 mV (n = 19). This change was not statistically significant (P > 0.05, one sample f-test). Hence, gabazine did not have significant action on membrane properties that could account for its antagonist effects. Application of gabazine (10 pM) abolished the IPSC in 5 neurons (Fig. 3.27,4). Gabazine had negligible effects on IPSCs in 4 neurons, which were later abolished by application of 2 p M strychnine (Fig. 3.275). Gabazine reduced the IPSCs by 25-75% in 9 neurons, which were later abolished by co-application of gabazine and strychnine (Fig. 3.27C). Chapter 3. Results - 7 1 -Figure 3.27. Sensitivities of inhibition evoked from medial lemniscus to gabazine and strychnine in three neurons. A, gabazine (Gbz, 10 U.M) reversibly abolished the IPSC. B, gabazine did not affect the IPSC, which was later abolished by application of strychnine (Str, 2 (IM). C, elimination of IPSC required co-application of gabazine and strychnine. Each IPSC represents an average of 10 responses. In the 13 neurons where gabazine did not eliminate the IPSC, we co-applied 3 to 5 concentrations of strychnine in a stepwise manner. Figure 3.28 shows the relationship between strychnine concentration and IPSC amplitude during co-application with gabazine. The decrease in IPSC amplitude caused by strychnine was concentration-dependent (P < 0.05, A N O V A ) . Under these conditions, the relationship was not a double sigmoid curve for responses to strychnine (Kolmogorov-Smirnov goodness of fit test). Rather, the curve was well-fitted by a single Hi l l function that approached a maximum at ~2 U.M with an approximate IC50 of 33 ± 2 nM. Hence, there was retention of an apparently high affinity segment of the concentration-response curve (cf. Fig. 3.26). The Hil l slope was 0.7 ± 0.2, implying that a single strychnine molecule could block the receptor-gated channel (cf. Shirasaki et al., 1991; Mori et al., 2001). This specific blockade was independent of gabazine-sensitive receptors, likely mediated by GlyR. Chapter 3. Results - 7 2 -Sham 0.001 0.01 0.1 1 10 Strychnine Concentration (pM) Figure 3.28. Concentration-response relationship for strychnine blockade of medial lemniscal IPSCs during continuous application of gabazine (10 pM). Dotted line points to the IC50 at 33 nM. Number of neurons tested at each concentration is shown in parentheses (total number of neurons was 13). Holding potential, Vh = -80 mV. 3.3.3. Slower component of synaptic inhibition In 9 out of 64 neurons (14%), blockade of G l y R and G A B A A receptor by co-application o f strychnine (1-10 p M ) and bicuculline (10-50 p M ) unmasked a late, slow hyperpolarization in current-clamp (n = 5), or an outward current in voltage-clamp (n = 4; Vh = -40 m V ; F ig . 3.29,4). This response had latency to onset of 60 ± 19 ms and duration of 1169 ± 88 ms in current clamp (Vh = -40 m V , n = 5). From recordings over the -85 to -40 m V range, we estimated the reversal potential at -81 ± 6 m V ( n = 7 ; £ K = -83 m V ; Fig . 3 .29Q. Application of CGP-35348 (100 nM, n = 4) reduced slow IPSCs by an average of 85 ± 5% ( V h = -40 m V ; F ig . 3.295). The long latency and duration, reversal potential in the vicinity of and sensitivity to antagonism by C G P 35348 implicated metabotropic G A B A B receptors. We did not further investigate the slow Chapter 3. Results -73 -response. Such neurons used for analysis of their early responses were clamped to Vh = -80 mV, minimizing a contribution of the slow GABAsergic component. A C 0 , Str + Bic + CGP-35348 Figure 3.29. Occasional late component of synaptic inhibition. Co-application of strychnine (Str) and bicuculline (Bic) unmasked a late slow hyperpolarization in current-clamp (A), or an outward current in voltage-clamp (B), at holding potential, Vh = -40. This response was abolished by application of CGP-35348 (100 nM), with an estimated reversal potential around Em C O -In summary, these data imply that medial lemniscal stimulation can evoke inhibitory postsynaptic responses due to a mixture of glycine and G A B A A receptor activations, as well as purely glycinergic or GABAergic inhibition. A minority of neurons exhibited an additional inhibition, mediated by G A B A R receptors. 3.4. Comparison of inhibition evoked from other sources 3.4.1. Susceptibility to antagonist Synaptic inhibition in the V B thalamus is mainly polysynaptic originating in the surrounding nuclei (Baldissera and Margnelli, 1979, Ohara and Lieberman, 1993). Hence, we compared inhibition evoked from the medial lemniscus, with the inhibition evoked from the caudal part of zona incerta, nRt, the ethmoid nucleus, and Pfc. The stimulation sites are shown in Figure 1.6 Chapter 3. Results - 74 -(schematic) and Figure 2.1 (bright field microscopy). We examined the sensitivity of synaptic inhibition evoked from the above locations to strychnine (1-2 pM) and bicuculline (25 pM) at V h = -80 m V . During the bath application of 1 m M kynurenic acid, stimulation of the caudal zona incerta evoked IPSPs in 7 tested ventrobasal T C neurons, as identified by reversal of IPSP polarities around Ec\ (cf. F ig . 3.30,4). The evoked IPSPs showed negligible sensitivity to strychnine (Fig. 3.305). Applied alone, bicuculline reversibly abolished these IPSPs in all neurons (cf. F ig . 3.305 & Q . Hence, inhibitory responses evoked from zona incerta were exclusively mediated by G A B A A receptors. A B 3mV 50 ms Figure 3.30. IPSPs evoked from the caudal zona incerta in three T C neurons. IPSPs were identified by reversal around Ec\ (A). They showed negligible sensitivity to 2 p M strychnine (Str; 5) . Applied alone, bicuculline 25 p M (Bic) reversibly abolished IPSPs (5 & C). Traces in A represent single recordings. Traces in 5 & C represent average of 10 recordings. G A B A A receptor predominatly mediated the inhibitory responses evoked from nRt. The IPSCs evoked by electrical stimulation of nRt were exclusively sensitive to bicuculline in 5 (71%) out of 7 tested neurons (Fig. 3.31,4). In 2 neurons (29%), strychnine partially blocked the IPSCs, which were abolished by subsequent co-application of strychnine and bicuculline (Fig. 3. 315). Hence, the inhibitory responses evoked from nRt were primarily mediated by G A B A A receptors, with occasional contribution by G lyR . Chapter 3. Results -75 -250 pA 100 msec Control Figure 3.31. IPSCs evoked from nRt in two T C neurons. A, IPSCs were often eliminated by bicuculline 25 JIM (Bic). B, occasionally, elimination of IPSC required co-application of strychnine 2 U.M (Str) and Bic . Traces represent average of 10 recordings. G l y R mediated responses were most prevalent when IPSPs were evoked from Eth. In 5 out of 7 tested neurons (71%), the IPSPs evoked by electrical stimulation of ethmoid (Eth) were exclusively sensitive to strychnine (Fig. 3.32). In 2 neurons (29%), strychnine decreased the amplitude of IPSPs by 10-90%, which were eliminated by subsequent co-application of strychnine and bicuculline. Hence, G l y R mainly mediated the inhibitory responses evoked from Eth. Bic + Str Figure 3.32. IPSP evoked from Eth in a T C neuron. Application of bicuculline 25 U.M (Bic) had negligible effect. Subsequent application of strychnine 2 |1M (Str) eliminated the IPSP. Traces represent average of 10 recordings. We evoked inhibitory responses from another neighbouring nucleus, Pfc. The majority of IPSPs evoked by stimulation of Pfc (3 out of 4) required co-application of strychnine and bicuculline for complete blockade (Fig. 3.33). In one neuron, the evoked IPSP was insensitive to strychnine. Chapter 3. Results This IPSP was abolished by bicuculline. mixed antagonist sensitivity. -76-Hence, Pfc evoked inhibitory responses mainly had a 4mV 50 ms Figure 3.33. IPSPs evoked from Pfc in two T C neurons. Application of bicuculline 25 p M (Bic) decreased the amplitude by 10-90%. Subsequent application of strychnine 2 p M (Str) eliminated the IPSP. Traces represent average of 10 recordings. There is a strong inhibitory input to the V B neurons of rat from interneurons projecting from the nRt (Cox et al., 1997; Gentet and Ulr ich , 2003). We sought to examine to what extent medial lemniscal evoked inhibitory responses depended on an intact nRt. We completely removed nRt in 11 slices and evoked inhibitory responses in ventrobasal neurons by electrical stimulation of medial lemniscus. In 4 neurons (36%), we were not able to evoke an inhibitory response. In slices with intact nRt, we were not able to evoke an inhibitory response from medial lemniscus only in 11 out of 83 tested neurons (13%; P < 0.05, C h i square test). The 7 IPSPs evoked from medial lemniscus in the absence of nRt displayed exclusive sensitivity to bicuculline in 2 (29%), exclusive sensitivity to strychnine in 1 (14%) and required co-application of bicuculline and strychnine for complete blockade in 4 (57%). Hence, while the V B thalamus receives major inhibitory inputs from the nRt, this nucleus likely receives G A B A A e r g i c and glycinergic inhibition from several other sources. In general, inhibitory responses evoked from Eth were more likely to be purely glycinergic, and those evoked from zona incerta (ZI) and nRt were more likely to be purely G A B A e r g i c (Fig Chapter 3 . Results -77-3.34/1). Since there was a variation in the ratio of glycinergic and GABAAerg i c components of the responses with mixed antagonist sensitivity (mixed responses, Fig. 3.34,4), the frequency of purely glycinergic or G A B A e r g i c responses did not fully represent the relative strength of glycinergic or G A B A A e r g i c inhibition evoked from a stimulation site. Hence, we further quantified the relative strength of glycinergic or G A B A A e r g i c inhibition as the ratio of sum of the amplitudes of all glycinergic or G A B A A e r g i c components evoked from a stimulation site to the sum of the amplitudes of all inhibitory responses evoked from that site (Fig. 3.345). Eth had relative glycinergic strength of > 80%, while ZI and nRt had negligible relative glycinergic strengths (0% and < 20%, respectively; Fig. 3.345). Hence, stimulation within Eth provided the strongest glycinergic input to V B ventrobasal thalamus. Figure 3.34. Sources of inhibition to ventrobasal thalamus. A, antagonist sensitivity o f inhibitory responses evoked within the caudal part of ZI (n = 7), nRt (« = 7), Eth (n = 7), and Pfc (n = 4) compared to those evoked from medial lemniscus (ml) with (n = 7) or without (n = 16) removal of nRt from the slices. The slices were exclusively sensitive to 1-2 pM strychnine (purely glycinergic), and exclusively sensitive to 25 pM bicuculline (purely G A B A A e r g i c ) or required co-application of strychnine and bicuculline for complete blockade (mixed). 5 , strength of G A B A A e r g i c and glycinergic inputs to ventrobasal thalamus evoked from variousstimulation sites. Neurons were held at Vh = -80 mV. Chapter 3. Results - 78 -3.4.2. Latency and latency fluctuations Inhibitory responses in T C neurons of the V B thalamus evoked from of medial lemniscal are considered to be disynaptic (Baldissera and Margnelli , 1979) whereas inhibitory responses of nRt and ZI origin are considered to be monosynaptic (Bartho et al., 2002; Gentet and Ulr ich , 2003). We sought to investigate the monosynaptic or polysynaptic nature of inhibitory responses by comparing latency and latency fluctuation in inhibitory responses evoked in the V B thalamus. Over all , the latency of medial lemniscal evoked IPSPs and IPSCs was 3.9 ± 0.5 ms in 23 examined neurons. The medial lemniscal inhibitory synaptic responses that were sensitive to both strychnine and bicuculline (mixed responses) had a mean latency to onset of 4.0 ± 0.5 ms (n = 21). The inhibitory synaptic responses with exclusive sensitivity to bicuculline, and the bicuculline-sensitive component of mixed responses, had a mean latency of 4.0 ± 0.4 ms (n = 8). The inhibitory synaptic responses with exclusive sensitivity to strychnine, and the strychnine-sensitive component of mixed responses, had a mean latency of 3.0 ± 0.4 ms (n = 15). The latencies of these 3 groups were not significantly different ( A N O V A , P > 0.05). Hence, the glycinergic and G A B A A e r g i c inhibition evoked from medial lemniscus could not be distinguished on a basis of latency to onset. Inhibitory synaptic responses that were evoked from nuclei surrounding the V B had a slightly but not significantly shorter latency compared to those that were evoked from medial lemniscus (Fig. 3.35). ZI , Pfc, Eth and nRt evoked inhibitory responses had average latencies of 2.8 ± 0.8 ms (n = 7), 3.4 ± 0.7 ms (n = 5), 2.6 ± 0.3 ms (« = 6) and 2.5 ± 0.3 (n = 7) respectively. These values did not significantly differ from each other, or from the latencies measured for the Chapter 3. Results -79-inhibitory synaptic responses evoked from medial lemniscus (P > 0.05, A N O V A ) . On the other hand, latency fluctuation for medial lemniscal evoked IPSPs and IPSCs was more than five times higher than the inhibitory synaptic responses evoked from the nuclei surrounding V B (Fig. 3.35; P < 0.05, A N O V A ) . Medial lemniscal evoked inhibitory responses displayed average latency fluctuation of 0.35 ± 0.03 (n = 23) while ZI, Pfc, Eth and nRt evoked inhibitory responses displayed average latency fluctuations of 0.05 ± 0.01 ms (n = 7), 0.05 ± 0.01 ms (n = 5), 0.05 ± 0.01 ms (n = 6) and 0.07 ± 0.01 (n = 7), respectively. A higher latency fluctuation likely represents a polysynaptic nature of medial lemniscal IPSPs and IPSCs. B 2 CD ml ZI Pfc Eth nRt B '•5 0.3 o £ 0.1 ml ZI Pfc Eth nRt Figure 3.35. Latency (A) and latency fluctuation (B) of inhibitory responses evoked from the ZI (n = 7), nRt (n = 7), Eth (n = 6), and Pfc (n = 5) compared to those evoked from m l (n = 23). The latency fluctuation of inhibitory synaptic responses evoked from medial lemniscus was significantly increased compared to other examined IPSPs and IPSCs ( P < 0.05, A N O V A ) . Chapter 3. Results - 80 -3.4.3. Rheobase and chronaxie The minimum strength and duration of stimulation required for evoking an inhibitory synaptic response from medial lemniscus were studied in seven neurons. The relationship between the minimum strength and the inverse duration of stimulation was well described by a linear function (Fig. 3.36; cf. Nowak and Bullier, 1998). We calculated the rheobasic current (x-intercept) and chronaxie (slope divided by x-intercept) from this relationship (see methods). OH 1 1 1 0 0.01 0.02 0.03 1/Duration (1/|xs) Figure 3.36. The relation between minimum strength and inverse duration of stimulation required to evoke an IPSP from medial lemniscus was well described by a linear function. The rheobasic current and chronaxie were calculated from this relationship (see methods). The average rheobasic current and chronaxie were 3.9 ± 0.9 u A and 245 ± 37 (is, respectively (n = 7). The rheobasic current for medial lemniscus was not significantly different from rheobasic current for ZI (4.0 ± 0.6 L I A , n = 7), Pfc (6.5 ± 1.2 L L A , n = 5), Eth (4.8 ± 1 | i A , n = 6), and nRt (2.3 ± 0.4 u A , n = 5, P > 0.05, A N O V A ) . However, the chronaxie for medial lemniscus was significantly higher than chronaxie for ZI (133 ± 31 LIS, n = 7), Pfc (109 ± 21 LIS, n = 5), Eth (66 ± 16 LIS, n = 6), and nRt (55 ± 15 LIS, n = 5, P < 0.05, A N O V A ) . On comparison using a post hoc multiple analysis, Eth and nRt showed the smallest, while medial lemniscus showed the largest chronaxie (P < 0.05; Fig . 3.37). Chapter 3. Results -81 -300 -i s - ' 200' 0 X 03 c 2 100' JZ O 1 1 10n 3 7 5 H CD CO 03 5H T - ° 1 + s T -j- o> LKL" 2.5H 1 i ml ZI Pfc Eth nRt ml ZI Pfc Eth nRt Figure 3.37. Comparison of chronaxie (left) and rheobase (right) for inhibitory responses evoked from medial lemniscus (n = 7), ZI (n = 7), Pfc (« = 5), Eth (n = 6), and nRt (n = 5). Chronaxie for medial lemniscus was significantly higher than for other stimulation sites. Eth and nRt displayed the lowest chronaxie on a Post hoc multiple comparison analysis (*tP < 0.05, A N O V A ) . To examine the possible heterogeneity of the pathways (cf. Jacobson, 1963), we compared the changes in the amplitude of inhibitory responses with the duration of stimulation for IPSPs and IPSCs evoked from various sites. The amplitude was normalized to the maximum amplitude to control for the between-neurons variations. The amplitude of inhibitory responses evoked from the medial lemniscus increased gradually with increasing stimulus duration (Fig. 3.38). In contrast, the amplitude of inhibitory responses evoked from the other stimulation sites increased sharply and reached the maximum amplitude at significantly lower stimulation durations compared to the medial lemniscal evoked responses (Fig. 3.38; P < 0.05, A N O V A ) . The gradual rise in the amplitude of inhibitory responses evoked from medial lemniscus implies that these responses are likely produced by a heterogeneous group of axons, some of which are activated at low stimulus durations while others are evoked as the stimulus duration increases. Chapter 3. Results -82--•- ml -o-ZI -*- Pfc ^ E t h -o- nRt 0 200 400 Stimulus duration (us) 600 Figure 3.38. The amplitude of inhibitory responses evoked from the medial lemniscus increased more gradually and reached the maximum amplitude at significantly higher stimulation durations compared to the responses evoked from other nuclei ( P< 0.05, A N O V A ) . 3.4.4. IPSP shape parameters To further to distinguish between the inhibitions evoked from the examined stimulation sites, we compared the shape parameters of IPSPs evoked from these sites. The parameters included 10-90% rise time, half-width duration, and shape index. The latter is defined as the slope of the line fitted to the scatter plot describing the relationship between 10-90% rise time and half-width duration of IPSPs. Shape index provides an estimate of the distance between the synapse and the recording electrode (Mason et al., 1991). Figure 3.39 shows an example of calculating shape index for IPSPs evoked from medial lemniscus. Chapter 3. Results -83-120 in £ 15 60 > ro I y = 6.6 x + 3 Shape index = 6.6 • 12 10 Rise time (ms) 15 Figure 3.39. Calculation of shape index from the relation between rise time and half-width for IPSPs evoked from ZI (n = 7). Table 3.1 summarizes these parameters for all the stimulation sites. The average rise time, half width, and shape index of IPSPs did not differ between stimulation sites (P > 0.05, A N O V A ) . However, IPSPs evoked from medial lemniscus displayed a higher coefficient of variation for all of the above parameters (Table 3.1). Hence, it is not possible to distinguish between the inhibitions evoked from the examined stimulation sites based on shape parameters. Table 3.1. IPSP shape parameters, expressed as mean ± S E M . Vh = -80 m V ; EQ\ = -53 m V . Numbers in brackets indicate coefficient of variation. Stimulation site n Rise time (ms) Hal f width (ms) Shape index ml 20 8.6 ± 1.3 T0.71 66 ± 2 4 ri-61 5.6 ± 0 . 8 [0.61 ZI 7 8.0 ± 1 . 1 [0.41 60 ± 12 T0.51 6.6 ± 0.4 T0.21 Pfc 4 7.9 ± 1.8 [0.41 73 ± 15 [0.41 6.4 ± 1.0 [0.31 Eth 7 7.3 ± 0 . 7 [0.31 60 ± 5 [0.21 6.0 ± 0 . 2 [0.11 In summary, these experiments for the first time uncovered glycinergic synaptic inhibition in ventrobasal thalamus. Synaptic inhibition evoked from medial lemniscus was heterogeneous. In the majority of neurons, synaptic inhibition was mediated by G A B A A and glycine receptors. Two minorities were exclusively mediated by either of the receptors. Occasionally, there was an additional G A B A B component. The medial lemniscal mediated synaptic inhibition was likely Chapter 3. Results -84-polysynaptic, resulting from co-transmission, rather than co-release, of G A B A and glycine. Among nuclei surrounding V B thalamus, stimulations within Eth displayed the highest relative glycinergic strength, likely of monosynaptic nature. Chapter 3. Results -85-Part III - Biophysical Properties of Synaptic Inhibitory Receptors 3.5. Kinetics of synaptic receptors 3.5.1. Evoked IPSCs Mixed IPSCs evoked from medial lemniscus had a rise time of 3.0 ± 0.6 ms at Vh = -80 mV (AZ = 11 neurons). We compared the rise times of glycinergic and GABAAergic components in the mixed IPSCs. The glycinergic and GABAAergic components of IPSCs were obtained by subtraction of the currents during glycine- and GABAA-receptor antagonism from their control (Fig. 3.40). The mean rise time for the glycinergic currents isolated by application of a GABA A antagonist in neurons held at Vh = -80 mV was 2.6 ± 0.5 ms (n = 7). This value was not statistically different from the mean rise time (3.0 ± 0.5 ms, n = 7) for GABAAergic currents isolated by strychnine application (/-test, P > 0.05). Figure 3.40. Glycinergic and GABAAergic components of IPSCs were obtained by subtraction of the currents during glycine- and GABAA-receptor antagonism from their controls. The glycinergic and GABAAergic components of IPSCs had distinct decays. Glycinergic currents more commonly displayed faster decay (Fig 3.41,4a) when compared to the GABAAergic currents. In addition, some glycinergic currents exhibited slower decay (Fig 3A\Aab). Moreover, glycinergic currents often displayed more complex decay kinetics than GABAAergic currents which all decayed monoexponentially with a normal distribution of decay Chapter 3. Results - 86 -time constants (Fig. 3.41,4 & Q . Eleven out of a total of 17 glycinergic IPSCs exhibited a monoexponential decay (cf. F ig . 3.41,4). The decay time constants for these IPSCs were not normally distributed (P < 0.05, Kolmogorov-Smirnov goodness-of-fit test), and likely represented 2 populations. One population had a short decay time constant, TstrCshon) = 10 ± 1.4 ms (n = 8), whereas the other had a long decay time constant, Tstr(iong) = 70 ± 4.0 ms in 3 neurons ( A N O V A , P < 0.01). These values remained stable over a period of 1.5 h, indicating stationarity of decay kinetics. Four IPSCs decayed with a biexponential time course (cf. F ig . 3.415). The decay time constants ( i i and 12) for biexponential IPSCs had means of 13 ± 2.1 ms and 93 ± 10 ms, respectively (n = 4). These values did not differ from xstr(short) and Tstr(iong), obtained from monoexponential fits (t-test, P > 0.05). On pooling the data obtained from mono- and biexponential fits, xstr(short) was 12 ± 1.1 ms (n = 12) and xstr(iong) was 80 ± 6.8 ms (« = 7), as shown in the frequency histogram of Figure 3.41 C. These time constants differed from XGABA = 22 ± 1.5 ms (H = 18; F ig . 3.41 Q , as wel l as from each other ( A N O V A , P < 0.01). Hence, three populations of inhibitory synaptic receptors with distinct decay time constants are present in the V B thalamus. Chapter 3. Results -87-Figure 3.41. Resolved glycinergic and GABAAerg ic currents had distinct decays. A, glycinergic and GABAAerg ic currents were peak aligned and scaled to the same amplitude for comparison of time courses in different neurons. The glycinergic current decayed faster in the neuron of Aa, and slower in the neuron of Ab, when compared to the G A B A A e r g i c currents. B, the decay phase of the glycinergic current in another neuron was well-fitted by the sum of two exponential terms (upper trace, smooth curve). The lower record separately shows these fits and the time constants. C, frequency distribution histograms o f decay time constants for glycinergic and G A B A A e r g i c currents. The arrowheads indicate mean values. Vh = -80 m V , EC\ = -53 m V . A and B show averages of 10 IPSCs. Chapter 3. Results - 88 -3.5.2. Spontaneous IPSCs We sought to substantiate the decay time constants measured from evoked IPSCs by repeating the measurements from spontaneous IPSCs. We investigated sIPSCs in whole-cell recordings with C s - Q X patch solution (Ea = 0 mV) at the holding potential, Vh = -60 m V (n = 7 neurons) during kynurenic acid (1 mM) blockade of ionotropic glutamate receptors (Fig. 3A2A). We recorded an average of 3.8 ± 0.9 events per second with average amplitude o f - 3 3 ± 0.5 p A . There was no correlation between the amplitude, the rise time, and the decay time constants of the detected events (Fig. 3.425). The polarity of the detected events reversed at 1.8 ± 1.7 m V (n - 6). The reversal potential was not significantly different from calculated Ea (~0 mV) confirming their C f dependence (P > 0.05, student /-test). Hence, the detected events represented sIPSCs rather than random noise. A Rise time (ms) Decay time constant (ms) Rise time (ms) Figure 3.42. sIPSCs recorded in the V B thalamus. A, typical recording at holding potential, V h = -60 m V . B, there was no correlation between the amplitude, the rise time, and the decay time constants of the detected sIPSCs. Chapter 3. Results - 8 9 -Application of strychnine (1 pM) decreased the frequency of sIPSCs to 0.8 ± 0.3 events per second in seven neurons (P < 0.05, paired f-test). Co-application application o f strychnine and gabazine (10 pM) abolished the sIPSCs (Fig. 3.43,4). The average amplitude of the sIPSCs decreased slightly to -26.3 ± 1.0 p A following the application of strychnine. The recorded sIPSCs had an average rise time of 0.9 ± 0.02 ms before and 1.1 ± 0.05 ms after the application of strychnine. The rise time of sIPSCs was significantly lower than the rise time of medial lemniscal evoked IPSCs (P < 0.05, A N O V A ) . Hence, medial lemniscal evoked IPSCs likely represented multiple quantal events. Control Str Str + Gbz 100pA 10s B T~10ms T~20ms x~ 110 ms c 0 25 50 75 100 125 150 Decay time constant (ms) Decay time constant (ms) Figure 3.43. Kinetics of sIPSCs. A , sIPSCs were recorded in C s - Q X patches at the holding potential, V h = -60 mV. Bath application of strychnine (Str, 1 pM) decreased the frequency of sIPSCs. Co-application of strychnine and gabazine (Gbz, 10 pM) abolished the sIPSCs. B, when sIPSCs were peak aligned and scaled to the same amplitude for comparison of time courses, three distinct time courses were discernible. C, frequency distribution histograms of decay time constants for sIPSCs were described with Gaussian functions. Three populations were observed with average decay time constants of 11 ± 0 . 1 ms, 22 ± 0.1 ms, and 74 ± 2.4 ms. Following bath application of strychnine, only one population of sIPSCs with average decay time constants o f 22 ± 0.2 ms remained. The arrowheads indicate average values. Vh = -60 m V , EQ\ = 0 m V . Chapter 3. Results -90-We aligned the recorded sIPSCs at their peaks and scaled them to the same amplitudes for the comparison of time courses. Three distinct time courses were easily discernible (Fig 3.435). A group of very fast sIPSCs completely decayed in less than 100 ms, while a group of very slow sIPSCs required a few hundreds of milliseconds to return to baseline. A third group of sIPSCs decayed between 100 and 200 milliseconds. We quantified the decay kinetics of sIPSCs by fitting the decay phase of the currents with single and double exponential functions (Fig 3.435). The majority of sIPSCs were well described by a single exponential function. In < 6% of sIPSCs, a double exponential function was required to adequately describe the sIPSCs. The frequency distribution histogram of decay time constants for sIPSCs before the application of strychnine was well described by the sum of three Gaussian functions (Fig. 3 .43Q. The three populations of sIPSCs displayed average decay time constants of 11 ± 0 . 1 ms, 22 ± 0.1 ms, and 74 ± 2.4 ms. Bath application of strychnine abolished the fastest and the slowest populations of sIPSCs, leaving one population of sIPSCs with average decay time constants of 22 ± 0.2 ms (Fig. 3 .43Q. Hence, glycinergic sIPSCs had fast and slow decay kinetics. GABAAerg i c sIPSCs, on the other hand, displayed single decay kinetic with intermediate duration. We compared the decay time constants obtained from sIPSCs to those obtained from IPSCs. The decay time constants of glycinergic (Fig. 3.44,4) and G A B A A e r g i c (Fig. 3.445) sIPSCs matched their respective evoked IPSCs (P > 0.05, A N O V A ) . Hence, the decay time constants were reproducible with different techniques and likely represented genuine findings and decay kinetics of synaptic receptors. Chapter 3. Results -91 -150 n C/3 E, S 1 0 0 CO c 8 cu I >s CD O CD Q 50 H Short (6) (12) Y////A sIPSC IPSC Long sIPSC IPSC B 150 E 100 H CD E >^ CD O CD Q 50 H (7) (18) si PSC IPSC Figure 3.44. Mean decay time constants for glycinergic (A) and G A B A A e r g i c (B) sIPSCs did not differ from evoked IPSCs (P > 0.05, A N O V A ) . Numbers of neurons are indicated in parentheses. Decay kinetics and antagonist sensitivity of IPSCs did not necessarily predict the decay kinetics and antagonist sensitivity of sIPSCs evoked from medial lemniscus in the same neuron. sIPSCs with all three types of decay kinetics and sensitivities to both antagonist were recorded in all 7 tested neurons. GABAAerg ic IPSCs with decay time constants similar to sIPSCs were evoked from medial lemniscus in all of these neurons. However, no glycinergic IPSCs was recorded in one neuron. In another, the glycinergic evoked IPSC decayed with a biexponential time course with time constants close to the short and the long decay time constants of observed sIPSCs. In the remaining five neurons, evoked IPSCs displayed monoexponential time courses with time constants close to either the short or the long decay time constants of observed sIPSCs. Table 3.2 summarizes this information. Hence, the decay kinetics of glycinergic IPSCs evoked from medial lemniscus does not always include both slow and fast kinetics of sIPSCs. In other words, the glycinergic sIPSCs likely originate from more than one pathway. Chapter 3. Results - 9 2 -Table 3.2. Comparison of the average decay time constants of evoked and spontaneous glycinergic IPSCs recorded in the same neurons. Numbers are expressed in milliseconds. spontaneous IPSC evoked IPSC Short Long Short Long 1 12 110 None 68 2 11 87 7 None 3 12 90 None 53 4 11 82 19 82 5 11 88 None None 6 13 86 None 130 7 12 75 None 52 We examined the possibility of cross-desensitization between synaptic G A B A A e r g i c and glycinergic receptors (cf. L i et al., 2003). If cross desensitization occurs between two synaptic receptors, one would expect that the blockade of one receptor would affect the decay time constants of the synaptic currents mediated by the other receptor. We compared the decay time constant of G A B A A e r g i c IPSCs before and after glycinergic blockade. The G A B A A e r g i c IPSCs before glycinergic blockade were isolated from mixed IPSCs by subtraction of residual IPSCs after application of 25 LLM bicuculline from the control IPSCs. The G A B A A e r g i c IPSCs after glycinergic blockade were isolated from mixed IPSCs by application of 2 LIM strychnine. G A B A A e r g i c IPSCs evoked from medial lemniscus had decay time constants of 25 ± 4 ms (n =10) in the absence of strychnine and 24 ± 3 ms (n =10) at the presence of strychnine (Fig. 3.45). There was no difference between these two decay time constants (P > 0.05, /-test). G A B A A e r g i c sIPSCs also showed similar decay time constants of 22 ± 0.1 ms before, and 22.3 ± 0.2 ms after application of strychnine (Fig. 3.45; n = 7; P > 0.05, paired /-test). Hence, blockade of glycine receptors had negligible effect on the decay time constant of GABAAerg i c IPSCs. Chapter 3. Results - 93 -50 1 Control Str Control Str IPSCs sIPSCs Figure 3.45. Decay time constants of G A B A A e r g i c IPSCs (n = 10) and sIPSCs (n = 7) were not affected by application of 2 pM strychnine (Str; P > 0.05, paired f-test). We compared the decay time constant of glycinergic IPSCs before and after G A B A A e r g i c blockade. The glycinergic IPSCs before GABAAerg ic blockade were isolated from mixed IPSCs by subtraction of residual IPSCs after application of 2 p M strychnine from the control IPSCs. The glycinergic IPSCs after GABAAerg ic blockade were isolated from mixed IPSCs by application of 25 p M bicuculline. Short glycinergic IPSCs evoked from medial lemniscus had decay time constants of 14 ± 3 ms (n = 5) before, and 12 ± 2 ms (n = 5) after application o f bicuculline (P > 0.05, f-test). Long glycinergic IPSCs evoked from medial lemniscus had decay time constants of 89 ± 4 ms (n = 4) before, and 84 ± 2 ms (n = 4) after application of bicuculline (P > 0.05, f-test). Hence, blockade of G A B A A receptors had negligible effect on the decay time constants of glycinergic IPSCs. 3 . 6 . Permeability of synaptic receptors We used non-stationary fluctuation analysis to estimate unitary current from whole-cell membrane currents. There is little agreement on some operational details of non-stationary Chapter 3. Results - 9 4 -fluctuation analysis such as the optimal bin width. These operational details may affect the accuracy of non-stationary fluctuation analysis. Hence, we first examined examine the potential methodological sources of error in non-stationary fluctuation analysis. 3.6.1. Methodological sources of error in non-stationary fluctuation analysis We used computer-simulated responses to systematically examine the potential sources of error in non-stationary fluctuation analysis of synaptic currents. Figure 3.46,4 shows an example o f simulated currents using the classic algorithm with input parameters of = 1 p A , N = 500, x l = 0.5 ms, x2 = 10 ms and sampling rate of 10 kHz under optimal conditions (no noise). A comparison of the rise and decay time constants obtained from single exponential fits of the currents to X i and x 2 confirmed the fidelity of the model (Fig. 3.465). We performed non-stationary fluctuation analysis on the currents, divided into 400 bins of equal length. The ratio iJU, which provides assessment of accuracy, was consistently below 0.5 (cf. F ig . 3.46C). Estimates of N were consistently twice the input N . This difference yielded an accurate maximum open probability ( P o m a x = I m a x / N . / e where I m ax is the peak current amplitude). However, the estimates of unitary current and N were associated with substantial error, which required further study. Chapter 3. Results -95-A Amplitude (pA) Figure 3.46. Computer simulated synaptic currents using the classic algorithm. Synaptic currents were stimulated with z, = 1, N = 500, %\= 0.5 ms, T 2 = 10 ms and sampling rate of 10 kHz. A, single channel currents were simulated using a binomial random number generator and open probability of Po(t> = (1- e'^.e" 1 ^ 2 , where t is time and %\ and T 2 were rise and decay time constants respectively. B, sum of 500 simulated single channels produced the synaptic current. A comparison of the rise and decay time constants obtained from single exponential fits of the currents to i l and T2 confirmed the fidelity of the model. Ten simulated synaptic currents were grouped to represent ten successive recordings from one virtual neuron. C When currents were divided into 400 bins for non-stationary fluctuation analysis, estimate of iJU obtained showed a >50% error. Estimate of N was twice the input N. Error bars indicate standard error of mean. We determined whether the error in ie and N is related to the conditions of simulation. Using the above input parameters, we simulated 6 groups of 5 neurons, each neuron having different numbers of currents (10, 20, 40, 60, 80, or 100). The estimated iJU did not significantly change Chapter 3. Results 96 -with the number of simulated currents in each virtual neuron (Fig. 3.47,4; P > 0.05 A N O V A ) . The error in ie also did not depend on the values chosen for N , Ti, and x 2 . Increases in the input unitary current up to 16 times did not change the estimates of iJU (Fig. 3.475; P > 0.05 A N O V A ) . Changes in N (Fig. 3.47C), Xi (data not shown), and x 2 (Fig. 3.47D) did not affect the estimates of iJU (P > 0.05 A N O V A ) . Similarly, changes in input parameters did not affect the error in estimating N (data not shown). Hence, the conditions of simulation could not account for the observed errors in ie and N . A 1 0.75 ^ 0.5 0.25 5 10 Unit current amplitude (pA) c 1 0.75 0.25 -I I -500 1000 1500 Number of channels (N) 50 100 150 Decay time constant (ms) 2000 200 Figure 3.47. Changes in input parameters did not affect the error in unitary current estimation. Altering the number of synaptic currents per virtual neuron (A), unit current amplitude, (5), the number of synaptic channels, N ( Q and the decay time constant (D) did not affect iJU. Error bars indicate standard error of mean (n = 5 for each group). For all comparisons P > 0.05, A N O V A . On the other hand, changes in bin width had marked effects on the iJU ratio. A s shown in Figure 3.48,4, ijii approached a value of 1 when bin width equalled the sampling interval of 0.1 ms, i.e., Chapter 3. Results -97-the minimum value. A n increase in bin width above 0.1 ms increased the error, and iJU approached zero at very large bin widths (Fig. 3.485). A s expected, negligible error occurred when bin width equalled the sampling interval over a range of 0.1 to 0.8 ms (Table 3.3). Hence, the ratio of bin width to sampling interval, rather than the absolute value of bin width, determined the error in ie. Bin width (ms) Figure 3.48. Changes in the bin width significantly affected the estimated unitary current. A, When currents were divided into segments with a bin width equal to the sampling rate, iJU and estimate of N indicated negligible error. B, iJU changed with the bin width (n = 5, P < 0.05, A N O V A ) . Error bars indicate standard error of mean. Chapter 3. Results -98-Table 3.3. Estimates of iJU with altering bin width and sampling rates (n = 5). B i n width/Sampling interval 0. 1 Sampling interval (ms) 0.2 0.4 0.8 1 0.99 ± 0 . 0 1 0.98 ± 0 . 0 1 0.99 ± 0.02 0.98 ± 0.02 2 0.49 ± 0.01 0.49 ± 0 . 0 1 0.51 ± 0 . 0 1 0.49 ± 0.01 4 0.25 ± 0 . 0 1 0.25 ± 0 . 0 1 0.25 ± 0 . 0 1 0.25 ± 0.01 8 0.13 ± 0 . 0 1 0.12 ± 0 . 0 1 0.13 ± 0 . 0 1 0.13 ± 0 . 0 1 16 0.06 ± 0 . 0 1 0.06 ± 0 . 0 1 0.06 ± 0 . 0 1 0.07 ± 0 . 0 1 32 0.03 ± 0 . 0 1 0.03 ± 0 . 0 1 0.03 ± 0 . 0 1 0.04 ± 0 . 0 1 A t every sampling interval, the classic algorithm re-assigns the state of a single channel. Therefore, the duration of a stationary segment is equal to the sampling interval. Hence, we raised the question whether this duration or the sampling rate determined the bin width required for achieving accurate estimates of ie. We simulated synaptic currents using the "stationary-segments algorithm", where the stationary segments were longer than the sampling interval. We simulated synaptic currents with stationary segments of 1 to 6 ms and input parameters of = 1 p A , N = 200, Ti= 0.5 ms, X\= 10 ms and a sampling rate of 10 kHz . A n example of single channel current simulated with stationary segments of 1 to 6 ms is shown in Figure 3.49,4 (compare to Fig . 3.46,4 where there is no stationary segment). Chapter 3. Results -99-A Illl 10 ms flULMJULJl 1 ms e 0.75 ..0.5H 0 .25H 0.1 - J 1—I M I N I 1 1—I I I I I 11 I 1—I I M i l l 1 10 100 Bin width (ms) Figure 3.49. Interplay between bin width and the length of stationary segments affected the iJU ratio. A, an example of computer simulated single channel current with stationary-segments algorithm. The input parameters were stationary-segments length of 1 ms, sampling rate of 10 kHz , U = 1 p A , T i = 0.5 ms, T 2 = 10 ms. B, synaptic currents produced from the sum of single channels currents (N = 200) were analyzed with varying bin width. iJU significantly changed with the bin width 5, P < 0.05, A N O V A ) . When the bin width was reduced to below the length of the stationary segment (1 and 6 ms), the estimate of unitary current reached a plateau with negligible error. Error bars indicate standard error of mean. Changes in the bin width also affected unitary current estimates obtained from currents simulated using the stationary-segments algorithm (P < 0.05 A N O V A ; Fig . 3.495). A t large bin widths, there were large errors in the estimate of unitary current. When the bin width was reduced to below the length of the stationary segment, the estimate of unitary current reached a plateau, of negligible error (Fig. 3.495). The standard error of mean was decreased at smaller bin widths, indicating an increased reliability of the estimates. The estimates of N were inversely related to the estimates of unitary current (data not shown). Thus, to obtain accurate estimates from non-stationary fluctuation analysis of synaptic currents, the bin width must be smaller than the length of the stationary segments for single channels. Chapter 3. Results -100-It is not certain whether random noise would change the relationship between the accuracy of non-stationary noise analysis and the bin width. Hence, we repeated the experiments after adding random Gaussian noise to the simulated synaptic currents. The noise had peak amplitude equal to the input unitary current. A n example of a simulated synaptic current of this kind is shown in Figure 3.51X4. A comparison of the rise and decay time constants obtained from single exponential fits of the currents to X\ and T 2 confirmed the fidelity of the model. When currents were divided into 400 bins for non-stationary fluctuation analysis, the iJU ratio consistently displayed ~50% error (cf. Fig . 3.505). The iJU ratio did not change with the number of synaptic currents per virtual neurons (P > 0.05 A N O V A ; Fig . 3.505). Changes in the bin width affected the iJU ratio obtained from non-stationary fluctuation analysis of currents with added noise (P < 0.05 A N O V A ; Fig . 3.50C). A t large bin widths, there were large errors in the estimate of unitary current. When the bin width was reduced below the length of the stationary segment, the iJU ratio reached a plateau with negligible error (Fig. 3 .50Q. The standard error of mean was decreased at smaller bin widths, indicating an increased reliability of the estimates. These observations indicated that bin width remains a major source of error in non-stationary fluctuation analysis of simulated synaptic currents, regardless of the presence of added noise. Chapter 3. Results - 101 -Figure 3.50. Effect of added Gaussian noise on unitary current estimates. A, shows an example of computer simulated synaptic current with added Gaussian noise. Comparison of the rise and decay time constants obtained from single exponential fits of the currents to Ti and %2 confirmed the fidelity of the model. B, altering the number of synaptic currents in the range o f ten to 100 currents per virtual neuron did not affect the iJU ratio. C, the iji; ratio significantly changed with the bin width (n = 5, P < 0.05, A N O V A ) . A t the bin width below the length of the stationary segment (1 and 6 ms), the estimate of unitary current reached a plateau with negligible error. Error bars indicate standard error of mean. We examined the applicability of our findings on simlPSCs to IPSCs. We determined unitary current estimated from non-stationary fluctuation analysis of pharmacologically isolated G A B A A e r g i c IPSCs in ventrobasal thalamus. The IPSCs were recorded with C s - Q X patch Chapter 3. Results -102-solutions at the holding potential of Vh = -60 mV. Bath application of 1 mM kynurenic acid and 2 p M strychnine isolated the GABAAerg ic IPSCs from ionotropic glutamatergic E P S C s and glycinergic IPSCs. Figure 3.51^4 and 3.5IC show examples of the recorded evoked and spontaneous GABAAerg ic IPSCs. Unitary currents were estimated using a wide range of bin widths and were normalized to the estimated unitary current at the smallest bin width (in0rm)- The true value of synaptic unitary current is not known. Hence, the estimated unitary current at the smallest bin width may or may not reflect an accurate synaptic unitary current. However, normalizing the estimated unitary currents to the estimated unitary current at the smallest bin width would be useful for examining the relation between bin width and the estimated unitary currents. B i n width affected the estimate of unitary current in both evoked and spontaneous IPSCs. A s the bin width increased, inorm decreased and the standard error of mean for estimates of unitary current greatly increased (Fig. 3.515 and 3.5\D). A t very large bin widths, inorm approached zero, reflecting low accuracy. The large standard error at these bin widths reflected a low reliability of the estimates. A t small bin widths, inorm had a small standard error, representing a high reliability. The estimate of unitary current reached a plateau at about the bin width of 2 ms for both evoked and spontaneous G A B A A e r g i c IPSCs (Fig. 3.515 and 3.51Z)). A similar relationship was observed for isolated glycinergic IPSCs recorded at the holding potential of Vh = -80 m V (Fig. 3.52). The estimate of unitary current reached a plateau at about the bin width of 3 ms for glycinergic IPSCs. Hence, we set the bin width to 1.5 ms in our future analysis. Chapter 3. Results - 103-Figure 3.51. Applicability of the findings to GABAAerg ic IPSCs. Non-stationary fluctuation analysis was performed on purely G A B A A e r g i c evoked (A) and spontaneous ( Q IPSCs recorded in ventrobasal thalamus (Vh = -60 mV) . Unitary currents were estimated using a wide range of bin widths and were normalized to the estimated unitary current at the smallest bin width (imrm). Uorm significantly changed with the bin width for evoked (B; n = 7, P < 0.05, A N O V A ) and spontaneous IPSCs (D;n = l,P< 0.05, A N O V A ) . A t smaller bin widths, inorm reached a plateau and the standard error of the ratio was greatly reduced. Error bars indicate standard error of mean. Dashed line in D indicates that the standard error extends beyond the size of the figure. Chapter 3. Results - 104-1.5i 1.25J 0.1 1 10 100 Bin width (ms) Figure 3.52. Applicability of the findings to glycinergic IPSCs. Non-stationary fluctuation analysis was performed on purely glycinergic IPSCs recorded in ventrobasal thalamus (Vh = -80 m V ) . Unitary currents were estimated using a wide range of bin widths and normalized to the estimated unitary current at the smallest bin width (i„0rm)- inorm significantly changed with the bin width (n = 7, P < 0.05, A N O V A ) . A t smaller bin sizes, inorm reached a plateau and the standard error of the ratio was greatly reduced. Error bars indicate standard error of mean. 3.6.2. Evoked IPSCs We confirmed the stability of quantal release in the recorded IPSCs before subjecting them to non-stationary noise analysis. The amplitude of evoked IPSCs did not significantly change with the stimulus number (Fig 3.53), which implied no change in quantal release at stimulation frequencies below 0.5 Hz . We grouped every 3 successive responses and calculated the coefficients of variation for each triplet. The coefficients of variation did not change with triplet number (Fig. 3.53), confirming stable quantal release (cf. Scheuss and Neher 2001). Chapter 3. Results - 105 -^ -300 < CD -g -200 CL ra -100 O W a. 0 c -0.15 o 2 -0.10 | -0.05 £ CD O o 0 I i i i i i m • i — i — i i i i i — i 2 3 4 5 6 7 8 9 10 Stimulus number i i i i i i i i i 1 I I i -i i i i 7 I I I l i n i i i i i — i i i 1 2 3 4 5 6 7 8 Triplet number Figure 3.53. Stability of quantal release at stimulation frequencies < 0.5 Hz . The amplitude of successive IPSCs did not change with the stimulus number in a train of 10 stimuli. .Every three successive currents were grouped into a triplet (brackets) and the coefficient of variation was calculated for the triplets. The coefficient of variation did not change with triplet number, confirming the stability of quantal release, n = 26, Vh = -80 m V , Ec\ = -53 m V . We estimated the single channel currents activated during glycinergic and G A B A A e r g i c IPSCs. Figure 3.544 shows short- and long-duration glycinergic IPSCs evoked by 10 successive stimuli to neurons held at -80 m V . Analysis of IPSCs yielded variance-to-mean current relationships that were well-described by a quadratic function (Fig. 3.545). Linear least squares fits to the initial part of the variance-to-mean current relationships yielded estimates o f ia, consistent with those obtained from quadratic fits. The estimates of I'CI for the short-duration glycinergic IPSC averaged -0.6 ± 0.2 p A (n = 12) whereas J'CI for the long-duration IPSC averaged -0.8 ± 0.2 p A (n = 8). These estimates did not differ from each other, or from estimates of z'ci from G A B A A e r g i c IPSCs (-0.6 + 0.1 p A , n = 17; P > 0.05, A N O V A ) . Chapter 3. Results - 106-Figure 3.54. Estimation of unitary channel currents from non-stationary fluctuation analysis of glycinergic IPSCs. A , 10 successive, short- and long-duration glycinergic currents in two neurons. B, current variance was plotted as a function of mean current. Variance-mean relationships were fitted with a2( t) = zci-^mean(t) - / 2mean(t)/N + oth2(t> (dark curves), z'ci was also obtained as the slopes of the initial part of the variance-average current relations, fitted by linear regression. The two methods yielded the same values of z'ci- The estimates were ic\ = 0.9 p A for short-duration currents (left), and ic\ = 0.5 p A for long-duration currents (right). The gray lines show 95% confidence intervals. Vh = - 80 m V , EQ\ = -53 m V . For comparison of /cis obtained from IPSCs and sIPSCs under differing holding potentials and EQ\S, we used the G H K equation to convert the values to C f permeability, Pci. Calculation of CI" permeability is contingent upon the channels being permeable solely to CI". We confirmed this by examining the current-voltage relationship for resolved glycinergic and G A B A A e r g i c currents. These currents reversed at £ci (Fig. 3.55), implicating CI" dependence with negligible contribution from other ion species. Using the Goldman-Hodgkin-Katz constant-field equation, Chapter 3. Results - 107 -we converted elementary currents to estimates of CI" channel permeability (Pci). The short- and long-duration, glycinergic IPSCs yielded P C i values of 1.6 ± 0.5 x IO" 1 3 cm 3/s (« = 12) and 1.7 ± 0.4 x 10"1 3 cm 3/s (n = 8), respectively. The G A B A A e r g i c IPSCs yielded a value of P a = 1.5 ± 13 3 0.3 x 10" cm /s (n = 17). These values were not significantly different from each other (P > 0.05, A N O V A ) . Glycinergic -200 J Figure 3.55. Current-voltage relationship for resolved glycinergic and G A B A A e r g i c currents. These currents reversed at Ecu implicating CI" dependence with little or no contribution from other ion species. Thalamic neurons possess a variety of voltage-gated channels including low threshold calcium channels, sodium channels, and potassium channels (Turner et al., 1997, Wan et al., 2003). While activation of these channels during voltage clamp is unlikely, we sought to confirm the pure CI" nature of the estimated permeabilities by applying drugs to block voltage-dependent currents. In 7 neurons, we used extracellular application of N i 2 + to block low threshold C a 2 + Chapter 3. Results -108-spikes that reflected a T-type C a 2 + current. We also used intracellular application of QX-314 and C s + to block voltage-dependent N a + and K + currents. We evoked IPSCs by electrical stimulation of medial lemniscus. We did not observe glycinergic IPSCs in these neurons, possibly due to the pre- and postsynaptic blocking actions of N i 2 + on C a 2 + channels and glycine receptors (cf. D o i et al. 1999). Under these conditions, GABAAerg ic IPSCs reversed polarity within 2 m V of Ea (Fig. 3.56). For these currents, we calculated a unitary current of 5.3 ± 2.3 p A ( A V = 60 m V ) , which pertained to a conductance of 88 ± 38 pS and Pci = 1.7 ± 0.8 x l O " 1 3 cm 3/s. The calculated Pci at this condition did not differ from Pci estimated in the absence of intracellular application of C s + and QX-314 (/-test, P > 0.05). 50 pA 50 ms Figure 3.56. Reversal potential for IPSCs recorded with C s - Q X patch solution. The recorded IPSCs reversed polarity within 2 m V of EC\ (0 mV) . IPSCs are average of 10 traces. We attempted to isolate purely glycinergic or purely G A B A A e r g i c IPSCs with no contribution by ions other than CI" by repeating the previous experiments using a +20 m V pre-pulse from the Chapter 3. Results -109-holding potential of -60 m V to block T-type C a 2 + current, rather than using N i 2 + . The duration of pre-pulse was set to 1 s, which is five times the inactivation time constant of T-type C a current at V h = -40 m V (Steriade et al., 1997). Hence, > 99% of T-type C a 2 + channels were l ikely inactivated. We evoked IPSCs by electrical stimulation of medial lemniscus. Unexpectedly, IPSCs evoked with this pre-pulse had higher amplitude, but similar rise and decay times compared to the IPSCs evoked without the pre-pulse in the same neurons (Fig. 3.57.4). The average amplitude of IPSCs increased from 795 ± 58 p A to 1041 ± 95 p A following the pre-pulse (n = 5, P < 0.05, paired /-test). When purely glycinergic and purely G A B A A e r g i c components of the mixed IPSCs were isolated, the increase in amplitude was only evident in the glycinergic component of the IPSCs (Fig. 3.575 and Q . The average amplitude of glycinergic IPSCs increased from 139 ± 54 p A to 429 ± 51 p A following the pre-pulse (n = 5, P < 0.05, paired /-test). The average amplitude of G A B A A e r g i c IPSCs was 655 ± 70 p A before and 608 ± 120 p A following the pre-pulse (n = 5,P> 0.05, paired /-test). 500 pA 100 ms Figure 3.57. IPSCs recorded with C s - Q X patches and blockade of T-type C a 2 + current using a 1 s long +20 m V pre-pulse from the holding potential of -60 m V . IPSCs evoked with the pre-pulse had higher amplitude, but similar rise and decay times (A). This increase in amplitude was due to an increase in glycinergic (B) rather than G A B A A e r g i c component of the IPSC. Traces are average of ten recordings. Chapter 3. Results - 110-For these currents, we calculated unitary current, conductance and Pci using non-stationary noise analysis. Table 3.4 summarizes the results. Unitary current, conductance, and Pci measured at this condition did not differ from those measured in the absence of intracellular application of C s + and QX-314 (paired f-test, P > 0.05). Table 3.4. Properties of glycinergic and GABAAerg ic channels before and after application of a pre-pulse to remove T-type Ca current. Data is presented as mean ± S E M . Vh = -60 m V ; Ea = 0 m V ; n = 5 neurons. I'd (pA) Y (PS) Pci (xl0" 1 3 cm 3 /s) Glycinergic Before pre-pulse After pre-pulse 6.1 ± 0 . 3 4.6 ± 0 . 8 101 ± 5 76 ± 13 2 ± 0 . 1 1.5 ± 0 . 3 GABAAerg ic Before pre-pulse 3.6 ± 2 . 2 60 ± 3 6 1.2 ± 0 . 7 After pre-pulse 3.7 ± 1.7 62 ± 2 8 1.2 ± 0 . 6 3.6.3. Spontaneous IPSCs Estimates of ia from multi-quantal IPSCs can undergo distortion from fluctuations in transmitter release on a trial-to-trial basis, as well as from other factors (cf. Diamond and Jahr, 1995). We sought to overcome this limitation by calculating f C i from sIPSCs recorded at -60 m V with Ea = 0 m V (n = 7 neurons), which have a predominantly mono-quantal nature. Figure 3.58 shows an example of calculating z'ci from sIPSCs. Chapter 3. Results - I l l -Short-duration sIPSC 10 ms 60-i 0 -I 1 1 1 0 15 30 45 'mean (?) (PA) Figure 3.58. Estimation of unitary channel currents from non-stationary fluctuation analysis of glycinergic sIPSCs. Current variance was plotted as a function of mean current. Variance-mean relationships were fitted with rj2(t) = /ci-Anean(t) - ^2mean(t)/N + GtMt) (dark curve). j"ci was also obtained as the slopes of the initial part of the variance-average current relations, fitted by linear regression. The two methods yielded the same values of z'ci- The estimated ic\ was -3.3 p A in this example. The gray lines show 95% confidence intervals. Vh = - 60 m V , Ec\ = 0 m V . For short-duration sIPSCs, we found a mean ia - -3.5 ± 0.6 p A . For long-duration sIPSCs, the mean z'ci was -2.8 ± 0.8 pA. These values did not differ from each other or from the value of -2.5 ± 0.5 p A obtained for G A B A A e r g i c sIPSCs (P > 0.05, A N O V A ) . Current-voltage relationships for the ic\s calculated from glycinergic and G A B A A e r g i c sIPSCs were linear and showed apparent reversal potentials near Ea (Fig. 3.59), implicating CI" dependence with little or no contribution from other ion species. The mean single channel conductances for fast (58 ± 10 pS, n = 6), and slow (46 ± 14 pS, n = 6) glycinergic sIPSCs did not differ from each other, or from the conductance (41 ± 9 pS, n = 7) from G A B A A e r g i c sIPSCs (P > 0.05, A N O V A ) . Chapter 3. Results -112-A 1.5- < a. i -60 -40 -20 20 I 0 V„ (mV) i - 1 .5 -i - 3 --4 .5 -B -60 -40 -20 1.5' 0 - 1 .5 -- 3 --4 .5 -< Q . i 20 Vn (mV) Figure 3.59. Current-voltage relationships for Zeis estimated from sIPSCs. Both glycinergic (A) and GABAAerg i c (B) ic\S reversed at 0 m V (Ec\) in n = 3 glycinergic, and n = 4 G A B A A e r g i c currents. Error bars indicate standard errors. We calculated P C i values of 0.8 ± 0.2 x 10"1 3 cm 3/s (n = 6) and 0.9 ± 0.3 x IO" 1 3 cm 3/s (n = 6) for short- and long-duration, glycinergic sIPSCs. The GABAAerg ic IPSCs yielded a mean value of 13 3 Pci = 1.1 ± 0.2 x 10" cm /s (n = 6). These values were not significantly different from each other (P > 0.05, A N O V A ) . Mean P C i estimated from sIPSCs did not differ from those estimated for evoked IPSCs (P > 0.05, A N O V A ) . Hence, there was a good agreement between spontaneous and evoked IPSCs in estimation of Pci (Fig. 3.60). 3.0' o 2.0-J x « o 1.0' o Q_ 0.0' (6) X (12) (8) T JL sIPSC IPSC sIPSC IPSC Short Long Figure 3.60. CI" permeabilities of G l y R estimated from evoked and spontaneous IPSCs were consistent. Chapter 3. Results - 113-3.7. Biophysical properties of extrasynaptic inhibitory receptors We compared the biophysical properties of synaptic channels estimated from whole-cell IPSCs and sIPSCs with the biophysical properties of extrasynaptic channels estimated from single channel recordings. 3.7.1. Kinetic properties of extrasynaptic receptors Application of glycine, taurine, and P-alanine induced inward, single channel currents in membrane patches (Fig. 3.61). Prior to agonist application, the outside-out patches infrequently displayed spontaneous currents at -60 m V . Glycine (20 pM) activated currents in 11 out of 27 patches. In separate experiments, taurine (20 pM) activated currents in 12 out of 28 patches, and p-alanine (20 pM) activated currents in 8 out of 30 patches. The means of P 0 for activations by glycine (0.029 ± 0.018, n = 11), taurine (0.029 ± 0.031, n = 10), and p-alanine (0.042 ± 0.013, n = 7) were not different (P > 0.05, A N O V A ) . P 0 had a tendency to decline during agonist application and hence, we did not test more than one agonist on individual patches. When strychnine (1 pM) was co-applied with glycine, taurine, or P-alanine, single channel currents occurred very infrequently (overall P 0 <0.001). A n observed reversal potential near EC\ and sensitivity to strychnine implicated glycine receptors in the agonist-evoked currents. To compare the burst durations of agonist-induced currents, we calculated the critical time, x c , from channel closed time distributions. The closed time distributions were well described by the sum of 4 exponentials, as exemplified for glycine in Figure 3.615. The calculated values of x c did not differ between channels that displayed short- or long-duration bursts (P > 0.05, Chapter 3. Results A N O V A ) . The mean T c for glycine (7.3 ± 0.8 ms, n = 11), taurine (8.3 ± 0.8 ms, n = alanine (8.9 ± 0.9 ms, n = 7) did not differ (P > 0.05, A N O V A ) . - 114 -10), and p-The currents activated by the p-amino acids displayed either short or long-duration bursts (Fig. 3.61,4). Glycine activated short-duration bursts in 10 out of 11 patches, and long-duration bursts only in 1 patch. Taurine-activated currents were characterized by short-duration bursts of openings in 6 out of 10 patches, and long-duration bursts in the remaining 4 patches. P-alanine activated short-duration bursts in 4 out of 7 patches, and long-duration bursts in the remaining 3 patches (Fig. 3.61,4). Chapter 3. Results - 115-Glycine Short-Duration Bursts Taurine p-Alanine Long-Duration Bursts Taurine I ' W W n-Alanine Log closed time (s) -3 -2 -1 Log burst duration (s) -2 -1 0 Log burst duration (s) Gly Tau [WUa Tau p-Ala Short Long Figure 3.61. Glycine receptor agonists activate short- or long-duration bursts of channel openings. A, representative currents obtained from separate membrane patches. Channel openings appear as downward deflections from baseline (inward currents). Arrows indicate currents associated with a substate conductance. The insert under the long-duration bursts activated by P-alanine shows a segment of the main trace with higher amplification and time-base. Current levels indicative of closed channel (c), substate conductance (s, arrow) and full open conductance (o) states are indicated by dashed lines. B, distribution of closed times (n = 2,742 events) for glycine-activated currents from an additional patch showing fit by the sum of four exponential terms, with areas and time constants (arrowheads). C, distributions of burst durations for taurine-activated currents in two additional patches showing short- (n = 552 bursts) and long-duration bursts (n = 506 bursts). Each distribution was well-fitted by the sum o f 3 exponentials, short-duration bursts. Arrowheads indicate time constants. D, mean durations of short- and long-duration bursts activated by glycine receptor agonists. Numbers in parentheses indicate patches. Mean durations of short-duration bursts activated by taurine and P-alanine were significantly different from long-duration bursts activated by the same agonist (*tP < 0.05, A N O V A ) . Agonists were applied at 20 U.M to outside-out patches. V h = -60 m V , Ea = 0 m V . A s exemplified by taurine (Fig. 3 .61Q, burst-duration distributions were well described by the sum o f 3 exponentials. The mean burst duration for glycine-activated channels was 19 ± 4 ms, whereas the sole long-duration burst averaged 87 ms. The taurine-activated short-duration bursts had a mean of 26 ± 4 ms, whereas long-duration bursts averaged 88 ± 8 ms. The P-alanine Chapter 3. Results -116-activated short-duration bursts had a mean of 21 ± 4 ms, whereas long-duration bursts averaged 137 ± 5 ms. The average lifetimes of short-duration bursts did not depend on the nature of the agonist (cf. F ig . 3.6ID). The average lifetimes of long-duration bursts activated by taurine or |3-alanine did not differ (P > 0.05, A N O V A ) . For the P-amino acids, short and long bursts differed significantly from each other in duration, and likely represented 2 populations (cf. F ig . 3.6ID, P < 0.05, A N O V A ) . 3.7.2. Permeability of extrasynaptic receptors A l l 3 agonists evoked extrasynaptic currents of small and large amplitude (arrowheads, Figs. 3.61/1 and 3.62/1 and B). Smaller currents were seen only in the presence of larger amplitude currents. Hence, the small currents likely reflected openings to a substate conductance, appropriately 70% of the full conductance. Amplitude distributions for the currents were wel l described by the sum of 2 Gaussian terms (Fig. 3.62A). The current-voltage (I-V) relationships were linear over the range of 0 to -60 m V (Fig. 3.625). From these relationships, the mean conductances from short-duration bursts were 15 ± 1 pS (n = 10), 21 + 2 pS (n = 6) and 22 ± 3 pS (n = 4) for glycine, taurine and p-alanine, respectively. The mean conductances from long-duration bursts were 19 pS (n = 1), 33 ± 1 pS (n = 4) and 30 ± 4 pS (n = 3) for glycine, taurine and P-alanine, respectively. These conductances were not different (P > 0.05, A N O V A ) . Table 3.5 summarizes the biophysical properties of extrasynaptic channels, and shows the Pcis calculated from single channel currents. The properties of GABAAerg ic synaptic channels were obtained from K i m et al. (2004). Chapter 3. Results - 117-Glycine Taurine p-Alanine O Sub-conductance Gly Tau p-Ala Tau p-Ala Figure 3.62. Conductance of single channels activated by glycine receptor agonists. A, amplitude distributions for agonist-activated currents (Vh = -60 m V ) were fitted by the sum o f 2 Gaussian terms (smooth curves). Means (arrowheads) were -1.2 p A and -0.96 p A for glycine (n = 1473 transitions), -1.6 p A and -1.2 p A for taurine (n = 2465 transitions), and -1.7 p A and -1.3 p A for (3-alanine (n = 1144 transitions). B, single channel I -V relationships for the full and substate conductances of glycine-activated receptor channels. Data were obtained from 12 patches. Linear regression fits indicated mean values of 13 pS and 17 pS for the substate and full conductances, respectively. In all panels, agonists were applied at 20 pM to outside-out patches with Ea = 0 m V . C, comparison of Pci values for short and long duration channel bursts, activated by glycine receptor agonists. Numbers in parentheses indicate number o f patches. There were no significant differences in values obtained (P >0.05, A N O V A ) . Chapter 3. Results - 1 1 8 -Table 3.5. Biophysical properties of extrasynaptic inhibitory receptors in ventrobasal thalamus. Data is represented as mean ± SEM. Numbers in parentheses represent the number of patches. Burst duration (ms) z'ci (pA) Y (pS) Pci (xl0"13cm3/s) Fast glycinergic Glycine (10) Taurine (6) P-alanine (4) 19 + 4 26 ±4 21 ±4 0.8 ±0.1 1.1 ±0.1 1.0 ±0.1 15± 1 21 ± 2 22 ± 3 0.3 ±0.01 0.4 ± 0.04 0.4 ± 0.06 Slow glycinergic Glycine (1) Taurine (4) p-alanine (3) 87 88 ±8 137 ± 5 0.8 1.6 ±0.1 1.6 ±0.3 19 33 ± 1 30 ± 4 0.4 0.7 ± 0.03 0.6 ±0.08 GABAAergic GABA (8) 17.8 ±0.8 0.9 ±0.1 16 ± 1 0.3 ±0.01 3.8. Comparison of synaptic and extrasynaptic inhibitory receptors Synaptic and extrasynaptic glycine receptors showed similar kinetics (Fig. 3.63,4). The short and long time constants for glycinergic IPSC and sIPSC decays were similar to short and long channel burst durations activated by glycine, taurine, and P-alanine (ANOVA, P > 0.05). The time constants for GABAAergic IPSC and sIPSC decays were similar to channel burst durations activated by GABA (ANOVA, P > 0.05). Synaptic channels had higher Pci than extrasynaptic channels (cf. Fig. 3.635). The Pci calculated from short glycinergic IPSCs and sIPSCs were higher than the Pci calculated from short burst durations activated by glycine, taurine, and p-alanine. The Pci calculated from long glycinergic IPSCs and sIPSCs were higher than the Pci calculated from long burst durations activated by glycine, taurine, and P-alanine (Fig. 3.635). Chapter 3. Results - 119-2 . 150 '100 81 g S .E-o O S 50 Q (S) I (12) (10) <6) 8 sIPSC IPSC Gly Tau p-Ala Short 3.0-1 2.0 E LO o . o (6) (12) sIPSC IPSC Gly Tau p-Ala Short T (6) m sIPSC IPSC Gly Tau p-Ala Long I T (6) (8) _ I H I sIPSC IPSC Gly Tau p-Ala Long Figure 3.63. Synaptic and extrasynaptic GlyRs show similar kinetics, but dissimilar CI" permeabilities. A, short and long time constants for IPSC and sIPSC decays were similar to short and long channel burst durations ( A N O V A , P > 0.05). B, synaptic channels had significantly higher P C i than extrasynaptic channels ( A N O V A , *P < 0.05). Numbers o f neurons or membrane patches are shown in parentheses. Gly , glycine; Tau, taurine; P - A l a , P-alanine. In view of the similar Pcis from evoked and spontaneous IPSCs, as well as the similar extrasynaptic Pcis for the agonists, we pooled the data into synaptic and extrasynaptic categories. Table 3.6 shows the distinct nature of Pcis of synaptic and extrasynaptic channels. The short-and long-duration, glycinergic synaptic channels yielded Pcis that were higher than the estimates from short- and long-duration bursts activated by the agonists (P < 0.05, unpaired f-test). The G A B A A e r g i c synaptic channels yielded Pcis that were higher than the estimates from extrasynaptic channels (P < 0.05, unpaired f-test) obtained from K i m et al. (2004). Chapter 3. Results - 120 -Table 3.6. Chloride permeabilities of synaptic and extrasynaptic glycine and GABA A receptors. Data are means ± SEM and number of observations in parentheses. Pci (xl0"13cm3/s) Synaptic (18) Fast glycinergic 1.4 ±0.3* Extrasynaptic (20) 0.4 ± 0.02 Synaptic (13) Slow glycinergic 1.4 ±0.3* Extrasynaptic (8) 0.6 ± 0.05 z-""1 A T> A Synaptic (25) GABAAergic 1.3 ±0.2* Extrasynaptic (8) 0.3 ±0.01 *Significant different from corresponding synaptic value (P < 0.05, unpaired f-test). 3.9. Effect of A M B D on IPSCs We sought to examine the possibility of taurine mediating long-duration glycinergic IPSCs in ventrobasal thalamus. We compared the antagonism of strychnine (2 pM) with that of putative taurine antagonist, AMBD (250 pM) on IPSCs evoked from medial lemniscus. This concentration of AMBD is shown to block > 95% of taurine-induced currents in mice spinal cord (Mathers, 1993). In five tested neurons held at -80 mV, bath application of AMBD inhibited the similar portion of the IPSC amplitude regardless of the decay time constant (Fig. 3.64). Application of AMBD blocked 41 ± 15% of the IPSC amplitude. Following recovery, strychnine blocked 32 ± 14% of the amplitude. The percentage of IPSCs blocked by AMBD and strychnine were not different (P > 0.05, paired f-test). Hence, the IPSCs were equally sensitive to 250 p M AMBD and 2 p M strychnine. Given the possible non-specific action of AMBD (cf. Mathers, 1993), we did not further pursue these investigations. Chapter 3. Results - 121 -I | ' IH | I»* » l » 50 pA 50 ms Figure 3.64. Application of A M B D (250 [IM) blocked a biexponential glycinergic medial lemniscal IPSC with short- and long-duration decay time constants. Upon recovery, strychnine (Str, 2 LIM) blocked a similar proportion of the IPSC. Traces represent an average of 10 responses. In summary, the major finding of this chapter was the diverse kinetics of the glycinergic inhibition. The observations were consistent with the activation of 2 kinetically distinct populations of glycine receptors. The 2 populations of glycine receptors with fast and slow kinetics appeared to be segregated under separate nerve terminals. The observation of monophasic sIPSCs further substantiated the previous conclusion that glycine and G A B A are likely co-transmitted, rather than co-released, in V B thalamus. The kinetics of synaptic glycine and G A B A A receptor channels mirrored the kinetics of extrasynaptic receptor channels. Synaptic channels displayed higher CI" permeations than their extrasynaptic counterparts. This observation was likely genuine and not due to vagaries in fluctuation analysis. - 122 -Chapter 4 DISCUSSION This thesis provides, for the first time, convincing evidence for glycinergic inhibition in the ventrobasal nuclei of rat thalamus. Furthermore, the thesis characterizes the biophysical properties of glycinergic in relation to GABAAerg ic inhibition evoked by somatosensory nerve stimulation. The results of these studies provide insight into the provenance of glycinergic inhibition in ventrobasal thalamus. 4.1. Functional glycine receptors in ventrobasal thalamus We demonstrated that glycine receptor (GlyR) subunits are expressed in V B thalamus. There was low to moderate immunopositive staining for both oci and ot2 subunits in the ventrobasal nuclei (Fig. 3.2). These subunits were evident on or near somata, as in the control medulla (cf. Arak i et al., 1988). During development, the oti subunit progressively replaces the a.2 in brainstem and spinal cord (Becker at al., 1988; Singer et al., 1998), whereas the 0C2 subunit l ikely predominates in the cortico-thalamocortical system of the adult rat (Malosio et al., 1991; Chattipakorn and McMahon , 2002). In ventrobasal thalamus, the expression of both subunits with punctate and diffuse distributions may correspond to synaptic and extrasynaptic glycine receptors. We confirmed the functional nature of a subunits by demonstrating the effects of glycine agonists on ventrobasal neurons. Glycine, taurine, and p-alanine, but not D- or L-serine, depolarized neurons and decreased their input resistance. These effects were due to an increased conductance for CI" as shown by their reversal potentials near Ea (Fig. 3.12). The effects of Chapter 4. Discussion - 123 -these amino acids were reversible upon washout. Strychnine completely blocked the increases in conductance induced by glycine and taurine, and greatly reduced the increase due to P-alanine (Fig. 3.14). Glycine, taurine, and P-alanine had similar efficacies, apparent from their maximal responses (cf. F ig . 3.15). It is noteworthy that to quantify the effects of agonists we used input resistance rather than the frequency of action potential firing and changes in resting membrane potential. This choice was made due to spatial considerations related to the quantification of the changes in resting membrane potential and action potential frequency. In addition, the all-or-none nature of action potentials limits the interpretation of action potential frequency. It is unlikely that the effects of glycine and taurine on ventrobasal neurons were mediated by receptors other than GlyRs . Bath application of a glutamate antagonist, kynurenic acid, did not block the effects of p-amino acids. Hence, P-amino acid-induced changes did not involve N M D A or A M P A receptors (cf. Parsons et al., 1998). Ligand gated chloride channels are limited to G l y R , G A B A A and G A B A C receptors (reviewed by Jentsch et al., 2002). While strychnine at high concentration non-specifically blocked the effects of G A B A on ventrobasal neurons, G A B A A receptor blockade did not occur at the low concentrations that we used to block the effects of p-amino acids (cf. Fig . 3.19). Furthermore, application of a G A B A A antagonist did not affect the decrease in input resistance induced by p-amino acids. It is unlikely that the bicuculline-insensitive G A B A c receptors mediated these effects because application of bicuculline abolished the effects of G A B A , indicating an absence of G A B A C receptors in ventrobasal thalamus. Hence, GlyRs likely mediated the effects of P-amino acids on ventrobasal neurons. Chapter 4. Discussion - 124 -In contrast to glycine and taurine, the actions of (3-alanine in the thalamus were not limited to strychnine-sensitive activation of GlyRs . The same concentration of strychnine that abolished the effects of other (3-amino acids, did not completely block the effects of (3-alanine. Co -application of strychnine, bicuculline and kynurenic acid did not completely block the effects of 3-alanine (cf. F ig . 3.14). Hence, the residual effects were not mediated by conventional G l y R , G A B A A receptors (cf. W u et al., 1993), or ionotropic glutamate receptors. The effects were mediated by CI" channels, as evident from the reversal potential calculated from voltage-current relationships. The only known ligand-gated CI" channel insensitive to strychnine and bicuculline is the G A B A c receptor (Chebib and Johnston, 2000) . However, the effects of G A B A were abolished by application of bicuculline, demonstrating an absence of the bicuculline-insensitive G A B A c receptors in ventrobasal thalamus. Hence, the residual effect of (3-alanine was either mediated by a yet to be discovered receptor, or a distinct binding site on a G l y R or G A B A A receptor variant insensitive to strychnine and bicuculline (cf. Kuhse et al., 1990). Concentration-response relationships provided further insight into the effects of (3-amino acids on ventrobasal neurons. The concentration-response relationships for the effects of glycine and taurine on membrane input resistance were consistent with the literature (Lynch et al., 1997; Moorhouse et al., 1999; Farroni and M c C o o l , 2004) . The EC50S of glycine and taurine found in the present studies were between those reported for homomeric receptors assembled from recombinant oci and 0C2 G l y R subunits. Ventrobasal thalamic neurons stained positive for both (Xi and ct2 subunits. Hence, the calculated E C 5 u s likely reflected an activation of a mixture of G l y R s containing oci, OC2 or both subunits. The calculated hillslope of ~2 for the concentration response relationships likely reflects two binding sites for glycine and taurine on each receptor (cf. Lynch Chapter 4. Discussion - 125 -et al., 1997; Moorhouse et al., 1999; Farroni and M c C o o l , 2004). P-alanine, however, displayed a concentration response relationship with hillslope of ~1 . This is in contrast to the previously reported investigations indicating similar slopes for P-alanine and other G l y R agonists (cf. Lynch et al., 1997; Moorhouse et al., 1999; Farroni and M c C o o l , 2004). Certain point mutations can produce variants of the G l y R subunit that are un-responsive to P-alanine (Rajendra et al., 1995). Another possibility is that the unique effect of P-alanine is mediated by another, currently unknown, receptor or a strychnine- and bicuculline-insensitive binding site on G l y R or G A B A A receptors. A t the cellular level, our findings uncovered a distinctive distribution of GlyRs . Functional GlyRs were likely limited to type-I T C neurons. Neurons that responded to glycine were typically larger as judged by their higher input capacitance, and had lower input resistance compared to the non-responding neurons. These characteristics match the morphological and electrophysiological characteristics of type-I T C neurons (cf. Turner et al., 1997). Type-I T C neurons project to layers I V and V of somatosensory cortex. The distinct anatomical connections of type-I and type-II T C neurons imply distinct functional roles (Penny et al. 1982; cf. Castro-Alamancos and Connors, 1997). Hence, glycinergic inhibition in thalamus likely serves a specific physiological function, undefined as yet. The effects of G A B A on the membrane properties of ventrobasal neurons were expected. G A B A A receptors are known to mediate inhibition in ventrobasal thalamus (Zhang et al., 1997; Jia et al., 2005). The EC 5 o for GABA-induced decrease in input resistance found in the present study was similar to EC 5 nS reported in thalamic literature (Gibbs et al., 1996, Jia et al., 2005). Chapter 4. Discussion - 126 -Surprisingly, we calculated a hillslope that was three times greater than the hillslope reported for G A B A A receptors in rat nRt (Gibbs et al., 1996). The source of this difference is uncertain, but may be due to receptor subunit differences between ventrobasal thalamus and nRt (cf. Zhang et al., 1997). Our immunohistochemical investigations of glycine and taurine were inconclusive, despite testing two different batches of antibodies over a wide range of dilutions (Fig. 3.4). These results were likely due to the ubiquitous nature of these amino acids in the brain (Kontro et al., 1980). A possible lack of selectivity for the commercially available antibodies may also have contributed to these results. 4.2. Synaptic inhibition in ventrobasal thalamus A significant finding in this study was the demonstration that medial lemniscal stimulation evoked glycinergic IPSPs and IPSCs in ventrobasal thalamus. The responses, unmasked by blockade of a concomitant glutamatergic excitation, reversed at EQ\ and were sensitive to G l y R blockade by strychnine. The evoked inhibitory responses were dependent on action potentials. Hence, we concluded that the isolated potentials were likely IPSPs and IPSCs requiring an action potential-evoked release of transmitter, and not non-synaptic transmission due to a glial release of P-amino acids (cf. Flint et al., 1998) or gap junctions (cf. Condorelli et al., 2000; Connors and Long, 2004). The pharmacological identification of glycinergic IPSPs in the thalamus relies on specific antagonism at GlyRs by strychnine. Strychnine actions on medial lemniscal IPSPs and IPSCs Chapter 4. Discussion - 127 -displayed a complex pharmacological character. Over a wide range of strychnine concentrations, we initially observed that a double sigmoid curve fitted the concentration-response relationship (Fig. 3.26). This prompted us to suspect a non-specific action of strychnine on G A B A A receptor (cf. Shirasaki et al., 1991). We distinguished selective from unselective actions on IPSCs, by applying various concentrations of strychnine during continuous G A B A A receptor blockade with gabazine. This strategy suppressed the upper sigmoid curve, revealing a relationship with a single IC50 of 33 nM (Fig. 3.28). This value is typical for strychnine antagonism of glycine receptors in other regions of the C N S (Krishtal et al., 1988; Tokutomi et al. 1989; Jonas et al., 1998). Hence, prior blockade of G A B A A receptors uncovered the specificity of strychnine for glycine receptors in low concentrations. The specificity of strychnine at low concentrations was also evident in separate experiments. Strychnine eliminated medial lemniscal inhibition in - 2 0 % of neurons where G A B A A antagonists were ineffective (Fig. 3 . 2 5 Q . Strychnine had negligible effects on an additional ~ 3 0 % of neurons where G A B A A antagonists eliminated the inhibition (Fig. 3 .255) . Other indicators for selectivity included negligible effects of strychnine on slow, G A B A B receptor-mediated inhibition and firing or intrinsic membrane properties (cf. Shirasaki et al., 1991). Strychnine at low concentrations also antagonized the depressant effects of exogenously applied glycine, taurine, and P-alanine, but not G A B A . Hence, low concentrations of strychnine produced antagonism at GlyRs , activated by medial lemniscal stimulation. The picture emerging from these studies was that ventrobasal T C neurons receive a heterogeneous inhibitory input. In a majority of these neurons ( - 7 0 % ) , combined blockade by Chapter 4. Discussion - 128 -strychnine and a G A B A A antagonist eliminated IPSPs and IPSCs. Two minorities exhibited nearly exclusive sensitivity to either antagonist. Occasionally, there was an additional G A B A B component, minimized by clamping neurons near Ea- The diverse nature of inhibitory input to ventrobasal T C neurons is in line with the complex function of higher order relay neurons in processing the somatosensory information (cf. Sherman, 2005). Physiological anatomy of the glycinergic inhibition in the ventrobasal thalamus indicated that medial lemniscal inhibition was likely di- or poly-synaptic. Medial lemniscus-evoked IPSPs displayed considerably higher latency fluctuation compared to IPSPs evoked from other stimulation sites. The difference in latency fluctuation is unlikely due to differences in axonal propagation. During development, medial lemniscus is among the first thalamic inputs to be myelinated (cf. Jacobson, 1963). It is unlikely that the higher latency fluctuation reflected lower degree of myelination in medial lemniscus. We observed similar latency of onset for IPSCs evoked from all stimulation sites with slightly, indicative of similar axonal propagation and myelination. Hence, increased latency fluctuation likely reflected di- or poly-synaptic pathway in medial lemniscal inhibition (cf. Baldissera and Margnelli , 1979). The provenance of medial lemniscal inhibition remains unclear. Polysynaptic transmission would require one or more intermediate excitatory limbs. Our preliminary attempts to uncover the excitatory neurotransmitter of this polysynaptic pathway by pharmacological blockade of several excitatory neurotransmitters were not successful. Fibers in the medial lemniscus may have activated inhibitory interneurons in ventrobasal thalamus (cf. Rainey and Jones, 1983) or projecting interneurons located in nuclei surrounding V B . While local circuit interneurons exist Chapter 4. Discussion - 129 -in the ventrobasal thalamus of cat and monkey, few interneurons exist in rat ventrobasal thalamus (Harris and Hendrickson, 1987, Ohara and Lieberman, 1993). Hence, it is more l ikely that the inhibitory limb of the pathway utilizes cell bodies that lie outside ventrobasal thalamus. Ventrobasal thalamus received inhibition from several surrounding nuclei. These inhibitory responses were likely monosynaptic, and displayed a spectrum of antagonist sensitivity. Inhibition evoked within the caudal zona incerta and nRt displayed the highest relative GABAAerg i c strength. In contrast, stimulations within the ethmoid nucleus produced IPSPs with the highest relative glycinergic strength. Glycinergic cell bodies have been demonstrated in the ethmoid nucleus of mice (Zeilhofer et al., 2005). Several bundles of axons also pass through the ethmoid nucleus, giving it the sieve-like appearance responsible for its name (Paxinos and Watson, 1986). The observed inhibition required stimulus parameters consistent with activation of axons rather than cell bodies (cf. Nowak and Bullier, 1998). While it is probable that the recorded glycinergic IPSPs originate in Eth, we cannot rule out the possibility that the electrical stimulation activated axons passing through this nucleus. A n unambiguous differentiation between the two possibilities would require simultaneous recording from interconnected neurons of Eth and V B nuclei. Stimulus parameters revealed a high threshold for medial lemniscal evoked glycinergic inhibition in ventrobasal thalamus. The higher chronaxie of medial lemniscal evoked IPSPs may relate to a heterogeneous nature of this pathway, or poor myelination of axons at this age (cf. Jacobson, 1963). Since medial lemniscus is myelinated before the other examined pathways projecting to thalamus (Jacobson, 1963), it is more likely that the higher chronaxie of medial Chapter 4. Discussion -130 -lemniscus reflected a greater diversity of fiber types activated by medial lemniscal stimulation. These observations imply relatively sporadic high-threshold inputs, consistent with the paucity of immunoreactivity and low-level m R N A signals for glycine receptors (Araki et al., 1988; Malosio et al., 1991). Our demonstration of low to moderate subunit staining compared to medulla also implicated weak glycinergic innervation. The observation that combined antagonism of G A B A A and glycine receptors eliminated synaptic inhibition is consistent with co-release of glycine-like amino acids and G A B A , or co-transmission by glycinergic and G A B A e r g i c pathways (cf. Jonas et al., 1998; Donato and Nistr i , 2000; Dumoulin et al., 2001). It is unlikely that glycine receptors on nerve terminals simply regulated the release of G A B A (Turecek and Trussell, 2001), because G A B A A receptor blockade should have eliminated all fast IPSPs. Bicuculline or gabazine application eliminated only a minority of the fast inhibitory responses. The present results are compatible with co-transmission by glycinergic and G A B A A e r g i c pathways, rather than co-release of glycine-like amino acids and G A B A . A n appreciable number of neurons showed exclusively glycinergic or G A B A e r g i c responses to medial lemniscal stimulation, consistent with co-transmission by independent pathways. Further evidence obtained from sIPSCs supports this inference. If co-release of glycine-like amino acids and G A B A were to occur (Jonas et al. 1998), we would expect a prevalence of multi-phasic sIPSCs, converting on strychnine application to mono-phasic G A B A A e r g i c currents (cf. Dumoulin et al. 2001). In contrast, the majority of sIPSCs showed a mono-phasic decay, with or without strychnine application. The longer rise times of evoked IPSCs compared to sIPSCs further Chapter 4. Discussion -131-indicated that the evoked IPSCs result from activation of multiple pathways with temporal dispersion. We conclude that, i f present in thalamic inhibition, co-release was a less common occurrence than co-transmission. 4.3. Biophysical characteristics of inhibition in ventrobasal thalamus Another major finding of these studies was glycinergic sIPSCs and glycinergic components of mixed IPSCs, decaying with fast or slow kinetics. Most sIPSCs (>94%) exhibited mono-exponential decays with fast (11 ms) or slow (74 ms) time constants. The fast and slow time constants of sIPSCs, which largely represented mono-quantal packets of transmitter, matched the fast (12 ms) and slow (80 ms) time constants of evoked IPSCs. This finding provided assurance that spontaneous and evoked IPSCs were attributable to the same glycine-receptor populations. The observations were consistent with the activation of two kinetically distinct populations of glycine receptors. The similarity of decay time constants measured from evoked and spontaneous IPSCs provides assurance that the decay time constants accurately reflected IPSC kinetics. There are limitations to interpretation of decay time constants measured from evoked IPSCs, particularly when the IPSCs are of a mixed nature. For example, decay time constant measured from an IPSC can be erroneously long i f there is asynchronous release of multiple quanta. Errors of this kind are unlikely with spontaneous IPSCs (sIPSCs) where the events are predominantly mono-quantal. A n important finding was that fast or slow synaptic currents occurred separately in different neurons. While the higher rise time of evoked IPSCs compared to sIPSCs may raise the Chapter 4. Discussion - 132 -possibility that there was spillover of transmitter to peri-synaptic receptors, the occurrence of slow IPSCs with mono-exponential decay independent of fast IPSCs opposes this hypothesis (cf. Chery and De Koninck, 1999). Slow glycinergic sIPSCs did not display longer rise times than the fast sIPSCs indicating that the longer rise time of evoked IPSCs likely results from asynchronous release of neurotransmitters. We did not observe cross-desensitization between glycinergic and GABAAerg ic currents, further indicating an independence of the two inhibitory systems. Based on these observations on spontaneous and evoked IPSCs, we suggest that the receptor populations are predominantly localized under separate nerve terminals. Our observations of fast and slow mono-quantal sIPSCs contrast with the literature on co-release. In embryonic zebrafish ( A l i et al., 2000), sIPSCs have a bi-exponential decay due to co-localization of receptors with fast and slow kinetics at the same synaptic sites. Spontaneous IPSCs decay mono-exponentially with a fast (4-8 ms) time constant in rat spinal neurons (Chery and De Koninck, 1999; Gonzales-Forero and Alvarez, 2005) and with a slow (~63 ms) time constant in mouse retinal ganglion cells (Tian et al., 1998). Apparently, thalamic neurons in juvenile rats have a predominant ability to segregate 2 populations of glycine receptors with fast and slow kinetics. The fast and slow kinetics of the synaptic currents were likely due to structurally distinct receptor populations. The ai and a 2 receptor subunits determine synaptic decays of fast and slow IPSCs (Takahashi et al., 1992; Singer and Berger, 1999). Given their very long burst duration, activated receptors containing a 2 , but not ai subunits (Mangin et al., 2003), may account for the long decay tails of 2 atypical IPSCs in this study. Co-assembly of ai and a 2 subunits likely Chapter 4. Discussion - 133 -occurs in developing neurons, where slow IPSCs are common (Takahashi et a l , 1992; A l i et al., 2000). Hence the slow kinetics may be attributable to activated ai/012 receptors that persisted in thalamic neurons. Slow glycinergic inhibition contrasts with metabotropic G A B A e r g i c inhibition (Browne et al., 2001), mostly suppressed in our recordings. Another possibility is that post-translational phosphorylation of glycine receptor channels (Agopyan et al. 1993), produced diverse kinetics. The kinetics of extrasynaptic glycine receptor channels resembled the decays of glycinergic currents. The average lifetimes of short- and long-duration bursts activated by glycine, taurine, and p-alanine were close to decay time constants for fast and slow IPSCs. A s with glycinergic IPSCs, the time constant of GABAAerg ic IPSCs approximated the burst duration of extrasynaptic G A B A A channels. The multiple congruencies in kinetics seem unlikely to have occurred by chance, although burst duration may depend on high agonist concentrations (cf. Twyman and Macdonald, 1991; Beato et al., 2002; Lewis et al., 2003). Channel burst durations often determine synaptic decay in both glycinergic and GABAAerg ic systems (cf. Takahashi et al. 1992; Singer and Berger 1999). Hence, the kinetics of synaptic channels likely mirrored the kinetics of extrasynaptic receptor channels. Synaptic channels exhibited higher P Q S than extrasynaptic channels. This observation was l ikely genuine and not due to vagaries in fluctuation analysis. Errors in fluctuation analysis tend to underestimate rather than overestimate the unitary current (cf. Benke et al. 2001). The Pci estimates remained unaltered when measured with an optimized space-clamp with C s + containing patch electrodes. Analysis of evoked and spontaneous IPSCs yielded similar P a s . Chapter 4. Discussion -134 -We established the optimal bin width that ensures fidelity in estimating the elementary current amplitude for glycinergic and G A B A A e r g i c IPSCs. We conclude that the higher P C is estimated for synaptic channels compared to extrasynaptic channels does not result from methodological errors. The difference between synaptic and extrasynaptic Pcis is consistent with the literature. The unitary conductance obtained from sIPSCs was in the same range as in other preparations under similar recording conditions (cf. Poncer et al., 1996; Singer and Berger, 1999). The synaptic G A B A A e r g i c channels had a higher conductance compared to extrasynaptic channels, as found elsewhere (Yeung et al., 2003). The low conductances of extrasynaptic glycine receptors were compatible with embryonic receptor channels (Rajendra et al., 1997) and extrasynaptic receptors on hippocampal neurons (Fatima-Shad and Barry, 1995). Given these considerations, we suggest that the conductance differences were genuine. Extrasynaptic receptors, usually thought as high conductance homomers (Lynch, 2004), in this case may have reduced conductance, reflecting post-translational modification (cf. Caraiscos et al., 2002). It is unlikely that the higher synaptic Pci estimates are due to contamination by other ion species. The reversal potentials for IPSCs, sIPSCs and estimated / Cis were consistently near £ci (Figs. 3.24, 3.55 and 3.59), implicating CI" dependence with little or no contribution from other ion species (cf. Bormann et al. 1987). Pharmacological isolation of CI" currents using N i 2 + , an antagonist of T-type C a 2 + channels, and QX-314 and C s + , antagonists of voltage-dependent N a + and K + currents, did not affect the estimated Pcis. Similar P a s and single channel conductances were obtained when T-type C a 2 + channels were electrophysiologically inactivated using a +20 Chapter 4. Discussion - 135 -m V pre-pulse. However, glycinergic IPSCs evoked using this pre-pulse protocol had higher amplitudes, but similar rise and decay times (Fig. 3.57). The amplitude increase may have been due to a presynaptic effect of the pre-pulse, since the biophysical properties G l y R were not affected. It is not clear how the pre-pulse would have affected presynaptic neurons. One possibility is that thalamic neurons are electrically connected via gap junctions (Condorelli et al., 2000; Connors and Long, 2004). Depolarization of presynaptic neurons via this electrical connection may have increased C a 2 + concentrations in synaptic boutons resulting in an increased neurotransmitter release. Despite the differences in Pci, the striking similarities between IPSC decay and extrasynaptic channel burst duration imply that glycine, taurine, and P-alanine each could mediate inhibition. When applied at the same concentration, glycine, taurine and P-alanine activated channels with comparable open probabilities. The abilities of P-amino acids, relative to glycine, to activate long-duration bursts was greater at extrasynaptic receptors than with most receptor variants (cf. Flint et al. 1998; Martin & Siggins, 2002). The infrequent incidence of long-duration bursts after application of glycine in single channel recording could have occurred by chance, or may reflect a neurotransmitter role for taurine and P-alanine in ventrobasal thalamus. Application of a putative taurine antagonist, A M B D , inhibited the fast and slow glycinergic IPSCs alike. Although taurine may mediate both IPSCs, we cannot exclude the possibility of non-specific blockade of glycine- or p-alanine-activated receptors by A M B D (cf. Mathers, 1993). Hence, the three P-amino acids remain neurotransmitter candidates for thalamic inhibition. Chapter 4. Discussion - 136 -The estimation of P C i involved methodological improvements to the non-stationary fluctuation analysis. The length of the stationary segment determined the accuracy of non-stationary fluctuation analysis. Analysis of simlPSCs showed that the estimated elementary current was more accurate when bin width was shorter than the duration of the stationary segment. The length of the stationary segment revealed itself as the bin width at which the curve describing the relationship between unitary current amplitude and bin width reached a plateau. For estimates of unitary current from evoked and spontaneous GABA A ergic IPSCs, we achieved a plateau at the bin width of ~2 ms. This is in agreement with the finding that channel kinetics determine the accuracy of non-stationary fluctuation analysis (Benke et al., 2001). The mean open time of G A B A A channels in the same neurons was ~9 ms (Kim et al., 2004). Moreover, changes in the decay time constant of simlPSCs, which directly correlate to mean open time, did not affect the estimates. While mean open time may indirectly affect the length of the stationary segment, it is not the only determining factor. Interestingly, evoked and spontaneous GABAAergic IPSCs achieved a plateau at the same bin widths. This consistency raises the possibility that the length of the stationary segment is an inherent characteristic of ionotropic channels. The length of stationary segment, i f reproduced in other systems, may introduce a novel electrophysiological tool for investigations of synaptic receptors. Bin width is a major determinant of accuracy in non-stationary fluctuation analysis. The number of simlPSCs within the range of 10-100 currents per virtual neuron did not affect the accuracy of estimates of unitary current. Under simulated conditions, 10 synaptic currents are sufficient to accurately estimate the unitary current (cf. Traynelis et al., 1993). The accuracy of non-stationary fluctuation analysis was not significantly affected by other examined parameters Chapter 4. Discussion - 137 -including decay time constant, unitary current amplitude, and presence of noise. Consistent with the observations of Benke et al. (2001), estimates of the number of channels were inversely related to the estimates of unitary current. This inverse relationship resulted in an accurately estimated maximum channel open probability (Pomax) regardless of all examined parameters. Hence, optimization of bin width improves the accuracy of unitary current and number of channels estimated by non-stationary fluctuation analysis. Empirical optimization of the bin width provides an improvement in non-stationary fluctuation analysis of synaptic currents. This method as applied to simulated IPSCs is simpler than the heuristic approach used previously for heartbeats (Fukuda et al., 2004). There are limitations to our model, which does not incorporate access resistance and synapse geometry (Benke et al., 2001), or cable filtering (Robinson et al. 1991). In contrast to the simulations, synaptic channel behavior usually involves substate conductances and intricate gating kinetics (cf. Kim et al., 2004). Although GABAAergic and glycinergic IPSCs are kinetically different, analysis of both current types yielded similar curves that plateaued at distinct bin width maxima. However, the absolute accuracy of the estimates of unitary synaptic current is not entirely clear, despite the plateau values observed at small bin widths. The above considerations and adjustments to the model would likely further improve accuracy of unitary currents involved in neuronal signaling. We have not reported the number of postsynaptic receptors (N) and P O m a x from non-stationary noise analysis. Non-stationary noise analysis involved aligning IPSCs in time by peak and scaling the average to individual IPSCs for the calculation of IPSC variance. This manipulation is necessary to avoid presynaptic sources of error in z'ci estimates (cf. Traynelis et al., 1993). Chapter 4. Discussion - 138 -However, the accuracies of N and P o m a x were greatly reduced with this method (Traynelis et al., 1993). Hence, the estimated N and Po m a x were not reliable and omitted from this thesis. 4.4. Functional implications Our findings revealed some aspects of the complex and diverse nature of inhibition in ventrobasal thalamus. Originally, the function of thalamus was viewed as a simple relay of sensory information from periphery to cortex. Findings over the past two decades have dramatically changed this view. Thalamus is now known to provide an integrated relay of information to cortex. For example, thalamus modulates the sensory information according to behavioral demands. Thalamus participates in cortical processing of information via its interactions with cortex and various other C N S areas (Sherman, 2005). We recorded an intricate inhibition in somatosensory thalamus evoked from multiple anatomical sources, mediated by various neurotransmitters and receptors and displaying a variety of biophysical properties. These observations underscore the complex role of thalamus in processing the somatosensory information. The functional implication of multiple synaptic inhibitory modalities in thalamus is currently speculative. The fast synaptic kinetics would allow rapid phasic transfer of information for somatotopic representations, particularly for rapidly adapting skin receptors (cf. Tsumoto and Nakamura 1974). Slow IPSP decays affect hyperpolarization-activated currents, remove C a 2 + channel inactivation, and promote low threshold C a 2 + bursting (cf. Steriade et al. 1997). The glycinergic IPSC components were kinetically distinct from GABAAerg i c IPSCs (~22 ms; cf. Dumoulin et al. 2001). The co-occurrence of fast and slow glycinergic IPSPs with intermediate Chapter 4. Discussion - 139 -G A B A A e r g i c IPSPs would confer fine-tuning of inhibitory transmission by modulation of voltage-dependent currents in somatosensory thalamus. The higher P C i o f synaptic receptors ensures high transmission efficacy. The observed biophysical properties of extrasynaptic glycine receptors are unusual, and appear favorable for tonic detection of background P-amino acids. When applied at the same concentration, glycine and the p-amino acids activated extrasynaptic receptors with equivalent open probabilities, but dissimilar channel permeabilities for CT. These findings contrast with observations in recombinant systems, where glycine is the most potent agonist (Lynch, 2004), and all three agonists activate CI" channels with similar unit conductances (Lewis et al., 2003). In the thalamus, the higher potency and channel permeability observed with the P-amino acids accord with extrasynaptic detection of background concentrations of taurine, due to synaptic spillover (Barbour and Hausser, 1997) or glial release (Flint et al., 1998). For example, a decrease in extracellular osmolality releases taurine from glial cells, which reduces neuronal excitability (Flint et al., 1998; M o r i et al., 2002). B y actions on extrasynaptic receptors, taurine may mediate homeostatic information, in addition to modulating sleep rhythms. The lower CI" permeability may suit extrasynaptic receptors for the detection of ambient P-amino acids, tonic inhibition, and receptor modulation (cf. Berger et al., 1998; Flint et al., 1998; M o r i et al., 2002). Our data have implications for the nature of the neurotransmitters mediating medial lemniscal inhibition in thalamus. Exogenous glycine usually produced channel openings that were consistent with mediation of rapidly decaying IPSCs. In approximately one-half of the patches, taurine and P-alanine activated bursts of brief duration, in accordance with mediation of fast Chapter 4. Discussion - 140 -IPSCs. Hence, glycine and the P-amino acids may contribute to receptor activations producing fast inhibition. The kinetic congruence between the long-duration bursts activated by taurine and P-alanine, and the slow IPSCs implicates the p-amino acids as mediators of slow synaptic inhibition. This inhibition may represent a transient phase in post-natal development, or a feature destined for adulthood. Although there is little information about P-alanine release, taurine increases in the extracellular space of cat thalamus during slow-wave sleep (Kekesi et al., 1997). During this sleep stage, discharge patterns based on low threshold Ca 2 + spiking promote network rhythmicity at low frequency (Steriade et al. 1997). The prolonged IPSPs due to P-amino acids would remove Ca 2 + inactivation, promoting low threshold spikes. Given these considerations, p-amino acids may have a role in the modulation of the sleep-wake cycle. 4.5. Significance Our findings have revealed for the first time, a co-existence of fast and slow ionotropic inhibition mediated by glycine receptors in the thalamus. The proposed roles for p-amino acids in synaptic and extrasynaptic signaling in the diencephalon departs from the traditional hegemony of transmission mediated by GABA A and G A B A B receptors. Glycinergic inhibition in the thalamus may account for some physiological and clinical observations. For example, strychnine has an ability to induce network oscillatory activity in the thalamus in vitro (Ran et al., 2004) and reduces the threshold for sensory evoked convulsions in vivo (Sherrington, 1947). Hence, an imbalance in glycinergic and GABAergic transmission in the thalamus may be involved in the pathophysiology of epileptic disorders. These findings open frontiers in the research and development of anti-epileptic drugs. Since thalamotomy is a penultimate surgical procedure for Chapter 4. Discussion - 141 -clinically intractable pain disorders (reviewed by Ohye, 1998), the findings may also have significance for the development of analgesic drugs. 4.6. Conclusions The exceptional finding in these studies was the demonstration of glycinergic inhibition in ventrobasal thalamic nuclei. We observed glycine receptor a-subunits in confocal images, and inhibitory responses to exogenous glycine agonists in thalamocortical neurons that mimicked the inhibition. Strychnine selectively antagonized synaptic inhibition in ventrobasal thalamic neurons evoked by electrical stimulation of the medial lemniscus. Medial lemniscal inhibition, apparently activated through non-glutamatergic pathways, frequently involved G A B A A receptors. Glycine-like amino acids and G A B A acted jointly to mediate this inhibition. During medial lemniscal stimulation, the activation of receptors sensitive to strychnine antagonism inhibited neurons in ventrobasal thalamic nuclei. The finding that this synaptic inhibition involved glycine receptors was significant, in view of the dominance of G A B A A receptors in the literature on the forebrain. Interestingly, we found that co-transmission by glycine and G A B A A receptors mediated the synaptic inhibition. In other neurons, glycine or G A B A A receptors exclusively mediated the inhibition. These observations may reflect the heterogeneity of pathways activated by medial lemniscal stimulation. Stimulation within nuclei surrounding ventrobasal thalamus revealed a spectrum of GABA A ergic, mixed or glycinergic inputs. The strongest glycinergic input was evoked within the ethmoid nucleus. We conclude that glycine-like amino acids, as well as G A B A , participate in synaptic inhibition of thalamocortical neurons. Chapter 4. Discussion - 142-These studies have revealed some unusual facets of inhibitory transmission in thalamus. Spontaneous IPSCs mediated by glycine- and GABA A - receptors were mono-phasic, showing fast, intermediate, and slow decays. IPSCs evoked by medial lemniscal stimulation also showed fast, intermediate, or slow decays, alone or in combination. Currents with intermediate decays were due to co-transmission by a GABAA-receptor mediated pathway. Strychnine antagonized the fast and slow synaptic currents, mediated by glycine-like amino acids. Extrasynaptic receptors, activated by glycine agonists in membrane patches, displayed the predicted short- or long-duration burst openings. A significant finding was the dual kinetics of synaptic and single channel currents, implicating functional diversity in glycine receptors at a juvenile stage of rat development. Thalamocortical neurons segregate ionotropic glycine receptors showing fast and slow decay kinetics. Co-transmission by G A B A A receptors showing intermediate kinetics, and known metabotropic G A B A B receptors, may facilitate postsynaptic discrimination of inputs in neurons of somatosensory nuclei. The different CI" permeability of synaptic and extrasynaptic receptors may suit their specific physiological functions. The novel finding that sensory information is co-transmitted by glycinergic and G A B A e r g i c pathways indicates that thalamic inhibition is a complex, integrative system for processing of somatosensory information. Given the known effect of strychnine in reducing the threshold for sensory-evoked convulsions in vivo, an imbalance in glycinergic and G A B A e r g i c transmission in the thalamus may be involved in the pathophysiology of epileptic disorders. - 143 -Abbreviations aCSF Artificial cerebrospinal fluid A M B D 6-aminomethyl-3-methyl, 1-4H-1,2,6-benzothiadiazine-1,1 -diazide hydrochloride A M P A a-amino-3-hydroxy-5methyl-4-isoxazoleproprionate A N O V A Analysis of variance A P V 2-amino-5-phosphono-valerate A T P Adenosine-5' -triphosphate G T P guanosine-5' -triphosphate c A M P Cycl ic 3' 5' -adenosine-monophosphate c G M P Cycl ic 3' 5' -guanosine-monophosphate C S F Cerebrospinal fluid C N Q X 6-cyano-7-nitroquinoxaline C N S Central nervous system D C Direct current EC50 Concentration of drug that produces a half-maximal effect Ea Equilibrium potential for CI" E E G Electroencephalogram EK Equilibrium potential for K+ Eth Ethmoid nucleus E G T A Ethylene glycol-bis-(P-aminoethyl ether) N,N,N'N'-tetraacetic acid EPSP Excitatory postsynaptic potential E P S C Excitatory postsynaptic current G A B A y-aminobutyric acid Abbreviations - 144 -G A D Glutamate decarboxylase G l y R Glycine receptor G l y T Glycine transporter h hour H & E Haematoxylin and Eosin H E P E S N-[2-hydroxyethyl]piperazine-N'-[2-ethanesulfonic acid] Hipp Hippocampus H z Hertz (s 1 ) ic internal capsule IC50 Concentration of a drug that produces a half-maximal inhibition / A H P Afterhyperpolarization current IT LOW threshold C a current IPSP Inhibitory postsynaptic potential IPSC Inhibitory postsynaptic current L T S L o w threshold C a 2 + spike min Minute ml Medial lemniscus N . A . Numerical aperture N M D A N-mefhyl-D-aspartate nRt Nucleus reticularis thalami p H Hydrogen concentration; pH-log[H + ] Pci CI" channel permeability Pfc Parafasicular nucleus Abbreviations - 145 -Po Posterior/Pulvinar thalamus. Pomax Maximum open probability Rj Input resistance S E M Standard error of the mean x Time constant; time required to reach (1-1/e) of a steady state value TauT Taurine transporter T A G 6-aminomethyl-3-methyl, 1-4H-1,2,6-benzothiadiazine-1,1 -diazide hydrochloride T C Thalamocortical T T X Tetrodotoxin Vh Holding membrane potential V m Resting membrane potential V B Ventrobasal nuclei of thalamus V P L Ventral posterior lateral thalamic nucleus V P M Ventral posterior medial thalamic nucleus - 146-Bibliography Agopyan N , Tokutomi N , Akaike N . Protein kinase A-mediated phosphorylation reduces only the fast desensitizing glycine current in acutely dissociated ventromedial hypothalamic neurons. Neuroscience 1993, 56: 605-615. Akaike N , Kaneda M . Glycine-gated chloride current in acutely isolated rat hypothalamic neurons. JNeurophysiol 1989, 62: 1400-1409. A l i D W , Drapeau P, Legendre P. Development of spontaneous glycinergic currents in the Mauthner neuron of the zebrafish embryo. J Neurophysiol 2000, 84: 1726-1736. Ali t to H J , Usrey W M . Corticothalamic feedback and sensory processing. Curr Opin Neurobiol 2003, 13: 440-445. Alvarez O, Gonzalez C , Latorre R. Counting channels: a tutorial guide on ion channel fluctuation analysis. Adv Physiol Educ 2002, 26: 327-341. Arak i T, Yamano M , Murakami T, Wanaka A , Betz H , Tohyama M . Localization of glycine receptors in the rat central nervous system: an immunohistochemical analysis using monoclonal antibody. Neurosci 1988, 25: 613-624. Asano T, U i M , Ogasawara N . Prevention of the agonist binding to y-aminobutyric acid B receptors by guanine nucleotides and islet-activating protein, pertussis toxin, in bovine cerebral cortex. Possible coupling of the toxin-sensitive GTP-binding proteins to receptors. J Biol Chem 1985, 260: 12653-12658. Aubrey K R , Vandenberg RJ . N[3-(4'-fluorophenyl)-3-(4'-phenylphenoxy)propyl]sarcosine (NFPS) is a selective persistent inhibitor of glycine transport. Br J Pharmacol 2001, 134: 1429-1436. Baldissera F , Margnelli M . Disynaptic IPSPs of lemniscal origin in relay-cells of thalamic ventrobasal nuclear complex of the cat. Brain Res 1979, 163: 171-175. Barakat L , Wang D , Bordey A . Carrier-mediated uptake and release of taurine from Bergmann glia in rat cerebellar slices. J Physiol 2002, 541: 753-767. Barbaresi P, Manzoni T. Direct synaptic connections between corticothalamic axon terminals and interneurons in the cat ventrobasal complex. Neurosci Lett 2003, 339: 211-4. Barbour B , Hausser M . Intersynaptic diffusion of neurotransmitter. Trends in Neurosci 1997, 20: 377-384. Bartho P, Freund T F , Acsady L . Selective G A B A e r g i c innervation of thalamic nuclei from zona incerta. J Neurosci 2002, 16: 999-1005. Bibliography - 147-Beato M , Groot-Kormelink PJ, Colquhoun D , Sivilotti L G . Openings of the rat recombinant a l homomeric glycine receptor as a function of the number of agonist molecules bound. J Gen Physiol 2002, 119: 443-466. Becker C M , Hoch W , Betz H . Glycine receptor heterogeneity in rat spinal cord during postnatal development. EMBO J1988, 7: 3717-3726. Benke T A , Luthi A , Palmer M J , Wikstrom M A , Anderson W W , Isaac JT, Collingridge G L . Mathematical modelling of non-stationary fluctuation analysis for studying channel properties of synaptic A M P A receptors. J Physiol 2001, 537: 407-420. Berger A J , Dieudonne S, Ascher P. Glycine uptake governs glycine site occupancy at N M D A receptors of excitatory synapses. JNeurophysiol 1998, 80: 3336-3340. Bettler B , Kaupmann K , Mosbacher J, Gassmann M . Molecular structure and physiological functions of G A B A B receptors. Physiol Rev 2004, 84: 835-867. Blomqvist A , Ericson A C , Craig A D , Broman J. Evidence for glutamate as a neurotransmitter in spinothalamic tract terminals in the posterior region of owl monkeys. Exp Brain Res 1996, 108: 33-44. Bormann J, Hamil l OP, Sakmann B . Mechanism of anion permeation through channels gated by glycine and y-aminobutyric acid in mouse cultured spinal neurons. J Physiol (Lond) 1987, 385: 243-286. Bormann J, Rundstrom N , Betz H , Langosch D. Residues within transmembrane segment M 2 determine chloride conductance of glycine receptor homo- and hetero-oligomers. EMBO J 1993, 12:3729-3737. Bowery N G , H i l l D R , Hudson A L , Doble A , Middlemiss D N , Shaw J, Turnbull M . (-)Baclofen decreases neurotransmitter release in the mammalian C N S by an action at a novel G A B A receptor. Nature 1980, 283: 92-94. Brauner-Osborne H , Krogsgaard-Larsen P. Functional pharmacology of cloned heterodimeric G A B A B receptors expressed in mammalian cells. Br J Pharmacol 1999, 128: 1370-1374. Burton H , Jones, E G . The posterior thalamic region and its cortical projection in New Wor ld and Old Wor ld monkeys. JComp Neurol 1976, 168: 249-301. Campbell V , Berrow N , Dolphin A C . G A B A B receptor modulation of C a 2 + currents in rat sensory neurones by the G protein G 0 : antisense oligonucleotide studies. J Physiol 1993, 470: 1-11. Caraiscos V B , M i h i c SJ, MacDonald JF, Orser B A . Tyrosine kinases enhance the function of glycine receptors in rat hippocampal neurons and human a i P receptors. J Physiol (Lond) 2002, 539: 495-502. Bibliography - 148 -Castro-Alamancos M A , Connors B W . Thalamocortical synapses. Prog Neurobiol 1997, 51: 581-606. Chattipakorn SC, McMahon L L . Pharmacological characterization of glycine-gated chloride currents recorded in rat hippocampal slices. J Neurophysiol 2002, 87: 1515-1525. Chebib M . G A B A c receptor ion channels. Clin Exp Pharmacol Physiol 2004, 31: 800-804. Chebib M , Johnston G A . Stimulation of [ 3 H ] G A B A and [ 3H] P-alanine release from rat brain slices by cj's-4-aminocrotonic acid. JNeurochem 1997, 68: 786-794. Chebib M , Johnston G A R . G A B A activated ion channel: Medicinal chemistry and molecular biology. JMed Chem 2000, 43: 1427-1447. Chery N , De Koninck Y . Junctional versus extrajunctional glycine and G A B A A receptor-mediated IPSCs in identified lamina I neurons of the adult rat spinal cord. J Neurosci 1999, 19: 7342-7355. Colquhoun D , Sakmann B . Fast events in single-channel currents activated by acetylcholine and its analogues at the frog muscle end-plate. J Physiol 1985, 369: 501-557. Condorelli D F , Belluardo N , Trovato-Salinaro A , Mudo G . Expression of Cx36 in mammalian neurons. Brain Res Brain Res Rev 2000, 32: 72-85. Connors B W , Long M A . Electrical synapses in the mammalian brain. Annu Rev Neurosci 2004, 27: 393-418. Cox C L , Huguenard JR, Prince D A . Nucleus reticularis neurons mediate diverse inhibitory effects in thalamus. Proc Natl Acad Sci USA 1997, 94: 8854-8859. Crabtree JW. Organization in the somatosensory sector of the cat's thalamic reticular nucleus. J Comp Neurol 1996, 366: 207-222. Cubelos B , Gimenez C, Zafra F. Localization of the G L Y T i glycine transporter at glutamatergic synapses in the rat brain. Cereb Cortex 2005, 15: 448-459. Curtis D R , Duggan A W , Felix D , Johnston G A . G A B A , bicuculline and central inhibition. Nature 1970, 226: 1222-1224. Curtis D R , Hosl i L , Johnston G A R , Johnston IH. The hyperpolarization of spinal motoneurones by glycine and related amino acids. Exp Brain Res 1968, 5: 235-258. Curtis D R , Leah JD, Peet M J . Lack of specificity of a "taurine antagonist". Brain Res 1982, 244: 198-199. Bibliography - 149-Darstein M , Landwehrmeyer G B , K l i n g C , Becker C M , Feuerstein TJ . Strychnine-sensitive glycine receptors in rat caudatoputamen are expressed by cholinergic interneurons. Neuroscience 2000, 96: 33-39. David-Watine B , Goblet C , de Saint Jan D , Fucile S, Devignot V , Bregestovski P, K o r n H . Cloning, expression and electrophysiological characterization of glycine receptor a subunit from zebrafish. Neuroscience 1999, 90: 303-317. Debarbieux F , Brunton J, Charpak S. Effect of bicuculline on thalamic activity: A direct blockade of /AHP in reticularis neurons. JNeurophysiol 1998, 79: 2911-2918. De Koninck Y , Mody I. Noise analysis of miniature IPSCs in adult rat brain slices: properties and modulation of synaptic G A B A A receptor channels. J Neurophysiol 1994, 71: 1318-1335. De Saint Jan D , David-Watine B , Korn H , Bregestovski P. Activation of human a i and a 2 homomeric glycine receptors by taurine and G A B A . J Physiol 2001, 535: 741-755. Diamond JS, Jahr C E . Asynchronous release of synaptic vesicles determines the time course of the A M P A receptor-mediated E P S C . Neuron 1995, 15: 1097-1107. D o i , A , Kishimoto, K , Ishibashi, H . Modulation of glycine-induced currents by zinc and other metal cations in neurons acutely dissociated from the dorsal motor nucleus of the vagus of the rat. Brain Res 1999, 816: 424-430. Donato R, Nistri A . Relative contribution by G A B A or glycine to Cl'-mediated synaptic transmission on rat hypoglossal motoneurons in vitro. J Neurophysiol 2000, 84: 2715-2724. Dong C J , Werblin FS. Use-dependent and use-independent blocking actions of picrotoxin and zinc at the G A B A c receptor in retinal horizontal cells. Vision Res 1996, 36: 3997-4005. Dudeck O, Lubben S, Eipper S, Knorle R, Kirsch M , Honegger J, Zentner J, Feuerstein TJ . Evidence for strychnine-sensitive glycine receptors in human amygdala. Naunyn Schmiedebergs Arch Pharmacol2003,368: 181-187. Dumoulin A , Triller A , Dieudonne S. IPSC kinetics at identified G A B A e r g i c and mixed G A B A e r g i c and glycinergic synapses onto cerebellar Golgi cells. J Neurosci 2001, 21: 6045-6057. El-Beheiry H , Pui l E . Unusual features of G A B A response in layers I V - V neurons of neocortex. Nerurosci Lett 1990, 119- 83-85. Eulenburg V , Armsen W , Betz H , Gomeza J. Glycine transporters: essential regulators of neurotransmission. Trends Biochem Sci 2005, 30: 325-333. Farroni JS, M c C o o l B A . Extrinsic factors regulate partial agonist efficacy of strychnine-sensitive glycine receptors. BMC Pharmacol 2004, 4: 16. Bibliography - 150-Fatima-Shad K , Barry P H . Heterogenous current responses to G A B A and glycine are present in post-natally cultured hippocampal neurons. Brain Res 1995, 704: 246-255. Feldman SG, and Kruger L . A n axonal transport study of the ascending projection of medial lemniscal neurons in the rat. J Comp Neurol 1980, 192: 427-454. Flint A C , L i u X , Kriegstein A R . Nonsynaptic glycine receptor activation during early neocortical development. Neuron 1998, 20: 43-53. Frostholm A , Rotter A Glycine receptor distribution in mouse C N S : autoradiographic localization of [ 3H]strychnine binding sites. Brain Res Bull 1985, 15: 473-486. Fucile S, de Saint Jan D , David-Watine B , Korn H , Bregestovski P. Comparison of glycine and G A B A actions on the zebrafish homomeric glycine receptor. J Physiol 1999, 517: 369-383. Fukuda K , Eugene Stanley H , Nunes Amaral L A . Heuristic segmentation of a nonstationary time series. Phys Rev E Stat Nonlin Soft Matter Phys 2004, 69: 021108. Gentet L J , Ul r ich D . Strong, reliable and precise synaptic connections between thalamic relay cells and neurons of the nucleus reticularis in juvenile rats. J Physiol 2003, 546: 801-811. Gibbs J W 3 r d , Schroder G B , Coulter D A . G A B A A receptor function in developing rat thalamic reticular neurons: whole cell recordings of GABA-mediated currents and modulation by clonazepam. J Neurophysiol 1996, 76: 2568-79. Girard Y , Atkinson JG, Haubrich D R , Williams M , Yarbrough G G . Aminomethyl-1,2,4-benzothiadiazines as potential analogues of y-aminobutyric acid. Unexpected discovery of a taurine antagonist. JMed Chem 1982, 25: 113-116. Gonzales-Forero D , Alvarez FJ . Differential postnatal maturation of G A B A A , glycine receptor, and mixed synaptic currents in Renshaw cells and ventral spinal interneurons. J Neurosci 2005, 25: 2010-2023. Gre i f G J , Sodickson D L , Bean B P , Neer EJ , Mende U . Altered regulation of potassium and calcium channels by G A B A n and adenosine receptors in hippocampal neurons from mice lacking G a G . J Neurophysiol 2000, 83: 1010-1018. Grenningloh G , Pribilla I, Prior P, Multhaup G , Beyreuther K , Taleb O, Betz H . Cloning and expression of the 58 kd (3 subunit of the inhibitory glycine receptor. Neuron 1990, 4: 963-970. Grenningloh G , Rienitz A , Schmitt B , Methfessel C , Zensen M , Beyreuther K , Gundelfinger E D , Betz H . The strychnine-binding subunit of the glycine receptor shows homology with nicotinic acetylcholine receptors. Nature 1987, 328: 215-220. Bibliography - 151 -Grenningloh G , Schmieden V , Schofield PR, Seeburg P H , Siddique T, Mohandas T K , Becker C M , Betz H . a subunit variants of the human glycine receptor: primary structures, functional expression and chromosomal localization of the corresponding genes. EMBO J1990, 9: 771-776. G u Y , Huang L Y . Cross-modulation of glycine-activated CI" channels by protein kinase C and cAMP-dependent protein kinase in the rat. J Physiol 1998, 506: 331-339. Hackam A S , Wang T L , Guggino W B , Cutting GR. The N-terminal domain of human G A B A receptor rhol subunits contains signals for homooligomeric and heterooligomeric interaction. J Biol Chem 19971, 272: 13 750-757. Hami l l OP, Bormann J, Sakmann B . Activation of multiple-conductance state chloride channels in spinal neurones by glycine and G A B A . Nature 1983, 305: 805-808. Handford C A , Lynch JW, Baker E , Webb G C , Ford J H , Sutherland G R , Schofield PR. The human glycine receptor P subunit: primary structure, functional characterisation and chromosomal localisation of the human and murine genes. Mol Brain Res 1996, 35: 211-219. Harris R M , Hendrickson A E . Local circuit neurons in the rat ventrobasal thalamus - a G A B A immunocytochemical study. Neuroscience 1987, 21: 229-236. Harvey RJ , Schmieden V , V o n Hoist A , Laube B , Rohrer H , Betz H . Glycine receptors containing the (u subunit in the embryonic sympathetic nervous system, spinal cord and male genital ridge. EurJNeurosci 2000, 12: 994-1001. Herkenham M . Laminar organization of thalamic projections to the rat neocortex. Science 1980, 207:532-535 Herrero M T , Barcia C , Navarro J M . Functional anatomy of thalamus and basal ganglia. Childs Nerv Syst 2002, 18: 386-404. Hevers W , Luddens H . The diversity of G A B A A receptors. Pharmacological and electrophysiological properties of G A B A A channel subtypes. Mol Neurobiol 1998, 18: 35-86. H i l l D R , Bowery N G , Hudson A L . Inhibition of G A B A B receptor binding by guanyl nucleotides. JNeurochem 1984, 42: 652-657. Hirai T, Schwark H D , Y e n C T , Honda C N , Jones E G . Morphology of physiologically characterized medial lemniscal axons terminating in cat ventral posterior thalamic nucleus. J Neurophysiol 1988, 60: 1439-1459. Horikoshi T, Asanuma A , Yanagisawa K , Anzai K , Goto S. Taurine and P-alanine act on both G A B A and glycine receptors in Xenopus oocyte injected with mouse brain messenger R N A . Brain Res 1988, 464: 97-105 Bibliography - 152 -Hosie A M , Akagi H , Ishida M , Shinozaki H . Actions of 3-[2'-phosphonomethyl [ l , l ' -biphenyl]-3-yl] alanine ( P M B A ) on cloned glycine receptors. Br J Pharmacol 1999, 126: 1230-1236. Ilinsky IA, Ambardekar A V , Kultas-Ilinsky K . Organization of projections from the anterior pole of the nucleus reticularis thalami (NRT) to subdivisions of the motor thalamus: light and electron microscopic studies in the rhesus monkey. J Comp Neurol 1999, 409: 369-384. Imboden M , De Saint Jan D , Leulier F , Korn H , Goblet C , Bregestovski P. Isolation and characterization of an oi2-type zebrafish glycine receptor subunit. Neuroscience 2001, 103: 799-810. Jacobson S. Sequence of myelinization in the brain of the albino rat. A . Cerebral cortex, thalamus and related structures. J Comp Neurol 1963, 121: 5-29. Jentsch TJ , Stein V , Weinreich F , Zdebik A A . Molecular structure and physiological function of chloride channels. Physiol Rev 2002, 82: 503-68. Jia F , Pignataro L , Schofield C M , Yue M , Harrison N L , Goldstein P A . A n extrasynaptic G A B A A receptor mediates tonic inhibition in thalamic V B neurons. J Neurophysiol 2005, 94: 4491-501. Johnston G A . G A B A A receptor pharmacology. Pharmacol Ther 1996, 69: 173-198. Jonas P, Bischofberger J, Sandkuhler J. Corelease of two fast neurotransmitters at a central synapse. Science 1998, 281: 419-424. Jones E G . The anatomy of sensory relay functions in the thalamus. Prog Brain Res 1991, 87: 29-52. Jursky F , Nelson N . Localization of glycine neurotransmitter transporter ( G L Y T 2 ) reveals correlation with the distribution of glycine receptor. JNeurochem 1995, 64: 1026-1033. Kandel E R , Schwarz J H , Jessel T M . (Eds) Principles of Neural Science, 4th edn, 1991, M c G r a w - H i l l Inc., U S A . Kekesi K A , Dobolyi A , Salfay O, Nyitrai G , Juhasz G . Slow wave sleep is accompanied by release of certain amino acids in the thalamus of cats. Neuroreport 1997, 8: 1183-1186. K h o m S, Baburin I, Timin E N , Hohaus A , Sieghart W , Hering S. Pharmacological properties of G A B A A - receptors containing yi subunits. Mol Pharmacol 2005, [Epub ahead of print]. K i m H S , Wan X , Mathers D A , Pui l E . Selective GABA-receptor actions of amobarbital on thalamic neurons. Brit J Pharmacol 2004, 143: 485-494. Kirsch J, Betz H . The postsynaptic localization of the glycine receptor-associated protein gephyrin is regulated by the cytoskeleton. J Neurosci 1995, 15: 4148^1156. Bibliography -153-Kontro P, Mamela K M , Oja SS. Free amino acids in the synaptosome and synaptic vesicle fractions of different bovine brain areas. Brain Res 1980, 184: 129-141. Krauthamer G , Mcguinness C , Gottesman L . Unit responses in the ventrobasal thalamus ( V P L ) of the cat of bradykinin injected into somatic and visceral arteries. Brain Res Bull 1977, 2: 299-306. Krishtal O A , Osipchuk Y u V , Vrublevsky S V . Properties of glycine-activated conductances in rat brain neurones. Neurosci Lett 1988, 84: 271-276. Kuhse J, Kuryatov A , Maulet Y , Malosio M L , Schmieden V , Betz H . Alternative splicing generates two isoforms of the 0C2 subunit of the inhibitory glycine receptor. FEBS Lett 1991, 283: 73-77. Kuhse J, Schmieden V , and Betz H . A single amino acid exchange alters the pharmacology of neonatal rat glycine receptor subunit. Neuron 1990, 5: 867-873. Kuner T, Augustine GJ . A genetically encoded ratiometric indicator for chloride: capturing chloride transients in cultured hippocampal neurons. Neuron 2000, 27: 447-459. Langosch D , Laube B , Rundstrom N , Schmieden V , Bormann J, Betz H . Decreased agonist affinity and chloride conductance of mutant glycine receptors associated with human hereditary hyperekplexia. EMBO J1994, 13: 4223-4228. Legendre P. The glycinergic inhibitory synapse. Cell Mol Life Sci 2001, 58: 760-793. Lewis T M , Schofield PR, McCle l lan A M . Kinetic determinants of agonist action at the recombinant human glycine receptor. J Physiol (Lond) 2003, 549: 361-374. Lewis T M , Sivilotti L G , Colquhoun D , Gardiner R M , Schoepfer R, Rees M . Properties of human glycine receptors containing the hyperekplexia mutation cci(K276E), expressed in Xenopus oocytes. J Physiol 1998, 507: 25-40. L i Y P , Lombardini JB. Guanidinoethanesulfonic acid-inhibitor of G A B A uptake in rat cortical synaptosomes. Brain Res 1990, 510: 147-149. L i Y , W u L J , Legendre P, X u T L . Asymmetric cross-inhibition between G A B A A and glycine receptors in rat spinal dorsal horn neurons. J Biol Chem 2003, 278: 38637-38645. Lund R D , Webster K E . Thalamic afferents from the dorsal column nuclei. A n experimental anatomical study in the rat. J Comp Neurol 1961 di, 130: 301-312. Lund R D , Webster K E . Thalamic afferents from the spinal cord and trigeminal nuclei. A n experimental anatomical study in the rat. J Comp Neurol 1967b, 130: 313-328. Bibliography - 154-Luque J M , Erat R, Kettler R, Cesura A , Da Prada M , Richards J G . Radioautographic evidence that the G A B A A receptor antagonist S R 95531 is a substrate inhibitor of M A O - A in the rat and human locus coeruleus. Eur JNeurosci 1994, 6: 1038-1049. Liischer C , Jan L Y , Stoffel M , Malenka R C , Nico l l R A . G protein-coupled inwardly rectifying K + channels (GIRKs) mediate postsynaptic but not presynaptic transmitter actions in hippocampal neurons. Neuron 1997, 19: 687-695. Lynch J W . Molecular structure and function of the glycine receptor chloride channel. Physiol /tev2004, 84: 1051-1095. Lynch J W , Rajendra S, Barry P H , Schofield PR . Mutations affecting the glycine receptor agonist transduction mechanism convert the competitive antagonist, picrotoxin, into an allosteric potentiator. J Biol Chem 1995,270: 13799-13806. Lynch J W , Rajendra S, Pierce K D , Handford C A , Barry P H , Schofield P R . Identification of intracellular and extracellular domains mediating signal transduction in the inhibitory glycine receptor chloride channel. EMBO J1997, 16: 110-120. M a W , Peschanski M , Besson J M . The overlap of spinothalamic and dorsal column nuclei projections in the ventrobasal complex of the rat thalamus: a double anterograde labeling study using light microscopy analysis. J Comp Neurol 1986, 245: 531-540. Malosio M L , Grenningloh G , Kuhse J, Schmieden V , Schmitt B , Prior P, Betz H . Alternative splicing generates two variants of the oti subunit of the inhibitory glycine receptor. J Biol Chem 1991a, 266: 2048-2053. Malosio M L , Marqueze-Pouey B , Kuhse J, Betz H . Widespread expression of glycine receptor subunit m R N A s in the adult and developing rat brain. EMBOJ 1991b, 10: 2401-2409. Malouf A T , Robbins C A , Schwartzkroin P A . Phaclofen inhibition of the slow inhibitory postsynaptic potential in hippocampal slice cultures: a possible role for the GABAe-mediated inhibitory postsynaptic potential. Neurosci 1990, 35: 53-61. Mangin J M , Baloul M , De Carvahlo P, Rogister B , Rigo J M , Legendre P. Kinetic properties of the a.2 homo-oligomeric glycine receptor impairs a proper synaptic functioning. J Physiol (Lond) 2003, 553: 369-386. Marvizon J C , Vazquez J, Garcia Calvo M , Mayor F Jr, Ruiz Gomez A , Valdivieso F , Benavides J. The glycine receptor: pharmacological studies and mathematical modeling of the allosteric interaction between the glycine- and strychnine-binding sites. Mol Pharmacol 1986, 30: 590-597. Mascia M P , Wick M J , Martinez L D , Harris R A . Enhancement of glycine receptor function by ethanol: role of phosphorylation. Br J Pharmacol 1998, 125: 263-270. Bibliography - 155 -Mason A , N i c o l l A , Stratford K . Synaptic transmission between individual pyramidal neurons of the rat visual cortex in vitro. J Neurosci 1991, 11: 72-84. Mathers D A . Effects of the benzothiadiazine T A G on channel activation at mammalian glycine receptors. Neurosci Lett 1993, 149: 23-26. Matthews G , Ayoub G S , Heidelberger R. Presynaptic inhibition by G A B A is mediated via two distinct G A B A receptors with novel pharmacology. J Neurosci 1994, 14: 1079-1090. Matzenbach B , Maulet Y , Sefton L , Courtier B , Avner P, Guenet J L , Betz H . Structural analysis of mouse glycine receptor subunit genes. Identification and chromosomal localization of a novel variant. J Biol Chem 1994, 269: 2 6 0 7 - 2 6 1 2 . McCormick D A , Ba i T. Sleep and arousal: thalamocortical mechanisms. Annu Rev Neurosci 1997, 20 : 185-215. Meyer G , Kirsch J, Betz H , Langosch D . Identification of a gephyrin binding motif on the glycine receptor (3 subunit. Neuron 1995, 15: 563-572. Michaud JC, Mienvil le J M , Chambon JP, Biziere K . Interactions between three pyridazinyl-G A B A derivatives and central G A B A and glycine receptors in the rat, an in vivo microiontophoretic study. Neuropharmacology 1986 25 : 1197-1203. Mishima K . Facilitatory and inhibitory processes in the thalamic ventrobasal nucleus of the rat. Jpn J Physiol 1992, 42 : 193-210. Mitrofanis J. Some certainty for the "zone of uncertainty"? Exploring the function of the zona incerta. Neuroscience 2005 130: 1-15. Moorhouse A J , Jacques P, Barry P H , Schofield PR. The startle disease mutation Q 2 6 6 H , in the second transmembrane domain of the human glycine receptor, impairs channel gating. Mol Pharmacol 1999, 55: 386-395. M o r i M , Gahwiler B H , Gerber U . p-alanine and taurine as endogenous agonists at glycine receptors in rat hippocampus in vitro. J Physiol 2002 , 539: 191-200. Morishita R, Kato K , Asano T. G A B A B receptors couple to G proteins G 0 , G 0 * and Gn but not to G i 2 . FEBSLett 1990, 2 7 1 : 2 3 1 - 2 3 5 . Mott D D , Lewis D V . The pharmacology and function of central G A B A B receptors. Int Rev Neurobiol 1994, 36: 9 7 - 2 2 3 . Nelson N . The family of N a + / C l " neurotransmitter transporters. J Neurochem 1998, 71 : 1785-1803. Bibliography - 156 -Nikol ic Z , Laube B , Weber R G , Lichter P, Kioschis P, Poustka A , Mulhardt C , Becker C M . The human glycine receptor subunit CC3. Glra3 gene structure, chromosomal localization, and functional characterization of alternative transcripts. J Biol Chem 1998, 273: 19708-19714. Nishizaki T, Ikeuchi Y . Activation of endogenous protein kinase C enhances currents through oti and 0C2 glycine receptor channels. Brain Res 1995, 687: 214-216. Nowak L G , Bullier J. Axons, but not cell bodies, are activated by electrical stimulation in cortical gray matter. I. Evidence from chronaxie measurements. Experimental Brain Research 1998, 118:477-488. Nunez E , Lopez-Corcueral B , Vazquezl J, Gimenez C, Aragon C. Differential effects of the tricyclic antidepressant amoxapine on glycine uptake mediated by the recombinant G l y T i and GlyT2 glycine transporters. Br J Pharmacol 2000, 129: 200-206. Ohara PT, Lieberman A R . Some aspects of the synaptic circuitry underlying inhibition in the ventrobasal thalamus. JNeurocytol 1993, 22: 815-825. Ohye C . Stereotactic treatment of central pain. Stereotact Fund Neurosurg 1998, 70: 71-6. Okamoto K , Kimura H , Sakai Y . Evidence for taurine as an inhibitory neurotransmitter in cerebellar stellate interneurons: selective antagonism by T A G (6-aminomethyl-3-methyl-4H,l,2,4-benzothiadiazine-1,1-dioxide). Brain Res 1983, 265: 163-168. Olpe H R , Karlsson G , Pozza M F , Brugger F, Steinmann M , Van Riezen H , Fagg G , Ha l l R G , Froestl W , Bittiger H . CGP-35348: a centrally active blocker of G A B A B receptors. Eur J Pharmacol 1990, 187: 27-38. Padjen A L , Mitsoglou G M , Hassessian H . Further evidence in support of taurine as a mediator of synaptic transmission in the frog spinal cord. Brain Res 1989, 488: 288-296. Pape H C , McCormick D A . Electrophysiological and pharmacological properties of interneurons in the cat dorsal lateral geniculate nucleus. Neuroscience 1995, 68: 1105-1125. Parsons C G , Danysz W , Hesselink M , Hartmann S, Lorenz B , Wollenburg C , Quack G . Modulation of N M D A receptors by glycine - introduction to some basic aspects and recent developments. Amino Acids 1998, 14: 207-216. Paus T. Functional anatomy of arousal and attention systems in the human brain. Prog Brain Res 2000, 126: 65-77. Paxinos G , Watson C. The Rat Brain in Stereotaxic Coordinates, 2nd edn, 1986, Academic Press, North Ryde. Penny G R , Itoh K , Diamond IT. Cells of different sizes in the ventral nuclei project to different layers of the somatic cortex in the cat. Brain Res 1982, 242: 55-65. Bibliography - 157-Piechotta K , Weth F, Harvey RJ , Friauf E . Localization of rat glycine receptor 1 and 2 subunit transcripts in the developing auditory brainstem. J Comp Neurol 2001,438: 336-352. Pol l in B , Joulin Y , Amsallem B , Rokyta R, Cesaro P. Reciprocal interactions between intralaminar and lateral thalamic nuclei in rats. Physiol Res 1997, 46: 467-474. Poncer JC, Durr R, Gahwiler B H , Thompson S M . Modulation of synaptic G A B A A receptor function by benzodiazepines in area C A 3 of rat hippocampal slice cultures. Neuropharmacol 1996,35: 1169-1179. Power B D , Kolmac CI , Mitrofanis J. Evidence for a large projection from the zona incerta to the dorsal thalamus. J Comp Neurol 1999, 404: 554-565. Pribil la I, Takagi T, Langosch D , Bormann J, Betz H . The atypical M 2 segment of the (3 subunit confers picrotoxinin resistance to inhibitory glycine receptor channels. EMBO J 1992, 11: 4305-4311. Ragozzino D , Woodward R M , Murata Y , Eusebi F, Overman L E , Mi led i R. Design and in vitro pharmacology of a selective y-aminobutyric acidc receptor antagonist. Mol Pharmacol 1996, 50: 1024-1030. Rainey W T , Jones E G . Spatial distribution of individual medial lemniscal axons in the thalamic ventrobasal complex of the cat. Exp Brain Res 1983, 49: 229-246. Rajendra S, Lynch JW, Pierce K D , French C R , Barry P H , Schofield PR. Mutation of an arginine residue in the human glycine receptor transforms beta-alanine and taurine from agonists into competitive antagonists. Neuron 1995, 14: 169-175. Rajendra S, Lynch JW, Schofield PR. The glycine receptor. Pharmacol Ther 1997, 73: 121-146. Rampon C , Luppi P H , Fort P, Peyron C, Jouvet M . Distribution of glycine-immunoreactive cell bodies and fibers in the rat brain. Neuroscience 1996, 75: 737-755. Ran I, Mathers D A , Pui l E . Pentobarbital induces thalamic oscillations in brain slices, modulated by G A B A and glycine receptors. Neuropharmacology 2004, 47: 985-993. Rausell E , Jones E G . Chemically distinct compartments of the thalamic V P M nucleus in monkeys relay principal and spinal trigeminal pathways to different layers of the somatosensory cortex. JNeurosci 1991, 11: 226-237. Ries C R , Pui l E . Mechanism of anesthesia revealed by shunting actions of isoflurane on thalamocortical neurons. J Neurophysiol 1999, 81: 1795-1801. Robinson H P , Sahara Y , Kawai N . Nonstationary fluctuation analysis and direct resolution of single channel currents at postsynaptic sites. Biophys J1991, 59: 295-304. Bibliography - 158-Roger M , Cadusseau J. Afferents to the zona incerta in the rat: a combined retrograde and anterograde study. J Comp Neurol 1985, 241: 480-492. Rognan D , Boulanger T, Hoffmann R, Vercauteren DP , Andre J M , Durant F , Wermuth C G . Structure and molecular modeling of G A B A A receptor antagonists. J Med Chem 1992, 35: 1969-1977. Roux M J and Supplisson S. Neuronal and glial glycine transporters have different stoichiometrics. Neuron 2000, 25: 373-383. Ruiz-Gomez A , Morato E , Garcia-Calvo M , Valdivieso F, Mayor F Jr. Localization of the strychnine binding site on the 48-kilodalton subunit of the glycine receptor. Biochemistry 1990, 29: 7033-7040. Salt T E and Eaton S A . Postsynaptic potentials evoked in ventrobasal thalamus neurones by natural sensory stimuli. Neurosci Lett 1990, 114: 295-299. Salt T E , Eaton S A . Functions of ionotropic and metabotropic glutamate receptors in sensory transmission in the mammalian thalamus. Prog Neurobiol 1996, 48: 55-72. Schmieden V , Betz H . Pharmacology of the inhibitory glycine receptor: agonist and antagonist action of amino acids and piperidine carboxylic acid compounds. Mol Pharmacol 1995, 48: 919-927. Schmieden V , Kuhse J, Betz H . Agonist pharmacology of neonatal and adult glycine receptor a subunits: identification of amino acid residues involved in taurine activation. EMBO J 1992, 11: 2025-2032. Schmieden V , Kuhse J, Betz H . A novel domain of the inhibitory glycine receptor determining antagonist efficacies: further evidence for partial agonism resulting from self-inhibition. Mol Pharmacol 1999, 56: 464-472. Scheuss V , Neher E . Estimating synaptic parameters from mean, variance, and covariance in trains of synaptic responses. Biophys J 2001, 81: 1970-1989. Seutin V , Scuvee-Moreau J, Dresse A . Evidence for a non-GABAergic action of quaternary salts o f bicuculline on dopaminergic neurones. Neuropharmacology 1997, 36: 1653-7. Sherman S M . Thalamic relays and cortical functioning. Prog Brain Res 2005, 149: 107-126. Sherrington C , Co-ordination in the simple reflex. In: Sherrington, C , (Ed), The Integrative Act ion of the Nervous System. 1947, Yale University Press, New Haven, pp. 107-114. Bibliography - 159-Shirasaki T, Klee M R , Nakaye T, Akaike N . Differential blockade of bicuculline and strychnine on G A B A - and glycine-induced responses in dissociated rat hippocampal pyramidal cells. Brain Res. 1991,561:77-83. Siebler M , Pekel M , Koller H , Muller H W . Strychnine-sensitive glycine receptors in cultured primary neurons from rat neocortex. Brain Res Dev Brain Res 1993, 73: 289-292. Sieghart W , Sperk G . Subunit composition, distribution and function of G A B A A receptor subtypes. Curr Top Med Chem 2002, 2: 795-816. Sigworth FJ . The variance of sodium curent fluctuations at the node of Ranvier. J Physiol 1980, 307: 97-129. Simon J, Wakimoto H , Fujita N , Lalande M , Barnard E A . Analysis of the set of G A B A A receptor genes in the human genome. J Biol Chem 2004, 279: 41422-41435. Singer J H , Berger A J . Contribution of single-channel properties to the time course and amplitude variance of quantal glycine currents recorded in rat motoneurons. J Neurophysiol 1999, 81: 1608-1616. Singer J H , Talley E M , Bayliss D A , Berger A J . Development of glycinergic synaptic transmission to rat brain stem motoneurons. J Neurophysiol 1998, 80: 2608-2620. Sivilotti L , Nistri A . G A B A receptor mechanisms in the central nervous system. Prog Neurobiol 1991,36: 35-92. Smythies J. The functional neuroanatomy of awareness: with a focus on the role of various anatomical systems in the control of intermodal attention. Conscious Cogn 1997, 6: 455-481. Steriade M . Sleep, epilepsy and thalamic reticular inhibitory neurons. Trends Neurosci 2005, 28: 317-324. Steriade M , Jones E G , McCormick D A . (Eds) Thalamus, 1997, Elsevier Science Ltd. , U K . Steriade M , McCormick D A , Sejnowski TJ . Thalamocortical oscillations in the sleeping and aroused brain. Science 1993, 262: 679-685. Takahashi T, Momiyama A , Hirai K , Hishinuma F, Akagi H . Functional correlation of fetal and adult forms of glycine receptors with developmental changes in inhibitory synaptic receptor channels. Neuron 1992, 9: 1155-1161. Takahashi Y , Shirasaki T, Yamanaka H , Ishibashi H , Akaike N . Physiological roles of glycine and y-aminobutyric acid in dissociated neurons of rat visual cortex. Brain Res 199'4, 640: 229-235. Bibliography - 160-Tappaz M L . Taurine biosynthetic enzymes and taurine transporter: molecular identification and regulations. Neurochem Res 2004, 29: 83-96. Tennigkeit F, Schwarz D W , Pui l E . G A B A B receptor activation changes membrane and filter properties of auditory thalamic neurons. Hear Res 1998, 122: 18-24. Tian N , Hwang T N , Copenhagen D R . Analysis of excitatory and inhibitory spontaneous synaptic activity in mouse retinal ganglion cells. J Neurophysiol 1998, 80: 1327-1340. Tokutomi N , Kaneda M , Akaike N . What confers specificity on glycine for its receptor site? Br J Pharmacol 1989, 97: 353-360. Tombol T. Short neurons and their synaptic relations in the specific thalamic nuclei. Brain Res 1967, 3: 307-326. Traynelis SF, Silver R A , Cull-Candy SG. Estimated conductance of glutamate receptor channels activated during EPSCs at the cerebellar mossy fiber-granule cell synapse. Neuron 1993, 11: 279-289. Tsumoto T. Characteristics of the thalamic ventrobasal relay neurons as a function of conduction velocities of medial lemniscal fibers. Exp Brain Res 1974, 21: 211-224. Turner JP, Anderson C M , Will iams SR, Crunelli V . Morphology and membrane properties of neurones in the cat ventrobasal thalamus in vitro. J Physiol 1997, 505: 707-726. Twyman R E , Macdonald R L . Kinetic properties of the glycine receptor main and sub-conductance states of mouse spinal cord neurons in culture. J Physiol (Lond) 1991, 435, 303-331. Ueno S, Bracamontes J, Zorumski C, Weiss DS , Steinbach J H . Bicuculline and gabazine are allosteric inhibitors of channel opening of the G A B A A receptor. J Neurosci 1997, 17: 625-634. Vaello M L , Ruiz-Gomez A , Lerma J, Mayor F Jr. Modulation of inhibitory glycine receptors by phosphorylation by protein kinase C and cAMP-dependent protein kinase. J Biol Chem 1994, 269: 2002-2008. van Groen T, Kadish I, Michael Wyss J. Role of the anterodorsal and anteroventral nuclei of the thalamus in spatial memory in the rat. Behav Brain Res 2002, 132: 19-28. von Krosigk M , Ba i T, McCormick D A . Cellular mechanisms of a synchronized oscillation in the thalamus. Science 1993, 261: 361-364. Wan X , Mathers D A , Pui l E . Pentobarbital modulates intrinsic and GABA-receptor conductances in thalamocortical inhibition. Neuroscience 2003, 121: 947-958. Bibliography - 161 -Wan X , Pui l E . Pentobarbital depressant effects are independent of G A B A receptors in auditory thalamic neurons. JNeurophysiol 2002, 88: 3067-77. Welker W l . Principles of organization of the ventrobasal complex in mammals. Brain Behav Evol 1973, 7: 253-336. Werman R, Davidoff R A , Aprison M H . Inhibitory action of glycine on spinal neurons of the cat. J Neurophysiol 1968, 31: 81-95. Wermuth C G , Bourguignon JJ, Schlewer G , Gies JP, Schoenfelder A , Mel ik ian A , Bouchet M J , Chantreux D , Molimard JC, Heaulme M , et al. Synthesis and structure-activity relationships of a series of aminopyridazine derivatives of y-aminobutyric acid acting as selective G A B A A antagonists. J Med Chem 1987, 30: 239-249. Weyand T G , Boudreaux M , Guido W . Burst and tonic response modes in thalamic neurons during sleep and wakefulness. J Neurophysiol 2001, 85: 1107-18. Wierenga C J , Wadman W J . Functional relation between interneuron input and population activity in the rat hippocampal cornu ammonis 1 area. Neuroscience 2003, 118:1129-1139. Wingrove P B , Thompson S A , Wafford K A , Whiting PJ. Key amino acids in the y subunit of the y-aminobutyric ac id A receptor that determine ligand binding and modulation at the benzodiazepine site. Mol Pharmacol 1997, 52: 874-881. W u FS, Gibbs TT, Farb D H . Dual activation of G A B A A and glycine receptors by (3-alanine: inverse modulation by progesterone and 5 alpha-pregnan-3 alpha-ol-20-one. Eur J Pharmacol 1993,246:239-46. Yarbrough G G , Singh D K , Taylor D A . Neuropharmacological characterization of a taurine antagonist. J Pharmacol Exp Ther 1981, 219: 604-613. Y e n C T , Jones E G . Intracellular staining of physiologically identified neurons and axons in the somatosensory thalamus of the cat. Brain Res 1983, 280: 148-154. Yeung J Y T , Canning K J , Zhu G , Pennefather P, MacDonald JF, Orser B A . Tonically activated G A B A A receptors in hippocampal neurons are high-affinity, low-conductance sensors for extracellular G A B A . Mol Pharmacol 2003, 63: 2-8. Young A B , Snyder S H . Strychnine binding associated with glycine receptors of the central nervous system. Proc Natl Acad Sci USA 1973, 70: 2832-2836. Y u S, Ho IK. Effects of G A B A antagonists, SR 95531 and bicuculline, on G A B A A receptor-regulated chloride flux in rat cortical synaptoneurosomes. Neurochem Res 1990 15: 905-910. Zarbin M A , Wamsley JK , Kuhar M J . Glycine receptor: light microscopic autoradiographic localization with [ 3H] strychnine. J Neurosci 1981, 1: 532-547. Bibliography - 162-Zeilhofer H U , Studler B , Arabadzisz D , Schweizer C , Ahmadi S, Layh B , Bos l M R , Fritschy J M . Glycinergic neurons expressing enhanced green fluorescent protein in bacterial artificial chromosome transgenic mice. J Comp Neurol 2005, 482: 123-141. Zhang Z W , Deschenes M . Intracortical axonal projections of lamina V I cells of the primary somatosensory cortex in the rat: a single-cell labeling study. J Neurosci 1997, 17: 6365-6379. Zhang SJ, Huguenard JR, Prince D A . G A B A A receptor-mediated CI" currents in rat thalamic reticular and relay neurons. J Neurophysiol 1997, 78: 2280-2286. Appendix A: Matlab codes for simulation of synaptic currents. - Filename: PSCSimulationClassic.m % Name: PSCSimulationClassic.m % Description: Simulates PSC according to the "classic algorithm" and input parameters. % % input parameters are: DecayTimeCnst (Decays time constant); % RiseTimeCnst (Rise time constant); Noofchanels (Number of Channels); % Current (Unitary Current), SamplingRate (Sampling Rate), % NoOfTrials (Number of T r i a l s ) , Repet (Number of Virtual Neurons) % % Uses a probability distr ibution given by: % p(t)=eA(t/RiseTimeCnst)*(l-eA(t/DecayTimeCnst)) % to assign open or closed states to the channels. % % * NoOfTrials reflects how many columns are in each f i l e . % * Repet is a lable for virtual "neuron" % * Gaussian random noise is added by AddNoise.m code. % function PSCSimulationClassic(DecayTimeCnst, Noofchanels, Current, . . . SamplingRate, NoOfTrials, Repet) clear a l l ; cd 'E:\PSCSi mulati on ' ; RiseTimeCnst = 0 .5; % Create the probability distr ibut ion: % The code for probDistr function is probDistr.m t = 0:SamplingRate:(8*DecayTimeCnst); PO = probDistr(RiseTimeCnst, DecayTimeCnst, SamplingRate); NoOfSamples = length(t); out(Noofsamples, Noofchanels) = 0; for a = 1 : NoOfTrials fpr in t f ( ' T r i a l : ') ; fprintf( '%d' ,a); f p r i n t f ( ' \ n ' ) ; for b = 1 : Noofchanels Out(:, b) = 0; % Assign open or closed values to the channels based on the PO for c = 1 : NoOfSamples Out(c, b) = binorndd, PO(c), 1 ,1); end end %Do I want random noise or not? i f 1 % Create Random noise with 9 9 . 7 % (within 3 SD) of the % observations lying in between -0 .5 and 0.5 pA. % The code for RandomNoise function is RandomNoise.m RandomNoise(NoOfSamples) = 0; for c = 1 : NoOfSamples RandomNoise(c) = randn()/6; end % Add the RandomNoise to each single channel value: for c = 1 : NoOfSamples Out(c) = out(c) + RandomNoise(c); end end % Place the result of one t ra i l into the Matrix that's going to be written on the HDD. for b = 1 : NoOfSamples RealOut(b,a)= Current*sum(Out(b,:)); end end Appendix A - 164-% O u t p u t d i r e c t o r y : O u t p u t D i r = ' E:\PSCSimu1ation \ D a t a ' ; % S a v e t h e d a t a t o a f i l e ; OutputName = s t r c a t ( ' O u t _ C l a s s i c a l _ T a u _ ' , n u m 2 s t r ( D e c a y T i m e C n s t ) , . . . ' _ N o C h n _ ' , n u m 2 s t r ( N 0 0 f C h a n e l s ) , ' _ C u r _ ' , n u m 2 s t r ( c u r r e n t ) , . . . ' _ S R _ ' , n u m 2 s t r ( S a m p l i n g R a t e ) , ' _ T R L _ ' , n u m 2 s t r ( N o O f T r i a l s ) , . . . '_cen_', num2str ( R e p e t ) . ' . t x t ' ) ; f p r i n t f C w r i t i n g t o f i l e : ' ) ; f p r i n t f ( O u t p u t N a m e ) ; f p r i n t f ( ' \ n ' ) ; f i d = f o p e n ( s t r c a t ( O u t p u t D i r , ' \ ' , O u t p u t N a m e ) , ' w ' ) ; f o r c = 1 : N o o f S a m p l e s f o r d = 1 : N o O f T r i a l s f p r i n t f ( f i d , '%f ' , R e a l o u t ( c . d ) ) ; end f p r i n t f ( f i d , ' \ n ' ) ; end f c l o s e ( f i d ) ; end F i 1 ename: P S C S i m u i a t i o n s t a t i o n a r y . m % Name: P S C S i m u l a t i o n S t a t i o n a r y . m % D e s c r i p t i o n : S i m u l a t e s PSC a c c o r d i n g t o a s t h e " s t a t i o n a r y - s e g m e n t s a l g o r i t h m " % and i n p u t p a r a m e t e r s . % % i n p u t p a r a m e t e r s a r e : D e c a y T i m e C n s t ( D e c a y s t i m e c o n s t a n t ) ; % R i s e T i m e C n s t ( R i s e t i m e c o n s t a n t ) ; N o o f c h a n e l s (Number o f C h a n n e l s ) ; % C u r r e n t ( U n i t a r y C u r r e n t ) , S a m p l i n g R a t e ( S a m p l i n g R a t e ) , % N o O f T r i a l s (Number o f T r i a l s ) , R e p e t (Number o f v i r t u a l N e u r o n s ) % % U s e s a p r o b a b i l i t y d i s t r i b u t i o n g i v e n b y : % p ( t ) = e A ( t / R i s e T i m e c n s t ) * ( l - e A ( t / D e c a y T i m e c n s t ) ) % t o a s s i g n open o r c l o s e d s t a t e s t o t h e c h a n n e l s . % % * N o O f T r i a l s r e f l e c t s how many c o l u m n s a r e i n e a c h f i l e . % * R e p e t i s a l a b l e f o r v i r t u a l " n e u r o n " % * G a u s s i a n random n o i s e i s added by A d d N o i s e . m c o d e . % f u n c t i o n P S C S i m u l a t i o n S t a t i o n a r y ( D e c a y T i m e C n s t , N o o f c h a n e l s , C u r r e n t , . . . S a m p l i n g R a t e , N o O f T r i a l s , R e p e t ) c l e a r a l l ; c d ' E : \ P S C S i m u l a t i o n ' ; R i s e T i m e C n s t = 0 . 5 ; % C r e a t e t h e p r o b a b i l i t y d i s t r i b u t i o n : % The c o d e f o r p r o b D i s t r f u n c t i o n i s p r o b D i s t r . m t = 0 : S a m p l i n g R a t e : ( 8 * D e c a y T i m e C n s t ) ; PO = p r o b D i s t r ( R i s e T i m e c n s t , D e c a y T i m e C n s t , S a m p l i n g R a t e ) ; N o O f S a m p l e s = l e n g t h ( t ) ; o u t ( N o o f s a m p l e s , N o o f c h a n e l s ) = 0; f o r a = 1 : N o O f T r i a l s f p r i n t f ( " T r i a l : ' ) ; f p r i n t f ( ' % d ' , a ) ; f p r i n t f ( ' \ n ' ) ; % F o r e a c h s i n g l e c h a n n e l : f o r b = 1 : N o o f c h a n e l s O u t ( : , b ) = 0; % s a m p l e a t 10 x t h e s a m p l i n g r a t e and a s s i g n open o r c l o s e d v a l u e s % A s s i g n open o r c l o s e d v a l u e s t o t h e c h a n n e l s b a s e d on t h e PO f o r c = 1 : 10 : N o O f S a m p l e s O u t ( c , b ) = b i n o r n d d , P O ( c ) , 1 ,1); e n d ; % R e s a m p l e a t minimum s a m p l i n g r a t e and c r e a t e s t a t i o n a r y s e g m e n t s % f o r t h e open c h a n n e l s . % C l o s e d c h a n n e l s a r e a l r e a d y s t a t i o n a r y b e c a u s e d e f a u l t i s c l o s e d Appendix A - 165-for c = 1 : NoOfSamples i f (c > 10 && Out(c.b) == 1) for d = 1 : 9 Out(c-d, b) = 1; end end end % DO I want random noise or not? i f 1 % Create Random noise with 99.7% (within 3 SD) of the % observations lying in between -0.5 and 0.5 pA. % The code for RandomNoise function i s RandomNoise.m RandomNoise(NoOfSamples) = 0; for c = 1 : NoOfSamples RandomNoise(c) = randn()/6; end % Add the RandomNoise to each single channel value: for c = 1 : NoOfSamples Out(c) = out(c) + RandomNoise(c); end end end plot(sum(Out(b,:))); % Place the result of one t ra i l into the Matrix that's going to be written on the HDD. for b = 1 : NoOfSamples RealOut(b,a)= Current*sum(out(b,:)); end end i f 1 % Output directory: OutputDir = 'E:\PSCCSimulation\Data'; % Save the data to a f i l e ; OutputName = strcat('Out_Stat_Tau_', num2str(DecayTimeCnst),... '_NoChn_',num2str(N00fChanels), '_Cur_' , num2str(Current), . . . '_SR_', num2str(SamplinqRate), '_TRI ' , num2str(NoOfTrials), . . . ' _ C e l l _ ' , num2str(Repet), ' . t x t ' ) ; fpr in t f ( 'writing to f i l e : ' ) ; fprintf(OutputName); f p r i n t f ( ' \ n ' ) ; f id = fopen(strcat(OutputDir, ' \ ' , OutputName), 'w'); for c = 1 : NoOfSamples for d = 1 : NoOfTrials fpr int f ( f id , "%f ' , RealOut(c.d)); end fpr int f ( f id , ' \ n ' ) ; end fc lose( f id ) ; end Filename: probDistr.m % Name: PSCSimulationStationary.m % Description: This function returns a sequence of probabil i t ies according to exponential % rise and decay for which the time constant i s given as an input parameter. % input parameters are Rise Time Constant (RiseTimeCnst), Decay Time Constant % (DecayTimeCnst) and Sampling Rate (SamplingRate) function OutMatrix = probDistr(RiseTimeCnst, DecayTimeCnst, SamplingRate) % The time span is 8 X Decay Time Constant (to make sure i t the synaptic current % has completely decayed. t = 0:SamplingRate:(8*DecayTimeCnst) ; Appendix A - 166 -A = (1 - e x p ( - t / R i s e T i m e C n s t ) ) ; B = expC - t /DecayTimeCnst ) ; OutMatr ix = A . * B; Fi lename: AddNoise.m % Name: AddNoise % D e s c r i p t i o n : % Reads i n a l l f i l e s from an input d i r e c t o r y . % Adds Gaussian random n o i s e . % wr i t es i t back to an output d i r e c t o r y . c l e a r a l l ; i n p u t D i r = ' E : \ u p d a t e N o i s e \ D a t a \ ' ; cd ( inputDi r ) ; m p u t F i l e s = d i r ; m p u t F i l e s = i n p u t F i l e s ( 3 : l e n g t h ( i n p u t F i l e s ) ) ; OutputDir = ' E : \ u p d a t e N o i s e \ D a t a A n a l \ ' ; cd E : \ U p d a t e N o i s e \ ; % Read the da ta : f o r a = 1 : l e n g t h ( i n p u t F i l e s ) f i d = f o p e n ( ( s t r c a t ( i n p u t D i r , i n p u t F i l e s ( a ) . n a m e ) ) , ' r ' ) ; RawData = f s c a n f ( f i d , '%d ' ) ; f c l o s e ( f i d ) ; n = 1ength(RawData); % use the nameAnalysis subrout ine to i d e n t i f y the number o f t r i a l s % The code f o r nameAnalysis subrout ine i s nameAnalysis.m N o O f T r i a l s = nameAnalysis ( m p u t F i l e s (a).name, 'TRL') NoOfRows = n / N o O f T r i a l s ; Data = reshape(RawData, N o O f T r i a l s , NoOfRows); Data = D a t a ' ; % Adding n o i s e : % use the nameAnalysis subrout ine to i d e n t i f y the number o f channels % The code f o r nameAnalysis subrout ine i s nameAnalysis.m Noofchanels = nameAna lys is ( lnputF i les (a ) .name, 'NoChn' ) ; % - Add noise to each column accord ing to the number o f channels f o r b = 1 : NoOfTr ia ls f o r c = 1 : NoOfRows f o r d = 1 : Noofchanels Data (c .b ) = Data(c .b) + randn()/6; end end end i f 1 % save i t to a f i l e ; OutputName = s t r c a t ( o u t p u t D i r , ' N o i s e _ u p d a t e d ' , l n p u t F i l e s ( a ) . n a m e ) ; f p r i n t f C w r i t i n g to f i l e : ' ) ; fpr in t f (OutputName); f p r i n t f ( ' \ n ' ) ; f i d = fopen((OutputName), 'w ' ) ; f o r b = 1 : NoOfRows f o r c = 1 : NoOfTr ia ls f p r i n t f ( f i d , '%d ' , D a t a ( b . c ) ) ; end f p r i n t f ( f i d , ' \ n ' ) ; end f c l o s e ( f i d ) ; end end Appendix A - 167-^ Filename: nameAnalysis.m % Name: nameAnalysis.m % Description: % Search through the input string for the item string and then return % the characters after item str ing. % The characters required to be returned are between underscores (_). function returnvalue = nameAnalysis (inputstring, Item) StartPos = f indstr( lnputstr ing, item); % Now move the cursor over the item characters one more to account for the f i r s t _ StartPos = StartPos + length(ltem)+l; % Cut out the beginning of the Inputstring so the new beginning starts % at the number of interest Inputstring = lnputString(StartPos:length(lnputString)); % Look for the next _ (which indicates where the number ends) UndrPos = f indstr( lnputstr ing, '_ ' ) ; undrpos = undrpos(l); % The number of interest in string form i s : No = lnputString(l:l)ndrPos-l); % Convert the string into a double precision number: returnvalue = str2num(No) ; Appendix B : Matlab codes for non-stationary fluctuation analysis Filename: F luctuat ionAna1ysis . n l % Name: F l u c t u a t i o n A n a l y s i s . m % D e s c r i p t i o n : % Creates an a n a l y s i s f i l e f o r each input f i l e that d e t a i l s the r e s u l t s o f performing % n o n - s t a t i o n a r y F l u c t u a t i o n a n a l y s i s (Sigworth) f o r vary ing b i n - s i z e s . % % * For b i o l o g i c a l da ta , the code reads i t as f l o a t and mi r ro rs i t on y a x i s . % % * The r e s u l t s o f are normal ized to the value o f the u n i t a r y cur ren t f o r b in s i z e 1. % % * Data w i l l be saved i n f i l e s that conta in the l a b e l _Norm2_. % % * The data i s adjusted f o r b a s e l i n e . % % * Number o f t r i a l s i s read from the name o f the f i l e . % % * Streaml ined the code c l e a r a l l ; i n p u t D i r = ' E : \ N o i s e A n a 1 y s i s \ D a t a \ ' ; cd ( I n p u t D i r ) ; i n p u t F i l e s = d i r ; i n p u t F i l e s = l n p u t F i l e s ( 3 : l e n g t h ( i n p u t F i l e s ) ) ; OutputDir = ' E : \ N o i s e A n a l y s i s \ D a t a A n a l \ ' ; cd E : \ N o i s e A n a l y s i s \ ; Avrg = 0; f o r a = 1 : l e n g t h ( i n p u t F i l e s ) f i d = f o p e n ( ( s t r c a t ( l n p u t D i r , i n p u t F i l e s ( a ) . n a m e ) ) , ' r ' ) ; RawData = f s c a n f ( f i d , ' % f ) ; f c l o s e ( f i d ) ; n = 1ength(RawData); % The code f o r nameAnalysis f u n c t i o n i s nameAnalysis.m N o O f T r i a l s = nameAnalys is ( InputF i les (a ) .name, ' T R L ' ) ; NoOfRows = n / N o O f T r i a l s ; Data = reshape(RawData, N o O f T r i a l s , NoofRows); Data = D a t a ' ; % *TO ad jus t the b a s e l i n e , the lowest negat ive number i s found and then data i s s h i f t e d % up so t h i s number w i l l be 0. % % For each column: f o r b = 1 : NoOfTr ia ls % Step 1. F ind lowest po in t ( t h a t ' s smal ler than 0). Min = 0; f o r c = 1 : NoofRows i f (Min > Data(c , b)) Min = Data(c , b ) ; end end % Step 2. I f Min i s negat ive s h i f t every th ing up by Min. i f Min < 0; f o r c = 1 : NoofRows Data (c .b ) = Data (c .b ) - Min; end end end BinMult = [2, 4, 8, 16, 32, 64, 128, 256]; % Analyze each Data set us ing d i f f e r e n t b in s i z e s . % The code f o r binFunc f u n c t i o n i s binFunc.m M a t r i x O l = Data; Appendix B - 169-Matrix02 = b inFunc(Data ,2 , N o O f T r i a l s ) ; Matr ix04 = binFunc(Data,4, N o O f T r i a l s ) ; Matrix08 = binFunc(Data,8, N o O f T r i a l s ) ; Matrixl6 = binFunc(Data,16, N o O f T r i a l s ) ; Matrix32 = binFunc(Data,32 , N o O f T r i a l s ) ; Matrix64 = binFunc(Data,64, N o O f T r i a l s ) ; Matrixl28 = binFunc(Data,128, N o O f T r i a l s ) ; Matrix256 = binFunc(Data ,256, N o O f T r i a l s ) ; % F i l l in the f i r s t column o f the output f i l e with the va lues o f the % b in s i z e . OutMat r ix(l,l) = 1; OutMatr ix(2 : length(BinMult)+l,l) = BinMul t ; % F i l l i n the res t o f the columns with the r e s u l t s o f the q u a d r a t i c f i t % The code f o r AvgCovReg f u n c t i o n i s AvgCovReg.m OutMatrix(l,2:4) = AvgCovReg(MatrixOl); OutMatrix(2,2:4) = AvgCovReg(Matrix02); OutMatrix(3,2:4) = AvgCovReg(Matrix04); OutMatrix(4,2:4) = AvgCovReg(Matrix08); OutMatrix(5,2:4) = AvgCovReg(Matrixl6); OutMatrix(6,2:4) = AvgCovReg(Matrix32); OutMatrix(7,2:4) = AvgCovReg(Matrix64); OutMatrix(8,2:4) = AvgCovReg(Matrixl28); 0utMatrix(9,2:4) = AvgCovReg(Matrix256); % Normalize the va lues i n the 3rd column o f the output matr ix . % Normalize the other va lues to value o f cur rent at b in s i z e 1. O u t M a t r i x ( : , 3) = OutMatr ix ( :, 3 ) /OutMatr ix(l, 3); i f 1 % save i t to a f i l e ; OutputName = s t r c a t ( o u t p u t D i r , ' A n a l _ N o i s e 5 _ ' . I n p u t F i l e s ( a ) . n a m e ) ; f p r i n t f ( ' w r i t i n g to f i l e : ' ) ; f p r i n t f ( i n p u t F i T e s ( a ) . n a m e ) ; f p r i n t f C \ n ' ) ; f i d = fopen((OutputName), 'w ' ) ; f o r c = 1 : length(BinMult)+l f o r d = 1 : 4 f p r i n t f ( f i d , '%d ', O u t M a t r i x ( c . d ) ) ; end f p r i n t f ( f i d , ' \ n ' ) ; end f c l o s e ( f i d ) ; end end Fi lename: binFunc.m % Name: binFunc.m % D e s c r i p t i o n : Bins the inputMatr ix i n t o BinMult b ins and returns the new matr ix as OutputMatr ix . % % B i n n i n g : % % *No o f t r i a l s i s read as a v a r i a b l e , (see F l u c t u a t i o n A n a l y s i s . m code) . f u n c t i o n OutputMatr ix = b inFunc( inputMat r ix , B inMul t , N o O f T r i a l s ) Dimen = s i z e ( i n p u t M a t r i x ) ; NoofRows = Dimen( l ) ; f o r a = 1 : BinMult : NoofRows - BinMult + 1 f o r b = 1 : NoOfTr ia ls SumOfBin = 0; f o r c = 1 : BinMult SumOfBin = SumOfBin + lnputMat r ix (a + c - 1, b ) ; end Outpu tMat r i x ( (a+B inMul t - l ) /B inMul t ,b ) = SumofBin/BinMult ; end end Fi lename: AvgCovReg.m % Name: F l u c t u a t i o n A n a l y s i s . m % D e s c r i p t i o n : c a l c u l a t e s an n X 2 matrix con ta in ing the var iance and mean ampl i tude . % Performs a l e a s t squares quadra t i c regress ion on the two columns o f the matr ix % and returns the c o e f f i c i e n t s a , b, c o f axA2 + bx + c . Appendix B - 170-f u n c t i o n Coef = AvgCovReg(lnData) Temp = s i z e ( l n D a t a ) ; NoOfRows = Temp( l ) ; A v g C o v ( : , l ) = mean( lnData,2) ; f o r a = 1:NoOfRows AvgCov(a,2) = va r ( inData (a end : ) ) ; % The code f o r quadReg f u n c t i o n i s quadReg.m Coef = quadReg(Avgcov); Fi lename: quadReg.m % Name: quadReg.m % D e s c r i p t i o n : Perform a l e a s t squares quadra t i c r e g r e s s i o n on the input se t o f d a t a . f u n c t i o n Coef = quadReg(inputMatr ix) XDrv (:, 1) = i n p u t M a t r i x ( : ,1) . * i n p u t M a t r i x ( : , 1 ) ; XDrv (:, 2) = i n p u t M a t r i x ( : , 1 ) ; XDrv (:, 3) = 1; coe f = i n v ( X D r v ' * X D r v ) * x p r v ' * i n p u t M a t r i x ( : , 2 ) ; Fi lename: nameAnalysis.m % Name: nameAnalysis.m % D e s c r i p t i o n : % Search through the input s t r i n g f o r the item s t r i n g and then re turn % the c h a r a c t e r s a f t e r item s t r i n g . % The c h a r a c t e r s requi red to be returned are between underscores (_). f u n c t i o n re turnva lue = nameAnalys is ( InputSt r ing , Item) Star tPos = f i n d s t r ( l n p u t S t r i n g , Item); % Now move the cursor over the item charac te rs one more to account f o r the f i r s t _ S ta r tPos = Star tPos + 1ength( i tem)+l; % Cut out the beginning of the i n p u t S t r i n g so the new beginning s t a r t s % at the number o f i n t e r e s t i n p u t S t r i n g = l n p u t S t r i n g ( S t a r t P o s : 1 e n g t h ( l n p u t S t r i n g ) ) ; % Look f o r the next _ (which i n d i c a t e s where the number ends) undrPos = f i n d s t r ( l n p u t s t r i n g , ' _ ' ) ; undrpos = UndrPos ( l ) ; % The number o f i n t e r e s t i n s t r i n g form i s : NO = l n p u t S t r i n g ( l : u n d r P o s - l ) ; % Convert the s t r i n g i n t o a double p r e c i s i o n number: re turnva lue = str2num(No); 

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0092823/manifest

Comment

Related Items