Noradrenergic Modulation of the Masseteric Reflex in Behaving Cats. I. Pharmacological Studies Lnes L

The Journal of Neuroscience, January 1990, 10(l): 91-98

Noradrenergic Modulation of the Masseteric Reflex in Behaving Cats. I. Pharmacological Studies lnes L. Stafford and Barry L. Jacobs Program in Neuroscience, Department of Psychology, Princeton University, Princeton, New Jersey 08544

The masseteric (jaw closure) reflex was utilized as a model tomical studieslocalized NE-containing cell bodies, pathways, system for assessing functional changes in central norepi- and axon terminals in the CNS (Dahlstrom and Fuxe, 1964; nephrine (NE) neurotransmission. This monosynaptic reflex Moore and Bloom, 1979). More recent experiments have elu- was chosen because of its simple and well-defined circuitry, cidated the effects of NE on its postsynaptic target neurons (see and because its motor component receives a dense NE in- Rogawski, 1985, for a review). However, despite extensive re- nervation. Previous experiments in our laboratory described searchimplicating NE in a wide variety of behavioral and phys- NE modulation of this reflex in the anesthetized rat. The iological processes(Amaral and Sinnamon, 1977; Foote et al., present experiments examine the effects of NE on this re- 1983) a clear and consistent understandingof the role of central sponse in the unanesthetized, behaving cat. The masseteric NE has remained elusive. reflex was elicited by electrical stimulation of the mesen- This may be attributable to several factors. First, many of the cephalic trigeminal nucleus, and the response was recorded early methods usedto study NE in intact animals were of ne- via electrodes permanently implanted in the masseter mus- cessity gross or imprecise, typically employing lesionsor sys- cle. The amplitude of the reflex response was measured temic pharmacology. Second,when more specificmethods such before and at various intervals following microinfusion (0.5 as neurotoxins (e.g., 6-hydroxydopamine) are used, there is a ~1) of NE or of various NE agonists directly into the motor lag between the time of the neural destruction and the time that trigeminal nucleus (MoV). Microinfusions of NE (0.125-5.0 the behavioral or physiological measurementis taken. This al- pg) produced dose-dependent increases in the amplitude of lows for regrowth and/or compensatory changesto occur, ob- the elicited reflex response. These effects were evident within scuring or altering the effects attributable to the lesion. Third, 1 min postinfusion and lasted up to 30 min; in all cases, the the results of many precise studiesare confounded by the ne- response amplitude returned to baseline levels. The in- cessityof conducting them either under anesthesiaor in reduced crease seen in response to 0.5 Pg NE was blocked by pre- preparations. The generality of theseresults for determining the treatment with the (Y-1 -adrenergic antagonist prazosin, but normal physiological or behavioral role for NE is, therefore, not by pretreatment with the serotonin (5-HT) antagonist limited. Finally, NE may not serve an easily specifiable single methysergide. Methysergide did, however, completely block role in the CNS, but may be involved in different functions in the increase in the amplitude seen in response to microin- different systems(motor, sensory,hormonal, etc.). fusion of 5-HT. Infusion of the a-1-adrenergic agonist phen- Converging results from several different experimental par- ylephrine also increased the amplitude of the reflex re- adigms suggestthat central NE modulates motor output in a sponse. By contrast, infusion of the &adrenergic agonist facilitatory fashion. At the cellular level, iontophoretic appli- isoproterenol had no effect, whereas clonidine, a presyn- cation of NE onto spinal (White and Neuman, 1980, 1983) or aptic a-2-adrenergic agonist, decreased its amplitude. Nei- brain-stem (McCall and Aghajanian, 1979, 1980) motor neu- ther saline infusion nor infusion of NE outside of the motor rons has been shown to facilitate the responseto GABA or to nucleus had any effect on the amplitude of the reflex re- afferent input without affecting spontaneousactivity. These response. These data indicate that NE augments the masse- sults are consistent with more general findings reported by teric reflex response by a direct action in the MoV, probably Woodward and colleagues(Woodward et al., 1979) examining via a-l -adrenergic receptors. Studies in the following paper the effectsof iontophoretic NE on forebrain and cerebellartarget examine whether physiological activation of this NE input neurons.In thesestudies NE was shown to increasethe response also facilitates the reflex response. of target neurons to afferent input, both excitatory and inhibitory, by preferentially inhibiting background activity to a greater Initial biochemical and pharmacologicalstudies demonstrated degreethan evoked activity. This modulatory action has been the existenceofnorepinephrine (NE) in the CNS and established interpreted asNE enhancementofthe signal/noiseratio, thereby its role as a central nemotransmitter (Euler, 1956). Later ana- facilitating neuronal information processing. Other studies at a greater level of neural complexity have more directly examined the functional effects of NE transmis- Received Mar. 7, 1989; revised June 6, 1989; accepted June 23, 1989. This work was funded by the U.S. Air Force (AFOSR 87-0301-A) and the sion. For example, Hino et al. (1984) recording from the ventral NIMH (MH-23433). Thanks to Drs. David Morilak and Casimir Fomal for nu- root of rats, report that chlorpromazine inhibits mono- and merous helpful discussions and advice and to Christine Metzler for excellent polysynaptic spinal reflexes elicited by stimulating the dorsal technical support. Correspondence should be addressed to Ines L. Stafford-Segert, Brain Research root, suggestingthat descendingNE input exerts a tonic facili- Institute, Center for Health Sciences 43-385, UCLA, Los Angeles, CA 90024. tatory effect on this response.In addition, locus coeruleusstim- Copyright 0 1990 Society for Neuroscience 0270-6474/90/O 1009 l-08$02.00/0 ulation has been shown to produce an NE-mediated increasein 92 Stafford and Jacobs + NE Modulation in Behaving Cats. I. Pharmacology the amplitude of the monosynaptic spinal reflex (Strahlendorf and Babb, 1973) and by a variety of stimuli (Sauerland et al., et al., 1980; Fung and Barnes, 198 1). More recent studies show 1967; Chase et al., 1970; Chase, 1980; Wyrwicka et al., 1982). that iontophoretic application of NE on spinal motoneurons Finally, the reflex can be easily and reliably elicited experimen- results in a slow, long-lasting depolarizing potential which de- tally in behaving animals, allowing examination under phys- creases the threshold for action potential production by intra- iological conditions. cellular depolarizing current (Fung et al., 1988). Although they In the first group of experiments, the massetericreflex was make an important contribution, these types of studies are, elicited in behaving cats by means of electrical stimulation of nonetheless, somewhat limited because they were conducted the cell bodies of the sensory neurons located in MesV. Phar- either in anesthetized or spinalized animals and, therefore, can- macologicalagents, locally infuseddirectly into MoV, were used not address the consequences of NE on functional output. to manipulate NE synaptic transmission. The results of these Studies carried out in intact organisms by Davis and col- studies(1) characterize NE modulation of the massetericreflex leagues (reviewed in Davis, 1984) have come closest to under- in the unanesthetized cat; (2) specify the NE receptor subtype standing the functional role of NE in behavior. Their data sug- mediating the effect; and (3) demonstrate neurochemical spec- gest that NE facilitates the acoustic startle reflex in rats. They ificity by examining the effect of NE following pretreatment with find that acoustic startle is increased by drugs that increase the systemic antagonist drugs. Studies described in the accompa- availability of central NE by blocking its reuptake and is de- nying paper demonstrate that thesepharmacological results can creased by lesions ofNE cell bodies in the locus coeruleus (Davis be generalized to physiological activation of NE neurons, pro- et al., 1977). In addition, they report that acoustic startle is duced by various environmental conditions. In addition, they increased following intrathecal administration of NE or the (Y-l- establishthat a causal relationship exists between increasedNE adrenergic agonist phenylephrine (Astrachan and Davis, 198 1). activity and augmentation of the reflex response. In experiments such as these, however, the relative complexity ofthe neural circuitry underlying the behavior studied precludes Materials and Methods determining the specific site of action of NE within the brain Surgical procedure. Adult cats weighing between 2.5 and 4.0 kg were stem and spinal cord. anesthetized with sodium oentobarbitol(35 mdkn, i.p.). Masseter mus- In summary, NE appears to facilitate sensorimotor processes, cle activity was recorded with the uninsulated%psof2’flexible, Teflon- most likely via an LY-1 -adrenergic mechanism. This modulation coated stainless steel electrodes (0.001 in. diameter), inserted 1 cm apart has been fairly well described at a cellular level by numerous into the thickest portion of both masseter muscles, and soldered into iontophoretic studies but is still not well understood at a be- place. The remaining insulated portions of the wires were led subcu- taneously to a small incision in the scalp and externalized. After the havioral level for the reasons outlined above. A fuller under- cheek incisions were sutured, the animal was placed in a stereotaxic standing of the functional role of brain NE systems would be frame. achieved by the use of a simple neuronal system whose circuitry The scalp incision was lengthened along the midline and the tem- received dense NE innervation. In addition, it would be desir- poralis muscle retracted. A stainless steel screw was placed in the frontal bone above the sinus to serve as an indifferent electrode. Another screw, able if this system could be studied in vivo, in intact animals, placed in the skull over the parietal cortex and in contact with the in the absenceofanesthesia or restraint, and under physiological overlying tissue, served as a ground electrode. Two additional screws conditions. were threaded into the skull to anchor the connector assembly. Insulated In a previous paper from this laboratory (Morilak and Jacobs, wires from all electrodes were later soldered to a standard 25-pin connector. 1985), it was proposedthat the masseteric,orjaw-closing, reflex Rhodes Nexus-200 bipolar electrodes, insulated except at the tips, provided a model systemto study the functional effectsof changes were bilaterally implanted in MesV for eliciting the masseteric reflex in NE synaptic transmission. In thosestudies, we reported that according to the technique of Chase et al. (1968). One electrode at a systemic administration of drugs that increasedNE’s synaptic time was lowered at a 30” angle off the vertical (anterior approach) to action augmentedthe amplitude of the evoked reflex response; a starting point -8 mm dorsal to MesV (P -2.5, L +2.5, V -0.5). Precise placement was achieved by mapping the boundaries of the nu- reciprocally, drugs that decreasedthe synaptic action of NE cleus through electrical stimulation while monitoring masseter muscle suppressedthe response. Destruction of the NE input to the activity. As the electrode was slowly advanced toward the target, stim- motor trigeminal nucleus(MoV), by meansofthe direct infusion ulation was applied (500 PA double pulses, 0.5 msec duration, delivered of the neurotoxin 6-hydroxydopamine, significantly decreased at a frequency of 1 Hz) and masseter muscle activity simultaneously the facilitatory effect produced by drugs that increaseNE’s syn- displayed on an oscilliscope. When a response (i.e., mass action EMG potential) was observed on the oscilloscope (as well as visually), stimulus aptic action. These experiments demonstrated that NE modu- duration and intensity were decreased and, if necessary, electrode place- lated this simple behavioral responseand specified the site of ment was adjusted until the response was optimized. The response was the modulation in intact, anesthetizedrats. The present 2 groups judged to be reflexive in nature if it met the following criteria: latency of studieswere designed to investigate the modulatory effects greater than 3 msec following stimulation; slight variability in amplitude; and sensitivity to current level and duration. of NE on the massetericreflex in the behaving cat and to go A similar method was used to implant bilateral, 22-gauge stainless beyond these pharmacological studies to physiologically rele- steel guide cannulae in the MoV nuclei (P -4.5, L +3.7, V -5.0). Each vant conditions. individual cannula was lowered at a 30” angle off the vertical (posterior The massetericreflex was chosen for several reasons.First, it approach) to an intermediate point dorsal to MoV. Final placement was is a simplemonosynaptic reflex with a well-understood circuitry: determined by stimulating through a temporary electrode which ex- tended 1 mm beyond the tip of the guide and making the appropriate stretch receptors in the massetermuscle, and with cell bodies adjustments until the best response to minimal current was obtained. in the mesencephalictrigeminal nucleus (MesV), synapse di- The response was judged to be mediated via direct stimulation of mo- rectly on a-motoneurons in MoV which innervate the masseter toneurons in MoV if it met the following criteria: constant latency less muscle.Second, there is a denseNE innervation of MoV from than 2 msec following stimulation; invariant amplitude; and insensi- tivity to changes in current intensity or duration once a threshold level lateral tegmental NE neurons(Levitt and Moore, 1979; Vomov was reached. The cannulae were cemented at a final point 0.5 mm below and Sutin, 1983; Grzanna et al., 1987). Third, the reflex has the coordinates determined to provide the best response since infused been shown to be modulated acrossthe sleep-wakecycle (Chase liquid spreads upward from the tip in an oval surrounding the cannula The Journal of Neuroscience, January 1990, IO(l) 93 barrel. The stimulating wire was removed and replaced with a 28-gauge the subject, the injector was left in place for the duration of the exper- stylet extending 1 mm beyond the tip of the cannula. imental session (we found that any drug leakage or diffusion was below The leads from the masseteric recording and stimulating electrodes threshold for producing a physiological effect). The test drugs (various were soldered to the connector. The cannulae were housed inside a doses of NE, phenylephrine, isoproterenol, clonidine, or 5-HT) were capped plastic tube and the entire assembly affixed to the skull with administered in a random order; successive infusions were separated dental cement and covered with acrylic. At least 1 week of recovery was by at least 3 d. In subjects with bilateral cannulae, the contralateral side allowed prior to experimentation. During this time, the subjects were could be tested the next day. adapted to the experimental chamber, which was a sound-attenuated Five subjects were pretreated either with the 01-1 -adrenergic antagonist square wooden box (65 x 65 x 95 cm), fitted with a clear Plexiglas prazosin (5 mg/kg, i.p.) or with the serotonergic antagonist methysergide door. (0.5 mg/kg, i.p.) 45 min prior to local infusions of 0.5 pg NE. These Reflex elicitation in behaving animals. The reflex was elicited by de- drugs and doses were chosen because previous studies ofNE modulation livering a square wave pulse to MesV through bipolar electrodes. Stim- of the masseteric reflex in the anesthetized rat (Morilak and Jacobs, ulation was isolated from ground (Grass model SIU5 stimulus isolation 1985) showed that prazosin blocked the increase in the masseteric reflex unit) and delivered as constant current (Grass model CCU 1A constant normally produced by yohimbine, which increases NE release via in- current unit coupled to a Grass S48 stimulator). Optimal response was hibition of presynaptic NE autoreceptors (Starke and Altman, 1973); usually obtained with twin pulse stimulation, pulse duration of 0.05- conversely, studies in the facial motor nucleus show that methysergide 0.10 msec, interpulse interval of 3-4 msec. Current intensity ranged is effective in blocking the facilitating action of 5-HT, but not that of from 50 to 300 PA. NE (McCall and Aahaianian. 1980). We chose to utilize the 0.5-ua dose The signal from the masseter recording electrodes was band-pass of NE because initial iesults’showed that this dose reliably augmented filtered (half-amplitude at 0.3 Hz and 3.0 kHz) and amplified (Grass the masseteric reflex in all subjects and was still on the rising phase of P5 preamp) before being displayed on a storage oscilloscope for direct the dose-response curve. For this series of experiments, the last prein- measurement and stored on tape for subsequent analysis. A subset of fusion values were taken as the baseline and tests were conducted only studies was performed blind. at the low current level (since current level did not seem to differentially Baseline refrpx determinations. Various pulse durations were first tested affect the response to microinfusion of NE). All subjects were also tested in order to determine that producing the most stable responses from with 0.5 pg NE alone in order to provide for within-subject analysis. trial to trial. Next, keeping the optimal duration constant, threshold and Again, the order of treatments was randomly determined. maximal response amplitudes were determined by adjusting current Saline infusions (0.5 ~1 of 0.9% physiological saline) were tested in 6 levels, beginning with a subthreshold level and gradually increasing until subjects in order to control for possible nonspecific effects. In addition, the maximal response was elicited. A baseline current-response curve the pH of the saline solution was adjusted to match that of the 1.O-pg was obtained by electrically stimulating MesV neurons at a rate of 0.1 solution of NE (which produced the largest increase in the reflex re- Hz at each of 3 current levels, beginning with the lowest current that sponse) in order to control for possible pH-related effects. In 4 animals, produced a consistent, suprathreshold response (usually 50 PV ampli- an injection cannula deliberately placed outside of MoV during surgery tude). The middle and highest currents were ones which doubled and allowed us to test for site specificity; this was accomplished by first tripled, respectively, the amplitude of the lowest (if necessary, the cur- localizing MoV and then moving 2 mm above that stereotaxic coor- rent level was adjusted so as not to elicit maximal responses). In general, dinate, making the final placement rostra1 and dorsal to MoV, above observable jaw closure was never elicited by the low current, occasion- the supratrigeminal nucleus. The effects of infusions of 1 .O pg NE were ally elicited by the middle current, and always produced by the high then tested in this control site as described above. current. Eight stimuli were delivered at each current level and the largest Histology. At the conclusion of all experiments, the animals were and smallest response for each level discarded. The mean response overdosed with Nembutal and perfused with normal saline followed by amplitude was then calculated for each current level, resulting in a 10% formalin. The brains were extracted and blocked; coronal sections baseline current-response curve. (40 pm thick) were cut and counterstained with cresyl violet. MoV This procedure was used to generate baseline profiles for each subject cannulae positions were determined by examination of the cannula used in the following experiments. The criteria ensured that similar tracks. Only those subjects with cannulae located in MoV were included response amplitudes would be generated for each subject (- 50 pV, - 100 in data analysis. rV, and - 150 PV at the low, middle, and high current levels, respec- Drugs and doses. Drugs used were NE hydrochloride (Sigma); clo- tively) although the specific values for pulse duration and current levels nidine hydrochloride (Sigma); phenylephrine hydrochloride (Sigma); varied between subjects. A subject was first considered eligible for testing isoproterenol hydrochloride (Sigma); prazosin hydrochloride (Pfizer); once current-response baselines were consistent for at least 3 consecutive 5-hydroxytryptamine bimaleate (Sigma); and methysergide maleate days. From then on, whenever possible, the initial parameters estab- (Sandoz). All drugs, with the exception of prazosin, were dissolved in lished for a given subject continued to be used for all experiments physiological saline; prazosin was dissolved in heated propylene glycol. involving that subject. Doses are expressed as the salt. Solutions were made fresh prior to each Drug infusions. On the day of an experiment, the subject was placed session. in the experimental chamber and prepared for a local infusion. The Data analysis. Effects of a single dose were first analyzed using an stylet was removed and replaced with a preloaded 28-gauge stainless analysis of variance (ANOVA) with current and time as repeated mea- steel injector connected to a Hamilton lo-r1 syringe via PE 20 tubing. sures; post-hoc comparisons between individual time points were then In order to leave the subject undisturbed, the syringe was driven by a analyzed with Newman-Keuls tests. For the antagonism tests, the dif- remotely controlled syringe pump set to deliver 0.5 ~1 over a 1-min ference score between the amplitude following NE alone and that fol- period. Previous studies indicated that slow infusions of this amount lowing NE with antagonist pretreatment was calculated for each subject. of dye. NE. or serotonin (5-HT) via small-aaune cannulae diffuse 0.3- These scores were then entered into a within-subject ANOVA with 3 0.8 mm from the injection site (Meyers et-al.: 1971), thus restricting levels of the independent variable. most, if not all, of the drug to MoV, which in the cat is larger than 1 mm in diameter in all dimensions. Since level ofarousal has been shown to affect the amplitude of the reflex, all testing was conducted during Results quiet waking (QW) behavioral states. These were defined as periods in which there were no gross bodily movements; infrequent or no large Long-term stability eye movements; and a stable tonic EMG (as indicated by the presence The evoked response amplitude was variable for the first week of spontaneous background masseter activity). This issue is also ad- postsurgery and then remained fairly stable from session to dressed in the Discussion section of the accompanying paper. session throughout the duration of the experimental procedure. Once the subject was judged to be in a QW state, baseline current- In agreement with a previous report (Soja et. al., 1987), we found response profiles were determined. The baseline was considered to be that the variability of individual responses within a given run stable if 2 consecutive runs produced similar reponse amplitudes within each current (i.e., < 10% variability); the mean amplitude at each current of 8 trials was minimal (approximately 10%). In general, once was taken as the baseline value. At this point, the drug infusion was stimulation parameters were determined for a given subject begun and, allowing 1 min for diffusion, the reflex test procedure re- prior to the start of experimentation, they continued to produce peated 1,5, 15,30, and 60 min later. Once again, in order not to disturb the same magnitude of response. In some cases, however, it was 94 Stafford and Jacobs * NE Modulation in Behaving Cats. I. Pharmacology

x.t 900 - I 600700- - Current Level ---o--’ low . . . . .- l -.. mid ) g 600 - + high ‘. E _ ‘\ E 80. .- 500 - ‘. ‘, a 60- 2 400 - ‘. !I % ‘. m 300- %. ‘, f.. ‘, m %* \ 200 - -.’ 09 100 h -I 0’ c 01 s .I25 .25 .5 1.0 5.0 0 BL 1 6 15 30 60 2 Dose of NE (pg) 2 TIME POST-NE INFUSION (min) Figure 2. Mean percent of baseline (LX) amplitude of the masseteric Figure 1. Mean masseteric reflex response amplitude (pV f SEM) reflex 5 min postinfusion for all the NE doses tested, at the low (open measured before (baseline, BL) and at various time points following circles), middle (closed circles), and high (closed squares) currentlevels. infusion of 0.125 fig NE into MoV on 2 separate test sessions, 4 weeks The reflex response amplitude was significantly increased by 0.25 /Ig apart, in the same subject. There was no significant difference in response NE at the low current(n = 6 subjects);the 0.5 pg (n = 6) and 1.0pg (n between the first NE test (open circles) and the second (closed circles). = 7) doses significantly increased the reflex response amplitude at all current levels. The 0.125 pg (n = 7) and 5.0 pg (n = 6) doses of NE had no significant effect on the masseteric response amplitude. necessaryto adjust the parametersused (usually increasingcur- rent or duration) in order to maintain a consistent baseline amplitude. This changecould reflect decreasesin the efficacy of of baseline amplitude. This current effect pattern was evident the stimulating electrode over time due, for example, to metal acrossall dosestested [F (2,46) = 66.3, p = -C 0.0011. depositsor gliosis. Second,magnitude of the facilitation waspositively correlated In addition to reliability of baselineresponses, the response to dose of NE, with lower dosesproducing significantly lessof seen following a given microinfused drug dose did not differ an increasein the response[F (4,23) = 3.6, p < 0.0 11.Looking acrossrepeated tests within a given subject [F (5,lO) = 0.249, at resultsfrom the low current, the net increasein amplitude 5 p > 0.051.That is, the samedose produced equivalent increases min after infusions of the lowest dosetested (0.125 Fg NE) was in responseamplitude on different sessions.Representative data 24.9% (k13.8 SEM); 0.25 pg NE produced a 53.2% (k20.9) from one subject are shown in Figure 1. First, note that the increase;0.5 Kg NE increasedthe responseby 209% (? 129.1); baselineamplitudes did not differ acrosssessions, which were the peak effect was seenfollowing infusionsof 1.O pg NE (543.9% 4 weeks apart: baselinewas 55.3 PV on the first test and 48.9 increase, ? 336). Similar resultswere produced at the mid- and KV on the second,the same stimulation parameterswere used high currents, although, as describedabove, the magnitude of during both tests. The first NE infusion (0.125 pg) produced a the increaseswas smaller. However, the highestdose tested, 5.0 50.6% increase in the amplitude of the reflex in min 1 (from pg NE, had no effect on the amplitude of the reflex response, 55.3 FV to 83.3 PV); the secondtime the samedose was tested, producing only a nonsignificant 20.8% (22 1.8) change. This a comparable increase in the amplitude of the reflex was pro- lack of effect at a high NE doseis consistentwith resultsreported duced (from 48.9 PV to 105 rV in min 1). Similar resultswere for iontophoretic NE experiments (Rogawskiand Aghajanian, seenat the other time points: 55% increasein min 5 on test 1 1980) and may reflect, as those results suggest,depolarization vs 68% increaseon test 2; 19% increasein min 15 on test 1 vs block. 21.0% on test 2. In both instances, the massetericreflex am- In addition to magnitude, the time course of the effects was plitude returned to baselineby min 30, indicating that the time also correlated with dose.That is, the shortest-lastingincrease coursewas also unchangedfrom the first test. was the smallestin amplitude (i.e., following infusions of 0.125 pg NE), whereas higher dosesproduced correspondingly larger NE dose-response relationships and longer-lastingincreases. The resultsof ANOVAs with time Infusions of NE into MoV resulted in a dose-dependentfacili- as the repeated measurerevealed that the amplitude of the re- tation of the reflex amplitude which was manifested as an in- sponsewas significantly increasedabove baselinefollowing increasein the evoked massetermuscle response,as illustrated in fusions of 0.125 pg NE [F (5,30) = 4.4, p < 0.005]; 0.25 pg NE Figure 2. This graph presentsthe facilitation (expressedas mean [F(5,25) = 2.6,~ < 0.051; 0.5 pg NE [F(5,25) = 3.8,~ < 0.011; percent baseline) observed 5 min postinfusion for all the NE and 1.0 pg NE [P (5,30) = 6.6, p < O.OOl]. The highest dose dosestested at the 3 current levels. Inspection of the figure tested, 5.0 pg NE, had no significant effect [F (5,15) = 1.7, p > reveals, first, that the magnitude of the increasewas inversely 0.051. Post hoc comparisons between individual time points related to current; that is, the largestfacilitation was observed using Newman-Keuls tests revealed that, at all current levels, at the lowest current level. For example, at the low current, the 0.125 pg NE significantly increased the amplitude of the re- evoked amplitude of the reflex was increasedto 643.9% of base- sponsefor only 1 min postinfusion; the responsehad returned line by infusions of 1.0 rg NE; at the midcurrent, the increase to baselineby 5 min after the infusion and remained so for the was 393.7% of baseline,and at the highestcurrent level, 235.9% duration of the test. The next higher dose,0.25 rg, significantly The Journal of Neuroscience, January 1990, 70(l) 95

4 =I 400- Current Level .= --g-’ 1.0 UII PE . . . . .-. 1 .o LlFl IS0 le + .25 pg CLON E < 300- Q) 5007 c m- 400 - i! z = 200- 3 300- s m 200- zi loo- 2s I BL 1 5 15 30 60 : ‘OL------c OJ n 2 BL 1 5 15 30 60 2 Time Post lpg NE (min) I Time Post Infusion (min) Figure 3. Mean percent ofbaseline (BL) amplitude in masseteric reflex measured at various time points following infusion of 1.0 pg NE into Figure 4. Comparison of mean percent of baseline (IX) amplitude of MoV. Mean baseline values were 3 1.8 rV (t 3.8 SEM) at the low current, the masseteric reflex response at various time points following infusion 105.5 + 14.9 PV at the middle current, and 266.3 k 57.8 rV at the of either the o-1-adrenergic agonist phenylephrine (PE, 1.0 pg; n = 7 high current (n = 7 subjects). NE significantly increased the reflex response amplitude at the l-, 5-, and 15-min time points at all current subjects), the or-2-adrenergic agonist clonidine (CLON, 0.25 pg; n = 6), or the @-adrenergic agonist isoproterenol (ISO, 1.O ~a; n = 6). Baseline levels; in addition, at the high current, the reflex response amplitude was significantly increased at the 30-min time point. amplitude values were 47.3 PV (k4.63 SCM) for the-PE group, 41.9 + 6.52 I.LVfor CLON. and 52.7 k 6.15 uV for the IS0 -erouu. . PE infusions significantly increased the reflex response amplitude at all time points, CLON significantly decreased the amplitude. increasedthe responseat the 1- and 5-min time points, but only at the low current level; no significant changewas seenat the other current levels. After infusion of 0.5 pg NE, the response Antagonist tests remained significantly elevated for 5 min at all currents. The Systemic administration of both prazosin, an (Y-1 -noradrenergic longest-lastingincrease was seenwith 1.0 Mg NE; at this dose, antagonist (5 mg/kg, i.p.) and the serotonergicantagonist meth- the responseremained significantly above baselinefor 15 min ysergide (0.5 mg/kg, i.p.) decreasedthe baselineamplitude of at the low and middle currents and for 30 min at the high current the massetericreflex. Neither antagonist produced any dramatic (Fig. 3). drug-induced state changes.However, methysergide pretreatment did not prevent the increase normally seenfollowing in- Agonist tests fusionsof 1.O MgNE, whereasprasozin pretreatment completely NE agonistswere tested in order to pharmacologically classify blocked the increase(Fig. 5). One minute after infusion of 1.0 the receptor subtype mediating the NE facilitation of the reflex. pg NE, the amplitude of the massetericreflex was increasedto The results at the low current level are presented in Figure 4. 582% of baselinein untreated subjectsand by 594% in the same Infusions of the (Y-1 -adrenergic agonistphenylephrine produced subjectswhen pretreated with methysergide;it decreasedto 90.5% a moderate, but significant, increase in the reflex amplitude at of baselinefollowing prazosin pretreatment. An ANOVA com- the 2 dosestested, 0.5 rg [F(5,25) = 3.6, p < 0.011 and 1.0 pg paring the responseto 1.0 pg NE following prazosin or meth- [F (4,24) = 8.1, p < 0.0031. The increase was evident at all ysergide to that following NE alone showsthat the responseto current levels by 1 min postinfusion and remained elevated for NE was significantly different following prazosin, but not meth- up to 60 min. Phenylephrine seemedto be lesspotent than NE ysergide, pretreatment [F (2,8) = 4.72, p < 0.051. This same in that the increaseseen after phenylephrine infusionswas small- dose of methysergide was, however, effective in blocking an er than that after NE, particularly at the low current: 266.4% of increaseseen in responseto infusions of 5-HT. In another group baseline5 min postphenylephrine vs 643.4% of baseline5 min of subjects,0.125 PLg5-HT increasedthe amplitude of the reflex post-NE. to 300.6% of baselinevalues. When the samedose of 5-HT was Infusions of the ol-2-adrenergicagonist clonidine (0.25 pug) infused following methysergide(0.5 mg/kg, i.p.), the increasein produced a long-lasting, significant depressionof the masseteric the amplitude of the reflex was markedly attenuated (116% of reflex [F (5,15) = 3.7-9, p < 0.021. By the first minute, the baseline).This demonstratesthat the inability of methysergide amplitude was reduced to 70.7% of baselineat the low current, to block an increase to NE was pharmacologically specific. 89.9% at the midcurrent, and 84.1% at the high current. The decreasewas seen l-5 min postinfusion and lasted for 60-90 Controls min. The peak suppressionwas seen 15 min after clonidine, Consistent with a previous report (Soja et al., 1987), infusion 54% of baselineat the low current level, 60.2% at the midcurrent of 0.5 ~1 of saline, same pH (5.5) as a 1.0~pgNE solution, did level, and 63.3% of baselineat the high current level. not changethe amplitude of the massetericreflex at any time Infusions of the P-adrenergic agonist isoproterenol, by con- point tested [F (4,20) = 1.02,p > 0.05; Fig. 6, top]. In addition, trast, had no effect on the amplitude of the massetericreflex at infusion of 1.0 pg NE in control sites 2 mm above MoV had either of the 2 dosestested, either 0.5 pg [F (4,20) = 1.02, p > no significant effect on the reflex response[F (5,15) = 0.94, p 0.051 or 1.0 pg [F(4,20) = 0.94, p > 0.051. > 0.05; Fig. 6, bottom]. 96 Stafford and Jacobs * NE Modulation in Behaving Cats. I. Pharmacology

1400 150- Current Level :xg:. :; C 1200 125- + high .= z x 1000 -c- =K-: I T 600 E B .-z 600 m" 50-

: 400 s 25- a 5

8 200 2 ' BL 1 5 15 30 60 5 0 Time Post Saline Infusion (min) Tik Post Inkion (mini5

Figure 5. Within-subject comparison of mean percent of baseline (BL) amplitude of the masseteric reflex response following infusion of 0.5 pg NE alone or following pretreatment 45 min earlier either with the a-adrenergic antagonist prazosin (PRAZ, 5.0 mg/kg, i.p.) or with the 5-HT antagonist methysergide (0.5 mg/kg, i.p.). Baseline amplitudes for the 3 conditions were 45.5 WV (+5.7 SEM) for NE alone, 26.8 + 5.1 PV for NE/METHY, and 35.3 f 10.1 rV for NUPRAZ. Tests were conducted at the low current level; n = 5 subjects for each condition. The response to NE following methysergide pretreatment did not differ from that following NE alone. The response to NE following prazosin pretreatment differed significantly from that in the other conditions. 8 25 Discussion 5 J , The resultsof the present study show that microinfusion of NE ' ' BL 1 5 15 30 60 into the MoV increasesthe amplitude of the evoked masseteric reflex response in the intact, unanesthetized cat in a dose-de- Time Post NE in Control Site (min) pendent fashion. The facilitation observed was evident within Figure 6. Top, Mean percent of baseline (BL) amplitude of the mas- 1 min of NE infusion and, at the higher doses,lasted for up to seteric reflex measured at various time points following infusion of 0.5 30 min. This long duration most likely reflectsan overwhelming ~1 of 0.9% saline, pH 5.5, into MoV. Mean baseline values (n = 6 of the NE uptake system by the presenceof artificially high NE subiects) were 35.6 (+3.7 SEM) at the low current. 102.6 f 7.6 uV at levels produced by the infusions. the-middle current, and 229 f ‘30.8 pV at the high current. Saline did not significantly affect the reflex amplitude at any time point. Bottom, Although the facilitatory effect was observed at each of the 3 Mean percent of baseline amplitude of the masseteric reflex amplitude current levels used, the degreeof NE facilitation was inversely measured at various timepoints following infusion of 0.5 ~1 of 1.0 pg related to current strength (i.e., the largestpercent increasewas NE in a control site 2 mm above MoV. Tests were conducted at the evident at the low current, the next larger at the middle current, low current, and mean baseline values (n = 4 subjects) were 63.1 (+4.4 and the smallest at the high current). In addition, the highest SEM). There was no significant effect at any time point. NE dose tested (5 rg) had no effect on the amplitude of the reflex responseat any current level. Consistent with thesefind- ings, Rogawskiand Aghajanian (1980) report that NE facilita- by the P-agonist isoproterenol, suggestingthat the facilitation tion of lateral geniculate neurons was produced with low ion- was mediated via postsynaptic (Y-1 -adrenergic receptors. Infu- tophoretic currents. Although increasing the ejection current sions of the a-2-adrenergic agonist clonidine, which decreases within a limited range strengthened the response,currents in NE releasevia autoreceptoractivation (Starkeand Altman, 1973) excessof that range resulted in a depressionor lack of effect, decreasedthe amplitude of the massetericreflex. This finding, similar to what we observed herewith the highestNE dose.The together with the observation that prazosin pretreatment also diminished responseseen at high dosesor currents could reflect decreasesthe baselineamplitude, is consistent with this general depolarization block or tachyphylaxis. Although this was not hypothesis and further suggeststhat tonic NE releasemay play tested directly in the present experiment, Rogawski and Agha- a role in modulating the reflex. Thus, theseresults demonstrate janian (1980) report that lateral geniculate neurons did exhibit that pharmacologically induced changesin NE transmissionin a widening and a decreasedamplitude of the extracellular action MoV increasethe amplitude of the massetericreflex in behaving potential, suggestingthat overdepolarization can occur in re- cats and extend the resultswe previously reported in the anes- sponseto the application of NE with supramaximal currents. thetized rat (Morilak and Jacobs, 1985). The increasenormally seenfollowing NE infusion was blocked The mechanismby which NE increasesthe amplitude of the by pretreatment with the specific a-1 receptor antagonist pra- massetericreflex was not directly tested in the present experi- zosin, but not by pretreatment with the serotonergic antagonist ment. However, the relative lack of intemeuronsin MoV (Sessle, methysergide. Consistent with this, the NE facilitation was 1977; Card et. al., 1986) suggeststhat the facilitatory effects of mimicked by the (Y-1 -adrenergic agonistphenylephrine, but not NE are causedeither by an action on the terminals of afferent The Journal of Neuroscience, January 1990. IO(l) 97

fibers or by a direct action on MoV neurons. A presynaptic under which NE neurons are activated-such as stress, arousal, action is unlikely, however, since ultrastructural studies of MoV or interaction with novel environmental stimuli (Foote et al., have, to date, failed to reveal axoaxonic connections involving 1983; Jacobs, 1986; Abercrombie and Jacobs, 1987). This hy- NE neurons (Card et. al., 1986; Card, personal communication). pothesis predicts that conditions which produce physiological Our results are consistent with a postsynaptic action on MoV release of NE should also enhance behavioral responses such as motoneurons since the NE effect was mimicked by phenyl- the masseteric reflex, and that this enhancement would be cor- ephrine, an a-l receptor agonist, and blocked by pretreatment related to the strength and duration of activation. This hypoth- with prazosin, an alpha-l antagonist. Moreover, infusions of esis was tested in the experiments presented in the following NE in a control site outside of MoV had no effect on the am- paper. plitude of the reflex response, indicating that the facilitation was a result of the action of NE in MoV. References Indeed, there is much evidence that NE facilitation of neu- Abercrombie, E. D., and B. L. Jacobs (1987) Single-unit response of ronal activity, in a number of other target areas, is due to a noradrenergic neurons in the locus coeruleus of freely-moving cats. postsynaptic action at alpha receptors. Electrical stimulation of I. Acutely presented stressful and non-stressful stimuli. J. Neurosci. 7: 2837-2843. NE neurons in the locus coeruleus of decerebrate cats facilitates Amaral, D. G., and H. M. Sinnamon (1977) The locus coeruleus: the monosynaptic response of lumbar motoneurons to afferent Neurobiology of a central noradrenergic nucleus. Prog. Neurobiol. 9: (dorsal root) inputs; this effect is partially blocked by systemic 147-196. administration of the a-adrenergic antagonist phenoxybenza- Astrachan, D. I., and M. Davis (198 1) Spinal modulation of the acoustic startle response: The role of norepinephrine, serotonin, and do- mine (Strahlendorf et al., 1980). Intracellular studies show that pamine. Brain Res. 206: 223-228. this spinal motoneuron facilitation is associated with a small Card, J. P., J. N. Riley, and R. Y. Moore (1986) The motor trigeminal membrane depolarization, which is also partially blocked by nucleus in the rat: Analysis of neuronal structure and the synaptic phenoxybenzamine (Fung and Barnes, 198 1). Similarly, in anes- organization of noradrenergic afferents. J. Comp. Neurol. 250: 469- thetized rats, the facilitation of the masseteric reflex produced 484. Chase, M. H. (1980) The motor functions of the reticular formation by electrical stimulation of the area of the lateral lemniscus (the are multifaceted and state-determined. In The Reticular Formation source of the NE input to MoV in the rat) is also blocked by Revisited, J. A. Hobson and M. A. B. Brazier, eds., pp. 449472, ol-adrenergic antagonists (Vomov and Sutin, 1986). Iontophor- Raven, New York. etic application of NE on facial motor neurons (McCall and Chase, M. H., and M. Babb (1973) Masseteric reflex response to reticular stimulation reverses during active sleep compared with wake- Aghajanian, 1979; VanderMaelen and Aghajanian, 1980), spin- fulness or quiet sleep. Brain Res. 59: 421-426. al motor neurons (White and Neuman, 1980), and lateral ge- Chase, M. H., D. J. McGinty, and M. B. Sterman (1968) Cyclic vari- niculate neurons (Rogawski and Aghajanian, 1980) also pro- ations in the amplitude of a brainstem reflex during sleep and wake- duces a small membrane depolarization which is blocked by fulness. Experientia 24: 4 148. a-adrenergic antagonists. The NE effect in these previous studies Chase, M. H., Y. Nakamura, and S. Torti (1970) Afferent vagal modulation of brainstem somatic reflex activity. Exp. Neurol. 27: 534- was considered modulatory in that the small depolarization did 544. not produce neuronal activation but, rather, decreased the Dahlstrom, A., and K. Fuxe (1964) Evidence for the existence of threshold for activation by other afferent inputs. monoamine-containing neurons in the central nervous system. I. In other target areas, NE has been shown to produce en- Demonstration of monoamines in the cell bodies of brain stem neurons. Acta Physiol. Stand. 62 (Suppl. 247): 39-85. hancement of inhibitory inputs. Iontophoretic application of Davis, M. (1984) The mammalian startle response. In Neural Mech- NE, at doses that have no effect on spontaneous activity, aug- anisms of Startle Behavior, R. C. Eaton, ed., pp. 241-263, Plenum, ments GABA or synaptically induced inhibition in cerebellar New York. Purkinje cells (Yeh and Woodward, 1983) and in somatosensory Davis, M., J. M. Cedarbaum, G. K. Aghajanian, and D. S. Gendelman cortex cells (Waterhouse and Woodward, 1980). Thus, NE ap- (1977) Effects of clonidine on habituation and sensitization of acoustic startle in normal, decerebrate and locus coeruleus lesioned rats. plication appears capable of enhancing both excitatory and in- Psychopharmacology 51: 243-253. hibitory afferent inputs. The net effect of NE enhancement com- Euler, U. S. von (1956) Noradrenaline, Charles C Thomas, Springfield, bined with the reported inhibition of background activity is to IL. increase the “signal-to-noise” ratio, thereby facilitating infor- Foote, S. L., F. E. Bloom, and G. Aston-Jones (1983) Nucleus locus coeruleus: New evidence of anatomical and physiological specificity. mation transfer through the circuitry. Physiol. Rev. 63: 844-9 14. Our results support previous neurophysiological studies in- Fung, S. J., and C. D. Barnes (198 1) Evidence of facilitatory coeru- dicating that NE facilitates the motoneuron response to excitato- losuinal action in lumbar motoneurons of cats. Brain Res. 216: 299- ry inputs and extend this observation from a cellular to a be- 3ii. havioral level. We demonstrated that pharmacological increases Fung, S. J., S. R. White, and C. D. Barnes (1988) Norepinephrine facilitates spinal motoneuron excitability recorded intracellularly in in the level of NE transmission increase the responsivity of this decerebrate cats. Sot. Neurosci. Abstr. 14: 2 14. simple behavioral response in the unanesthetized, behaving cat. Grzanna, R., W. K. Chee, and E. W. Akeyson (1987) Noradrenergic The present results are consistent with the hypothesis that NE projections to brainstem nuclei: Evidence for differential projections acts to enhance transmission through reflex circuits, thereby from noradrenergic subgroups. J. Comp. Neurol. 263: 76-9 1. Hino, M. H., H. Ono, and H. Fukada (1984) Supraspinal tonic influ- facilitating the motor response mediated by that circuit. More- ence on spinal reflexes and involvement in the effect of chlorprom- over, they validate the utility of the masseteric reflex as a model azine. Gen. Pharmacol. 15: 155-l 58. for studying NE transmission in the CNS of behaving animals. Jacobs, B. L. (1986) Single-unit activity of locus coeruleus neurons in These findings are relevant not only within the context of behaving animals. Prog. Neurobiol. 27: 183-l 94. motor control, but to a more general hypothesis concerning the Levitt, P., and R. Y. Moore (1979) Origin and organization of brainstem catecholamine innervation in the rat. J. Comp. Neurol. 186: role of NE in the normal functioning of an organism. A current 505-528. view posits a role for NE in modulating an organism’sreactivity McCall, R. B., and G. K. Aghajanian (1979) Serotonergic facilitation to stimuli. Thus, a response is likely to be enhanced at conditions of facial motoneuron excitation. Brain Res. 169: 1 l-27. 99 Stafford and Jacobs - NE Modulation in Behaving Cats. I. Pharmacology

McCall, R. B., and G. K. Aghajanian (1980) Pharamacological char- and C. D. Barnes (1980) Facilitation of the lumbar monosynaptic acterization of serotonin receptors in the facial motor nucleus: A reflexes by locus coeruleus stimulation. Neuropharmacology 19: 225- microiontophoretic study. Eur. J. Pharm. 65: 175-183. 230. Meyers, R. D., M. Tytell, A. Kawa, and T. A. Rudy (197 1) Microin- VanderMaelen, C. P., and G. K. Aghajanian (1980) Intracellular stud- jection of ZH-acetylcholine, 14C-serotonin, and ‘H-norepinephrine into ies showing modulation .of facial motoneuron excitability by sero- the hypothalamus of the rat: Diffusion into tissue and ventricles. tonin. Nature 287: 346-347. Physiol. Behav. 7: 743-75 1. Vomov. J. J.. and J. Sutin (1983) Brainstem nroiection to the normal Moore, R. Y., and F. E. Bloom (1979) Central catecholamine neuron and noradrenergically hyperinnervated trigeminal motor nucleus. J. systems: Anatomy and physiology of the norepinephrine and epi- Comp. Neurol. 214: 198-208. nephrine systems. Annu. Rev. Neurosci. 2: 113-l 68. Vomov, J. J., and J. Sutin (1986) Noradrenergic hyperinnervation of Morilak, D. A., and B. L. Jacobs (1985) Noradrenergic modulation the motor triaeminal nucleus: Alterations in membrane properties of sensorimotor processes in intact rats: The masseteric reflex as a and responsesto synaptic input. J. Neurosci. 6: 30-37. model system. J. Neurosci. 5: 1300-l 306. Waterhouse, B. D., and D. J. Woodward (1980) Interaction of nor- Rogawski, M. A. (1985) Norepinephrine. In Neurotransmitter Actions eninenhrine with cerebrocortical activity evoked by stimulation of a in the Vertebrate Nervous System, M. A. Rogawski and J. L. Barker, sbmatosensory afferent pathway in the rat. Exp. Neurol. 67: 1 l-34. eds., pp. 241-284, Plenum, New York. White, S. R., and R. S. Neuman (1980) Facilitation of spinal moto- Rogawski, M. A., and G. K. Aghajanian (1980) Activation of lateral neuron excitability by 5-hydroxytryptamine and noradrenaline. Brain geniculate neurons by norepinephrine: Mediation by an a-adrenergic Res. 188: 119-127. _ - - -- receptor. Brain Res. 182: 345-359. White, S. R., and R. S. Neuman (1983) Pharmacological antagonism Sauerland, E. K., T. Knauss, Y. Nakamura, and C. D. Clemente (1967) of facilitatorv but not inhibitorv effects of serotonin and norepineph- Inhibition of monosynaptic and polysynaptic reflexes and muscle tone rine on excitability of spinal mbtoneurones. Neuropharmacology 22: by electrical stimulation of the cerebral cortex. Exp. Neurol. 17: 159- 489-494. 171. Woodward, D. J., H. C. Moises, B. D. Waterhouse, B. J. Hoffer, and Sessle, B. J. (1977) Identification of alpha and gamma motoneurons R. Freedman (1979) Modulatory actions of norepinephrine in the and effects of stimulation of amygdala, cerebellum, and cerebral cor- central nervous system. Fed. Proc. 38: 2 109-2 116. tex. Exp. Neurol. 54: 303-322. Wyrwicka, W., M. H. Chase, and C. D. Clemente (1982) Changes in Soja, P. J., D. M. Finch, and M. H. Chase (1987) Effect of inhibitory the masseteric reflex during hypothalamic aphagia in cats. Exp. Neu- amino acid antagonists on masseteric reflex suppression during active rol. 77: 194-207. sleep. Exp. Neurol. 96: 178-193. Yeh, H. H., and D. J. Woodward (1983) Beta-l adrenergic receptors Starke, K., and P. Altman (1973) Inhibition of noradrenergic trans- mediate noradrenergic facilitation of Purkinje cell responses to gam- mission by clonidine: An action on pre-junctional a-receptors. Neu- ma-aminobutyric acid in cerebellum of rat. Neuropharmacology 22: ropharmacology 12: 339-347. 629-639. Strahlendorf, J. C., H. K. Strahlendorf, R. E. Kingsley, J. Gintautus,