Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
PageArticles 1 of 39 in PresS. Am J Physiol Regul Integr Comp Physiol (May 23, 2007). doi:10.1152/ajpregu.00912.2006 Are pacemaker properties required for respiratory rhythm generation in adult turtle brainstems in vitro? Stephen M. Johnson, Liana M. Wiegel, David J. Majewski Department of Comparative Biosciences School of Veterinary Medicine University of Wisconsin Madison, Wisconsin 53706 Abbreviated title: Pacemaker properties and turtle breathing in vitro Corresponding Author: Stephen M. Johnson, M.D., Ph.D. Assistant Professor Department of Comparative Biosciences School of Veterinary Medicine University of Wisconsin 2015 Linden Drive Madison, Wisconsin 53706 Phone: (608) 261-1104 Fax: (608) 263-3926 e-mail: [email protected] Copyright © 2007 by the American Physiological Society. Page 2 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 2 ABSTRACT The role of pacemaker properties in vertebrate respiratory rhythm generation is not well understood. To address this question from a comparative perspective, brainstems from adult turtles were isolated in vitro and respiratory motor bursts were recorded on hypoglossal (XII) nerve rootlets. The goal was to test whether burst frequency could be altered by conditions known to alter respiratory pacemaker neuron activity in mammals (e.g., increased bath KCl or blockade of specific inward currents). While bathed in artificial cerebrospinal fluid (aCSF), respiratory burst frequency was not correlated with changes in bath KCl (0.5-10.0 mM). Riluzole (50 µM; persistent Na+-channel blocker) increased burst frequency by 31 ± 5% (P<0.05) and decreased burst amplitude by 42 ± 4% (P<0.05). In contrast, flufenamic acid (FFA, 20-500 µM; Ca2+-activated cation channel blocker) reduced and abolished burst frequency in a dose- and time-dependent manner (P<0.05). During synaptic inhibition blockade with bicuculline (50 µM; GABAA channel blocker) and strychnine (50 µM; glycine receptor blocker), rhythmic motor activity persisted and burst frequency was directly correlated with extracellular KCl (0.5-10.0 mM; P=0.005). During synaptic inhibition blockade, riluzole (50 µM) did not alter burst frequency while FFA (100 µM) abolished burst frequency (P<0.05). These data are most consistent with the hypothesis that turtle respiratory rhythm generation requires Ca2+-activated cation channels but not pacemaker neurons, which thereby favors the group-pacemaker model. During synaptic inhibition blockade, however, the rhythm generator appears to be transformed into a pacemaker-driven network that requires Ca2+-activated cation channels. Key words: control of breathing, respiratory control, reptile, pacemaker, chelonian Page 3 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 3 INTRODUCTION The mechanisms underlying vertebrate respiratory rhythm generation are controversial and not well understood. Network models propose that respiratory neurons are reciprocally coupled via inhibitory synaptic connections that work to establish exclusive bursts of activity associated with inspiration, postinspiration, and expiration (35, 41, 42, 46). The hybrid pacemaker-network model proposes that pacemaker neurons embedded within a neural network play a necessary role in rhythm generation (5, 6, 13, 16, 24, 36, 38, 52). A third model is the group-pacemaker model, which postulates that inspiratory bursts are an emergent property of nonpacemaker neurons that are interconnected by chemical and electrotonic excitatory synaptic interactions (10, 11, 14, 15, 39). Thus, the role of pacemaker properties (whether in neurons or networks) and synaptic inhibition are discriminating features of these models. Most data supporting the hybrid pacemaker-network and group-pacemaker models are derived from perinatal and juvenile mammals whereas network models are supported primarily by data from adult mammals. The maturation network-burster hypothesis (43) attempts to harmonize these findings by stating that the respiratory control system is pacemaker-driven in young animals, but becomes network-driven in adults due to increasing synaptic inhibition and activation of specific ionic conductances (17, 18, 31, 32, 33). However, in some nonmammalian adult vertebrate preparations, there is evidence suggesting that pacemaker properties may be required for respiratory rhythm generation. For example, in brainstems isolated from adult lampreys (44) and adult turtles (23), rhythmic motor activity persists during synaptic inhibition blockade. This persistent rhythmic activity in isolated adult turtle brainstem (and brainstemspinal cord) preparations is hypothesized to be respiratory-related because motor bursts are produced on a spinal expiratory nerve (23), and abolished by characteristic respiratory Page 4 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 4 depressants such as µ-opiate receptor activation (23) and high pH/low CO2 conditions (unpublished observations). These data suggest that expiratory pacemaker neurons or clusters of expiratory neurons with pacemaker properties drive the rhythm, rather than non-respiratory neurons taking over control of the respiratory control system and producing a seizure-like pattern (4, 43). To test the hypothesis that pacemaker properties are involved in turtle respiratory rhythm generation, brainstems from adult turtles were isolated and tested under in vitro conditions. Specifically, we tested whether depolarization via increased bath KCl increases the frequency of spontaneous respiratory bursts of motor activity, since the frequency of endogenous bursting in respiratory pacemaker neurons is voltage-dependent (5, 53). We also tested whether riluzole or flufenamic acid (FFA) altered respiratory burst frequency produced by turtle brainstems. Riluzole blocks persistent Na+ currents and FFA blocks Ca2+-activated cation currents in respiratory neurons in the mammalian medulla (11, 33, 59). Both currents are hypothesized to be required for pacemaker activity in mammalian inspiratory neurons (33, 59), although their obligatory role for respiratory rhythm generation has been challenged (11). In separate experiments under conditions of synaptic inhibition blockade, the effects of altered bath KCl, riluzole, and FFA were examined to test whether synaptic inhibition blockade transforms the respiratory rhythm generator into a pacemaker-driven network (45, 47, 48, 49, 50). A preliminary report of this work was published in abstract form (20). MATERIALS AND METHODS All procedures were approved by the Animal Care and Use Committee at the University of Wisconsin-Madison School of Veterinary Medicine. Adult red-eared slider turtles (Trachemys Page 5 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 5 scripta, n = 161, weight = 696 ± 10 g) were obtained from commercial suppliers and kept in a large open tank where they had access to water for swimming, and heat lamps and dry areas for basking. Room temperature was set to 27-28°C with light 14 hr/day. Turtles were fed ReptoMin® floating food sticks (Tetra, Blacksburg, VA) 3-4 times per week. Turtle brainstem preparations. Turtles were intubated and anesthetized with 5% isoflurane (balance O2) until limb withdrawal to noxious foot pinch was eliminated. Turtles were rapidly decapitated and decerebrated. Brainstems were removed and pinned down in a recording chamber (13 ml volume) with the ventral surface facing upwards (Fig. 1A). Brainstems were superfused (4-6 ml/min) with artificial cerebrospinal fluid (aCSF) at 23ºC containing HEPES (N[2-hydroxyethyl]piperazine-N’-[2-ethane-sulfonic acid]) buffer as follows (in mM): 100 NaCl, 23 NaHCO3, 10 Glucose, 5 HEPES (sodium salt), 5 HEPES (free acid), 2.5 CaCl2, 2.5 MgCl2, 1.0 K2PO4, and 1.0 KCl (bubbled with 5% CO2-95% O2). The pH of solution in the reservoir was measured periodically with a calomel glass pH electrode (Digi-Sense , Cole-Parmer Inst. Co., Vernon Hills, IL) and averaged 7.30 ± 0.01 during data collection. All brainstems were allowed to equilibrate for 2-5 hr prior to initiating an experimental protocol. To record respiratory motor bursts, glass suction electrodes were attached to hypoglossal (XII) nerve rootlets (Fig. 1A). Signals were amplified (10,000x) and band-pass filtered (1-500 Hz) using a differential AC amplifier (model 1700, A-M Systems, Everett, WA) before being rectified and integrated (time constant = 200 ms) using a moving averager (MA-821/RSP, CWE, Inc., Ardmore, PA). Analysis was performed using Axoscope (Axon Instruments, Foster City, CA) and MiniAnalysis software (Synaptosoft, Decatur, GA). All drugs used in this study were obtained from Sigma/RBI Aldrich (St. Louis, MO) and include: (+)-bicuculline (GABAA receptor antagonist), flufenamic acid Page 6 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 6 (FFA; blocks Ca2+-activated cation currents), riluzole (blocks persistent Na+ channels), and strychnine (glycine receptor antagonist). Riluzole and FFA were dissolved in 100 µl of dimethyl sulfoxide (DMSO) and then dissolved slowly in 0.5-1.0 l of HEPES-buffered solution before being applied to turtle brainstems. Control experiments with equivalent amounts of DMSO in HEPES-buffered solution showed that DMSO had no effect on respiratory burst frequency and amplitude (data not shown). Experimental protocols. To determine the effects of bath KCl on burst frequency, baseline data were recorded for 20 min in aCSF (2 mM KCl) before switching to aCSF with altered KCl for 60 min and then washing out with aCSF (2 mM KCl) for 60 min. To determine the effects of riluzole and FFA, baseline data were recorded for 20 min before switching to aCSF with riluzole (50 µM) or FFA (20-500 µM) for 2 hr before washing out with aCSF. To determine the effects of KCl, riluzole, and FFA during synaptic inhibition blockade, bicuculline (50 µM) and strychnine (50 µM) were added to the aCSF to block synaptic inhibition. After allowing for complete transformation of the motor output within 100 min (22), bath KCl was altered for 1 hr, or riluzole (50 µM) or FFA (100 µM) were added along with bicuculline and strychnine for 2 hr. Data analysis XII burst amplitude was measured at the highest point of integrated discharge trajectory in arbitrary units and normalized to the average amplitude recorded during the baseline period. Burst frequency was calculated as the number of bursts per min. All measurements were averaged into 20-min bins and reported as mean ± S.E.M. For statistical inferences, linear regression, one-way ANOVA, or two-way ANOVA with repeated measures design were used Page 7 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 7 (Sigma Stat, Jandel Scientific Software, San Rafael, CA). Post-hoc comparisons were made using Dunnett’s or Student-Newman-Keul’s test. P-values < 0.05 were considered significant. RESULTS Increased bath KCl did not alter burst frequency in aCSF While bathed in aCSF, increasing bath KCl from 2.0 mM to 4.0-9.0 mM often transiently increased burst frequency within the first 10-20 min, but burst frequency was highly variable during the next 40 min (Figs. 2A, 2B, 2D). There were no changes in frequency when bath KCl was increased from 2.0 mM KCl to 5.0 mM (n=4) or 7.0 mM KCl (n=6; Fig. 2B). There was, however, a sustained 80-110% frequency increase was observed when bath KCl was increased from 2.0 mM to 9.0 mM KCl (n=3; P<0.05 for KCl- and time-dependent effects; Fig. 2B). At other increased bath KCl levels, burst frequency decreased below baseline (Fig. 2D). Increasing bath KCl to levels greater than 10.0 mM resulted in tonic activity and disruption of the respiratory rhythm (n=3; data not shown). Decreasing bath KCl from 2.0 to 0.5 mM (n=3) did not alter burst frequency. A graph of burst frequency after 20-40 min of increased KCl exposure (i.e., data at the 60-min time point in Fig. 2B) versus bath KCl revealed no correlation (P=0.066; r2=0.129; Fig. 2D). This time period was chosen for analysis because KCl-induced frequency changes often reached steady-state without the rhythm being disrupted by prolonged KCl exposures. Although there was an occasional initial burst frequency increase during the KCl application, there was no correlation for burst frequency versus bath KCl during the 0-20 min period (P=0.313; r2=0.041). Burst amplitude was highly variable with time-dependent decreases (5.0 and 7.0 mM KCl; n=4, 6, respectively) and increases (9 mM KCl; n=3) after 60 min (Fig. Page 8 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 8 2C). Overall, there was no correlation between bath KCl and burst amplitude after 60 min of increased KCl exposure (P=0.296; r2=0.129; Fig. 2E). Effects of riluzole and FFA in aCSF Bath-applied riluzole (50 µM; n=18) produced a time-dependent increase in burst frequency (Figs. 3, 4A, 4B) and decrease in burst amplitude (Figs. 3, 4D). In the presence of riluzole, frequency increased from 0.71 ± 0.06 bursts/min (baseline) to a maximum of 0.92 ± 0.07 bursts/min (P<0.05) at 80 min before reaching a steady-state frequency of 0.83-0.87 bursts/min for the next 40 min (P<0.05; Fig. 4A). In contrast, burst frequency in time control experiments (n=12) did not change; burst frequency was 0.52 ± 0.07 bursts/min at baseline and 0.52 ± 0.07 bursts/min at the 140-min time point (P<0.05; Fig. 4A). When graphed as percent change in frequency, riluzole steadily increased frequency by 31 ± 5 % within 80-min before reaching a steady-state increase of 23-26% (P<0.05 for drug and time-dependent effects; Fig. 4B). With respect to the number of bursts/episode, riluzole had no effect and was constant at 1.4 – 1.6 bursts/episode throughout the riluzole exposure (P>0.05; Figs. 3B; 4C). For burst amplitude, riluzole produced an immediate, time-dependent decrease such that amplitude was 58 ± 4 % of baseline after 120 min (P<0.05) compared to an 8 ± 3 % decline (P>0.05) in time control experiments (Fig. 4D). To rule out the caveat that riluzole was not penetrating the brainstems and altering rhythm-generating neurons deep within the tissue, riluzole (50 µM) was applied to turtle hemi-brainstems (i.e., brainstems completely transected along the midline). In these reduced preparations (n=2), riluzole produced a similar time-dependent frequency increase and amplitude decrease (data not shown). Page 9 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 9 In contrast to riluzole, FFA produced a time- and dose-dependent decrease in burst frequency with minor decreases in amplitude (Fig. 5A). Although bath-application of 20 µM FFA (n=6) had little effect on amplitude, FFA at higher concentrations (50, 100, 500 µM; n=5, 7, 8, respectively) decreased burst frequency to new steady-state levels within 40-100 min (P<0.05; Fig. 5B). The percent decrease in frequency at 140 min post-drug application was significantly decreased compared to time controls at FFA concentrations between 50-500 µM (Fig. 5C). Similar to riluzole, FFA had no effect on the number of bursts/episode compared to time controls (Fig. 5D). With respect to amplitude, biphasic, dose-dependent effects were observed. There was no change when brainstems were exposed to 20 µM FFA (Fig. 5E). During the 50 µM exposure, however, there was a transient increase of 25 ± 30% (P<0.05) after 40 min, but amplitude was decreased by 38-47% during the next 40 min (P<0.05). For the 100 µM exposure, amplitude was decreased by 17-33% (P<0.05) during 40-80 min of the FFA application before returning to baseline. The amplitude data after 100-120 min of 50-100 µM FFA application are highly variable because only 1-3 brainstems produced respiratory bursts at a low frequency. Increased extracellular KCl increased burst frequency during synaptic inhibition blockade As demonstrated previously (23), the motor output of the turtle respiratory neural control system was transformed by blocking fast inhibitory synaptic transmission via bath application of bicuculline (50 µM) and strychnine (50 µM), resulting in rhythmic bursts that had an increased frequency and amplitude, and a rapid onset/slow decrementing shape (Fig. 6A). Time control experiments (n=14) showed that after 100 min of exposure to bicuculline and strychnine, rhythmic bursts reached a steady-state frequency of 0.86 ± 0.10 bursts/min. After another 80 min of bicuculline and strychnine, burst frequency was 0.85 ± 0.09 bursts/min and amplitude Page 10 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 10 decreased by only 6 ± 4 % (P>0.05; data not shown). Thus, increased bath KCl was applied after establishing a 20-min baseline the 100-120 min period of bicuculline and strychnine application (Fig. 6A). During synaptic inhibition blockade, burst frequency was not altered when bath KCl was increased from 2.0 mM to 5.0 mM KCl (n=8; Fig. 6B). However, within 60-min of increasing bath KCl to 7 mM (n=7) or 8 mM KCl (n=7), burst frequency increased to 320 ± 70% and 380 ± 110% of baseline, respectively (P<0.05; Fig. 6B). Decreasing bath KCl to 0.5 mM (n=8) produced a slight 12 ± 7% decrease in burst frequency. Burst frequency was correlated (P=0.005; r2=0.167) with bath KCl after 60 min of KCl exposure (i.e., 80-min time point in Fig. 6B) (Fig. 6D). In contrast, burst amplitude generally decreased with time (Fig. 6A) in a dose-dependent manner with amplitude falling by 18 ± 12% and 27 ± 8% when bath KCl was 7 mM or 8 mM KCl, respectively (P<0.05 for a time effect after 60 min of increased KCl; Fig. 6C). After 60 min of KCl exposure, burst amplitude was inversely correlated with bath KCl (P=0.012; r2=0.139; Fig. 6E). Effects of riluzole and FFA during synaptic inhibition blockade To test whether the rhythmic bursts produced during synaptic inhibition blockade were altered by riluzole or FFA, turtle brainstems were exposed to bicuculline (50 µM) and strychnine (50 µM) for 100 min and then a 20-min baseline was established. Against a background of bicuculline and strychnine, either riluzole (50 µM; n=12) or FFA (100 µM; n=7) were applied. Under these conditions, riluzole had no effect on burst frequency (Fig. 7A, 8A). In contrast, FFA decreased burst frequency in a time-dependent manner from 0.96 ± 0.10 bursts/min (baseline) to 0.36 ± 0.08 bursts/min after 60 min (P<0.05) and 0.10 ± 0.05 bursts/min after 120 min (P<0.05; Page 11 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 11 Figs. 7B, 8A). Burst amplitude was not altered by either riluzole or FFA although there were significant time effects for all the data during the last 80 min of drug exposure (Fig. 8B). DISCUSSION This is the first study to show that FFA, but not riluzole, decreases respiratory burst frequency in an adult ectothermic vertebrate using in vitro isolated brainstems. This suggests that riluzole-sensitive persistent Na+ currents in pacemaker neurons are not required for rhythm generation in aCSF or during synaptic inhibition blockade. In contrast, the finding that FFA abolished burst frequency in aCSF as well as during synaptic inhibition blockade suggests that Ca2+-activated cation currents are required for rhythm generation. Since burst frequency was not correlated with bath KCl in aCSF, it appears unlikely that pacemaker neurons with FFA-sensitive currents are required for rhythm generation. Following network transformation via synaptic inhibition blockade, the frequency of the FFA-sensitive rhythmic activity was correlated with bath KCl, suggesting that synaptic inhibition “releases” pacemaker neurons or properties which then produce rhythmic motor activity. Taken together, we hypothesize that recurrent excitatory synaptic transmission involving postsynaptic Ca2+-activated cation currents are responsible for rhythm generation in turtles (i.e., group-pacemaker model). Pacemaker properties in vertebrate respiratory rhythm generation The maturation network-burster hypothesis (43) hypothesizes that the hybrid-pacemaker network model applies during the early postnatal period in mammals because glycine-mediated synaptic inhibition is weak, respiratory neurons are relatively depolarized, and persistent Na+ currents in pacemaker neurons are active. During maturation, however, neuronal membrane Page 12 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 12 potentials are hypothesized to become more negative and thereby permit a large repertoire of ionic conductances to contribute to rhythm generation. Furthermore, synaptic inhibition is hypothesized to become necessary for rhythm generation and the obligatory role of pacemaker neurons is diminished in adult animals. In support of this model, respiratory-related pacemaker neurons are found in primarily in vitro preparations from perinatal rodents in key putative rhythm-generating sites, such as the pre-Bötzinger Complex (preBötC; 22, 53, 57) and the parafacial respiratory group (pFRG; 1, 26, 27). Consistent with the maturation network-burster hypothesis, there is little evidence for respiratory-related pacemaker neurons in adult mammals (although the precise age at which the pacemaker-driven respiratory rhythm presumably transitions to a network-driven rhythm is not well defined). Interestingly, pre-I neurons in a perfused in situ preparation from 21-42 day old mice burst rhythmically in low Ca2+/high Mg2+ solution, but it was not clear whether synaptic transmission was completely blocked (28). In amphibians, the maturation network-burster hypothesis also appears to apply since the lung rhythm persists during synaptic inhibition blockade in isolated brainstems from tadpoles, but not from adult frogs (3, 17). In contrast, evidence suggests that the maturation network-burster hypothesis may not apply in all vertebrates. For example, respiratory-related motor bursts are produced by isolated adult lamprey brainstems during bath application of picrotoxin (GABAA antagonist) and strychnine (44). Also, as shown here (Fig. 6A) and in Johnson et al., (23), rhythmic motor activity persists in isolated adult turtle brainstems, and this rhythm is abolished by well-known respiratory depressants, such as µ-opiate receptor agonists (23) and high pH/low CO2 conditions (unpublished observations). This suggests that pacemaker properties, whether they are expressed in individual neurons or as an emergent network property, may be involved in respiratory rhythm Page 13 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 13 generation in some adult vertebrates. Since respiratory-related pacemaker neurons (in mammals) have intrinsic, voltagedependent bursting properties, action potential burst frequency is directly related to membrane potential (5, 53). Thus, under in vitro conditions, increasing bath KCl would be expected to depolarize pacemaker neurons, increase action potential burst frequency in pacemaker neurons, and thereby increase respiratory burst frequency within the network. There are several caveats associated with this approach. For example, networks without pacemaker neurons have voltagedependent conductances whose properties will be altered by depolarization. Also, increased bath KCl may increase neurotransmitter release, alter intrinsic membrane properties, alter the driving force through K+-permeable conductances, or alter the function of membrane ion pumps. Although correlating bath KCl with rhythm frequency is not specific, it is useful when combined with other experimental approaches, such as synaptic inhibition blockade. For example, the tadpole lung rhythm frequency was directly related to bath KCl levels (61) and persisted during synaptic inhibition blockade (3, 17), which is consistent with the network being driven by pacemaker neurons. In contrast, the adult frog lung rhythm was not correlated with bath KCl (61) and was abolished during synaptic inhibition blockade (3, 17), which is consistent with a nonobligatory role for pacemaker neurons. In this study, the finding that burst frequency in isolated turtle brainstems was not correlated with increased bath KCl is consistent with the hypothesis that pacemaker neurons with voltage-dependent properties are not required for rhythm generation. Can the turtle respiratory rhythm generator be transformed into a pacemaker-driven network? In mammalian in vitro brainstem preparations, the “switching” hypothesis states that the respiratory network can undergo a non-physiological switch from a network-driven to a Page 14 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 14 pacemaker-driven system when there is decreased synaptic inhibition and increased extracellular K+-ions (45, 47, 48, 49, 50, 55). Under these conditions, pacemaker neurons, which are not required for normal breathing, are proposed to become rhythmically active and force respiratory neurons downstream to oscillate rhythmically (4, 45, 47, 48, 49, 50). Accordingly, the precise nature of the motor output produced by reduced perinatal mammalian preparations is part of an ongoing debate as to what constitutes normal breathing (7, 12, 29, 34, 40, 54). In contrast to mammalian preparations, turtle brainstems in vitro are highly resistant to hypoxia, and have a low metabolic rate at physiologically relevant temperatures (e.g., room temperature). Thus, turtle brainstems are unlikely to have hypoxia-induced increases in extracellular H+ and K+ levels (58, 62) that are found in some mammalian in vitro brainstemspinal cord preparations. It should be noted that turtle brainstems used in this study were at pH=7.30, which is acidic compared to the normal blood pH (~7.60) in red-eared sliders. Whether or not low pH biases the turtle respiratory rhythm generator towards pacemaker- versus networkbased function is not known. In addition, isolated turtle brainstems produce appropriate phasic expiratory and inspiratory activity similar to intact turtles, thereby negating the mammalian “eupnea versus gasping” controversy (21). Nevertheless, blockade of synaptic transmission in turtle brainstems produces a rhythm that was directly correlated with extracellular KCl levels, which is consistent with the hypothesis that turtle brainstems, in a manner similar to mammalian preparations, can undergo a switch to a pacemaker-driven network. It should be noted that this correlation was statistically significant but only 17% of the variation was due to bath KCl. One difference is that only synaptic inhibition blockade appears to be necessary to make the switch in turtles whereas increased extracellular K+ levels may also be necessary in mammals. Thus, this capacity for switching to a pacemaker-driven network may be a conserved mechanism within the Page 15 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 15 respiratory systems of mature vertebrates. Role of riluzole- and FFA-sensitive conductances in respiratory rhythm generation In mammals, the ionic conductances hypothesized to be responsible for pacemaker properties of respiratory neurons include persistent Na+ currents (5, 9, 10, 25, 47, 48), Ca2+activated cation currents (33, 57), and intracellular Ca2+ oscillations (2). For example, persistent Na+ currents are required for voltage-dependent bursting in a subpopulation of respiratory pacemaker neurons in the preBötC (10, 33). Riluzole blocks respiratory activity in rhythmically active rodent slices (37, 48) and in P21 (and older) perfused in situ murine preparations (37), suggesting that persistent Na+ currents are required for respiratory rhythm generation. In contrast, riluzole did not change the frequency of respiratory motor output produced by rhythmically active rodent slices (10) or in 6-week old perfused in situ rat preparations (30, 56), although positive controls showing the penetrance of riluzole into the tissue were not shown. These data were interpreted as consistent with the hypothesis that persistent Na+ currents are required for gasping and not eupnea (30, 56). Another interesting observation is that rhythmic motor activity (presumed to be respiratory-related) persists during simultaneous blockade of both persistent Na+ currents (with riluzole) and Ca2+-activated cation currents (with FFA), which suggests that pacemaker neurons are not required for mammalian respiratory rhythm generation (11). Although some of the contradictory results may be due to different experimental conditions, protocols, preparations, and animal strains and species, clearly more work is required to resolve these questions. Since both riluzole and FFA block currents other than persistent Na+ and Ca2+-activated cation currents, respectively, the strongest finding in this paper was that riluzole did not abolish Page 16 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 16 respiratory burst frequency in aCSF or during synaptic inhibition blockade. For example, riluzole blocks NMDA-dependent responses in frog oocytes with an IC50 = 18.2 µM (8). Riluzole appeared to penetrate the tissue rapidly because the same effects were observed in aCSF for intact and hemi-brainstems. If persistent Na+ currents are present in turtle respiratory-related neurons, these currents do not appear to be essential for rhythm generation. On the other hand, it’s possible that persistent Na+ currents are not present in turtle respiratory neurons. Since persistent Na+ currents are hypothesized to be necessary for mammalian gasping (30, 56) and turtles do not appear to gasp during 6 hr of anoxia (unpublished observations), we hypothesize that the turtle respiratory neurons express low levels, if any, of persistent Na+ channels. In contrast, FFA blocked turtle respiratory rhythm generation in a time- and dose-dependent manner with 50 µM FFA producing significant decreases in burst frequency within 40 min of exposure. One interpretation is that FFA blocked Ca2+-activated cation currents in pacemaker neurons that are required for respiratory rhythm generation and that increased bath KCl elicits offsetting complex effects on frequency so that increased bath KCl does not increase burst frequency in aCSF. If so, intracellular recordings in the absence of synaptic transmission will be required to identify and characterize candidate pacemaker neurons. On the other hand, the lack of correlation between burst frequency and bath KCl in aCSF argues against the role of pacemaker neurons. Instead, Ca2+-activated cation currents may be expressed in non-pacemaker neurons and the positive feedback interaction of this current with excitatory glutamatergic synaptic currents may underlie respiratory rhythm generation (i.e., group pacemaker model; 10, 11, 14, 15, 39). In this case, Ca2+-activated cation currents would play a critical role in initiating and maintaining inspiratory bursts. However, other interpretations are plausible because FFA blocks other membrane currents, such as transient receptor potential currents (19) and L-type Ca2+ channels Page 17 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 17 (51) that may be necessary for rhythm generation. Also, FFA may inactivate other neurons that project to the turtle rhythm generator and provide critical neurotransmitters that maintain respiratory neurons at appropriate levels of excitability. Acknowledgements This work was supported by a National Science Foundation Grant (IOB 0517302). The authors also acknowledge excellent technical assistance provided by Robert Creighton. Page 18 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 18 REFERENCES 1. Arata A, Onimaru H, and Homma I. Effects of cAMP on respiratory rhythm generation in brainstem-spinal cord preparation from newborn rat. Brain Res 605: 193-199, 1993. 2. Baker RE, Ballantyne D, Bingmann D, Jones D, and Widman G. Rhythm generation in organotypic medullary cultures of newborn rats. Int J Dev Neurosci 13: 799-809, 1995. 3. Broch L, Morales RD, Sandoval AV, and Hedrick MS. Regulation of the respiratory central pattern generator by chloride-dependent inhibition during development in the bullfrog (Rana catesbeiana). J Exp Biol 205: 1161-1169, 2002. 4. Büsselberg D, Bischoff AM, Paton JFR, and Richter DW. Reorganisation of respiratory network activity after loss of glycinergic inhibition. Pflüegers Archives 441: 444-449, 2001. 5. Butera RJ, Rinzel J, and Smith JC. Models of respiratory rhythm generation in the preBötzinger complex. I. Bursting pacemaker neurons. J Neurophysiol 82: 382-397, 1999a. 6. Butera RJ, Rinzel J, and Smith JC. Models of respiratory rhythm generation in the preBotzinger complex. II. Populations Of coupled pacemaker neurons. J Neurophysiol 82: 398415, 1999b. 7. Champagnat J. What is eupnea. Respir Physiol Neurobiol 139: 91-95, 2003. Page 19 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 8. 19 Debono MW, LeGuern J, Canton T, Doble A, Pradier L. Inhibition by riluzole of electrophysiological responses mediated by rat kainite and NMDA receptors expressed in Xenopus oocytes. Eur J Pharmacol 235: 283-289, 1993. 9. Del Negro CA, Koshiya N, Butera RJ, and Smith JC. Persistent sodium current, membrane properties and bursting behavior of pre-Botzinger complex inspiratory neurons in vitro. J Neurophysiol 88: 2242-2250, 2002a. 10. Del Negro CA, Morgado-Valle C, and Feldman J. Respiratory rhythm: an emergent network property? Neuron 34: 821-830, 2002b. 11. Del Negro CA, Morgado-Valle C, Hayes JA, Mackay DD, Pace RW, Crowder EA and Feldman JL. Sodium and calcium current-mediated pacemaker neurons and respiratory rhythm generation. J Neurosci 25: 446-453, 2005. 12. Duffin J. A commentary on eupnoea and gasping. Respir Physiol Neurobiol 139: 105-111, 2003. 13. Feldman JL and Cleland CL. Possible roles of pacemaker neurons in mammalian respiratory rhythmogenesis. In: Cellular Pacemakers, edited by Carpenter DO, NY: Wiley, 1982, p. 101-119. 14. Feldman JL and Del Negro CA. Looking for inspiration: new perspectives on respiratory Page 20 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 20 rhythm. Nat Rev Neurosci 26: 232-242, 2006. 15. Feldman JL, Mitchell GS, and Nattie EE. Breathing: rhythmicity, plasticity, chemosensitivity. Annu Rev Neurosci 26: 239-66, 2003. 16. Funk GD and Feldman JL. Generation of respiratory rhythm and pattern in mammals: insights from developmental studies. Curr Opin Neurobiol 5: 778-785, 1995. 17. Galante RJ, Kubin L, Fishman AP, and Pack AI. Role of chloride-mediated inhibition in respiratory rhythmogenesis in an in vitro brainstem of tadpole, Rana catesbeiana. J Physiol (Lond) 492: 545-558, 1996. 18. Hayashi F and Lipski J. The role of inhibitory amino acids in control of respiratory motor output in an arterially perfused rat. Respir Physiol 89:47-63, 1992 19. Hill K, Benham CD, McNulty S, and Randall AD. Flufenamic acid is a pH-dependent antagonist of TRPM2 channels. Neuropharmacology 47: 450-460, 2004. 20. Johnson SM and Creighton RJ. Riluzole decreases frequency of rhythm induced during synaptic inhibition blockade in turtle brainstems in vitro. Soc Neurosci Abstr 503.9, 2003. 21. Johnson SM and Mitchell GS. NMDA-mediated bulbospinal respiratory drive is pH/PCO2insensitive in turtle brainstem-spinal cord preparation. Respir Physiol 113: 201-212, 1998. Page 21 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 21 22. Johnson SM, Smith JC, Funk GD, and Feldman JL. Pacemaker behavior of respiratory neurons in medullary slices from neonatal rat. J Neurophysiol 72: 2598-2608, 1994. 23. Johnson SM, Wilkerson JE, Wenninger MR, Henderson DR, and Mitchell GS. Role of synaptic inhibition in turtle respiratory rhythm generation. J Physiol (Lond) 544: 253-265, 2002. 24. Koshiya N and Smith JC. Neuronal pacemaker for breathing visualized in vitro. Nature 400: 360-363, 1999. 25. McCrimmon DR, Monnier A , Ptak K, Zummo G, Zhang Z, and Alheid GF. Respiratory rhythm generation: preBötzinger neuron discharge patterns and persistent sodium current. Adv Exp Med Biol 499: 147-152, 2001. 26. Onimaru H, Arata A, and Homma I. Firing properties of respiratory rhythm generating neurons in the absence of synaptic transmission in rat medulla in vitro. Exp Brain Res 76: 530-536, 1989. 27. Onimaru H, Arata A, and Homma I. Intrinsic burst generation of preinspiratory neurons in the medulla of brainstem-spinal cord preparations isolated from newborn rats. Exp Brain Res 106: 57-68, 1995. Page 22 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 22 29. Paton JF. Rhythmic bursting of pre- and post-inspiratory neurones during central apnoea in mature mice. J Physiol (Lond) 502: 623-639, 1997. 29. Paton JFR. Defining eupnea. Respir Physiol Neurobiol 139: 89, 2003. 30. Paton JF, Abdala AP, Koizumi H, Smith JC and St-John WM. Respiratory rhythm generation during gasping depends on persistent sodium current. Nat Neurosci 9: 311-313, 2006. 31. Paton JF, Ramirez JM, and Richter DW. Mechanisms of respiratory rhythm generation change profoundly during early life in mice and rats. Neurosci Lett 170: 167-170, 1994. 32. Paton JF and Richter DW. Role of fast inhibitory synaptic mechanisms in respiratory rhythm generation in the maturing mouse. J Physiol (Lond) 484: 505-521, 1995. 33. Pena F, Parkis MA, Tryba AK, and Ramirez JM. Differential contribution of pacemaker properties to the generation of respiratory rhythms during normoxia and hypoxia. Neuron 43: 105-117, 2004. 34. Ramirez JM and Lieske SP. Commentary on the definition of eupnea and gasping. Respir Physiol Neurobiol 139: 113-119, 2003. 35. Ramirez JM and Richter DW. The neuronal mechanisms of respiratory rhythm generation. Page 23 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 23 Curr Opin Neurobiol 6: 817-825, 1996. 36. Ramirez JM, Tryba AK, and Pena F. Pacemaker neurons and neuronal networks: an integrative view. Curr Opin Neurobiol 14: 665-674, 2004. 37. Ramirez JM, Viemari JC. Determinants of inspiratory activity. Respir Physiol Neurobiol 147: 145-157, 2005. 38. Ramirez JM, Zuperku EJ, Alheid GF, Lieske SP, Ptak K, and McCrimmon DR. Respiratory rhythm generation: converging concepts from in vitro and in vivo approaches? Respir Physiol Neurobiol 131: 43-56, 2002. 39. Rekling JC and Feldman JL. PreBötzinger complex and pacemaker neurons: hypothesized site and kernel for respiratory rhythm generation. Annu Rev Physiol 60: 385-405, 1998. 40. Richter DW. Commentary on eupneic breathing patterns and gasping. Respir Physiol Neurobiol 139: 121-130, 2003. 41. Richter DW, Ballanyi K, and Ramirez JM. Respiratory rhythm generation. In: Neural Control of the Respiratory Muscles. edited by Miller AD, Bianchi AL, Bishop BP, FL: CRC Press, 1997, p. 119-130. 42. Richter DW, Ballanyi K, and Schwarzacher S. Mechanisms of respiratory rhythm Page 24 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 24 generation. Curr Opin Neurobiol 2: 788-793, 1992. 43. Richter DW and Spyer KM. Studying rhythmogenesis of breathing: comparison of in vivo and in vitro models. Trends Neurosci 24: 464-472, 2001. 44. Rovainen CM. Generation of respiratory activity by the lamprey brain exposed to picrotoxin and strychnine, and weak synaptic inhibition in motoneurons. Neuroscience 10: 875-882, 1983. 45. Rybak IA, Paton JFR, Rogers RF, and St. John WM. Generation of the respiratory rhythm: state-dependency and switching. Neurocomputing 44-46: 605-614, 2002a. 46. Rybak IA, Paton JF, and Schwaber JS. Modeling neural mechanisms for genesis of respiratory rhythm and pattern. II. Network models of the central respiratory pattern generator. J Neurophysiol 77: 2007-2026, 1997. 47. Rybak IA, Shevtsova NA, Ptak K, and McCrimmon DR. Intrinsic bursting activity in the pre-Botzinger Complex: role of persistent sodium and potassium currents. Biol Cybern 90: 59-74, 2004. 48. Rybak IA, Shevtsova NA, St. John WM, Paton JF, and Pierrefiche O. Endogenous rhythm generation in the pre-Botzinger complex and ionic currents: modelling and in vitro studies. Eur J Neurosci 18: 239-257, 2003. Page 25 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 25 49. Rybak IA, St. John WM, and Paton JFR. Models of neuronal bursting behavior: implications for in vivo versus in vitro respiratory rhythmogenesis. In: Frontiers in Modeling and Control of Breathing, edited by Poon CS, Kazemi H, NY: Kluwer Academic/Plenum Publishers, 2001, p. 159-164. 50. Rybak IA, St. John WM, and Paton JFR. Potential switch from eupnea to fictive gasping after blockade of glycine transmission and potassium channels. Am J Physiol Regul Integr Comp Physiol 283: R721-R731, 2002b. 51. Shimamura K, Zhou M, Ito Y, Kimura S, Zou LB, Sekiguchi F, Kitramura K, Sunano S. Effects of flufenamic acid on smooth muscle of the carotid artery isolated isolated from spontaneously hypertensive rats. J Smooth Muscle Res 38: 39-50, 2002. 52. Smith JC, Butera RJ, Koshiya N, Del Negro C, Wilson CG, and Johnson SM. Respiratory rhythm generation in neonatal and adult mammals: the hybrid pacemakernetwork model. Respir Physiol 122: 131-147, 2000. 53. Smith JC, Ellenberger HH, Ballanyi K, Richter DW, and Feldman JL. Pre-Botzinger complex: a brainstem region that may generate respiratory rhythm in mammals. Science 254: 726-729, 1991. 54. St. John WM and Paton JFR. Defining eupnea. Respir Physiol Neurobiol 139: 97-103, Page 26 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 26 2003. 55. St. John WM, Rybak IA, and Paton JFR. Potential switch from eupnea to fictive gasping after blockade of glycine transmission and potassium channels. Am J Physiol Regul Integr Comp Physiol 283: R721-R731, 2002. 56. St. John, WM, Waki H, Dutschmann M, Paton JFR. Maintenance of eupnea of in situ and in vivo rats following riluzole: a blocker of persistent sodium currents. Respir Physiol Neurobiol Jul 1, 2006; doi:10.1016/j.resp.2006.04.018. 57. Thoby-Brisson M and Ramirez JM. Identification of two types of inspiratory pacemaker neurons in the isolated respiratory neural network of mice. J Neurophysiol l 86: 104-112, 2001. 58. Torgerson CS, Gdovin MJ, Kogo N, and Remmers JE. Depth profiles of pH and PO2 in the in vitro brainstem preparation of the tadpole Rana catesbeiana. Respir Physiol 108: 205213, 1997. 59. Viemari JC and Ramirez JM. Norepinephrine differentially modulates different types of respiratory pacemaker and nonpacemaker neurons. J Neurophysiol 95: 2070-2082, 2006. 60. Wang D, Grillner S, and Wallen P. Effects of flufenamic acid on fictive locomotion, plateau potentials, calcium channels, and NMDA receptors in the lamprey spinal cord. Page 27 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 27 Neuropharmacology 51: 1038-1046, 2006. 61. Winmill RE and Hedrick MS. Developmental changes in the modulation of respiratory rhythm generation by extracellular K+ in the isolated bullfrog brainstem. J Neurobiol 55: 278-287, 2003. 62. Xia Y, Jiang C, and Haddad GG. Oxidative and glycolytic pathways in rat (newborn and adult) and turtle brain: role during anoxia. Am J Physiol 262: 595-603, 1992. Page 28 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 28 FIGURE LEGENDS Fig. 1. Spontaneous respiratory motor output produced by isolated in vitro turtle brainstems. A: Drawing of turtle brainstem preparation with bursts of integrated respiratory motor output recorded by a suction electrode attached to hypoglossal (XII) nerve rootlets. B: Integrated XII discharge shows several episodic (i.e., doublet) bursts and one singlet burst. Fig. 2. Respiratory burst frequency is not correlated with bath [KCl] in aCSF. A: Integrated XII discharge shows effects of increasing bath [KCl] from 2.0 to 9.0 mM KCl for a brainstem bathed in aCSF. The dotted lines show expanded traces during baseline (left) and after >30 min of increased KCl exposure (right). Arrows in the expanded traces show individual XII respiratory bursts (note that frequency and episodic discharge was not altered). Time-dependent changes in burst frequency (B) and amplitude (C) are shown in response to increasing bath [KCl] from 2.0 mM (data at 20-min time point) to 5.0, 7.0, or 9.0 mM KCl (closed circles, open circles, and open squares, respectively). B: Burst frequency increased significantly in a sustained manner after increasing bath KCl to 9.0 mM, but not after increasing to 5.0 or 7.0 mM KCl. C: Burst amplitude mostly decreased with increased bath KCl, except for 60 min after increasing to 9.0 mM KCl. D,E: Linear regression analysis of burst frequency (D) and amplitude (E) for data at the 60-min time point in Figs. 2B and 2C, respectively, shows no correlation with bath [KCl]. Asterisk = P<0.05 compared to 5 mM KCl data; # = P<0.05 compared to baseline; & = P<0.05 for time-dependent effect at that data point; † = P<0.05 for drug effect; ‡ = P<0.05 for timedependent effect. Fig. 3. Riluzole does not abolish respiratory motor output. Traces of integrated XII motor Page 29 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 29 discharge are shown for two different brainstems bathed in aCSF and exposed to riluzole (50 µM, 2 hr). Riluzole increased frequency and decreased amplitude in brainstems producing either singlet (A) or episodic (B) discharge. Fig. 4. Effects of riluzole on respiratory variables for brainstems bathed in aCSF. A: Bath-applied riluzole (50 µM, 2 hr) significantly increased burst frequency (open circles) while frequency remained constant in time control experiments (closed circles). B: When frequency data were graphed as percent change from baseline, riluzole increased frequency during the first 80 min of application until a plateau was reached. C: Riluzole did not alter the number of bursts/episode. D: Riluzole caused a steady time-dependent decrease in burst amplitude. Asterisk = P<0.05 compared to time control data; # = P<0.05 compared to baseline; † = P<0.05 for drug effect; ‡ = P<0.05 for time-dependent effect. Fig. 5. Effects of FFA on respiratory variables for brainstems bathed in aCSF. A: Integrated XII traces from a brainstem during baseline (left panel), after 1 hr of bath-applied FFA (100 µM; middle panel), and after 2 hr of bath-applied FFA (right panel). B: FFA produced a dose- and time-dependent decrease in burst frequency at FFA concentrations between 50-500 µM. C: Frequency data at the 140-min time point in Fig. 5B are graphed as percent change from baseline at each dose. D: FFA did not alter the number of bursts/episode (data at all FFA concentrations were pooled together). The numbers in parentheses above the x-axis indicate the number of brainstems producing respiratory motor output at that time-point for the FFA experiments. E: FFA (20 µM) did not alter burst amplitude compared to time controls. FFA (50 µM) produced complex time-dependent increases and decreases in amplitude, while FFA (100 µM) mainly Page 30 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 30 decreased amplitude. Asterisk = P<0.05 compared to time control data; # = P<0.05 compared to baseline; † = P<0.05 for drug effect; ‡ = P<0.05 for time-dependent effect. Fig. 6. Burst frequency is correlated with bath [KCl] during synaptic inhibition blockade. A: Integrated XII discharge for a brainstem bathed in bicuculline and strychnine (50 µM each) and bath [KCl] was increased from 2.0 to 8.0 mM. Time-dependent changes in burst frequency (B) and amplitude (C) in response to increasing bath [KCl] from 2.0 mM (data at 20-min time point) to 5.0, 7.0, or 8.0 mM KCl (closed circles, open circles, and open squares, respectively). B: Burst frequency increased significantly in a sustained manner for the 7.0 and 8.0 mM KCl data; there was no change when bath KCl was increased to 5.0 mM. C: Burst amplitude decreased with time during prolonged KCl application. D,E: Linear regression analysis of burst frequency (D) and amplitude (E) for data at the 60-min time point in Figs. 6B and 6C, respectively (i.e., 20-40 min during KCl exposure). D: Burst frequency was directly correlated with bath [KCl], while burst amplitude (E) was indirectly correlated with bath [KCl]. & = P<0.05 for time-dependent effect at that data point; ‡ = P<0.05 for time-dependent effect. Fig. 7. FFA, but not riluzole, alters rhythmic motor bursts during synaptic inhibition blockade (i.e., bicuculline and strychnine at 50 µM each). (A) Riluzole (50 µM) did not alter burst amplitude or frequency; (B) FFA (100 µM) decreased burst frequency with little change in amplitude. Fig. 8. Effects of riluzole and FFA during synaptic inhibition blockade. A: While bathed in bicuculline and strychnine (50 µM each), riluzole (50 µM; open circles) did not alter burst Page 31 of 39 Pacemaker properties and turtle breathing in vitro Johnson, Weigel, Majewski 31 frequency compared to time controls (closed circles). In contrast, FFA (100 µM; open squares) decreased burst frequency in a time-dependent manner. B: Riluzole and FFA did not alter burst amplitude compared to time controls, but there was a time-dependent decrease for all the data. Asterisk = P<0.05 compared to time controls; # = P<0.05 compared to baseline; & = P<0.05 for time-dependent effect at that data point; † = P<0.05 for drug effect; ‡ = P<0.05 for timedependent effect. Page 32 of 39 Fig. 1 A V VI VII VIII B IX X XII C1 XII 60 s Page 33 of 39 Fig. 2 9 mM KCl A 5 min 60 s C KCl 400 5 mM KCl 7 mM KCl 9 mM KCl 300 & †‡ 200 100 0 0 Frequency change (% of baseline) D 20 40 60 Amplitude (normalized) 500 Time (min) E 600 P = 0.066 r 2= 0.129 200 0 0 2 4 6 Bath [KCl] (mM) 8 * 1.2 1.0 10 ‡ 0.8 5 mM KCl 7 mM KCl 9 mM KCl 0.6 # & 0.4 0 800 400 KCl 1.4 80 Amplitude (normalized) Frequency change (% of baseline) B 60 s 20 40 60 80 Time (min) 1.5 P = 0.295 r 2= 0.044 1.0 0.5 0.0 0 2 4 6 Bath [KCl] (mM) 8 10 Page 34 of 39 Fig. 3 baseline A Riluzole (50 µM, 2 hr) XII 0 B XII 60 s Page 35 of 39 Fig. 4 riluzole 1.2 control riluzole 1.0 * * * 0.8 B #* #* #* #* †‡ 0.6 0.4 0.2 0.0 20 40 60 130 # # # †‡ 120 110 100 80 100 120 140 riluzole 2.0 1.5 1.0 0.5 control riluzole 0.0 0 20 40 0 D Amplitude (normalized) Bursts/eposide # control riluzole 90 0 C riluzole 140 Frequency change (% of baseline) Frequency (bursts/min) A 80 100 120 140 Time (min) 40 60 80 100 120 140 riluzole 1.2 1.0 #* # * 0.8 0.6 0.4 0.2 #* #* # * #* control riluzole 0.0 60 20 0 20 40 60 80 100 120 140 Time (min) †‡ Page 36 of 39 Fig. 5 FFA (100 µM, 1 hr) baseline A FFA (100 µM, 2 hr) XII B 1.0 Frequency (bursts/min) 1 min 0.8 FFA C Percent change in frequency at 140 min post-FFA 100 0.6 80 # 0.4 control 20 µM 50 µM 100 µM 500 µM 0.2 0.0 0 # # # * #* #* 20 40 #* 60 80 60 #* * #* 100 #* #* #* #* #* #* 120 †‡ 1.5 1.0 control FFA 0.5 0.0 0 (26) (26) 20 40 (17) (14) 60 80 (12) (10) (10) 100 120 140 Time (min) Amplitude (normalized) Bursts/episode E FFA 2.0 * 20 0 140 Time (min) D * 40 * ol µM µM µM µM ntr co 20 50 100 500 FFA 1.5 * 1.0 * 0.5 control 20 µM 50 µM 100 µM 0.0 0 20 40 60 * * * * 80 100 120 140 Time (min) Page 37 of 39 Fig. 6 8 mM KCl A 5 min KCl 500 5 mM KCl 7 mM KCl 9 mM KCl 400 C & & & 300 ‡ 200 100 0 0 40 60 Time (min) P = 0.005 r 2= 0.167 600 400 200 0 0 2 4 6 [KCl] (mM) 1.2 8 & 1.0 ‡ 0.8 5 mM KCl 7 mM KCl 9 mM KCl 0.6 0.4 0 E 800 KCl 1.4 80 Amplitude (normalized) Frequency change (% of baseline) D 20 Amplitude (normalized) Frequency change (% of baseline) B 10 20 40 60 80 Time (min) 1.5 P = 0.012 r 2= 0.139 1.0 0.5 0.0 0 2 4 6 [KCl] (mM) 8 10 Page 38 of 39 Fig. 7 A Bicuculline/Strychnine Bicuculline/Strychnine + Riluzole (50 µM) B Bicuculline/Strychnine Bicuculline/Strychnine + FFA (100 µM) 60 s Page 39 of 39 Fig. 8 A Frequency (bursts/min) Bic/Strych Bic/Strych + Riluzole (50 µM) Bic/Strych + FFA (100 µM) Riluzole or FFA 1.4 1.2 1.0 0.8 #* 0.6 0.4 0.2 Amplitude (normalized) #* #* #* † ‡ 120 140 0.0 0 B #* 20 40 60 80 100 Riluzole or FFA 1.4 1.2 1.0 0.8 & 0.6 ‡ & & & & 0.4 0.2 0.0 0 20 40 60 80 Time (min) 100 120 140