Motoneuron Excitability and Muscle Spasms Are Regulated by 5-HT 2B and 5-HT2C Receptor Activity Katherine C. Murray, Marilee J. Stephens, Edmund W. Ballou, Charles J. Heckman and David J. Bennett J Neurophysiol 105:731-748, 2011. First published 27 October 2010; doi:10.1152/jn.00774.2010

You might find this additional info useful...

This article cites 104 articles, 38 of which can be accessed free at: http://jn.physiology.org/content/105/2/731.full.html#ref-list-1 Updated information and services including high resolution figures, can be found at: http://jn.physiology.org/content/105/2/731.full.html Additional material and information about Journal of Neurophysiology can be found at: http://www.the-aps.org/publications/jn

This infomation is current as of March 8, 2011. Downloaded from jn.physiology.org on March 8, 2011

Journal of Neurophysiology publishes original articles on the function of the nervous system. It is published 12 times a year (monthly) by the American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2011 by the American Physiological Society. ISSN: 0022-3077, ESSN: 1522-1598. Visit our website at http://www.the-aps.org/. J Neurophysiol 105: 731–748, 2011. First published October 27, 2010; doi:10.1152/jn.00774.2010.

Motoneuron Excitability and Muscle Spasms Are Regulated by 5-HT2B and 5-HT2C Receptor Activity

Katherine C. Murray,1 Marilee J. Stephens,1 Edmund W. Ballou,2 Charles J. Heckman,2 and David J. Bennett1 1Centre for Neuroscience, University of Alberta, Edmonton, Alberta, Canada; and 2Department of Physiology, Northwestern University, Chicago, Illinois

Submitted 8 September 2010; accepted in final form 27 October 2010

Murray KC, Stephens MJ, Ballou EW, Heckman CJ, Bennett DJ. that can disrupt residual motor function, cause debilitating Motoneuron excitability and muscle spasms are regulated by 5-HT2B pain, and interrupt sleep (Maynard et al. 1990). Treatment of J Neurophysiol and 5-HT2C receptor activity. 105: 731–748, 2011. spasms with conventional antispastic drugs such as baclofen is First published October 27, 2010; doi:10.1152/jn.00774.2010. Imme- diately after spinal cord injury (SCI), a devastating paralysis results often not adequate or not tolerated because of adverse side from the loss of brain stem and cortical innervation of spinal neurons effects such as lethargy and weakness (Dario and Tomei 2004). that control movement, including a loss of serotonergic (5-HT) inner- Here we explore whether activity in one or more of the many vation of motoneurons. Over time, motoneurons recover from dener- 5-HT receptor subtypes facilitates spasms and examine Downloaded from vation and function autonomously, exhibiting large persistent calcium whether selectively blocking these receptors may serve as a currents (Ca PICs) that both help with functional recovery and target for novel antispastic treatment. contribute to uncontrolled muscle spasms. Here we systematically The neuronal mechanisms underlying spasms have recently evaluated which 5-HT receptor subtypes influence PICs and spasms after injury. Spasms were quantified by recording the long-lasting been identified in a rat model of SCI (Bennett et al. 2004; reflexes (LLRs) on ventral roots in response to dorsal root stimulation, Harvey et al. 2006b; Li et al. 2004a; Murray et al. 2010), and in the chronic spinal rat, in vitro. Ca PICs were quantified by similar mechanisms have been shown to underlie spasms in intracellular recording in synaptically isolated motoneurons. Application humans (Gorassini et al. 2004; Norton et al. 2008). Briefly, jn.physiology.org of agonists selective to 5-HT2B and 5-HT2C receptors (including these studies have shown that, immediately after SCI, mo- BW723C86) significantly increased the LLRs and associated Ca PICs, toneuron excitability plummets; in particular, there is a loss of whereas application of agonists to 5-HT , 5-HT , 5-HT , or 5-HT 1 2A 3 4/5/6/7 low voltage-activated persistent calcium and sodium currents receptors (e.g., 8-OH-DPAT) did not. The 5-HT2 receptor agonist– induced increases in LLRs were dose dependent, with doses for 50% (persistent inward currents; Ca and Na PICs) that are normally effects (EC ) highly correlated with published doses for agonist crucial for enabling motoneurons to produce sustained mo-

50 on March 8, 2011 receptor binding (Ki) at 5-HT2B and 5-HT2C receptors. Application of toneuron discharges in response to synaptic inputs (Heckman selective antagonists to 5-HT2B (e.g., RS127445) and 5-HT2C et al. 2005; Hultborn et al. 2004; Li et al. 2004a). Many (SB242084) receptors inhibited the agonist-induced increase in LLR. motoneurons are acutely so impaired that they can only fire However, antagonists that are known to specifically be neutral antag- transiently and can no longer follow synaptic inputs with onists at 5-HT2B/C receptors (e.g., RS127445) had no effect when given by themselves, indicating that these receptors were not activated appropriate firing (Harvey et al. 2006c). However, over the by residual 5-HT in the spinal cord. In contrast, inverse agonists (such months after injury (chronic injury), PICs spontaneously re- as SB206553) that block constitutive activity at 5-HT2B or 5-HT2C turn, attaining high levels likely only seen in alert awake receptors markedly reduced the LLRs, indicating the presence of animals (Button et al. 2008; Harvey et al. 2006c; Hultborn et constitutive activity in these receptors. 5-HT2B or 5-HT2C receptors al. 2004; Li et al. 2004b). Theses PICs are crucial for the were confirmed to be on motoneurons by immunolabeling. In sum- recovery of motoneuron firing ability, enabling motoneurons to mary, 5-HT and 5-HT receptors on motoneurons become consti- 2B 2C produce sustained firing and associated muscle contractions in tutively active after injury and ultimately contribute to recovery of motoneuron function and emergence of spasms. response to transient inputs (Li et al. 2004a). However, these PICs are permanently facilitated and, whenever these voltage- dependent currents are activated by a brief synaptic depolar- INTRODUCTION ization, they are difficult to terminate because of the loss of After spinal cord injury (SCI), debilitating spasms often normal descending control of inhibitory inputs to motoneurons develop in the muscles innervated from the spinal cord below with SCI (Holstege and Bongers 1991; Jankowska and Ham- the injury, and this is the hallmark of the general spasticity mar 2002; Nielsen et al. 2007; Rekling et al. 2000). Thus, syndrome after SCI (Bennett et al. 2004; Gorassini et al. 2004; activity in PICs causes uncontrolled motoneuron firing and Kawamura et al. 1989; Kuhn and Macht 1948). Spasms involve associated muscle spasms (Bennett et al. 2004; Li et al. 2004a). simultaneous contractions of many muscles, lasting anywhere To make matters worse, after SCI, there is a general loss of from seconds to minutes, and are triggered by brief innocuous inhibition over sensory afferent transmission (disinhibition) sensory stimuli, such as cutaneous inputs. Upward of 80% of (Hochman et al. 2001; Lundberg 1982; Schmidt and Jordan spinal cord–injured humans have spasms and general spasticity 2000), leading to unusually large and prolonged polysynaptic excitatory postsynaptic potentials (EPSPs) on motoneurons Address for reprint requests and other correspondence: D. J. Bennett, (Baker and Chandler 1987; Bennett et al. 2004; Li et al. 5005-A Katz, Ctr. for Neuroscience, Univ. of Alberta, Edmonton, Alberta T6G 2004a). These EPSPs readily trigger the large PICs in chronic 2E1, Canada (E-mail: [email protected]). SCI, which in turn produce many seconds of uncontrolled www.jn.org 0022-3077/11 Copyright © 2011 The American Physiological Society 731 732 MURRAY, STEPHENS, BALLOU, HECKMAN, AND BENNETT motoneuron firing. Thus spasms are readily triggered by tran- 2004). All recordings were made from the whole sacrocaudal spinal sient sensory stimulation (Bennett et al. 2004). cord that was removed from the rat with an S2 sacral transection and Brain stem–derived serotonin (5-HT) plays a critical role in maintained in vitro (Li et al. 2004b). This transection was made just regulating normal levels of spinal cord excitability by several rostral to the chronic spinal injury, to not further damage the sacro- receptors (Hochman et al. 2001; Schmidt and Jordan 2000), caudal cord. All experimental procedures were approved by the University of Alberta Animal Care and Use Committee: Health including facilitating Ca PICs on motoneurons by activating Sciences. 5-HT2 receptors (Harvey et al. 2006a; Heckman et al. 2005; Perrier and Delgado-Lezama 2005) and by inhibiting sensory In vitro preparation afferent transmission by activating 5-HT1 receptors (Hochman et al. 2001; Millan 2002; Schmidt and Jordan 2000; Yoshimura Details of the in vitro experimental procedures have been described and Furue 2006). Thus, the immediate loss of motoneuron in previous publications (Harvey et al. 2006c; Li et al. 2004a,b; excitability and emergence of large EPSPs acutely after spinal Murray et al. 2010). Briefly, all the rats were anesthetized with transection is partly explained by a loss of 5-HT. The sponta- urethane (0.18 g/100 g; with a maximum dose of 0.45 g), and the neous recovery of motoneuron excitability (e.g., PICs) over the sacrocaudal spinal cord was removed and transferred to a dissection months after injury (chronic spinal rat) is more difficult to chamber containing modified artificial cerebrospinal fluid (mACSF). Spinal roots were removed, except the sacral S and caudal Ca explain, but recently Harvey et al. (2006b) suggested that it 4 1 ventral roots and the Ca1 dorsal roots. After 1.5 h in the dissection involves re-activation of 5-HT2 receptors. Furthermore, Mur- chamber (at 20°C), the cord was transferred to a recording chamber ray et al. (2010) have recently shown that this activation of the containing normal ACSF (nACSF) maintained near 24°C, with a flow 5-HT2 receptors occurs in the absence of 5-HT after injury; that rate Ͼ5 ml/min. A 1-h period in nACSF was given to wash out the is, the receptors become constitutively active, as described in residual anesthetic and mACSF before recording, at which time the Downloaded from other systems (Aloyo et al. 2009; Seifert and Wenzel-Seifert nACSF was recycled in a closed system with a peristaltic pump 2002). These two studies do not clearly define which specific (Harvey et al. 2006b). 5-HT2 receptor is involved, although both suggest it could be the 5-HT2C receptor, and Harvey (2006b) speculated that Ventral root reflex recording and averaging 5-HT2A receptors may be involved as well. Little is known about the role of other 5-HT receptors in motoneuron function Dorsal and ventral roots were mounted on silver-silver-chloride

wires above the nASCF of the recording chamber and covered with a jn.physiology.org and spasms after injury. Furthermore, because mostly nonse- 1:1 mixture of petroleum jelly and mineral oil (as for intracellular lective pharmacological agents (e.g., 5-HT2A/2B/2C receptor recording) for monopolar stimulation and recording (Li et al. 2004b). agonist DOI) have been used to study 5-HT modulation of We evoked ventral root reflexes with a low-threshold Ca1 dorsal root normal motoneuron function, it even remains unclear which stimulation (single pulse, 0.1 ms, 0.02 mA, corresponding to 3x receptors are specifically involved in normal animals, although afferent threshold, 3ϫT; afferent and reflex threshold are similar; these seem to include the Gq-protein coupled 5-HT2A, 5-HT2B, Bennett et al. 2004) using a constant current stimulator (Isoflex, and 5-HT receptors (Elliott and Wallis 1992; Harvey et al. Israel). This stimulation intensity (3ϫT) is compatible with activation on March 8, 2011 2C ␤ 2006a; Holohean and Hackman 2004; Perrier and Hounsgaard of low-threshold group I and II (A ) afferents. Because the Ca1 dorsal root innervates the distal third of the tail, which lacks large muscles 2003; Wang and Dun 1990), Gi coupled 5-HT1A receptors (Holohean et al. 1990; Perrier et al. 2003; Wang and Dun (Bennett et al. 2004), this stimulation activates largely cutaneous or joint afferents. However, there are small intrinsic muscles in the tail 1990), Gs-coupled 5-HT receptors like 5-HT7 receptors that with group Ia and II muscle afferents (Steg 1964), and thus to a lesser increase cAMP (Inoue et al. 2002; Larkman and Kelly 1997), extent, muscle afferents may be activated. The stimulation was re- and possibly ionotropic 5-HT3 receptors (Ziskind-Conhaim et peated five times at 10-s intervals for each trial. The ventral root al. 1993). recordings were amplified (ϫ2,000), high-pass filtered at 100 Hz, In this study, we use numerous 5-HT agonists and highly low-pass filtered at 3 kHz, and recorded with a data-acquisition selective antagonists to determine which receptors facilitate system sampling at 6.7 kHz (Axonscope 8, Axon Instruments). Ven- motoneuron function (Ca PICs) and associated long-lasting tral root reflexes were quantified using custom written software reflexes (spasms) after SCI, by correlating agonist potency at (Matlab, MathWorks, Natick, MA). That is, data were high pass facilitating the spasms with the published binding affinity of filtered at 800 Hz and rectified to allow averaging (Fig. 1A, bottom). these agonists to all 5-HT receptor types (Barnes and Sharp We quantified the long-latency, long-lasting reflex (LLR) by averag- ing the rectified response 500–4,000 ms after stimulus, a period 1999; Boess and Martin 1994). We determine whether spon- where the response is mainly determined by the motoneuron Ca PIC taneous activity in these receptors contributes to recovery of activity and not by stimulus evoked synaptic inputs, which end by PICs and emergence of spasms and examine whether this about 500 ms. Average ventral root activity was averaged for all five activity results from either residual 5-HT in the caudal spinal stimuli in a trial. This recording procedure was repeated at 12-min cord or constitutive receptor activity (Murray et al. 2010; intervals, and 5-HT receptor agonists were added immediately after Seifert and Wenzel-Seifert 2002). each recording, giving them time to fully act by the next recording session. Cumulative dose–response relations were computed by in- creasing agonist doses at these 12-min intervals (0.003, 0.01, 0.03. METHODS 0.1,..., 30 ␮M doses used). Antagonists took longer to act, and responses reached near steady state typically Ͼ20 min after applica- Recordings were made from motoneurons and associated ventral tion, at which time responses were averaged. The effect of agonists on roots of the sacrocaudal spinal cord of spastic adult rats with chronic the reflexes were reversible on washout of the agonist, but full SCI (3.5–5 mo old) (Bennett et al. 2004). Adult female rats were recovery to baseline only occurred after several hours, likely because transected at the S2 sacral level at about 2 mo of age (adult rat), and of the large size of the whole cord preparation. Thus washout of recordings were made after their affected muscles became spastic agonists was not feasible between doses of the agonists used in the (1.5–3 mo after injury), as detailed previously (Bennett et al. 1999, construction of dose–response relations.

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 5-HT2B RECEPTORS ON MOTONEURONS 733

Ventral root recording M KCl. Electrodes were beveled down from an initial resistance of A 40–80 to 26–32 M⍀ using a rotary beveller (Sutter BV-10). A LLR mediated by PIC stepper-motor micromanipulator (660, Kopf) was used to advance the electrodes through the ventral cord surface into motoneurons. Large 30-␮m forward and backward steps were initially required to breach the white matter (about 100 ␮m thick). Multiple tracks were made in the gray matter with 2-␮m steps, without coming completely out of the white matter, by moving the electrode sideways in the white matter before each new tract (to further weaken pia and improve stability). Penetrations were made with capacitance-over-compen- sation ringing. After penetration, motoneuron identification was LLR analysis window made with antidromic ventral root stimulation, and noting ventral horn location, input resistance and time constant (Ͼ6 ms for motoneu- rons) (Li and Bennett 2007; Li et al. 2007). Cells were always held ϽϪ + 15 uM isradipine below rest ( 70 mV) with a small continuous bias current to assure B that the Ca PICs were not even partly activated before testing, because the Ca PIC slowly deactivates if it is activated for a few minutes and Reflex not does not recover from this deactivation for many minutes (unpub- mediated by PIC lished data). Data were collected with an Axoclamp 2b intracellular amplifier (Axon Instruments, Burlingame, CA) running in discontin- 0.05 mV uous current clamp (DCC, switching rate 4–6 kHz, output bandwidth

3.0 kHz, sample rate of 6.7 kHz) or discontinuous single-electrode Downloaded from voltage clamp (SEVC; gain, 0.8–2.5 nA/mV) modes. 1 sec Slow triangular voltage ramps (3.5 mV/s voltage clamp) were applied to the motoneurons to measure their electrical properties, as detailed previously (Harvey et al. 2006c; Li et al. 2004a). The input

resistance (Rm) was measured during the voltage ramps over a 5-mV range near rest and subthreshold to PIC onset. Resting potential (Vm)

was recorded with 0-nA bias current, after the cell had been given jn.physiology.org C Motoneuron recording about 15 min to stabilize after penetration. The slow triangular voltage ramps were used to directly measure the PICs as follows (shown in Fig. 5A). During the upward portion of this ramp, the current response initially increased linearly with voltage in response to the passive leak conductance. A linear relation was fit in the region just below the PIC onset (5 mV below) and extrapolated to the whole range of the ramp

(leak current; thin line in Fig. 5A). At depolarized potentials above the on March 8, 2011 LLR and plateau PIC onset threshold, there was a downward deviation from the mediated by PIC extrapolated leak current, and the PIC was estimated as the difference between the leak current and the total current (initial peak current at arrow in Fig. 5A; leak-subtracted current). The onset voltage for the

PIC (Von) was defined as the voltage at which the slope of the current -72 mV response initially reached zero (Li and Bennett 2003). Finally, the excitatory postsynaptic potential (EPSP) and associated reflexes were

measured in motoneurons by stimulating the Ca1 dorsal roots (at 3xT, as in ventral root reflex recording) while applying hyperpolarizing bias currents to block the PICs (Fig. 1C, bottom). EPSP 20 mV -81 mV Drugs and solutions Dorsal root stim (3xT) 1 sec Two kinds of ACSF were used in these experiments: mACSF in the dissection chamber before recording and nACSF in the recording FIG. 1. Long-lasting reflexes (LLRs), the counterpart of muscle spasms, are chamber. The mACSF was composed of (in mM) 118 NaCl, 24 mediated by the persistent calcium currents (Ca PICs) in motoneurons in chronic spinal rats. A: LLR triggered by dorsal root stimulation (0.1 ms pulse, NaHCO3, 1.5 CaCl2, 3 KCl, 5 MgCl2, 1.4 NaH2PO4, 1.3 MgSO4,25 3xT) and recorded from the ventral roots (top). LLR quantified by rectifying D-glucose, and 1 kynurenic acid. nACSF was composed of (in mM) the ventral root activity (bottom) and averaging over window indicated. 122 NaCl, 24 NaHCO3, 2.5 CaCl2, 3 KCl, 1 MgCl2, and 12 D-glucose. 2ϩ B: elimination of LLR after blocking the L-type Ca channel with isradipine Both types of ACSF were saturated with 95% O2-5% CO2 and main- (15 ␮M). C: PIC-mediated plateau potential and sustained firing (LLR) evoked tained at pH 7.4. The drugs added to the ACSF were 5-HT, DOI[(Ϫ)(R)- by dorsal root stimulation (3xT) in motoneuron at rest (without injected DOI hydrochloride] (from Sigma-Aldrich), 2-methyl-5-HT, 8-OH- current; top). With a hyperpolarizing bias current to prevent PIC activation, the DPAT, AC90179, ␣-methyl-5-HT, BW723C86, clozapine, cisapride, cy- same stimulation only evoked polysynaptic excitatory postsynaptic potentials proheptadine, EMD386088, granisetron, isradipine, ketanserin, LP44, (EPSPs) that ended within 0.5 s (in gray shaded region), with no synaptic LY344864, methylergonovine, methysergide, MK212, prazosin, activity during the subsequent period where the LLR was computed (bottom). RS102221, RS127445, SB204741, SB206553, SB224289, SB242084, Intracellular recording tryptamine (all from Tocris), TTX (TTX-citrate; Alomone), and zolmi- triptan (kindly donated by Astra Zeneca). All drugs were first dis- Sharp intracellular electrodes were made from glass capillary tubes solved as a 10–50 mM stock in water before final dilution in ACSF, (1.5 mm OD, Warner GC 150F-10) using a Sutter P-87 micropipette with the exception of BW723C86, RS102221, RS127445, cisapride, puller and filled with a combination of 1 M potassium acetate and 1 clozapine, SB204741, and prazosin, which were dissolved in minimal

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 734 MURRAY, STEPHENS, BALLOU, HECKMAN, AND BENNETT

amounts of DMSO (final concentration in ACSF Ͻ0.04%; by itself, different drugs for a given receptor were taken from large studies or DMSO had no effect on the LLR in vehicle controls). During summary reviews (Boess and Martin 1994), usually using isolated intracellular recording, we usually first blocked the sodium currents cloned receptors. Also, high affinity agonist-preferring binding sites (fast and persistent) and synaptic transmission with 2 ␮M TTX and were always used, measured with radioactive agonists (usually [3H]- then examined the Ca PIC sensitivity to 5-HT ligands. The Ca PIC 5-HT), rather than radioactive antagonists that bind to a low-affinity

was verified to be L-type calcium channel mediated by its blockage by site (Egan et al. 2000; Knight et al. 2004). If rat receptor Ki values dihydropiridines (nimodipine or isradipine; 15 ␮M). were not available, human values were used instead, because these are similar for most receptors (Boess and Martin 1994). Immunolabeling DOI is one of the few 5-HT2 agonists that is very selective to only 5-HT2 receptors (it activates just 5-HT2A/2B/2C) (Boess and Martin Rats were killed with Euthanyl (Bimeda-MTC; 700 mg/kg) and 1994), and accordingly, its dose–response relation was a simple perfused intracardially with 100 ml saline followed by 400 ml 2% sigmoidal relation. In contrast, other agonists such as ␣-methyl-5-HT paraformaldehyde (PFA; in phosphate buffer at room temperature; bind nonselectively to other 5-HT receptors, although with much less over about 15 min). Spinal cords were postfixed in PFA for 90 min at affinity than to 5-HT2 receptors (Table 1). With these drugs, the 4°C, cryoprotected in 20% sucrose and 2% ethylene-glycol, frozen, dose–response relation was mostly sigmoidal, until higher doses were and cut on a cryostat in horizontal or transverse 25- to 40-␮m sections. reached, at which point the LLR tended to decrease with increasing Spinal cord sections were incubated overnight at 4°C with primary dose, consistent with a nonselective action on other receptors. To

antibodies: goat anti-5-HT2B receptors (1:500, Santa Cruz SC-15081, minimize the effect of this nonselective receptor activation, only the Santa Cruz) and mouse anti-SMI32 (1:1,000, Convance SMI-32R) in main part of the dose–response relation with increasing responses to

PBS. This specificity of this 5-HT2B receptor antibody has previously increasing dose (over 2 decades) was fit to the sigmoidal dose– been shown (Moran et al. 2008; Tadros et al. 2007). Antibody labeling response relation to obtain an estimate of the EC50. We used a Hill Downloaded from was visualized with fluorescent secondary antibodies (5-HT2B: don- slope of 1.0 (see above) to force the sigmoid to have a fixed width of key anti-goat conjugated with Alexa488, 1:400, Invitrogen A11055; about 2 decades (2 orders of magnitude in dose, as is expected for the SMI32: donkey anti-mouse conjugated with Alexa568, 1:500, Invit- response to a single receptor). This approach caused the low dose– rogen A11037) for 90 min at room temperature. Image acquisition responses of agonists (e.g., selective 5-HT2B agonists) to dominate the used a Zeiss LSM 510 laser-scanning confocal microscope: Plan- sigmoidal fit where they are most selective and assured that the EC50 Apochromat 63ϫ 1.4 oil objective, BP 505–520, BP 560–615, HFT was not affected (e.g., underestimated) by the nonselective action of 488/543 with lasers at 488 and 543 nm, acquisitions in multi-track other 5-HT receptors that may have affected (e.g., inhibited) the LLR ␮ mode, optical slice thickness 0.14 m. Controls in which the primary at high doses. jn.physiology.org antibody was preabsorbed with the antigen that it was raised against were used to verify the selectivity of the antibody labeling. Also, RESULTS controls in which the primary antibody was omitted were used to confirm that the secondary antibody produced no labeling by itself. Use of LLRs as a measure of Ca PICs and spasms in SMI32 labels nonphosphorylated neurofilament H and does not chronic spinal rats change with SCI (Anelli et al. 2007). We have previously shown that SMI32 labels motoneurons and not interneurons in the ventral horn, When the dorsal roots of the in vitro sacrocaudal spinal cord on March 8, 2011 although it also labels large cells elsewhere in the brain. Thus SMI32 were stimulated with a low-threshold current (3ϫT), a long- provides a fairly selective marker of motoneurons in the ventral horn lasting reflex (LLR; 2- to 10-s duration) was evoked in mo- (Anelli et al. 2007), and in particular, labels the intracellular space (neurofilament). toneurons, as recorded both from ventral roots (Fig. 1A) and from single motoneurons measured intracellularly (Fig. 1C, top). The many-second-long portion of this LLR is mediated Data analysis by a large persistent inward calcium current (Ca PIC) that Data were analyzed in Clampfit 8.0 (Axon Instruments) and Sig- activates a plateau potential and associated sustained firing in maplot (Jandel Scientific) and expressed as mean Ϯ SD. A Student’s motoneurons of chronic spinal rats (Li et al. 2004a), which we t-test was used to test for statistical differences before and after drug verified with two methods. First, directly blocking the Ca PIC applications, with a significance level of P Ͻ 0.05. A Kolmogorov- (with isradipine or nimodipine; 15 ␮M) largely eliminated the Smirnov test for normality was applied to each data set, with a P ϭ LLR (significantly reduced by 83.9 Ϯ 7.9%, n ϭ 9, P Ͻ 0.05; 0.05 level set for significance. Most data sets were found to be Fig. 1B), and second, hyperpolarizing motoneurons to prevent normally distributed, as is required for a t-test. For those that were not activation of the voltage-dependent Ca PIC eliminated the LLR normal, a Wilcoxon signed rank test was instead used with a signif- Ͻ and associated PIC-mediated plateau (Fig. 1C, bottom; n ϭ icance level of P 0.05. ϳ Standard sigmoidal curves were fit to the relation between agonist 10/10). In both cases, eliminating the Ca PIC left an 0.5-s- dose and reflex responses, with doses expressed in log units and with long response that was caused by a polysynaptic EPSP in the

a Hill slope of unity. The dose that produced 50% effect (EC50) was motoneurons (Fig. 1, B and C, bottom). For this study, we measured from the curve, and –log(EC50) was used to quantify the focused on the Ca PIC and use the LLR reflex measured at ϭ drug potency: pEC50 –log(EC50). Also, the maximum drug-induced times Ͼ0.5 s as an indirect estimate of this Ca PIC on response (efficacy) was computed from the curve (peak of curve). For motoneurons (Fig. 1A, bottom; rectified response averaged in a comparison to our computed potencies (pEC50), the binding affinity of 0.5- to 4-s window). This enabled us to efficiently screen many each drug at the rat 5-HT receptors was also reported, with values 5-HT drugs and doses with simple ventral root, rather than taken from the literature (Table 1). The binding of an agonist to a intracellular motoneuron, recordings. Previous work has shown receptor is expressed in terms of its Ki value (nM), which corresponds to the dose that produces 50% binding to that receptor (Knight et al. that the LLR we record in vitro also occurs in vivo in the awake 2004). This is typically measured by the agonist’s ability to displace chronic spinal rat, where it produces muscle spasms (Bennett et a standard radio-labeled ligand, such as [3H]-5-HT, from the receptor al. 2004; Murray et al. 2010). Thus the LLR recorded in vitro ϭ expressed in isolated cells. Binding affinity is computed as pKi is also a convenient assay of spasms, and we used it to assess –log(Ki) (Knight et al. 2004). When possible, binding affinities of which 5-HT receptors modulate spasms.

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 5-HT2B RECEPTORS ON MOTONEURONS 735

TABLE 1. 5-HT receptor agonists and their receptor binding affinity

Receptor Agonist Ki (nM) pKi Receptor Agonist Ki (nM) pKi

5-HT1A 5-HT 1.65 8.78Hoy 5-HT2B 2-Methyl-5-HT 316.23 6.50Boe 8-OH-DPAT 0.97 9.01Boe 5-HT 10.23 7.99Boe LP44 52.48 7.28Leo ␣-Methyl-5-HT 10.47 7.98Boe BW723C86 12.59 7.90Bax DOI 27.54 7.56Boe 5-HT1B 5-HT 24.49 7.61Hoy Methylergon 0.50 9.30Kni BW723C86 125.89 6.60Bax Tryptamine 112.20 6.95Boe Zolmitriptan 5.01 8.30Mar EMD386088 179.88 6.75Mat 5-HT2C 5-HT 10.99 7.96Ega ␣-Methyl-5-HT 2.69 8.57Kni BW723C86 125.89 6.90Bax 5-HT1D 5-HT 2.51 8.60Boe DOI 9.30 8.03Ega BW723C86 125.89 6.30Bax MK212 97.72 7.01Kni Zolmitriptan 0.63 9.20Mar Methylergon 4.57 8.34Kni EMD386088 109.90 6.96Mat Tryptamine 62.0 7.21Ega

5-HT1E 5-HT 6.16 8.21Boe 5-HT3 2-Methyl-5-HT 85.11 7.07Mil ␣-Methyl-5-HT 120.22 6.92Boe BW723C86 316.23 6.50Ken Methylergon 89.12 7.05Boe EMD386088 34.00 7.47Mat

MK212 29.00 7.54Gle Downloaded from 5-HT1F 5-HT 67.60 7.17Boe ␣-methyl-5-HT 181.97 6.74Boe 5-HT4 5-HT 6.31 8.20Adh LY344864 6.02 8.22Phe ␣-Methyl-5-HT 263.03 6.58Adh Methylergon 30.90 7.51Boe Cisapride 25.00 7.60Adh Zolmitriptan 63.09 7.20Mar 5-HT5A 5-HT 7.94 8.10Boe 5-HT2A 5-HT 5.75 8.24Boe 8-OH-DPAT 50.12 7.30Boe Eng ␣-Methyl-5-HT 127.05 6.90 jn.physiology.org BW723C86 251.18 6.60Bax 5-HT6 2-Methyl-5-HT 52.48 7.28B97 DOI 0.79 9.10Boe 5-HT 56.23 7.25Boe Methylergon 0.35 9.45Kni BW723C86 398.11 6.40Ken EMD386088 7.41 8.13Mat

5-HT7 5-HT 1.51 8.82Boe 8-OH-DPAT 34.67 7.46Boe LP44 0.22 9.66Leo on March 8, 2011

ϭϪ Agonists Ki and binding affinity (pKi log Ki) obtained from high affinity agonist radioligand bindings studies (Adham et al. 1996; Baxter 1996; Boess and Martin 1994; Boess et al. 1997; Egan et al. 2000; Engel et al. 1986; Glennon et al. 1989; Hoyer et al. 1985; Kennett et al. 1997a; Knight et al. 2004; Leopoldo et al. 2007; Martin et al. 1997; Mattsson et al. 2005; Milburn and Peroutka 1989; Phebus et al. 1997), with references abbreviated by the first few lettersof authors’ names, except for Boess et al. 1997, which is abbreviated as B97. Methylergonovine is abbreviated as methylergon. Each agonist is considered to activate Ͻ a receptor if Ki 400 nM and is listed with that receptor. Cisapride binds to other receptors but was only used after blocking all but 5-HT4 receptors (Table 2), and so is only listed with that receptor.AQ: Au please spell out DOI.

Only 5-HT2B and 5-HT2C receptors facilitate the LLR 5-HT2B about 10 times higher than 5-HT2C; Table 2) or 5-HT2C ␣ receptors (MK212; or agonists made selective by antagonists, e.g., Application of 5-HT2 receptor agonists ( -methyl-5-HT, ␣ DOI, BW723C86, methylergonovine, MK212, 2-methyl-5-HT, -methyl-5-HT acts as a selective 5-HT2C agonist when given tryptamine, and 5-HT) increased the LLR (Fig. 2), with in- with the 5-HT2C antagonist RS127445 present; Fig. 2C; Table 2) creasing doses generally producing larger responses over about had significantly lower efficacies. However, the sum of the effects a 100-fold change in dose (e.g., 10–1,000 nM for ␣-methyl- of selective 5-HT2B and 5-HT2C agonists was similar to the effect ␣ 5-HT; Fig. 2D). This agonist dose–response relation was well of broad-spectrum 5-HT2 agonists ( -methyl-5-HT and 5-HT, approximated by a sigmoidal curve (Fig. 2, C and D), from Table 2), suggesting that both 5-HT2B and 5-HT2C receptors are which we computed the agonist dose to produce 50% maximal required to maximally facilitate the LLR. ϭ effect (EC50; Fig. 2D), agonist potency [pEC50 –log(EC50)], Prior application of antagonists that blocked both 5-HT2B and efficacy (maximal response; Fig. 2C). All 5-HT2 agonists and 5-HT2C receptors (including SB206553, clozapine, cypro- tested exhibited a significant efficacy in facilitating the LLR heptadine, methysergide, or a combination of RS127445 and (Table 2), with efficacy reported as a percent of mean 5-HT- RS102221) inhibited the 5-HT2 agonist-induced increase in induced LLR, which was 0.393 Ϯ 0.19 mV. This 5-HT– LLR (Fig. 2, C and D; Table 2), eliminating responses to all but induced LLR was three times greater than the predrug LLR of the highest doses and significantly reducing the LLR induced 0.11 Ϯ 0.06 mV (100% in Table 2 represents a 3-fold change by the agonist dose that maximally facilitated the LLR before Ϯ Ϯ in LLR). Agonists that are known to bind to both 5-HT2B and the antagonists (SB206553 reduced LLR to 15.9 25.7, 4.1 Ͻ ␣ Ϯ ␣ 5-HT2C receptors (Ki 50 nM; -methyl-5-HT, DOI, 5-HT) 13.2, and 1.0 8.2% of preantagonist efficacy for -methyl- consistently had the highest efficacies, including 5-HT. Ago- 5-HT, DOI, and 5-HT, respectively). The antagonist applica- nists that bind relatively selectively to either 5-HT2B receptors tion shifted the dose–response relation to such high doses that (BW723C86, methylergonovine, 2-methyl-5-HT; i.e., Ki for the standard sigmoidal curve with 100-fold dose range could

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 736 MURRAY, STEPHENS, BALLOU, HECKMAN, AND BENNETT

A LLR in chronic spinal rat B + Agonist

0.2 mV

1 sec

FIG. 2. 5-HT2B and 5-HT2C receptor Dorsal root stimulation (3xT) agonists increase the LLRs in chronic spinal rats. A: LLR triggered by dorsal root stimu- lation (0.1 ms pulse, 3xT) and recorded from 5-HT2 agonist 5-HT2 agonist C D 1.5 the ventral roots. B: facilitation of LLR with application of the 5-HT2 receptor agonist, ␣-methyl-5-HT (0.1 ␮M), quantified over time window indicated in Fig. 1 (0.5–4 s). 0.4 C–F: facilitation of LLR with increasing Efficacy 1.0 + RS127445 doses of broad spectrum 5-HT2 receptor agonists (DOI and ␣-methyl-5-HT; C and D)

EC50 and relatively selective 5-HT2B (methyler-

0.2 LLR (mV) LLR (mV) 0.5 gonovine; E) and 5-HT2C (MK212; F) recep- tor agonists (increase over ϳ100-fold change Downloaded from + SB206553 + SB206553 in dose). Best-fit sigmoidal curves and subse-

quent estimation of EC50 are shown with thick 0.0 0.0 lines. Prior application of a single blocking Control 100 1000 10,000 Control 10 100 1000 dose of the selective 5-HT2B/2C receptor antag- onist SB206553 (5 ␮M) or the 5-HT recep- α-methyl 5-HT dose (nM) 2B DOI dose (nM) tor antagonist RS127445 (3 ␮M) inhibited these agonists induced changes in the LLR 5-HT2B agonist 5-HT2C agonist (bottom plots in C and D). Each plot shows the jn.physiology.org E F typical response from a single rat, with a dif- ferent rat for each condition, because agonists 0.6 are not feasible to wash out and repeat after 0.6 antagonist application (taking many hours to wash). 0.4 on March 8, 2011

LLR (mV) 0.4 LLR (mV) 0.2 EC50 EC50

0.2 0.0 Control 1110 001000 Control 100 1000 10,000 Methylergonovine dose (nM) MK212 dose (nM)

no longer be fit to the few responsive high doses tested, and the not altered significantly), although the efficacies were lowered usual EC50 measurements were thus not detectable (ND; Table (Table 2). Similarly, blocking all but the 5-HT2C receptor did 2). This thorough inhibitory action of these antagonists sug- not eliminate the ␣-methyl-5-HT–induced LLR or significantly gests that 5-HT2B and/or 5-HT2C receptors facilitate the LLR. shift the EC50 (Table 2). This is consistent with the involvement Furthermore, the finding that the facilitation of the LLR by of both the 5-HT2B and 5-HT2C receptors in facilitating the LLR; 5-HT itself is blocked by these combined 5-HT2B and 5-HT2C the EC50 was not significantly shifted with a block of either ␣ antagonists indicates that all other 5-HT receptors (including 5-HT2B or 5-HT2C receptors, because -methyl-5-HT binds with 5-HT2A) are likely not involved in facilitating the LLR or PIC. similar affinity to 5-HT2B and 5-HT2C receptors (Table 1). Prior application of the selective 5-HT2B receptor antagonist Application of selective agonists (or agonist/antagonist combi- SB204741 significantly inhibited the effect of the selective nations) for other 5-HT receptors (5-HT1, 5-HT2A, 5-HT3, 5-HT4, 5-HT2B agonist BW723C86 (decreasing the potency, doubling the 5-HT5, 5-HT6, 5-HT7) produced no significant facilitation of the ␮ EC50, and decreasing the efficacy; Table 2), consistent with the LLR even at high doses (3–10 M; Table 2), again suggesting involvement of 5-HT2B receptors. However, the response to that only 5-HT2B and 5-HT2C receptors facilitate the LLR. BW723C86 was not completely eliminated by SB204741, likely because of the weak action of BW723C86 on the 5-HT2C receptor Agonist potency is correlated with receptor binding affinity ϩ (We thus consider BW723C86 SB204741 a weak 5-HT2C ago- at 5-HT2B and 5-HT2C receptors nist in Table 3, described below.) Blocking all 5-HT2 receptors except the 5-HT2B receptor (with 5-HT2C and 5-HT2A antago- The effective 5-HT2 agonist doses required to facilitate the nists) did not eliminate the facilitation of the LLR by subsequent LLR (EC50 values and associated potencies, pEC50) varied by ␣ application of the 5-HT2 receptor agonist -methyl-5-HT (Table orders of magnitude between the different agonists (or agonist/ 2) and did not significantly shift the dose–response curve (EC50 antagonist combinations). However, this variation was largely

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 5-HT2B RECEPTORS ON MOTONEURONS 737

TABLE 2. Facilitation of LLR by 5-HT2B and 5-HT2C receptor agonists

Pretreatment With Selectivity/Receptors Activated with Ki EC50 Efficacy Ͻ ϭ Ϫ Agonist Antagonists 50 nM and (Ki 50–400 nM) Potency ( logEC50) (nM) (% of 5-HT effect)

2-Methyl-5-HT None (5-HT2B, 3, 6) 5.41 Ϯ 0.32 3852 70 Ϯ 38* 5-HT None 5-HT1-7 7.48 Ϯ 0.23 33 100 Ϯ 43* 5-HT SB206553† 5-HT1, 2A, 3–7 ND ND 1.0 Ϯ 8.2 ␣-Methyl-5-HT None 5-HT2A,2B,2C (5-HT1E, 1F; 5-HT4) 7.24 Ϯ 0.20 58 116 Ϯ 49* ␣-Methyl-5-HT SB206553 5-HT2A (5-HT1E, 1F; 5-HT4) ND ND 18.4 Ϯ 29.8 ␣-Methyl-5-HT RS102 ϩ AC901 5-HT2B (5-HT1E, 1F; 5-HT4) 7.12 Ϯ 0.31 76 77 Ϯ 30* ␣-Methyl-5-HT RS127 ϩ AC901 5-HT2C (5-HT1E, 1F; 5-HT4) 7.25 Ϯ 0.26 56 74 Ϯ 46* BW723C86 None 5-HT2B (5-HT1B, 1D, 2A, 2C, 3, 6) 6.53 Ϯ 0.31 295 49 Ϯ 48* BW723C86 SB204741 (5-HT1B, 1D, 2A,2C,3,6) 5.84 Ϯ 0.17‡ 1445 31 Ϯ 23* DOI None 5-HT2A, 2B, 2C 6.19 Ϯ 0.17 646 119 Ϯ 37* DOI SB206553 5-HT2A ND ND 5.8 Ϯ 15.7 Methylergonovine None 5-HT2A, 2B, 2C (5-HT1E, 1F) 7.96 Ϯ 0.25 11 69 Ϯ 28* MK212 None (5-HT2C, 3) 6.35 Ϯ 0.24 447 47 Ϯ 29* Tryptamine SB224289 5-HT2A (5-HT2B) 6.04 Ϯ 0.39 912 51 Ϯ 24*

Agonists with varying selectivity for the different 5-HT receptors were applied, sometimes after prior application of 5-HT receptor antagonists to effectively make the agonist action more selective (column 2, pretreatment, starting Ͼ 30 min before agonist). The receptors that can be activated by this agonist/antagonist

combination are indicated in column 3, with high binding affinity for the agonist listed first (with Ki about 1–50 nM), followed by moderate affinity (in brackets;

ϭ Downloaded from Ki 50–400 nM; see Table 1). Antagonist abbreviations and receptors blocked are as follows: RS127445 (RS127, 5-HT2B); RS102221 (RS102, 5-HT2C); ␣ AC90179 (AC901, 5-HT2A); SB206553 (5-HT2B/2C); SB204741 (5-HT2B), SB224289 (5-HT1B), prazosin ( 1 adrenergic); granisetron (5-HT3); methysergide ␮ ␮ ␮ (5-HT2), and clozapine (5-HT2). Antagonist doses are given in Table 4, except for prazosin (1 M), granisetron (0.3 M) and SB224289 (1 M). The efficacy (maximal effect) of these agonists in facilitating the LLR is indicated in column 6 (max mean rectified LLR, normalized by 5-HT efficacy). When a significant ϭϪ facilitation (efficacy) was detected, the mean agonist potency was computed, from which the EC50 was computed (potency logEC50). In addition to the agonists shown, no detectable facilitation (ND, P Ͼ 0.05) occurred with application of the agonists: 8-OH-DPAT, cisapride (RS), EMD386088, LP44, LY34864, ␮ and zolmitriptan when given up to 3–10 M doses, ruling out the involvement of the receptors that these drugs selectively activate [see Table 1 for Ki values; cisapride given after pretreatment with antagonists RS127445, RS1022221, granisetron and prazosin to make it selective to 5-HT4 receptors, which is abbreviated

cisapride (RS)]. Also after some antagonist pre-treatments (SB206553), there was no detectable (indicated ND) facilitation of the LLR at all but the highest doses, jn.physiology.org ϩ making it impossible to estimate EC50. between antagonists indicates they were given in combination. *Efficacy significantly different from zero (LLR significantly increased), P Ͻ 0.05. †Pretreatment with methysergide, clozapine, or RS127445 ϩ RS102221 also eliminated the facilitation of the LLR by 5-HT (ND), like SB206553. ‡Potency significantly lowered by SB204741, P Ͻ 0.05. Data shown as mean Ϯ SD, with n Ͼ 8 per drug (or combination).

accounted for by the differing binding affinity of these drugs to other receptors. This relative potency should account for all 5-HT2B and 5-HT2C receptors (pKi; see description of binding factors that affect potency (e.g., drug diffusion, receptor sig- affinity in METHODS). That is, we found that the agonist poten- naling) other than receptor binding. As shown in Table 3, the on March 8, 2011 cies in facilitating the LLR (pEC50 and associated EC50) were relative potency computed for all agonists of the 5-HT2B significantly correlated with the binding affinity of these ago- receptor was indeed very similar, all falling within the 99% ϭ ϭ ϭ nists (pKi) for 5-HT2B (r 0.93, slope 0.95, n 8 drugs confidence interval (Ϯ0.58) about the mean relative potency Ͻ ϭ ϭ combinations tested, P 0.05) and 5-HT2C (r 0.77, slope [–1.02 Ϯ 0.32 (SD)], and within 2 SD of this mean, which we 0.89, n ϭ 8, P Ͻ 0.05) receptors, as shown in Fig. 3, B and C. set as the similarity criterion. The relative potency of agonists ϭ ϩ Interestingly, a line with unity slope (pEC50 pKi C, where at the 5-HT2C receptor was likewise invariant, clustered re- C is a fixed value) was equally well fit to the data, with r ϭ markably near the mean value of –1.03 Ϯ 0.49 and not ϭ 0.92 and r 0.77 for 5-HT2B and 5-HT2C receptors, respec- significantly different from the relative potency at the 5-HT2B tively (Fig. 3, A and B), a useful finding that we use in the next receptor (Table 3). Again, most 5-HT2C agonists had relative section to model the potency–affinity relation to further predict potencies within the 99% confidence interval about the mean which receptors modulate the LLR. In contrast, there was no (Ϯ0.53), and within 2 SD of the mean, consistent with the significant correlation with the potency and the binding affinity interpretation that 5-HT2C receptors also play a role in modu- ϭ ϭ Ͼ for 5-HT2A receptors (r 0.31, n 8, P 0.05) or any other lating the LLR (see also Fig. 3, squares). The only exception receptor (Fig. 3), and points were scattered far from the unity was DOI, which had a response that was dominated by the slope line that fit the 5-HT2B and 5-HT2C receptors. 5-HT2B, rather than the 5-HT2C, receptor. These findings sug- gest that, after taking into account variable binding affinity, all Potency of each agonist can be quantitatively predicted from these 5-HT2 agonists appeared to diffuse equally well to its receptor binding affinity common targets (the 5-HT2B and 5-HT2C receptors) and have similar actions (facilitating the Ca PIC underlying the LLR). To directly compensate for the variable receptor binding Receptor binding affinities (pKi and Ki values) are typically affinity of different agonists, we computed the potency of each measured in homogenized membranes from isolated cells agonist in facilitating the LLR relative to its binding affinity at (transfected with cloned receptors, Boess and Martin 1994; ϭ Ϫ each receptor, which we term the relative potency pEC50 Knight et al. 2004). In contrast, we measured the functional pKi. As we just mentioned, this difference is a fixed constant effects of these agonists (the potency or EC50) in a whole (C ϭ relative potency) for the receptors involved in modulating spinal cord preparation, where there are substantial barriers to the LLR (5-HT2B and 5-HT2C), and thus we expect that it drug diffusion to the receptors. Furthermore, downstream ac- should be invariant when computed with pKi values for func- tions of receptor binding may further alter the potency (see tional receptors in the spinal cord (5-HT2B) but variable for DISCUSSION). Thus it is not unexpected that the EC50 exceeds the

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 738 MURRAY, STEPHENS, BALLOU, HECKMAN, AND BENNETT

TABLE 3. Relative potency of agonists at facilitating the LLR

Receptor Agonist pEC50 -pKi Receptor Agonist pEC50 -pKi

*5-HT1A 5-HT Ϫ1.30 Ϯ 0.33* 5-HT2B 2-Methyl-5-HT ؊1.09 ؎ 0.32 *5-HT (SB206) ND 5-HT ؊0.51 ؎ 0.33 *8-OH-DPAT ND ␣-Me-5-HT ؊0.74 ؎ 0.20 *LP44 ND ␣-Me-5-HT (RS102) ؊0.86 ؎ 0.31 *BW723C86 ؊1.37 ؎ 0.31 *5-HT1B 5-HT Ϫ0.13 Ϯ 0.33 DOI ؊1.37 ؎ 0.31 *5-HT (SB206) ND Methylergon ؊1.34 ؎ 0.25 *BW723C86 Ϫ0.07 Ϯ 0.31 Tryptamine (SB224) ؊0.91 ؎ 0.39 EMD386088 ND *Zolmitriptan ND 5-HT2C 5-HT ؊0.48 ؎ 0.33 *Me-5-HT ؊1.33 ؎ 0.20-␣ *5-HT1D 5-HT Ϫ1.12 Ϯ 0.33* ␣-Me-5-HT (RS127) ؊1.32 ؎ 0.26 *5-HT (SB206) ND BW723C86 (SB204) ؊1.06 ؎ 0.17 BW723C86 ϩ0.23 Ϯ 0.31 DOI ؊1.84 ؎ 0.17 *EMD386088 ND MK212 ؊0.66 ؎ 0.24 *Zolmitriptan ND Tryptamine (SB224) ؊1.17 ؎ 0.39

5-HT1E 5-HT Ϫ0.73 Ϯ 0.33* 5-HT3 2-Methyl-5-HT Ϫ1.66 Ϯ 0.32 5-HT (SB206) ND BW723C86 ϩ0.03 Ϯ 0.31

␣-me-5-HT ϩ0.32 Ϯ 0.20 EMD386088 ND Downloaded from Methylergon ϩ0.91 Ϯ 0.25 MK212 Ϫ1.19 Ϯ 0.24*

5-HT1F 5-HT ϩ0.31 Ϯ 0.33 5-HT4 5-HT Ϫ0.72 Ϯ 0.33* 5-HT (SB206) ND 5-HT (SB206) ND ␣-Me-5-HT ϩ0.50 Ϯ 0.20 ␣-Me-5-HT ϩ0.66 Ϯ 0.20 LY344864 ND Cisapride (RS) ND Methylergon Ϫ0.45 Ϯ 0.25

Zolmitriptan ND 5-HT5 5-HT Ϫ0.62 Ϯ 0.33* jn.physiology.org 5-HT (SB206) ND 5-HT2A 5-HT Ϫ1.76 Ϯ 0.33 8-OH-DPAT ND 5-HT (SB206) ND ␣-Me-5-HT ϩ0.34 Ϯ 0.20 5-HT6 2-Methyl-5-HT Ϫ1.87 Ϯ 0.32 ␣-Me-5-HT (SB206) ND 5-HT ϩ0.23 Ϯ 0.33 BW723C86 Ϫ0.07 Ϯ 0.31 BW723C86 ϩ0.13 Ϯ 0.31 DOI Ϫ2.91 Ϯ 0.17 EMD (SB216ϩ)ND Methylergon Ϫ1.49 Ϯ 0.25* on March 8, 2011 5-HT7 5-HT Ϫ1.34 Ϯ 0.33* 5-HT (SB206) ND 8-OH-DPAT ND LP44 ND

Relative potency computed as the difference between potency (Table 2) and affinity (Table 1): pEC50 -pKi. Pretreatment with antagonists used to make agonist more selective, indicated in brackets, abbreviated as in Table 2. ND, no detected inhibition of the reflex in Table 2. Relative potency not computed for methylergonovine at the 5-HT2C receptor, because this drug has a much higher affinity for 5-HT2B receptors, which dominate its response. Bold indicates Ϫ receptors with relative potency values mostly within confidence interval. *relative potency within 2 SD of 1.0 (or the mean for 5-HT2B receptors), the confidence interval for similarity, which is about 0.4–0.6. Ͻ Ki dose value (potency affinity) and the relative potency proach is that the lack of involvement of a given receptor in the (potency Ϫ affinity) ϳϭϪ1.0. Another way to interpret relative LLR can be proven by the potency measured from as little as potency arises from the law of differences of logarithms one agonist, if that potency is far from the affinity (relative ϭ Ϫ ϭϪ Ϫ Ϫ ϭ potency different from –1, and corresponding point far from C pEC50 pKi log(EC50) ( logKi) ϭ Ϫ the pEC50 pKi – 1 line in Fig. 3; e.g., 5-HT2A cannot be log(EC50 ⁄ Ki)(1) involved because DOI exhibits a relative potency of –3 at this ϭ –C Thus the ratio EC50/Ki 10 , which again should be a fixed receptor). Indeed, we found that the relative potency at all other number for a receptor involved in modulating the LLR. For the 5-HT receptors was highly variable, usually significantly dif- 1 Ͼ 5-HT2B receptor, the overall EC50/Ki was 10.0 , indicating that ferent from the relative potency at the 5-HT2B/2C receptors ( 2 ϭ the EC50 in the present whole sacral spinal cord preparation is SD away from –1.0; Table 3; and scattered away from pEC50 about 10 times the Ki value, which is likely because of pKi – 1 line in Fig. 3), thus further ruling out the involvement of common diffusion barriers. all but the 5-HT2B and 5-HT2C receptors. Occasionally, the Our consistent findings of fixed relative potencies for relative potency fell close to –1.0 for a particular agonist, like 5-HT2B and 5-HT2C receptors suggest that perhaps, in general, methylergonovine at the 5-HT2A receptor, but this was simply by we can predict the EC50 and potency of drugs at a particular chance, because the relative potency of other agonists at these ϭ ϫ receptor from the published Ki values (EC50 10 Ki and receptors was not consistently –1.0. For example, MK212 has an ϩ potency ϭ affinity – 1); receptors that do not consistently fit impossibly high relative potency of 0.36 at the 5-HT2A receptor, this prediction are unlikely to be functional (assuming similar with a corresponding EC50 less than Ki (Table 3; MK212 is a poor diffusion to each receptor). The power of this modeling ap- 5-HT2A agonist). Furthermore, these receptors with inconsistent

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 5-HT2B RECEPTORS ON MOTONEURONS 739

A 5-HT2B receptor B 5-HT2C 5-HT3 5-HT2A 5-HT4

8 methylergonovine 8

5-HT

a-methyl-5-HT FIG. 3. Potency of 5-HT receptor agonists at 7 7 facilitating the LLR is only related to binding to

5-HT2B and 5-HT2C receptors. A: 5-HT2B recep- ϭϪ BW723C86 tor agonist potency [pEC50 log(EC50)] for tryptamine DOI facilitating the LLR plotted against the agonist 6 6 binding affinity to that receptor (pKi), with Potency (-log EC50) each agonist name indicated next to its point. ␣-methyl-5-HT potency with and without

2-methyl-5-HT block of 5-HT2C receptor with RS102221 plot- ND ted. Thin line: linear correlation between po- tency and affinity. Dashed line: best fit line 6789 6789with unit slope (potency ϭ binding affinity ϩ C, where C ϳ –1). B–D: similar potency- Binding affinity (-log Ki) Binding affinity (-log Ki) affinity scatter plots for the remaining 5-HT receptors. Agonists used for each receptor are listed in Table 1, with agonists assumed to act

at a receptor only if K Ͻ400 nM. Potencies are Downloaded from C 5-HT1A 5-HT1D D 5-HT1F 5-HT6 i from Table 2. We used the potency of 5-HT1B 5-HT1E 5-HT5 5-HT7 BW723C86 after treatment with SB204741 8 for the 5-HT2C receptor plot to avoid dominant 8 5-HT2B receptor action of this drug. Thin line: correlation between agonist potency and affin-

ity for 5-HT2C receptors. Dashed line: best fit unity slope relation for the 5-HT receptor, 7 2C 7 copied into C and D to show lack of similar jn.physiology.org relation for other receptors. Other receptors had no significant correlation between potency and affinity (open circles; P Ͼ 0.05). ND and 6 gray zone: no detected effect of agonist on the Potency (-log EC50) 6 LLR. on March 8, 2011 ND ND 6789 6789 Binding affinity (-log Ki) Binding affinity (-log Ki) relative potencies often had substantial binding affinity for ago- of all other receptors tested). Satisfying this latter criterion nists (like zolmitriptan at the 5-HT1B receptor) that caused no was not always possible for nonselective antagonists like detectable facilitation of the LLR (ND; Table 3). Thus we again methysergide. conclude that only the 5-HT2B and 5-HT2C receptors are involved in facilitating the LLR. 5-HT2B and 5-HT2C receptor antagonists block the Antagonist dose determination action agonists

The experiments described above partly rely on using a To verify that we used 5-HT2 receptor antagonists at an blocking dose of selective 5-HT2 receptor antagonists before appropriate dose to block the 5-HT2B/2C receptors in the ex- adding agonists, to identify the receptors involved in facilitat- periments described above (e.g., Table 2), we directly tested ing the LLR. The antagonist dose that we used was always the efficacy of most antagonists in inhibiting a prior dose of the Ն ␣ ␮ chosen to be 10 times the published Ki value of that antag- 5-HT2 receptor agonist -methyl-5-HT (0.3 M). A submaxi- onist at the target receptor (e.g., 5-HT2; Table 4, column 2), to mal agonist dose (near agonist EC50; dose vs. time) was used account for drug diffusion into the whole cord preparation, as to assure that the agonist only activated 5-HT2 receptors. As described above for agonists (assuming that the EC50 of the shown in Fig. 3 and Table 4, all 5-HT2B/2C antagonists tested antagonist can be predicted from its Ki value, as for agonists: (selective and nonselective drugs) significantly decreased a ϭ ϫ Ն EC50 10 Ki) and typically set 100 times the Ki to prior agonist-induced increase in LLR. Importantly, the selec- maximally block the receptor. Furthermore, to be sure that we tive 5-HT2B receptor antagonists (RS127445 and SB204741) did not affect other receptors nonselectively, we tried to use an and 5-HT2C receptor antagonist (RS102221) each inhibited the antagonist dose that was less than about 10 times the Ki value LLR induced by the 5-HT agonist (Table 4), confirming that of that antagonist at other receptors known to bind this antag- both 5-HT2B and 5-HT2C receptors modulate the LLR and onist (Table 4, column 3 shows the receptor most likely to be associated Ca PIC. In contrast, the 5-HT2A receptor antagonist nonselectively activated, with the highest published Ki value AC90179 had no effect (Table 4).

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 740 MURRAY, STEPHENS, BALLOU, HECKMAN, AND BENNETT

TABLE 4. 5-HT2 receptor antagonists that inhibit the LLR

Antagonist Inhibition of Receptors Non selective Binding to Refences Antagonist Agonist-Induced LLR Antagonist Inhibition of

Antagonists Antagonized, Ki (nM) Receptor and, Ki (nM) for Ki dose (nM) (% change) Basal LLR (% change)

RS127445 5-HT2B (1.1) 5-HT2C (470) Knight04 3,000 Ϫ45.2 Ϯ 17.8* Ϫ1.9 Ϯ 5.3 SB242084 5-HT2B (45) 5-HT2A (850) Cussac02 3,000 Ϫ89.0 Ϯ 18.2* 0.3 Ϯ 26.9 5-HT2C (0.48) Knight04 Methysergide 5-HT2B (0.36) 5-HT1,2A,5,6,7 Knight04 10,000 Ϫ98.8 Ϯ 13.3* 1.0 Ϯ 56.3 5-HT2C (4.4) Boess94 SB204741 5-HT2B (51) 5-HT2C (2100) Cussac02 30,000 — Ϫ33.1 Ϯ 21.1* RS102221 5-HT2C (5.0) 5-HT2B (2900) Knight04 3000–10000 Ϫ40.9 Ϯ 21.3* Ϫ28.1 Ϯ 8.6* SB206553 5-HT 2B (5.5) 5-HT2A (2300) Cussac02 3000–10000 Ϫ79.1 Ϯ 28.2* Ϫ68.6 Ϯ 18.2* 5-HT2C (3.2) Knight04 Cyprohept 5-HT2B (1.5) 5-HT2A,5,6,7 Yoshio01 10,000 — Ϫ92.5 Ϯ 10.0* 5-HT2C (2.2) Bonhaus97 alpha1 (25) Boess94 Clozapine 5-HT2B (10) 5-HT2A,3,6,7 Knight04 10,000 — — 5-HT2C (13) Boess94 Ketanserin 5-HT2A (8.1) 5-HT2B (740) Knight04 10,000 — Ϫ52.8 Ϯ 9.9* 5-HT2C (62) Yoshio01 alpha1 (6.3)

AC90179 5-HT2A (0.20) 5-HT2C (2.9) Vanover04 30 0.1 Ϯ 6.6 1.4 Ϯ 9.7 Downloaded from

The receptors for which the antagonists are selective are shown in column 2, along with Ki values (parentheses). Receptors that the antagonists nonselectively bind at higher doses are shown in column 3, along with Ki values. Ki values were obtained from the following references, abbreviated in column 4 (Boess and Martin 1994; Bonhaus et al. 1999; Cussac et al. 2002; Knight et al. 2004; Vanover et al. 2004; Yoshio et al. 2001). Selectivity is defined as the ratio of Ki’s for ϭ two receptors (e.g., 470/1.1 427-fold selectivity of RS127445 at 5-HT2B over 5-HT2C receptors). For the nonselective antagonists methysergide, cyproheptadine, and clozapine, only Ki values for target receptors are shown, and other receptors affected only listed. Antagonists were used during LLR reflex ␣ ␮ testing at the doses indicated in column 5, with choice of dose based on Ki values. Following application of the 5-HT2 receptor agonist -methyl-5-HT (0.3 M), antagonists were applied (Fig 4A), and the inhibition of the agonist-induced LLR was measured (% change in LLR, relative to agonist-induced LLR; column jn.physiology.org

6; 0% no effect), with significant effects found for all 5-HT2B or 5-HT2C antagonists tested. Antagonists were also tested without agonists present to examine their negative efficacy (inverse agonist) action (e.g. Fig 4F), and the inhibition of the control/basal LLR measured (% change in LLR; column 7), with significant effects found only for inverse agonists. The first 3 antagonists listed are neutral 5-HT2B/2C receptor antagonists (no effect without agonist present, column 7), whereas the next 6 antagonists act as inverse agonists at 5-HT2B/2C receptors (shaded rows; negative efficacy without agonist present, column 7). Data shown as mean Ϯ SD, with n Ͼ 8 per drug (or combination). –, not tested. *Significant inhibition induced by antagonist, P Ͻ 0.05.

5-HT2B and 5-HT2C receptors are not activated by constitutive receptor activity (receptor activity in the absence on March 8, 2011 endogenous 5-HT of 5-HT), in addition to their conventional ability to antagonize 5-HT–induced receptor activity (Aloyo et al. 2009; Berg et al. Some of the 5-HT receptor antagonists had no effect when 2 1999; Chanrion et al. 2008; Seifert and Wenzel-Seifert 2002; given by themselves, including drugs selective to 5-HT2B receptors (RS127445; Fig. 4B) and 5-HT receptors Weiner et al. 2001; Westphal and Sanders-Bush 1994). Taken 2B/2C together with the lack of functional 5-HT in the chronic (SB242084; Fig. 4D) and the nonselective 5-HT2 receptor antagonist methysergide (Table 4, column 7). All these drugs spinal rat, this inhibition of the LLR by such inverse have previously been classified as neutral antagonists (Bonhaus agonists suggests that there is substantial constitutive activ- et al. 1999; Chanrion et al. 2008; Westphal and Sanders-Bush ity in 5-HT2B and 5-HT2C receptors that endogenously 1994). Neutral antagonists, by definition, only act to inhibit the facilitates the LLR. The magnitude of the inhibition of the action of 5-HT (endogenous or applied) or other agonists and LLR by SB204741 and RS102221 was similar (20–30% have no effect when given by themselves (no intrinsic effi- reduction in LLR; Table 4), and thus the endogenous facil- cacy), because they do not block constitutive receptor activity itation of the LLR is likely equally produced by constitutive (as we see in Table 4) (Chanrion et al. 2008; Seifert and activity in both 5-HT2B and 5-HT2C receptors. Furthermore, Wenzel-Seifert 2002). Thus the lack of effect of these drugs the inverse agonist SB206553, which selectively binds both (when given alone) shows that there is no functional residual 5-HT2B and 5-HT2C receptors (but no other receptors), 5-HT activating 5-HT2B or 5-HT2C receptors in the spinal cord reduced the PIC by approximately the amount predicted by after chronic spinal injury, consistent with our recent report for the sum of the action of the selective 5-HT2B and 5-HT2C 5-HT2C receptors (Murray et al. 2010). inverse agonist SB204741 and RS102221 (Table 4), sug- gesting that the action of 5-HT2B and 5-HT2C receptors is 5-HT2B and 5-HT2C receptors are constitutively active additive in facilitating the endogenous LLR (and associated LLR). The broad spectrum inverse agonists cyproheptadine The remaining 5-HT2 receptor antagonists that we used, including those that bind selectively to 5-HT receptors and ketanserin also inhibited LLR (Table 4). We have 2B previously reported this action of SB206553 and cyprohep- (SB204741), 5-HT2C receptors (RS102221 and ketanserin), or both (SB206553), significantly inhibited the LLR when they tadine, but at the time we were only able to conclude that were given by themselves (without prior agonist application; either 5-HT2B or 5-HT2C receptors may be constitutively Table 4; Fig. 4, F and H), thus suggesting that these com- active (Murray et al. 2010). Here we show that both recep- pounds acted as inverse agonists, which by definition can block tors are constitutively active.

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 5-HT2B RECEPTORS ON MOTONEURONS 741

A + 5-HT agonist B +RS127445 + RS127445 0.6 0.2 Control 0.3

LLR (mV) FIG. 4. 5-HT2B and 5-HT2C receptors are 0.0 0.0 constitutively active after spinal cord injury. 050100150 050100Plots of mean LLR sampled over time with agonists and antagonist applications. A and B: the neutral antagonist RS127445 (3 ␮M; C D + SB242084 + 5-HT2 agonist selective for 5-HT2B receptors) inhibited the increase in the LLR induced by 5-HT2 re- ␣ ␮ 0.6 + SB242084 0.2 ceptor agonist -methyl-5-HT (0.3 M; A), whereas when applied alone RS127445 had 0.3 no effect on the LLR (B). C and D: likewise, LLR (mV) the neutral antagonist SB242084 (3 ␮M; se- 0.0 0.0 lective for 5-HT2B and 5-HT2C receptors) in- 0 50 100 150 050100 hibited that same agonist induced LLR (C), whereas SB242084 had no effect when applied E + 5-HT2 agonist F alone (D). E–H: in contrast, the inverse ago- + RS102221 ␮ 0.2 nists RS102221 (3 M, selective for 5-HT2C ␮ + RS102221 0.6 receptors) and SB206553 (5 M, selective for 5-HT2B and 5-HT2C receptors) inhibited the

0.1 Downloaded from 0.3 agonist-induced LLR (E and G) and inhibited

LLR (mV) the spontaneously occurring LLR when given 0.0 0.0 alone (F and H). Considering that inverse ago- 0 50 100 150 0 50 100 150 nists block constitutive activity and agonist- induced activity, whereas neutral antagonists G + 5-HT2 agonist H only block agonist induced activity, these re- + SB206553 sults indicate that the 5-HT2B and 5-HT2C re- ceptors exhibit constitutive activity and are not

+ SB206553 jn.physiology.org 0.6 Control activated by endogenous residual 5-HT. See statistics in Table 4. 0.2 0.4 LLR (mV) 0.2 0.0 0 50 100 150 200 050100150 Time (min) Time (min) on March 8, 2011 Notably, cyproheptadine was the most effective of all measured the Ca PIC in motoneurons using slow voltage ramps inverse agonists tested, significantly exceeding the inhibi- to inactivate transient currents and in the presence of TTX to tion of the LLR produced by the 5-HT2B/2C inverse agonist synaptically isolate the motoneuron (by stopping spike-medi- SB206553 (Table 4). This is likely because cyproheptadine ated transmission) and block the Na PIC, which can otherwise ␣ potently inhibits both 5-HT2 receptors and 1 adrenergic obscure the Ca PIC, as previously described (Harvey et al. receptors (Westphal and Sanders-Bush 1994; Yoshio et al. 2006c). During this slow voltage ramp (under voltage-clamp ␣ 2001) (Table 4; 1 receptors, like 5-HT2 receptors, are conditions), the Ca PIC was activated at, on average, Ϫ49.6 Ϯ known to facilitate the LLR; Li et al. 2004b). Also, the ϭ 6.41 mV (Von, n 46 motoneurons), producing a downward 5-HT2C inverse agonist ketanserin produced significantly deflection in the recorded current of 1.98 Ϯ 1.07 nA, which we more inhibition of the LLR than the selective 5-HT2C took as an estimate of the Ca PIC amplitude (Fig. 5A, arrow; inverse agonist RS102221 (Table 4). Again, this is likely previously verified to be mediated by L-type calcium channels; because of the known potent binding of ketanserin to ␣1 nimodipine-sensitive) (Li et al. 2004c). Application of the receptors (Yoshio et al. 2001), in addition to its 5-HT ␣ ␮ ␮ 2C 5-HT2B/2C agonists -methyl-5-HT (0.3 M), DOI (1 M), receptors action (Cussac et al. 2002) (Table 4; ketanserin BW723C86 (1–2 ␮M), and 5-HT (1 ␮M) each significantly also binds potently to 5-HT2A receptors, but these are not increased the Ca PIC amplitude (by Ϫ0.71 Ϯ 0.82 nA, n ϭ involved in modulating the LLR; see above). Together these 11; Ϫ0.60 Ϯ 0.75, n ϭ 9; Ϫ0.42 Ϯ 0.43, n ϭ 15; and Ϫ0.58 data indicate that there is endogenous activity in ␣1 recep- Ϯ 0.68 nA, n ϭ 11; respectively; Figs. 4B and 5B), and tors, likely mediated by constitutive activity (consistent with Ϫ Ϯ lowered the onset threshold of the PIC, Von (by 7.9 8.0, Harvey et al. 2006b), as with 5-HT2 receptors; we address Ϫ5.6 Ϯ 3.8, Ϫ3.8 Ϯ 2.2, and Ϫ6.3 Ϯ 4.7 mV, respectively). this issue in a separate paper (M. Rank and D. J. Bennett., ␮ The 5-HT2B/2C antagonist SB206553 (5 M) significantly unpublished data). inhibited the facilitation of the Ca PIC by ␣-methyl-5-HT (by –1.77 Ϯ 0.48 nA; Fig. 5, C and D). Whereas the selective 5-HT agonist BW723C86 increased the Ca PIC, 5-HT2B and 5-HT2C receptors regulate the Ca PIC 2B in motoneurons its mean effect was one half that of the 5-HT2B/2C agonist ␣-methyl-5-HT (see above), and subsequent application of ␣ ␮ Considering that the LLR is mediated by the Ca PIC, it is the 5-HT2B/2C agonist -methyl-5-HT (0.3 M) after likely that the 5-HT2B and 5-HT2C receptors modulate the Ca BW723C86 further significantly facilitated the Ca PIC (PIC Ϫ Ϯ Ϫ Ϯ PIC, as they do to the LLR. To confirm this, we directly increased by 0.43 0.55 nA; and Von lowered by 3.3

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 742 MURRAY, STEPHENS, BALLOU, HECKMAN, AND BENNETT

A Chronic spinal rat (in TTX) B + α−methyl-5-HT

Von 20 mV -70 mV

Ca PIC 1 nA FIG. 5. 5-HT2B and 5-HT2C receptors fa- cilitate the motoneuronal Ca PIC, as they do the LLR. A: Ca PIC in motoneuron of chronic spinal rat, activated by slowly in- creasing the membrane potential under volt- 0 nA age clamp, and quantified at its initial peak, 5 sec where it produced a downward deflection in the recorded current (at arrow) relative to the

leak current (thin line). Von represents the voltage at which the Ca PIC was activated. α Horizontal dashed line indicates resting po- C + -methyl-5-HT D + α-methyl-5-HT + SB206553 tential. Recording in presence of TTX (2 ␮M) to block the Na PIC and synaptic input. ␣ Downloaded from B: the 5-HT2 receptor agonist -methyl- 5-HT (0.1 ␮M) increased the Ca PIC and

20 mV reduced Von. C and D: in another motoneu- ron, the selective 5-HT2B/2C antagonist SB206553 (5 ␮M) reduced the agonist in- duced increase in the Ca PIC. jn.physiology.org Ca PIC 2 nA

0nA on March 8, 2011

ϭ ␣ 3.7 mV, n 9; Fig. 6), suggesting involvement of both The 5-HT2 receptor agonists -methyl-5-HT, DOI, and 5-HT2B and 5-HT2C receptors. The agonists 8-OH-DPAT (3 BW723C86 did not significantly change the resting potential ␮M) and zolmitriptan (1 ␮M) had no significant effects on (changes: 2.4 Ϯ 9.2, Ϫ3.2 Ϯ 8.0, and Ϫ1.2 Ϯ 6.1 mV, motoneuron properties (Ca PIC unaffected; n ϭ 5, P Ͼ respectively), consistent with previous reports that resting 0.05), confirming that 5-HT1A/1B/1D/1F, 5-HT5, and 5-HT7 membrane potential is increased by 5-HT via a non-5-HT2 receptors have no influence on the Ca PIC. Together these receptor (Harvey et al. 2006a). The input resistance was sig- Ϯ results verify that both 5-HT2B and 5-HT2C receptors are nificantly increased by these 5-HT2 receptor agonists (by 2.7 involved in facilitating Ca PIC in motoneurons, whereas 2.4, 0.79 Ϯ 0.89, and 0.85 Ϯ 1.15 M⍀, corresponding to 69, other 5-HT receptors are not likely involved. 21, and 18% increases, respectively).

A cinorhC lanips tar ni( )XTT B + 68C327WB

Von 20mV Von -80mV

FIG. 6. 5-HT2B receptor activation alone facilitates 5 sec the Ca PIC on motoneurons. A: Ca PIC recorded in motoneuron of chronic spinal rat, quantified at arrow, as detailed in Fig. 5 (in TTX, 2 ␮M). B: the moderately ␮ Ca PIC selective 5-HT2B receptor agonist BW723C86 (1 M) increased the size of the Ca PIC and lowered its onset 1nA Ca PIC voltage.

0nA

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 5-HT2B RECEPTORS ON MOTONEURONS 743

5-HT2B receptors are seen on motoneurons describe are on the motoneurons themselves, because receptor with immunolabeling agonists increased the Ca PICs when synaptic transmission was blocked with TTX. In addition, immunolabeling confirms Considering that 5-HT2B and 5-HT2C receptor agonists in- crease the Ca PIC in the presence of TTX, it is likely that 5-HT2B (Fig. 7) and 5-HT2C (Murray et al. 2010) receptors on motoneurons of chronic spinal rats have both these receptors. motoneurons in our preparation. Considering that the Ca PIC is known to cause long-lasting Previously, we confirmed the presence of 5-HT2C receptors on these motoneurons using immunolabeling (Murray et al. 2010), motoneuron discharges evoked by brief stimuli (LLRs) (Li et al. 2004a,b), we used these LLRs as an assay of the Ca PIC to and here we repeat this for 5-HT2B receptors (Fig. 7). Mo- toneurons were labeled with an antibody to a type of neuro- efficiently estimate the dose–response relations for the Ca filament H (SMI32; Fig. 7, red) that we have previously shown PIC/LLR for numerous drugs and to perform a thorough labels motoneurons and not interneurons in the ventral horn analysis of all 5-HT receptors, ruling out the involvement of all (Anelli et al. 2007). Double labeling with a 5-HT2B receptor but the 5-HT2B and 5-HT2C receptors. We do not know why antibody (green) showed a high degree of colocalization with other receptors, including 5-HT1A, 5-HT2A, 5-HT3, and 5-HT7, SMI32 (yellow and orange), indicating the presence of this do not facilitate the LLR, considering that these receptors are receptor on motoneurons, in all chronic spinal (n ϭ 6/6; Fig 7, present in the spinal cord (Hochman et al. 2001; Millan 2002; A–C). As with the 5-HT2C receptor (Murray et al. 2010), a Schmidt and Jordan 2000) and have previously been suggested large fraction of the 5-HT2B receptor labeling was intracellular to have multiple actions on motoneurons, reflexes, or locomo- to the motoneurons (perinuclear), consistent with the tendency tion (Dougherty et al. 2005; Hammar et al. 2007; Holohean et for constitutively active receptor isoforms to become internal- al. 1995; Shay et al. 2005; Ziskind-Conhaim et al. 1993) (see Downloaded from ized (Chanrion et al. 2008). Similar 5-HT2B labeling was seen Introduction). The relatively selective 5-HT receptor agonist ϭ 3 on SMI32-labeled motoneurons in normal rats (n 7/7 rats; 2-methyl-5-HT (3-fold selectivity in Table 1) (Alexander et al. Fig. 7, D–F). However, we did not try to quantify the differ- 2004) and 5-HT receptor agonist DOI (10-fold selectivity ence between the labeling in normal and injured rats, because 2A over other 5-HT receptors; Table 1) did increase the LLR, but cords of injured rats generally perfuse less well with PFA, and 2 both these drugs worked at doses too high to be consistent with this affects the antibody staining. binding to these receptors and instead acted by their lower affinity binding to the 5-HT receptor (Table 3; relative jn.physiology.org DISCUSSION 2B potency near –1) (Wainscott et al. 1993). This underscores the Our results showed two novel concepts: first, the 5-HT2B importance of considering the nonselective lower affinity bind- receptor modulates motoneuron excitability (Ca PICs) after ing of putative agonists, with quantitative comparisons to SCI, and second, the 5-HT2B receptor plays an essential role in receptor binding affinity, which requires testing agonists at the spontaneous recovery of motoneuron function after SCI, by multiple doses and comparing across multiple agonists. Fi- exhibiting constitutive receptor activity. Constitutive 5-HT nally, in this study, we only examined the 5-HT receptors that on March 8, 2011 receptor activity has previously been shown in the CNS for facilitate the LLR and associated PIC; in a separate paper, we 5-HT2A (Harvey et al. 1999) and 5-HT2C (De Deurwaerdere et examine the receptors that inhibit sensory afferent transmission al. 2004; Murray et al. 2010) receptors but not previously for in the LLR pathway, which include the 5-HT1B receptor (K. C. 5-HT2B receptors. The 5-HT2B and 5-HT2C receptors that we Murray, unpublished data).

5-HT2B SMI32 Merged AB C d FIG. 7. 5-HT2B receptor localization on motoneurons of the rat sacrocaudal spinal cord. A: typical immunofluorescence label- n ing of 5-HT2B receptors (green) in the S4 sacral ventral horn of a chronic spinal rat, showing extensive receptor distribution, in- cluding labeling of the soma of motoneurons (with nuclei labeled with n), as visualized d with 0.14-␮m thin optical sections with con- n focal microscopy (n ϭ 6). B: double-labeling of the same tissue with SMI32 (red), a selec- tive marker of motoneurons in the ventral d horn, which specifically labels neurofilament DE F in the intracellular space. C: an overlay of

5-HT2B and SMI32-labeling shows extensive co-localization (yellow and orange), indicat-

ing 5-HT2B receptors on motoneuron soma n d (arrow) and dendrites (d). D–F: typical im-

munolabeling of 5-HT2B receptors SMI32 in the sacral ventral horn (below injury) of normal rats (n ϭ 7 rats). Note again the

5-HT2B receptors on motoneurons, with yel- low/orange co-localization with SMI32.

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 744 MURRAY, STEPHENS, BALLOU, HECKMAN, AND BENNETT

We did not expect a prominent role of the 5-HT2B receptor, Indeed, our previous work has shown that the in vitro dose of which is a relatively obscure receptor, first described in the 5-HT needed to affect motoneurons in spinal cords from stomach fundus and not widely expressed in the brain (Barnes normal rats is dramatically higher (by a factor of 30; 10 vs. 0.3 and Sharp 1999; Hoyer et al. 2002; Nichols and Nichols 2008; ␮M) than that in chronic spinal rats (Harvey et al. 2006a). We Wainscott et al. 1993). In the brain (and heart), this receptor is suggest that this supersensitivity in chronic spinal rats is mainly associated with developmental plasticity and pathology largely caused by the absence of SERT in chronic spinal rats, (knockouts of the 5-HT2B receptor gene is lethal, unlike for whereas in normal rats, SERT efficiently prevents much of the other 5-HT receptors) (Barnes and Sharp 1999; Nichols and exogenously applied 5-HT from reaching the 5-HT receptors. Nichols 2008). Outside of reports of 5-HT2B receptors modu- This lack of SERT after injury may also explain the supersen- lating frog N-methyl-D-aspartate (NMDA) receptor function on sitity seen to other 5-HT receptors agonists (Harvey et al. spinal motoneurons (Hackman and Holohean 2010; Holohean 2006a), because SERT is rather promiscuous, transporting and Hackman 2004), little is known about the role of this tryptamine-like compounds (Adkins et al. 2001; Henry et al. receptor in the adult spinal cord, even though it is present in the 2006), including many common agonists that we use, such as spinal cord (Hochman et al. 2001; Holohean et al. 1992; Millan ␣-methyl-5-HT, 2-Methyl-5-HT, 5-HT, 5-CT, and tryptamine 2002). Interestingly, 5-HT2B receptor gene expression (mRNA) appears in rat sacral motoneurons, and this expression persists (Adkins et al. 2001; Helfman et al. 1995; Henry et al. 2006) ␣ ␣ unchanged with long-term SCI (Wienecke et al. 2010). Consid- ( -methyl-5-HT is abbreviated -MST in Henry et al. 2006). ering its novelty, we used several approaches to prove the pres- In summary, our quantitative analysis of 5-HT receptor agonist potency and correlation to receptor binding affinity would not ence and functionality of the 5-HT2B receptor, including show-

have been possible had we not been studying the fully tran- Downloaded from ing that 1) 5-HT2B agonists consistently facilitated motoneuron excitability (LLR and PICs), even when they were made highly sected rat that lacks SERT, because SERT interferes with selective by a prior block of other receptors known to be agonists reaching the receptors in the normal spinal cord, activated by the agonists (e.g., 5-HT2C and 5-HT2A receptors effectively making the normal spinal cord much less sensitive for DOI), 2) selective 5-HT2B receptor antagonists overcame to 5-HT agonists compared with the spinal cord of chronic the agonist action, when given after a low dose of the agonist spinal rats. titrated to just activate 5-HT2 receptors, 3) agonist action was We found a remarkably consistent relation between agonist jn.physiology.org highly correlated with 5-HT2B binding affinity for the agonists potency (–logEC50) and receptor binding affinity (–logKi). at the 5-HT2B receptor and indeed agonist potency could be Basically, effective agonist doses (EC50) were consistently quantitatively predicted from binding affinity, and 4) 5-HT2B only about 10 times the binding Ki values of that agonist to receptors are on the motoneurons (immunolabeling). The same 5-HT2B receptors. Receptor binding affinities are typically methods were used to show that the 5-HT2C receptor was also measured in homogenized membranes from isolated cell cul- present and functional. In contrast, 5-HT2A receptors were not ture systems where agonists are applied directly to the recep- functional in modulating the PICs, contrary to our previous tors (Boess and Martin 1994), whereas in our whole spinal cord on March 8, 2011 erroneous conclusion that they modulated the PICs, based on preparations, agonists must diffuse substantial distances to data from the nonselective agonist DOI (Harvey et al. 2006a). reach receptors (see RESULTS). Likely, this diffusion barrier in The 5-HT2A receptors likely have other functions on motoneu- part helps explain the factor of 10 that relates EC50 to Ki rons, because they are on motoneurons and are upregulated values. Even disregarding the pia, diffusion in brain and spinal with injury (Kong et al. 2010). Interestingly, both 5-HT2B and cord tissue is somewhat restricted, where diffusion coefficients ϭ ϫ 5-HT2C receptors seem to have equally large effects, based on for small molecules are 4 times lower than in saline (D* 5 the action of selective agonists and antagonists. 10Ϫ7 cm2/s)(Nicholson and Tao 1993; Sykova and Nicholson An advantage of studying the action of 5-HT receptor 2008), because of the convoluted and small extracellular space agonists after chronic spinal transection or severe contusion is (only 20% of total volume) (Sykova and Nicholson 2008). that 5-HT terminals and associated high-affinity serotonin Considering that the motoneuron cell bodies in the S4–S3 sacral reuptake transporters (SERTs) are mostly eliminated in the spinal cord are about 0.25 mm below the surface and that we spinal cord by the injury (Hayashi et al. 2010; Murray et al. waited about 10 min for drugs to diffuse into the cord before 2010; Newton and Hamill 1988), so they cannot interfere with recording, the diffusion equation indicates that the drug con- the action of exogenously applied 5-HT or other agonists centration at the motoneuron cell bodies should be only about [although there may remain some minor low-affinity plasma 0.31 times the applied ACSF concentration (assuming D* ϭ membrane monoamine transport (PMAT)] (Engel et al. 2004). 5 ϫ 10Ϫ7 cm2/s, and 1D diffusion paths) (using Eq. 9 in SERT is almost exclusively located on 5-HT neurons in the Sykova and Nicholson 2008). If we also consider that the 5-HT brain (explaining its loss with injury), and it regulates the receptors may be on motoneuron dendrites that can be even amount of 5-HT in the synaptic cleft (Blakely et al. 1994). deeper in the tissue (say 0.4 mm), the concentration at these Accordingly, SERT dramatically affects the dose of exoge- deep dendrites should be about 1/10th the applied surface nously applied 5-HT needed to activate receptors in the syn- concentration. In summary, diffusion can contribute substan- apse, and serotonin reuptake inhibitors that block SERT (cita- tially to our factor of 10 difference between EC50 and Ki. lopram) decrease the dose of 5-HT or other 5-HT agonists Interestingly, had we waited longer times between the drug needed to facilitate motoneuron function by a factor of 30 application and recording, the drug concentrations in the cord (Elliott and Wallis 1992) (Bennett and Murray, unpublished would have been higher, and the EC50 and Ki would have been data). Thus the lack of synaptic SERT in the chronic spinal rat closer, although computing the EC50 would have been difficult allows 5-HT to reach receptors unimpeded, and our estimates because of slow rundown of reflexes over the hours needed to of drug potency (and EC50) therefore are much more accurate. construct such a dose–response relation.

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 5-HT2B RECEPTORS ON MOTONEURONS 745 Other factors like receptor reserve (pool of receptors) and Westphal and Sanders-Bush 1994). Therefore, a combination downstream G protein signaling (Boess and Martin 1994) may of these drugs is useful in identifying constitutive activity. also contribute to the higher EC50 values (lower potency) in Both inverse agonists and neutral antagonists block the action our system. Additionally, we compared our EC50 values to Ki of agonists, and thus the action of an inverse agonist alone is values obtained from competitive binding between our agonists not sufficient to prove the presence of constitutive receptor 3 and radiolabeled agonists (e.g., [ H]-5-HT; see METHODS). This activity, if there is any chance that the receptor is activated by generally labels a high affinity binding state that occurs in 5-HT (e.g., residual 5-HT in the spinal cord). Definitive proof response to receptor activation. Had we instead compared with of constitutive activity only comes by also showing that neutral the low affinity binding obtained with radiolabeled antagonists, antagonists have no effect (as in Table 4) (Seifert and Wenzel- which do not activate the receptor (Egan et al. 2000; Knight et Seifert 2002). al. 2004), the EC50 and Ki might have been closer. Brain stem–derived 5-HT normally serves a critical role in Assuming that diffusion is the predominant factor in increas- tuning the excitability of motoneurons, facilitating Ca PICs ing our observed EC50 and that diffusion affects all 5-HT that amplify and prolong responses to synaptic input (Heckman receptor activation similarly (similar receptor depths), our et al. 2005). Thus, when SCI eliminates this major source of observed factor-of-10 relation between EC50 and Ki may ulti- 5-HT, motoneurons are initially left in a depressed state and are mately help us predict doses of agonist and antagonists needed not able to produce adequate muscle contractions. However, to effectively activate any particular 5-HT receptor. Indeed, we over the weeks after injury, Ca PICs in motoneurons sponta- used this to choose doses of antagonists (Table 4) and inde- neously recover, to the point where they are permanently large, pendently verified those doses by their action alone or against and not only help with residual motor functions but also an agonist. We also used this simple relation to help rule out contribute to unwanted contractions (muscle spasms) (Bennett Downloaded from the involvement of receptors other than 5-HT2B and 5-HT2C; et al. 2004; Li et al. 2004a). Recently, we showed that a novel these other receptors had EC50/Ki ratios far from 10. With this form of plasticity in 5-HT2 receptors causes this recovery of Ca method and direct use of agonists to other receptors (8-OH- PICs: these receptors become spontaneously active in the DPAT, etc.), it is clear that no other 5-HT receptor is involved absence of 5-HT (constitutively active) (Murray et al. 2010). in modulating the Ca PIC and associated LLR. Different However, at the time, we were uncertain which particular agonists have variable functional efficacy (partial or full ago- 5-HT2 receptor modulates the Ca PIC. The present results jn.physiology.org nists) (Kenakin 1996), but we avoided this issue by focusing on resolve this issue, showing that both 5-HT2B and 5-HT2C binding affinity and potency rather than efficacy. However, a receptors are constitutively active after SCI, because selective caveat that must be noted is that we do not know for certain 5-HT2B and 5-HT2C inverse agonists that block this constitu- that all drugs that bind to a receptor act as an agonist at that tive activity decrease motoneuron excitability (LLRs and as- receptor, and indeed, compounds like 8-OH-DPAT have poor sociated Ca PICs), whereas neutral antagonists do not. As efficacy (partial agonists) at some receptors to which they bind discussed above, the lack of effect of neutral antagonists shows on March 8, 2011 (e.g., 5-HT7 for 8-OH-DPAT) (Eglen et al. 1997); thus we also that there is no residual 5-HT below the chronic injury that tested more effective agonists to these receptors (LP44). activates the 5-HT2 receptors. Thus, even though the inverse Both 5-HT2B and 5-HT2C receptors are Gq protein–coupled agonists we tested can block the action of 5-HT, in addition to receptors that generally activate the classic PLC (phospho- blocking constitutive receptor activity, their action can only be lipase C) pathways involved in the synthesis of inositol phos- attributed to constitutive 5-HT2B and 5-HT2C receptor activity. phates (IP) and mobilization of intracellular Ca2ϩ stores Considering that most known antagonists can act as inverse (Hoyer et al. 2002; Kenakin 1996; Lucaites et al. 1996; Mizuno agonists at 5-HT2B and 5-HT2C receptors in various model and Itoh 2009). These pathways facilitate PICs (and NMDA systems, whereas neutral antagonists are rare (Weiner et al. receptors) in spinal motoneurons (Holohean and Hackman 2001; Westphal and Sanders-Bush 1994), we were not sur- 2004; Mejia-Gervacio et al. 2004; Perrier et al. 2000). Thus it prised that the selective 5-HT2B and 5-HT2C receptor antago- is likely that the 5-HT2B and 5-HT2C receptors modulate the nists (SB204741 and RS102221, respectively) could act as PICs via IP pathways in our preparation. Both these receptor inverse agonists in inhibiting the LLR. However, we had not types are known to exhibit a substantial degree of constitutive anticipated such prominent constitutive activity in 5-HT2B activity (Seifert and Wenzel-Seifert 2002; Villazon et al. 2003; receptors, as shown by these compounds, because constitutive Weiner et al. 2001), usually observed in the form of IP activity in this receptor has not previously been shown in the production in the absence of 5-HT (or other agonists), which CNS, although it has been seen in the stomach fundus (Villa- produces a basal level of IP (Chanrion et al. 2008; Kennett et zon et al. 2003). al. 1997b; Knight et al. 2004). Thus, PICs may in part be The constitutive 5-HT2 receptor activity that we observed facilitated by this basal IP production (without 5-HT), and this after SCI arises from at least two mechanisms. First, there is an may explain the emergence of the large PICs seen after SCI increase in expression of 5-HT2C receptor isoforms that exhibit when there is an absence of 5-HT, as we discuss below. a high degree of constitutive activity (Murray et al. 2010). Importantly, the subclass of antagonists known as inverse Most 5-HT2C receptor isoforms have some probability of being agonists inhibit this basal IP production (or associated func- constitutively active, but certain isoforms, like the INI isoform, tions) in 5-HT2B and 5-HT2C receptors, including classic an- have a very high degree of constitutive activity (with activity tagonists such as cyproheptadine, clozapine, and ketanserin approaching that produced by maximal activation with 5-HT) (Chanrion et al. 2008; Villazon et al. 2003; Weiner et al. 2001; (Berg et al. 2008; Herrick-Davis et al. 1999; Marion et al. Westphal and Sanders-Bush 1994). In contrast, neutral antag- 2004; Niswender et al. 1999; Weiner et al. 2001), and these onists like methysergide and RS127445 have no effect when isoforms increase in expression with SCI (Murray et al. 2010). given alone (Bonhaus et al. 1999; Chanrion et al. 2008; Second, the total number of 5-HT2C receptors increases with

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 746 MURRAY, STEPHENS, BALLOU, HECKMAN, AND BENNETT

SCI (Hayashi et al. 2010), which increases overall constitutive Bennett DJ, Gorassini M, Fouad K, Sanelli L, Han Y, Cheng J. Spasticity activity because all isoforms exhibit some constitutive activity in rats with sacral spinal cord injury. J Neurotrauma 16: 69–84, 1999. (Berg et al. 2008; Herrick-Davis et al. 1999). We do not know Bennett DJ, Sanelli L, Cooke C, Harvey PJ, Gorassini MA. Spastic long-lasting reflexes in the awake rat after sacral spinal cord injury. J whether the 5-HT2B receptor isoform types or total receptor Neurophysiol 91: 2247–2258, 2004. numbers change with injury, although we do know that the Berg KA, Dunlop J, Sanchez T, Silva M, Clarke WP. A conservative, 5-HT2B receptor gene expression is unchanged with long-term single-amino acid substitution in the second cytoplasmic domain of the SCI (Wienecke et al. 2010). human Serotonin2C receptor alters both ligand-dependent and -independent In summary, we performed a systematic analysis of the 5-HT receptor signaling. J Pharmacol Exp Ther 324: 1084–1092, 2008. receptors that facilitate motoneuron PICs and associated LLRs Berg KA, Stout BD, Cropper JD, Maayani S, Clarke WP. Novel actions of inverse agonists on 5-HT2C receptor systems. Mol Pharmacol 55: 863–872, (spasms) after SCI and found that both 5-HT2B and 5-HT2C 1999. receptors are involved, whereas 5-HT1, 5-HT2A, and 5-HT3/4/5/6/7 Blakely RD, De Felice LJ, Hartzell HC. Molecular physiology of norepi- receptors are not. Although both 5-HT2B and 5-HT2C receptors are nephrine and serotonin transporters. J Exp Biol 196: 263–281, 1994. present before injury, they exhibit dramatic plasticity after chronic Boess FG, Martin IL. Molecular biology of 5-HT receptors. Neuropharma- spinal transection, presumably as a compensation for the near- cology 33: 275–317, 1994. complete loss of 5-HT innervation. That is, both receptors become Boess FG, Monsma FJ Jr, Carolo C, Meyer V, Rudler A, Zwingelstein C, constitutively active, producing large Ca PICs in the absence of Sleight AJ. Functional and radioligand binding characterization of rat 5-HT6 receptors stably expressed in HEK293 cells. Neuropharmacology 36: 5-HT (whereas acutely after injury, PICs are small). These 713–720, 1997. large Ca PICs are essential for the recovery of normal mo- Bonhaus DW, Flippin LA, Greenhouse RJ, Jaime S, Rocha C, Dawson M, toneuron firing ability after chronic injury (sustained firing and Van Natta K, Chang LK, Pulido-Rios T, Webber A, Leung E, Eglen amplification of synaptic input) (Harvey et al. 2006c; Heckman RM, Martin GR. RS-127445: a selective, high affinity, orally bioavailable Downloaded from et al. 2005) and thus must play an important role in all motor 5-HT2B receptor antagonist. Br J Pharmacol 127: 1075–1082, 1999. recovery, including locomotion (Murray et al. 2010). However, Button DC, Kalmar JM, Gardiner K, Marqueste T, Zhong H, Roy RR, Edgerton VR, Gardiner PF. Does elimination of afferent input modify the unlike with normal brain stem control over receptor activation, changes in rat motoneurone properties that occur following chronic spinal these receptors are permanently active, leading to permanently cord transection? J Physiol 586: 529–544, 2008. large PICs and ultimately hyperexcitability of motoneurons Chanrion B, Mannoury la Cour C, Gavarini S, Seimandi M, Vincent L, associated with muscle spasms. Thus controlling excess con- Pujol JF, Bockaert J, Marin P, Millan MJ. Inverse agonist and neutral stitutive activity in 5-HT receptors, especially 5-HT recep- antagonist actions of antidepressants at recombinant and native jn.physiology.org 2 2B 5-hydroxytryptamine2C receptors: differential modulation of cell surface tors, offers a new approach in managing spasticity after SCI. expression and signal transduction. Mol Pharmacol 73: 748–757, 2008. Cussac D, Newman-Tancredi A, Quentric Y, Carpentier N, Poissonnet G, ACKNOWLEDGMENTS Parmentier JG, Goldstein S, Millan MJ. Characterization of phospho- lipase C activity at h5-HT2C compared with h5-HT2B receptors: influence We thank L. Sanelli for expert technical assistance. of novel ligands upon membrane-bound levels of [3H]phosphatidylinositols. Naunyn Schmiedebergs Arch Pharmacol 365: 242–252, 2002. on March 8, 2011 Dario A, Tomei G. A benefit-risk assessment of baclofen in severe spinal GRANTS spasticity. Drug Saf 27: 799–818, 2004. Funding was provided by National Institute of Neurological Disorders and De Deurwaerdere P, Navailles S, Berg KA, Clarke WP, Spampinato U. Stroke Grant NS-47567, Canadian Foundation for Innovation, the Canadian Constitutive activity of the serotonin2C receptor inhibits in vivo dopamine Institutes of Health Research, and the Alberta Heritage Foundation for Medical release in the rat striatum and nucleus accumbens. J Neurosci 24: 3235– Research. 3241, 2004. Dougherty KJ, Bannatyne BA, Jankowska E, Krutki P, Maxwell DJ. Membrane receptors involved in modulation of responses of spinal dorsal DISCLOSURES horn interneurons evoked by feline group II muscle afferents. J Neurosci 25: 584–593, 2005. No conflicts of interest, financial or otherwise, are declared by the authors. Egan C, Grinde E, Dupre A, Roth BL, Hake M, Teitler M, Herrick-Davis K. Agonist high and low affinity state ratios predict drug intrinsic activity REFERENCES and a revised ternary complex mechanism at serotonin 5-HT(2A) and 5-HT(2C) receptors. Synapse 35: 144–150, 2000. Adham N, Gerald C, Schechter L, Vaysse P, Weinshank R, Branchek T. Eglen RM, Jasper JR, Chang DJ, Martin GR. The 5-HT7 receptor: orphan [3H]5-hydroxytryptamine labels the agonist high affinity state of the cloned found. Trends Pharmacol Sci 18: 104–107, 1997. rat 5-HT4 receptor. Eur J Pharmacol 304: 231–235, 1996. Elliott P, Wallis DI. Serotonin and L-norepinephrine as mediators of altered Adkins EM, Barker EL, Blakely RD. Interactions of tryptamine derivatives excitability in neonatal rat motoneurons studied in vitro. Neuroscience 47: with serotonin transporter species variants implicate transmembrane domain 533–544, 1992. I in substrate recognition. Mol Pharmacol 59: 514–523, 2001. Engel G, Gothert M, Hoyer D, Schlicker E, Hillenbrand K. Identity of Alexander SP, Mathie A, Peters JA. Guide to receptors and channels. Br J inhibitory presynaptic 5-hydroxytryptamine (5-HT) autoreceptors in the rat Pharmacol 141 Suppl. 1: S1–S126, 2004. brain cortex with 5-HT1B binding sites. Naunyn Schmiedebergs Arch Aloyo VJ, Berg KA, Spampinato U, Clarke WP, Harvey JA. Current status Pharmacol 332: 1–7, 1986. of inverse agonism at serotonin2A (5-HT2A) and 5-HT2C receptors. Phar- Engel K, Zhou M, Wang J. Identification and characterization of a novel macol Ther 121: 160–173, 2009. monoamine transporter in the human brain. J Biol Chem 279: 50042–50049, Anelli R, Sanelli L, Bennett DJ, Heckman CJ. Expression of L-type calcium 2004. channel alpha(1)-1.2 and alpha(1)-1.3 subunits on rat sacral motoneurons Glennon RA, Ismaiel AE, McCarthy BG, Peroutka SJ. Binding of following chronic spinal cord injury. Neuroscience 145: 751–763, 2007. arylpiperazines to 5-HT3 serotonin receptors: results of a structure-affinity Baker LL, Chandler SH. Characterization of postsynaptic potentials evoked study. Eur J Pharmacol 168: 387–392, 1989. by sural nerve stimulation in hindlimb motoneurons from acute and chronic Gorassini MA, Knash ME, Harvey PJ, Bennett DJ, Yang JF. Role of spinal cats. Brain Res 420: 340–350, 1987. motoneurons in the generation of muscle spasms after spinal cord injury. Barnes NM, Sharp T. A review of central 5-HT receptors and their function. Brain 127: 2247–2258, 2004. Neuropharmacology 38: 1083–1152, 1999. Hackman JC, Holohean AM. The effects of polyamine agonists and antag- Baxter GS. Novel discriminatory ligands for 5-HT2B receptors. Behav Brain onists on N-methyl-D-aspartate-induced depolarizations of amphibian mo- Res 73: 149–152, 1996. toneurons in situ. Brain Res 1325: 10–18, 2010.

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 5-HT2B RECEPTORS ON MOTONEURONS 747

Hammar I, Stecina K, Jankowska E. Differential modulation by monoamine Kennett GA, Wood MD, Bright F, Trail B, Riley G, Holland V, Avenell membrane receptor agonists of reticulospinal input to lamina VIII feline KY, Stean T, Upton N, Bromidge S, Forbes IT, Brown AM, Middlemiss spinal commissural interneurons. Eur J Neurosci 26: 1205–1212, 2007. DN, Blackburn TP. SB 242084, a selective and brain penetrant 5-HT2C Harvey JA, Welsh SE, Hood H, Romano AG. Effect of 5-HT2 receptor receptor antagonist. Neuropharmacology 36: 609–620, 1997b. antagonists on a cranial nerve reflex in the rabbit: evidence for inverse Knight AR, Misra A, Quirk K, Benwell K, Revell D, Kennett G, Bick- agonism. Psychopharmacology (Berl) 141: 162–168, 1999. erdike M. Pharmacological characterisation of the agonist radioligand Harvey PJ, Li X, Li Y, Bennett DJ. 5-HT2 receptor activation facilitates a binding site of 5-HT(2A), 5-HT(2B) and 5-HT(2C) receptors. Naunyn persistent sodium current and repetitive firing in spinal motoneurons of rats Schmiedebergs Arch Pharmacol 370: 114–123, 2004. with and without chronic spinal cord injury. J Neurophysiol 96: 1158–1170, Kong XY, Wienecke J, Hultborn H, Zhang M. Robust upregulation of 2006a. serotonin 2A receptors after chronic spinal transection of rats: an immuno- Harvey PJ, Li X, Li Y, Bennett DJ. Endogenous monoamine receptor histochemical study. Brain Res 1320: 60–68, 2010. activation is essential for enabling persistent sodium currents and repetitive Kuhn RA, Macht MB. Some manifestations of reflex activity in spinal man firing in rat spinal motoneurons. J Neurophysiol 96: 1171–1186, 2006b. with particular reference to the occurrence of extensor spasm. Bull Johns Harvey PJ, Li Y, Li X, Bennett DJ. Persistent sodium currents and repetitive Hopkins Hosp 84: 43–75, 1948. firing in motoneurons of the sacrocaudal spinal cord of adult rats. J Larkman PM, Kelly JS. Modulation of IH by 5-HT in neonatal rat motoneu- Neurophysiol 96: 1141–1157, 2006c. rones in vitro: mediation through a phosphorylation independent action of Hayashi Y, Jacob-Vadakot S, Dugan EA, McBride S, Olexa R, Simansky cAMP. Neuropharmacology 36: 721–733, 1997. K, Murray M, Shumsky JS. 5-HT precursor loading, but not 5-HT receptor Leopoldo M, Lacivita E, Contino M, Colabufo NA, Berardi F, Perrone R. agonists, increases motor function after spinal cord contusion in adult rats. Structure-activity relationship study on N-(1,2,3,4-tetrahydronaphthalen-1- Exp Neurol 221: 68–78, 2010. yl)-4-aryl-1-piperazinehexanamides, a class of 5-HT7 receptor agents. II. J Heckman CJ, Gorassini MA, Bennett DJ. Persistent inward currents in Med Chem 50: 4214–4221, 2007. motoneuron dendrites: implications for motor output. Muscle Nerve 31: Li X, Bennett DJ. Apamin-sensitive calcium-activated potassium currents 135–156, 2005. (SK) are activated by persistent calcium currents in rat motoneurons. J Helfman CC, Anderson GF, Barraco RA. Presynaptic displacement of Neurophysiol 97: 3314–3330, 2007. Downloaded from serotonin by alpha-methyl-5-hydroxytryptamine in the nucleus tractus soli- Li X, Murray K, Harvey PJ, Ballou EW, Bennett DJ. Serotonin facilitates tarius of the rat. Neurosci Lett 189: 65–68, 1995. a persistent calcium current in motoneurons of rats with and without chronic Henry LK, Field JR, Adkins EM, Parnas ML, Vaughan RA, Zou MF, spinal cord injury. J Neurophysiol 97: 1236–1246, 2007. Newman AH, Blakely RD. Tyr-95 and Ile-172 in transmembrane segments Li Y, Bennett DJ. Persistent sodium and calcium currents cause plateau 1 and 3 of human serotonin transporters interact to establish high affinity potentials in motoneurons of chronic spinal rats. J Neurophysiol 90: 857– recognition of antidepressants. J Biol Chem 281: 2012–2023, 2006. 869, 2003. Herrick-Davis K, Grinde E, Niswender CM. Serotonin 5-HT2C receptor Li Y, Gorassini MA, Bennett DJ. Role of persistent sodium and calcium

RNA editing alters receptor basal activity: implications for serotonergic currents in motoneuron firing and spasticity in chronic spinal rats. J jn.physiology.org signal transduction. J Neurochem 73: 1711–1717, 1999. Neurophysiol 91: 767–783, 2004a. Hochman S, Garraway SM, Machacek DW, Shay BL. 5-HT receptors and Li Y, Harvey PJ, Li X, Bennett DJ. Spastic long-lasting reflexes in the the neuromodulatory control of spinal cord function. In: Neurobiology of the chronic spinal rat, studied in vitro. J Neurophysiol 91: 2236–2246, 2004b. Spinal Cord, edited by Cope T. Washington, DC: CRC, 2001, p. 47–87. Li Y, Li X, Harvey PJ, Bennett DJ. Effects of baclofen on spinal reflexes and Holohean AM, Hackman JC. Mechanisms intrinsic to 5-HT2B receptor- persistent inward currents in motoneurons of chronic spinal rats with induced potentiation of NMDA receptor responses in frog motoneurones. Br spasticity. J Neurophysiol 92: 2694–2703, 2004c. J Pharmacol 143: 351–360, 2004. Lucaites VL, Nelson DL, Wainscott DB, Baez M. Receptor subtype and on March 8, 2011 Holohean AM, Hackman JC, Davidoff RA. Changes in membrane potential density determine the coupling repertoire of the 5-HT2 receptor subfamily. of frog motoneurons induced by activation of serotonin receptor subtypes. Life Sci 59: 1081–1095, 1996. Neuroscience 34: 555–564, 1990. Lundberg A. Inhibitory control from the brainstem of transmission from Holohean AM, Hackman JC, Davidoff RA. Modulation of frog spinal cord primary afferents to motoneurons, primary afferent terminals and ascending interneuronal activity by activation of 5-HT3 receptors. Brain Res 704: pathways. In: Brainstem Control of Spinal Mechanisms, edited by Sjolund 184–190, 1995. B, Bjorklund A. New York: Elsevier, 1982, p. 179–224. Holohean AM, Hackman JC, Shope SB, Davidoff RA. Serotonin1A facil- Marion S, Weiner DM, Caron MG. RNA editing induces variation in itation of frog motoneuron responses to afferent stimuli and to N-methyl- desensitization and trafficking of 5-hydroxytryptamine 2c receptor isoforms. D-aspartate. Neuroscience 48: 469–477, 1992. J Biol Chem 279: 2945–2954, 2004. Holstege JC, Bongers CM. A glycinergic projection from the ventromedial Martin GR, Robertson AD, MacLennan SJ, Prentice DJ, Barrett VJ, lower brainstem to spinal motoneurons. An ultrastructural double labeling Buckingham J, Honey AC, Giles H, Moncada S. Receptor specificity and study in rat. Brain Res 566: 308–315, 1991. trigemino-vascular inhibitory actions of a novel 5-HT1B/1D receptor partial Hoyer D, Engel G, Kalkman HO. Molecular pharmacology of 5-HT1 and agonist, 311C90 (zolmitriptan). Br J Pharmacol 121: 157–164, 1997. 5-HT2 recognition sites in rat and pig brain membranes: radioligand binding Mattsson C, Sonesson C, Sandahl A, Greiner HE, Gassen M, Plaschke J, studies with [3H]5-HT, [3H]8-OH-DPAT, (-)[125I]iodocyanopindolol, Leibrock J, Bottcher H. 2-Alkyl-3-(1,2,3,6-tetrahydropyridin-4-yl)-1H- [3H]mesulergine and [3H]ketanserin. Eur J Pharmacol 118: 13–23, 1985. indoles as novel 5-HT6 receptor agonists. Bioorg Med Chem Lett 15: Hoyer D, Hannon JP, Martin GR. Molecular, pharmacological and func- 4230–4234, 2005. tional diversity of 5-HT receptors. Pharmacol Biochem Behav 71: 533–554, Maynard FM, Karunas RS, Waring WP III. Epidemiology of spasticity 2002. following traumatic spinal cord injury. Arch Phys Med Rehabil 71: 566– Hultborn H, Brownstone RB, Toth TI, Gossard JP. Key mechanisms for 569, 1990. setting the input-output gain across the motoneuron pool. Prog Brain Res Mejia-Gervacio S, Hounsgaard J, Diaz-Munoz M. Roles of ryanodine and 143: 77–95, 2004. inositol triphosphate receptors in regulation of plateau potentials in turtle Inoue T, Itoh S, Wakisaka S, Ogawa S, Saito M, Morimoto T. Involvement spinal motoneurons. Neuroscience 123: 123–130, 2004. of 5-HT7 receptors in serotonergic effects on spike afterpotentials in Milburn CM, Peroutka SJ. Characterization of [3H]quipazine binding to presumed jaw-closing motoneurons of rats. Brain Res 954: 202–211, 2002. 5-hydroxytryptamine3 receptors in rat brain membranes. J Neurochem 52: Jankowska E, Hammar I. Spinal interneurones; how can studies in animals 1787–1792, 1989. contribute to the understanding of spinal interneuronal systems in man? Millan MJ. Descending control of pain. Prog Neurobiol 66: 355–474, 2002. Brain Res Brain Res Rev 40: 19–28, 2002. Mizuno N, Itoh H. Functions and regulatory mechanisms of Gq-signaling Kawamura J, Ise M, Tagami M. The clinical features of spasms in patients pathways. Neurosignals 17: 42–54, 2009. with a cervical cord injury. Paraplegia 27: 222–226, 1989. Moran A, Ortiz de Urbina AV, Martin ML, Garcia M, Rodriguez- Kenakin T. The classification of seven transmembrane receptors in recombi- Barbero A, Dorado F, San Roman L. Characterization of contractile nant expression systems. Pharmacol Rev 48: 413–463, 1996. 5-hydroxytryptamine receptor subtypes in the in situ autoperfused kidney in Kennett GA, Ainsworth K, Trail B, Blackburn TP. BW 723C86, a 5-HT2B the anaesthetized rat. Eur J Pharmacol 592: 133–137, 2008. receptor agonist, causes hyperphagia and reduced grooming in rats. Neuro- Murray KC, Nakae A, Stephens MJ, Rank M, D’Amico J, Harvey PJ, Li pharmacology 36: 233–239, 1997a. X, Harris RL, Ballou EW, Anelli R, Heckman CJ, Mashimo T, Vavrek

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org 748 MURRAY, STEPHENS, BALLOU, HECKMAN, AND BENNETT

R, Sanelli L, Gorassini MA, Bennett DJ, Fouad K. Recovery of motoneu- Shay BL, Sawchuk M, Machacek DW, Hochman S. Serotonin 5-HT2 ron and locomotor function after spinal cord injury depends on constitutive receptors induce a long-lasting facilitation of spinal reflexes independent of activity in 5-HT(2C) receptors. Nat Med 16: 694–700, 2010. ionotropic receptor activity. J Neurophysiol 94: 2867–2877, 2005. Newton BW, Hamill RW. The morphology and distribution of rat serotonin- Steg G. Efferent muscle innervation and rigidity. Acta Physiol Scand Suppl ergic intraspinal neurons: an immunohistochemical study. Brain Res Bull 225, 61: 1–53, 1964. 20: 349–360, 1988. Sykova E, Nicholson C. Diffusion in brain extracellular space. Physiol Rev Nichols DE, Nichols CD. Serotonin receptors. Chem Rev 108: 1614–1641, 88: 1277–1340, 2008. 2008. Tadros SF, D’Souza M, Zettel ML, Zhu X, Lynch-Erhardt M, Frisina RD. Serotonin 2B receptor: upregulated with age and hearing loss in mouse Nicholson C, Tao L. Hindered diffusion of high molecular weight compounds auditory system. Neurobiol Aging 28: 1112–1123, 2007. in brain extracellular microenvironment measured with integrative optical Vanover KE, Harvey SC, Son T, Bradley SR, Kold H, Makhay M, imaging. Biophys J 65: 2277–2290, 1993. Veinbergs I, Spalding TA, Weiner DM, Andersson CM, Tolf BR, Brann Nielsen JB, Crone C, Hultborn H. The spinal pathophysiology of spasticity– MR, Hacksell U, Davis RE. Pharmacological characterization of AC-90179 from a basic science point of view. Acta Physiol (Oxf) 189: 171–180, 2007. [2-(4-methoxyphenyl)-N-(4-methyl-benzyl)-N-(1-methyl-piperidin-4-yl)- Niswender CM, Copeland SC, Herrick-Davis K, Emeson RB, Sanders- aceta mide hydrochloride]: a selective serotonin 2A receptor inverse agonist. Bush E. RNA editing of the human serotonin 5-hydroxytryptamine 2C J Pharmacol Exp Ther 310: 943–951, 2004. receptor silences constitutive activity. J Biol Chem 274: 9472–9478, 1999. Villazon M, Enguix MJ, Tristan H, Honrubia MA, Brea J, Maayani S, Norton JA, Bennett DJ, Knash ME, Murray KC, Gorassini MA. Changes Cadavid MI, Loza MI. Different pharmacological properties of two equi- in sensory-evoked synaptic activation of motoneurons after spinal cord potent antagonists (clozapine and rauwolscine) for 5-HT2B receptors in rat injury in man. Brain 131: 1478–1491, 2008. stomach fundus. Biochem Pharmacol 66: 927–937, 2003. Perrier JF, Alaburda A, Hounsgaard J. 5-HT1A receptors increase excit- Wainscott DB, Cohen ML, Schenck KW, Audia JE, Nissen JS, Baez M, ability of spinal motoneurons by inhibiting a TASK-1-like Kϩ current in the Kursar JD, Lucaites VL, Nelson DL. Pharmacological characteristics of adult turtle. J Physiol 548: 485–492, 2003. the newly cloned rat 5-hydroxytryptamine2F receptor. Mol Pharmacol 43:

Perrier JF, Delgado-Lezama R. Synaptic release of serotonin induced by 419–426, 1993. Downloaded from stimulation of the raphe nucleus promotes plateau potentials in spinal Wang MY, Dun NJ. 5-Hydroxytryptamine responses in neonate rat motoneu- motoneurons of the adult turtle. J Neurosci 25: 7993–7999, 2005. rones in vitro. J Physiol 430: 87–103, 1990. Perrier JF, Hounsgaard J. 5-HT2 receptors promote plateau potentials in Weiner DM, Burstein ES, Nash N, Croston GE, Currier EA, Vanover KE, Harvey SC, Donohue E, Hansen HC, Andersson CM, Spalding TA, turtle spinal motoneurons by facilitating an L-type calcium current. J Gibson DF, Krebs-Thomson K, Powell SB, Geyer MA, Hacksell U, Neurophysiol 89: 954–959, 2003. Brann MR. 5-Hydroxytryptamine2A receptor inverse agonists as antipsy- Perrier JF, Mejia-Gervacio S, Hounsgaard J. Facilitation of plateau poten- chotics. J Pharmacol Exp Ther 299: 268–276, 2001. tials in turtle motoneurones by a pathway dependent on calcium and Westphal RS, Sanders-Bush E. Reciprocal binding properties of 5-hydroxy- jn.physiology.org calmodulin. J Physiol 528: 107–113, 2000. tryptamine type 2C receptor agonists and inverse agonists. Mol Pharmacol Phebus LA, Johnson KW, Zgombick JM, Gilbert PJ, Van Belle K, 46: 937–942, 1994. Mancuso V, Nelson DL, Calligaro DO, Kiefer AD Jr, Branchek TA, Wienecke J, Westerdahl AC, Hultborn H, Kiehn O, Ryge J. Global gene Flaugh ME. Characterization of LY344864 as a pharmacological tool to expression analysis of rodent motor neurons following spinal cord injury study 5-HT1F receptors: binding affinities, brain penetration and activity in associates molecular mechanisms with development of postinjury spasticity. the neurogenic dural inflammation model of migraine. Life Sci 61: 2117– J Neurophysiol 103: 761–778, 2010. 2126, 1997. Yoshimura M, Furue H. Mechanisms for the anti-nociceptive actions of the

Rekling JC, Funk GD, Bayliss DA, Dong XW, Feldman JL. Synaptic descending noradrenergic and serotonergic systems in the spinal cord. J on March 8, 2011 control of motoneuronal excitability. Physiol Rev 80: 767–852, 2000. Pharmacol Sci 101: 107–117, 2006. Schmidt BJ, Jordan LM. The role of serotonin in reflex modulation and Yoshio R, Taniguchi T, Itoh H, Muramatsu I. Affinity of serotonin receptor locomotor rhythm production in the mammalian spinal cord. Brain Res Bull antagonists and agonists to recombinant and native alpha1-adrenoceptor 53: 689–710, 2000. subtypes. Jpn J Pharmacol 86: 189–195, 2001. Seifert R, Wenzel-Seifert K. Constitutive activity of G-protein-coupled re- Ziskind-Conhaim L, Seebach BS, Gao BX. Changes in serotonin-induced ceptors: cause of disease and common property of wild-type receptors. potentials during spinal cord development. J Neurophysiol 69: 1338–1349, Naunyn Schmiedebergs Arch Pharmacol 366: 381–416, 2002. 1993.

J Neurophysiol • VOL 105 • FEBRUARY 2011 • www.jn.org