Neuromodulation of Vertebrate

doi: 10.1152/physiol.00013.201126:393-411, 2011. ;Physiology

Gareth B. Miles and Keith T. SillarNetworksNeuromodulation of Vertebrate Locomotor Control

You might find this additional info useful...

193 articles, 85 of which you can access for free at: This article citeshttp://physiologyonline.physiology.org/content/26/6/393.full#ref-list-1

2 other HighWire-hosted articles: This article has been cited by http://physiologyonline.physiology.org/content/26/6/393#cited-by

including high resolution figures, can be found at: Updated information and serviceshttp://physiologyonline.physiology.org/content/26/6/393.full

can be found at: Physiology about Additional material and informationhttp://www.the-aps.org/publications/physiol

This information is current as of April 7, 2013.

Physiol. Soc.. ESSN: 1548-9221. Visit our website at http://www.the-aps.org/. American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. ©2011 Int. Union Physiol. Sci./Am.

thephysiological developments. It is published bimonthly in February, April, June, August, October, and December by (formerly published as News in Physiological Science) publishes brief review articles on majorPhysiology

by guest on April 7, 2013

http://physiologyonline.physiology.org/D

ownloaded from

http://physiologyonline.physiology.org/content/26/6/393#cited-by

http://physiologyonline.physiology.org/

Neuromodulation of VertebrateLocomotor Control Networks

Vertebrate locomotion must be adaptable in light of changing environmental,

organismal, and developmental demands. Much of the underlying flexibility in

the output of central pattern generating (CPG) networks of the spinal cord

and brain stem is endowed by neuromodulation. This review provides a

synthesis of current knowledge on the way that various neuromodulators

modify the properties of and connections between CPG neurons to sculpt

CPG network output during locomotion.

Gareth B. Miles and Keith T. SillarSchool of Biology, University of St. Andrews,

St. Andrews, Scotland, United [email protected]

The neural networks that control locomotion invertebrates, so called central pattern generators(CPGs), are located primarily in the spinal cord.CPGs produce a basic rhythmic motor output inthe absence of sensory feedback, largely from theinterplay between the electrical properties of theconstituent neurons and the nature of the synapticinterconnections between them. However, thesecomponents of CPGs are subject to neuromodula-tion deriving from a wide range of sources bothintrinsic to the spinal cord and also projecting tothe spinal cord from other, extrinsic sources. Theneuromodulators involved are chemically diverse,ranging from simple amino acids to biogenicamines, peptides, and even a gas, nitric oxide. Forthe most part, these neuromodulators act on G-protein-coupled receptors to alter the concentration ofintracellular second messengers. The response prop-erties of individual CPG neurons can be dramaticallymodified in the presence of any one of this range ofneuromodulators. Growing evidence points to theexistence of complex interactions between differ-ent modulatory inputs to the motor system. Ineffect, the prevailing “cocktail” of neuromodula-tors endows locomotor CPGs with an almost infi-nite range of output configurations. This isnecessary to enable the short term flexibility andlong term adaptability required of locomotion dur-ing the life time of the individual.

Need for Flexibility in SpinalMotor Network Output

The precision, grace, and elegance with which an-imals navigate smoothly and efficiently throughtheir environment ultimately results from a com-plex set of interactions between the individualneural components, the CPG networks that driverhythmic firing in motoneuron, the biomechanicalproperties of peripheral propulsive structures, andthe integration of sensory information with centralmotor control networks. The ability of a neural

system to adapt in the face of external demands isoften essential for survival, and this ability in turnrequires that it generates a motor output that isinherently flexible. Neuromodulators represent animportant source of this flexibility at the level ofthe underlying neuronal networks of the centralnervous system and the motor output they pro-duce. Here, we review current evidence on thesources and effects of the various neuromodulatorsthat influence vertebrate locomotor control net-works and provide an account of how they achievetheir modulatory influences.

Role of Neuromodulation inModifying Centrally GeneratedActivity

The rhythmic firing patterns of motoneuron that un-derpin locomotory movements such as walking,swimming, or flying, result from the activity andinter-connections of premotor interneurons thatform complex networks residing in the spinal cordand brain stem (reviewed in Ref. 65). Neuromodula-tors change locomotor activity via one or both of tworoutes: modulation of the integrative electrical prop-erties of motoneuron and CPG interneurons, andmodulation of the synaptic connections between in-terneurons and motoneuron. As a broad rule ofthumb, it is possible to infer the locus of a neuro-modulators influence on the locomotor output bydetermining whether it affects the frequency and/orthe amplitude of the output. The hypothetical loco-motor rhythm shown in FIGURE 1 illustrates motornerve recordings from functionally antagonistic mo-torneuron pools in which the pattern of burstingactivity strictly alternates during rhythm generation.

If the frequency of the rhythm stays the same inthe presence of a neuromodulator but the amplitudeor duration of each motor burst changes, then theeffects are most likely occurring predominantly at thelevel of the motoneuron (or last order interneurons;Bi in FIGURE 1). Alternatively, if the burst properties

REVIEWSPHYSIOLOGY 26: 393–411, 2011; doi:10.1152/physiol.00013.2011

1548-9213/11 ©2011 Int. Union Physiol. Sci./Am. Physiol. Soc. 393



ownloaded from


are unaffected but the frequency is altered, then theneuromodulator is most likely affecting rhythm-generating circuitry without major influences ondownstream neurons (Bii in FIGURE 1). In manycases, both amplitude and frequency change,which suggests that the neuromodulator affectsmotoneurons and CPG neurons simultaneously(Biii in FIGURE 1).

Conceptually, there are two possible sources ofneuromodulator that can affect locomotor CPGs(89): extrinsic sources such as modulatory inputsthat descend from the brain and that affect CPGsremotely, and intrinsic modulators that are re-leased from cells that are embedded within theCPGs (89). In both cases, neuromodulators usuallyexert their effects by acting as agonists at metabo-tropic receptors, some of which increase excitabil-ity, whereas others decrease excitability.

Opposing Effects of Neuromodulatorson Sensory vs. Motor Systems

For monoamines, 5-hydroxytryptamine (5-HT)in particular, there is a strong inverse correlationbetween their effects on motor output and sensory

input: if motor output is facilitated, then sensoryinputs are diminished and vice versa. This seemslogical because at times when an increase in locomo-tor activity is necessary, a down-tuning of sensorytransmission will help to ensure that innocuousextrinsic sensory inputs do not interfere with on-going locomotor activity. How are these opposingeffects accomplished? The most likely explanationfor this functional dichotomy is that the receptorsto which 5-HT is coupled are expressed differen-tially in sensory and motor circuits such that re-ceptors in one pathway are positively coupled to asecond messenger and are negatively linked to thesame second messenger in the other pathway andvice versa. In addition, monoamines differentiallymodulate ascending sensory information relatingto locomotor movements (68). However, this re-view focuses on neuromodulation of motor sys-tems, and effects on sensory systems will not bediscussed further.

Role of Neuromodulation DevelopmentalNetwork Adaptation

Some of the neuromodulators that exert acutemodulation of locomotor activity also play important

FIGURE 1. Schematic showing rhythmic, alternating motor output generated by antagonistic motor poolsSchematic showing how rhythmic, alternating motor output generated by antagonistic motor pools (A) can be al-tered in amplitude (i), frequency (ii), or both (iii), depending on whether a given modulator acts at the level of themotoneurons (Bi), the CPG (Bii), or a combination of the two (Biii).

REVIEWS

PHYSIOLOGY • Volume 26 • December 2011 • www.physiologyonline.org394



ownloaded from


roles in the longer term during motor system de-velopment. 5-HT, for example, has been shown ina variety of vertebrate motor systems to have amajor influence on circuit assembly during earlydevelopment, and, in keeping with this role, thefibers that invade the spinal cord from the brainstem arrive there at the critical points in circuitmaturation when major changes in the motor out-put are taking place. In chick embryos, ablation ofserotonergic projections in ovo using neurotoxinsdramatically affects synaptogenesis of inputs tospinal motoneuron (130a). In Xenopus frog tad-poles, application of a 5-HT neurotoxin or 5-HTreceptor antagonist during embryogenesis pre-vents the normal postembryonic maturation ofswimming activity, suggesting that descending ra-phespinal projections are causal to locomotor net-work development (149). In rodents, as in lowervertebrates, descending projections of the raphesystem also invade the spinal cord at critical stagesin network assembly, and 5-HT profoundly modu-lates locomotor activity, inferring a prominent rolefor the amine in development (168).

Sources and Types ofNeuromodulator

The neuromodulators that have been shown toaffect vertebrate locomotion fall into three maincategories: the biogenic amines (5-HT, noradrena-line, dopamine, and trace amines), amino acidsnormally acting on metabotropic receptors (GABAB,mGluRs), and peptides (e.g., substance P). In addi-tion, many other molecules that do not fit neatlyinto these categories have also been shown to alterlocomotor activity including the purines (ATP andadenosine), d-serine, endocannabinoids, and ni-tric oxide. In the following sections, we summarizeevidence on the sources and effects of these neu-romodulators across a range of vertebrate locomo-tor networks.

Biogenic Amines

The sources of the amines depend partly on thespecies of vertebrate in question. Homologousgroups of aminergic neurons are located in thebrain stem of most if not all species, including theraphe nuclei (5-HT), the locus coeruleus (nor-adrenaline), and the substantia nigra (dopamine).Additional groups include in mammals the sero-tonergic parapyramidal region (PPR) of the brainstem and the A11 dopaminergic nucleus of thedorsoposterior hypothalamus. In lower verte-brates, notably the lamprey, there are additionalsources of amines located in the spinal cord itself(see below).

Evolution of 5-HT Signaling

5-HT is a phylogenetically ancient signaling mole-cule, distributed throughout the metazoans, whichplays key roles in the development and modulationof locomotor function in vertebrates and inverte-brates alike. In chordates like the lancelet, immunocy-tochemical evidence reveals serotonergic neurons inthe dorsal cerebral vesicle and in the spinal cordarranged in a ladder-like ventral chain close to thecentral canal (23). Lampreys belong to a primitivevertebrate group, the agnathans, which split awayfrom the main vertebrate line around 500 millionyears ago. Lampreys possess an intrinsic modula-tory plexus running as a strip spanning the ventralmidline, which serves as an important source of5-HT (183), dopamine (156), and other modulatorslike substance P.

5-HT in Mammalian Systems

Almost all 5-HT in the mammalian spinal cordoriginates from brain stem raphe nuclei and theparapyramidal region (reviewed in Ref. 99). A rangeof spinal neurons relevant to locomotor control inmammals receive serotonergic input, as identifiedby their close association with 5-HT-labeled syn-aptic boutons. These neuronal targets for 5-HTinclude spinal motoneurons (3), Renshaw cells(26), commissural interneurons of laminae VII andVIII (67), V0C interneurons (182), V2-derived in-terneurons (2), Hb9� interneurons (178), and neu-rons, predominantly in laminae VII and VIII,shown to be activated during locomotion via c-fosexpression (128). There is also widespread expres-sion of 5-HT receptors in the mammalian spinalcord, with 5-HT2 receptors dominating in the ven-tral horn (reviewed in Ref. 144).

5-HT can both activate and modulate mamma-lian locomotor networks. Activation of spinal loco-motor networks by 5-HT was first shown in therabbit using the 5-HT precursor 5-HTP (167). 5-HTwas later shown to also initiate locomotor activityin isolated spinal cord preparations obtained fromneonatal rats and mice (27, 36, 92, 112, 127) andspinalized adult rats and mice (55, 100, 165). Incontrast, 5-HT appears insufficient to initiate loco-motion in the spinalized cat (6, 7). However, 5-HTis not without effect on cat locomotion. In spinal-ized cats, activation of 5-HT receptors leads to anincrease in the amplitude and duration of bursts ofon-going locomotor-related EMG activity recordedfrom hindlimb muscles (6, 7). In addition, in catsrecovering from incomplete spinal transections,5-HT increases the regularity of stepping (21).5-HT also modulates on-going locomotor activityin rodents with the effects varying depending onthe receptor subtypes activated. The effects of5-HT2 and 5-HT7 receptor activation are generally

REVIEWS

PHYSIOLOGY • Volume 26 • December 2011 • www.physiologyonline.org 395



ownloaded from


facilitatory causing increases in the frequencyand/or amplitude of locomotor activity recordedfrom in vitro preparations (14, 46, 62, 108). In ad-dition, the activation of 5-HT2 and 5-HT7 receptorsstrengthens left/right and flexor/extensor alterna-tion during fictive locomotion (107, 132). Incontrast, the activation of 5-HT1 receptors has “in-hibitory” effects slowing the frequency of locomo-tor activity (25, 46). Taken together, these dataindicate that 5-HT can modulate motoneurons orlast-order interneurons to affect the magnitude ofthe final output of the locomotor network and in-terneurons involved in rhythm generation to affectthe timing and coordination of locomotor activity.

Neuromodulators affecting motoneurons oftentarget persistent inward currents (PICs). PICs canamplify synaptic inputs and give rise to plateaupotentials that allow for periods of sustained firing(reviewed in Ref. 74). PICs and their associatedplateau potentials are enhanced by 5-HT receptoractivation in the cat (37, 81), rat (15, 73, 106), andturtle (82, 134). This enhancement appears to in-volve the facilitation of Na� and Ca2� PICs alongwith a reduction in K� currents (16, 73, 82, 106,134). Thus direct or indirect modulation of PICs inmotoneurons is likely to contribute to the effects of5-HT on the amplitude of locomotor output fromspinal networks. Other modulatory effects of 5-HTon motoneuronal properties that may contributeinclude a reduction in the voltage sensitivity ofNMDA channels and subsequent facilitation ofNMDA receptor-dependent oscillations (110, 111)and modulation of inwardly rectifying K� channelsand Ih currents (94).

In addition to modulating motoneurons, 5-HTmodulates the intrinsic properties of spinal in-terneurons via several mechanisms that may un-derlie the effects of 5-HT on the frequency oflocomotor activity. The most common effect of5-HT on spinal interneurons is to depolarize theirresting membrane potential (24, 40, 184 –186). Inaddition, 5-HT decreases the action potentialthreshold in the majority of commissural interneu-rons (184, 185), an undefined population of ventralhorn interneurons (52), and interneurons that areactive during locomotion (40). Interestingly, 5-HThas also been shown to compress the range ofvoltage-thresholds for repetitive firing in ventralhorn interneurons by reducing firing thresholds inhigh threshold interneurons and increasing firingthresholds in low threshold interneurons (162).Other effects of 5-HT on interneurons include areduction in the magnitude of the AHP in mostcommissural interneurons (45, 184, 185) andlocomotor activity-related interneurons (40), aswell as modulation of Ih currents and enhance-ment of PICs in locomotor activity-related in-terneurons (41, 42). Although 5-HT has a range of

modulatory effects on the intrinsic properties ofspinal interneurons, the common consequence ofthis modulation is neuronal excitation. This sug-gests that excitation of either inhibitory or excit-atory CPG neurons, depending on the receptorsubtype involved, may underlie the mixed facili-tatory and inhibitory actions of 5-HT on locomotoractivity. Alternatively, novel inhibitory actions of5-HT on spinal interneurons may remain to bediscovered.

Finally, besides affecting locomotion via themodulation of intrinsic properties of spinal neu-rons, 5-HT may also exert its effects by modulatingsynaptic transmission. Evidence for 5-HT-mediatedmodulation of synaptic transmission in the ventralspinal cord includes the modulation of sensory anddescending inputs to spinal interneurons in the cat(67, 69, 86) and sensory input to motoneurons in therat (11).

Noradrenaline in Mammalian Systems

Noradrenergic input to the mammalian spinal cordoriginates from A5, A6 (locus coeruleus), and A7brain stem nuclei (34, 130, 174). Examples of mam-malian spinal neurons innervated by these brainstem sources of noradrenalin (NA) include mo-toneurons (61, 79, 138), commissural interneurons(67), and lamina VII premotor interneurons (116).There is also widespread expression of adrenergicreceptors in the mammalian spinal cord (60, 129,142, 143).

Like 5-HT, NA can both initiate locomotor activ-ity and modulate ongoing locomotor patterns. Inspinalized cats, activation of noradrenergic, partic-ularly �2, receptors powerfully stimulates locomo-tor activity (7, 31, 56, 91). In contrast to the cat, NAinduces only poor locomotor-like activity in isolatedrat spinal cord preparations (57, 93, 154). In bothchronic spinalized cats and isolated rat spinal cordpreparations, application of NA during ongoinglocomotor activity slows the rhythm (31, 93) andincreases the amplitude of EMG or ventral rootactivity (31, 93, 154). By using specific agonists, itcan be shown that the modulation of locomotoractivity by NA reflects the net effect of separatemechanisms activated by different adrenoceptorsubtypes. During locomotion in chronic spinalcats, �2-adrenoceptor activation increases step cy-cle duration, primarily by lengthening burst dura-tion in flexor muscles. In contrast, activation of�1-adrenoceptors has little effect on the timingof locomotion but instead increases the strength ofextensor muscle activity (31). These data suggestthat �2-adrenoceptor activation influences in-terneurons involved in rhythm generation,whereas �1-adrenoceptor activation modulatesmotoneuron output. Data from rodent studiesdemonstrate a similar separation in the effects

REVIEWS




ownloaded from


of �1- and �2-adrenoceptor activation. In isola-ted rat spinal cord preparations, activation of �2-adrenoceptors leads to a decrease in the frequencyof pharmacologically induced locomotor activitywith no effect on the amplitude of ventral rootbursts. In contrast, activation of �1-adrenoceptorsresults in an increase in locomotor frequency and adecrease in the amplitude of ventral root bursts(154). Interestingly, �1-adrenoceptor activation re-duces locomotor frequency and increases ventralroot burst amplitude when locomotion is inducedby stimulation of the cauda-equina in isolatedmouse spinal cord preparations (62). This may re-flect species differences or indicate that the modeof activation of locomotion influences the poten-tial modulatory roles of NA.

The dominant effect of NA on spinal neuronsseems to be an increase in excitability, most likelyvia activation of �1-adrenoceptors, whereas activa-tion of �2-adrenoceptors can inhibit activity (51,131, 161, 154, 176). NA shares many cellular mech-anisms of neuromodulation with 5-HT. In parallelwith 5-HT, NA activation can facilitate persistentinward currents and plateau potentials (35, 104,162), reduce inwardly rectifying K� currents (161),hyperpolarize the action potential threshold (52),and compress the range of voltage-thresholds forrepetitive firing in ventral horn interneurons (162).However, unlike 5-HT, NA appears to have littlemodulatory effect on the AHP (161).

Adrenoceptor activation, like serotonergic recep-tor activation, may also alter locomotor activity bymodulating synaptic connectivity with the spinalcord. One of the most striking examples of thiscomes from experiments utilizing the isolated ratspinal cord in which the application of NA un-masks recurrent excitatory pathways that can mod-ulate locomotor activity, increasing the frequency ofrhythmic activity (109). NA also increases locomotor-related synaptic drive to motoneurons in the iso-lated rat spinal cord (161), an effect that maycontribute to the increased motor output seenupon adrenoceptor activation. Finally, like 5-HT,NA can modulate sensory and descending inputsto spinal interneurons in the cat (67, 69, 86) andsensory input to motoneurons in the rat (11).

Dopamine in Mammalian Systems

In mammals, dopaminergic inputs, arising pre-dominantly from the hypothalamic A11 region (18,78, 137, 150), are distributed throughout the ven-tral horn of the spinal cord (80, 172, 181). In addi-tion, all subtypes of dopamine receptors (D1–D5)are also expressed within the ventral horn (187,188).

Involvement of dopaminergic input in locomo-tor control is supported by measurements of in-creased dopamine levels in the spinal cord during

locomotor activity (59). However, dopamine is notas effective as 5-HT or NA at initiating locomotion.In spinal cats, dopaminergic agonists fail to inducelocomotion (7). In contrast, in mice with spinalcord transections, locomotion can be induced byD1/D5 receptor agonists (102), suggesting that spe-cific activation of D1/D5 receptor subtypes mightinduce locomotor activity in the cat. In isolatedrodent preparations, the ability of dopamine re-ceptor activation to induce locomotion varies byspecies. In neonatal rat preparations, dopaminecan induce rhythmic locomotor-like activity, al-though it is much slower than that induced by5-HT and in some reports more irregular (10, 92).In contrast, in neonatal mouse preparations, theactivation of dopamine receptors alone is insuffi-cient to induce locomotor activity, but activation ofdopamine receptors (primarily D1 and D2) by en-dogenous dopamine is required for 5-HT-inducedlocomotor activity (112). Interestingly, at later devel-opmental time points in mice, toward 1 wk old andbeyond, exogenous dopamine is required along with5-HT and NMDA to elicit locomotor activity in iso-lated spinal cord preparations (86a, 124).

In both cat and rodents, dopamine can modu-late on-going locomotor activity. In the cat, dopa-mine receptor activation increases the amplitudeof flexor activity during locomotion (7), primarilysuggesting modulation of motoneurons. In compari-son, in isolated rodent spinal cord preparations, dopa-mine receptor activation slows locomotor rhythmsinitiated by 5-HT and NMDA while also increasingthe amplitude of locomotor-related output andmaking the activity more reliable (10, 62, 175).Thus, in rodents, dopamine is likely to modulatethe activity of interneurons involved in rhythmgeneration and motoneurons.

Analyses of the mechanisms by which dopaminereceptor activation modulates mammalian spinalneurons are at present limited to motoneurons andHb9� interneurons in isolated neonatal mouse spi-nal cord preparations (70, 71). In mouse motoneu-rons, dopamine application increases excitabilityby modulating potassium currents including Ca2�-dependent K� currents to reduce the mAHP and IA

to reduce the latency to the first spike (70). Inaddition, activation of the D1 subtype of dopaminereceptors modulates AMPA receptor-mediatedcurrents in motoneurons (71). In the case of Hb9�

interneurons, dopamine application has beenshown to be necessary but not sufficient to ind-uce oscillatory behavior, which may relate to thelocomotor rhythm (77, 178). In addition to modu-lating the intrinsic properties of motoneurons andHb9� interneurons, dopamine modulates synapticinput to motoneurons including those from sen-sory afferents (11, 25, 33) and may modulate

REVIEWS




ownloaded from


synaptic transmission between motoneurons andRenshaw cells (113, 146).

Biogenic Amines in Non-MammalianSystems

In contrast to studies of mammalian systems, re-search involving aquatic vertebrates has demon-strated that aminergic signaling pathways arefunctionally integrated, often providing comple-mentary modulatory effects (118). The modulatoryeffects of biogenic amines in these animals willtherefore be discussed together.

In the lamprey, the modulatory effects of 5-HThave been a topic of intense research and debate.Bath applied 5-HT has a well defined effect onlocomotor activity; the rhythm slows and thebursts intensify in a similar manner to other ver-tebrates. The fact that the 5-HT uptake inhibitorcitalopram has a similar effect to 5-HT indicatesthat endogenous release of the amine can regulateburst formation during locomotion. The mecha-nisms of action of 5-HT classically involve a reduc-tion of the slow after-hyperpolarization (sAHP)that follows the action potential in motoneuronmediated in large part by apamin-sensitive Ca2�-dependent K� (KCa) channels (reviewed in Ref.65a). The sAHP acts as a brake on motoneurondischarge by determining the time it takes for theneuron to recover and fire again. Hence, in reduc-ing the sAHP, 5-HT reduces spike accommoda-tion, and so, for a given excitatory input,motoneuron fire for longer and at a higher fre-quency. The serotonergic block of KCa thereforeneatly explains the increase in discharge fre-quency of motoneurons within each cycle ofswimming. Modulation of the sAHP in CPG in-terneurons can also affect their firing frequencyand consequently change the frequency of the lo-comotor rhythm. The same KCa channels also con-tribute to the plateau phase of the intrinsic NMDAreceptor-mediated membrane potential oscillationsreported in lamprey motoneurons when they areactivated by the calcium entry that follows NMDAreceptor activation (169a). KCa channel activationdrags the membrane potential from the depolar-ized plateau phase into a sufficiently more hyper-polarized region so that Mg2� ions can block theopen NMDA receptor ion channel and trigger thefalling phase of the oscillation. The serotonergicblock of KCa channels therefore prolongs the de-polarized plateau phase, allowing neurons tofire for longer in each cycle. Thus, through parallelactions of 5-HT on KCa channels and NMDAreceptors on spinal neurons, 5-HT modulates fre-quency, duration, and amplitude of locomotorbursts.

Given that KCa channels play a major role in themodulation of lamprey locomotion by 5-HT, it

would be anticipated that apamin, which blocksKCa channels, should have the same effects as 5-HTon swimming. However, even at high concentra-tions (10 �M) sufficient to block the sAHP andincrease motoneuron firing, apamin had no signif-icant effect on the frequency of NMDA-inducedfictive locomotion (123a). This result can only bepartially reconciled by the fact that the effects de-pend on the initial concentration of NMDA and thestarting frequencies of swimming that this elicits.Since 5-HT has also been shown to reduce thestrength of excitatory synaptic transmission via apresynaptic mechanism (145), a more completeexplanation is that, to mediate its full effects onlocomotion, 5-HT must have additional effects be-sides blocking KCa channels.

In zebrafish, the effects of 5-HT are developmen-tal stage specific. At early stages, 5-HT modulatesthe duration of quiescent periods, and thus thefrequency of swim bouts, without directly affectingthe firing of neurons active during swimming orthe various parameters of a swim cycle (19). Thisappears to be mediated by an effect on chloridehomeostasis (20). Once the zebrafish becomes freeswimming, and continuing into adulthood (58),5-HT has effects that parallel those in the lamprey,namely the cycle period lengthens. In this case,however, the mechanism appears to involve astrengthening of the mid-cycle glycinergic inhibi-tion, akin to the effect of NA in Xenopus tadpoles(118). In Xenopus tadpoles, 5-HT applied to younglarvae increases the duration of motor bursts buthas little effect on the locomotor frequency, pro-ducing a relatively fast, short, and intense versionof the fictive swimming rhythm. In contrast, NAhas the opposite effect of increasing the cycle pe-riod but has little or no effect on the duration ofmotor bursts. Which receptor subtypes are in-volved in the serotonergic modulation of locomo-tion? In both lamprey (76) and Xenopus tadpoles,there is evidence that 5-HT1a receptors are in-volved in some of the effects, and for the lampreyit has been reported that the effects of KCa chan-nels are largely mediated by this receptor and itsnegative coupling to adenylate cyclase (AC) (177).Also in the lamprey, it has been reported that theserotonergic presynaptic inhibition of excitatorytransmission is mediated by the 5-HT1d receptor(145), which also negatively couples AC. In mam-malian systems, the receptor subtypes differ in that5HT2 and 5HT7 receptors seem to play a promi-nent role (see above).

Dopaminergic effects on lower vertebrate locomo-tor centers are less well described. In zebrafish, do-pamine has an inhibitory effect on the frequency ofswim episodes during early development (163). At 3days postfertilization, zebrafish larvae generate asmall number of relatively long-duration swim

REVIEWS




ownloaded from


bouts. Dopamine, or the dopamine uptake inhibi-tor bupropion, abolishes swimming activity, an ef-fect mediated in the brain stem. Dopamine has noeffect on the integrative electrical properties of spi-nal neurons. Pharmacological blockade of D2receptors or activation of adenylate cyclase (adownstream target that is inhibited by D2 recep-tors) blocks the inhibitory effect of dopamine. Thesuppression of swim initiation appears to be tran-sient since, by 5 days postfertilization, dopamineuptake no longer affects the frequency of swimepisodes.

Conventional Fast ExcitatoryTransmitters

Conventional transmitters involved in the fast syn-aptic interactions that occur during locomotor net-work operation mediate their effects by activatingionotropic postsynaptic receptors (glutamate:NMDA, AMPA; GABA: GABAa; acetylcholine: nAChR;glycine: glyR). However, the same transmitters (withthe possible exception of glycine) also activatemetabotropic receptors located both pre- and post-synaptically to modulate ongoing activity. In manycases (e.g., group 2,3 mGluRs, mAChR, GABAB), thepresynaptic receptors mediate homosynaptic de-pression of transmitter release and thus function asnegative feedback autoreceptors. In certain cases,pharmacologically similar receptors are located onthe terminals of adjacent neurons where they canmediate heterosynaptic depression or facilitation oftransmitter release. However, other metabotropic re-ceptors for classically fast transmitters are locatedpostsynaptically, such as mGluR1 receptors, where arange of complex modulatory effects have beendescribed.

Glutamate (mGluRs)

Glutamate is best known for its role as the majorfast, excitatory neurotransmitter in the CNS, a roleit fulfils via the activation of ionotropic glutamatereceptors. However, glutamate can also act as aneuromodulator by binding to three groups (I–III)of G-protein-coupled metabotropic glutamate re-ceptors (5, 136). Given that the functioning of thelocomotor CPG is dependent on glutamatergic sig-naling between spinal interneurons (13, 160, 175),there are considerable intraspinal sources of glu-tamate. In addition, the locomotor CPG receivesglutamatergic input from both descending (66, 88)and sensory systems (117). In addition to the abun-dance of glutamatergic transmission affecting spi-nal motor circuitry, there is widespread expressionof mGluRs in the ventral horn (4, 103) that couldbe activated in parallel with ionotropic glutamatereceptors (44).

The effects of metabotropic glutamate receptors(mGluRs) within vertebrate motor systems havebeen most extensively studied in the lamprey (49,95) (FIGURE 2), with comparative data arising fromthe tadpole (28, 29) model system. In aquatic ver-tebrates, as in other species, group 2 and 3 mGluRsappear to be located primarily presynapticallywhere they mediate presynaptic inhibition of glu-tamate release and therefore function as negativefeedback controllers of excitatory transmission.Activation of these receptors generally slows fictivelocomotion. In contrast, group 1 mGluRs play avariety of different, mainly excitatory, roles, and inboth lamprey and tadpole their activation leads to anincrease in the locomotor frequency. In both species,there is evidence that this effect is due partly topostsynaptic modulation of neuronal excitability andpartly to presynaptic inhibition of glycinergic inhib-itory transmission. Group 1 mGluRs can be furthersubdivided into mGluR1 and mGluR5 subtypes. Inthe lamprey spinal cord, mGluR1 (but not mGluR5)receptors cause membrane potential depolariza-tion and excitation of spinal cord neurons byblocking a leak (mixed Na�-K�) current.

The effect requires activation of phospholipase C(PLC) and the release of calcium ions from intra-cellular stores. mGluR1s also enhance excitabilityof lamprey spinal neurons by facilitating currentflow through NMDA receptors that play a key rolein the rhythmic excitation that occurs in each cycleof locomotion. In parallel, mGluR5 receptors reg-ulate lamprey locomotion but via different mech-anisms that trigger oscillations in intracellularcalcium and have a net inhibitory effect on thefrequency of locomotion (90). In contrast, in frogtadpoles, both subtypes of group I mGluRs in-crease locomotor frequency (29). Experiments us-ing mGluR1- and mGluR5-specific antagonistsdemonstrate that endogenous activation of bothtypes of receptors contribute to intrinsic spinalmodulation of locomotor output in lamprey andfrog tadpoles (29, 90, 95). These data on bothyoung frog tadpoles and adult lampreys also infer aconserved mGluR-mediated modulatory mecha-nism that is established early in vertebrate devel-opment (tadpole) and evolution (lamprey), namelythe presynaptic inhibition of glycine release bygroup 1 mGluRs. Although this conserved modula-tory mechanism suggests a strategic positioning ofgroup 1 mGluRs on the synaptic terminals of in-hibitory interneurons of the spinal CPG, there isextensive evidence from the lamprey system thatactivation of group 1 mGluRs couples to an intra-cellular signaling cascade that triggers the releaseof the endocannabinoid 2-AG from postsynapticmembranes that activates presynaptic endocannab-oind receptors to reduce inhibitory transmission.

REVIEWS




ownloaded from


This pathway is reviewed in detail in a recent issue ofthis journal (50).

Activation of mGluRs also modulates, but doesnot initiate, locomotor activity in isolated rodentspinal cord preparations (85, 157–159). Analyseshave concentrated on the role of group I mGluRs,although group II and III mGluRs also modulate ratlocomotor activity (159). In the rat spinal cord, theapplication of group I mGluR agonists either dis-rupts locomotor activity completely or slows it(158). Seemingly paradoxically, group I-specificmGluR antagonists also decrease the frequency oflocomotor activity (158). The authors of these stud-ies suggest that this apparent discrepancy reflects

diffuse excitatory effects of exogenous agonistovershadowing more specific depressant effects ofendogenous group I mGluR activation. In contrastto these data in the rat but consistent with find-ings in lamprey (95) and Xenopus tadpoles (29),recent data obtained from isolated mouse spinalcord preparations have demonstrated an in-crease in locomotor activity on activation ofgroup I mGluRs (85). In addition to affecting thefrequency of locomotion, group I mGluR activa-tion also modulates the amplitude of locomotor-related motoneuron output in mice, indicatingmodulatory effects at the level of CPG interneu-rons and motoneurons.

FIGURE 2. Activation of mGluR1 receptors increases locomotor frequency in the lamprey by divergentneuromodulatory effectsGlutamate released from excitatory interneurons in the spinal cord activates locomotor network neurons via ionotropic AMPA and NMDA receptorsand simultaneously activates postsynaptic mGluR1 receptors. mGluR1s block a leak K� channel, enhance NMDA receptor-mediated currents, andtrigger release of the endocannabinoids, which bind to EC receptors on glycinergic interneurons to reduce mid-cycle inhibition. Figure was adaptedfrom Ref. 50 and used with permission.

REVIEWS




ownloaded from


Data concerning the cellular effects of mGluRson mammalian spinal neurons are limited to theeffects of group I mGluR activation on motoneu-rons (85, 114, 115). Application of group I mGluRagonists leads to depolarization of motoneuronalresting membrane potentials in both rats andmice. In rat motoneurons, this is associated withreports of no change (115) or an increase in resis-tance (114), whereas in mouse motoneuronsmGluR-dependent depolarization is associatedwith a decrease in resistance (85). In mouse mo-toneurons, group I mGluR activation also reducesthe amplitude of transient Na� currents, an ef-fect that seems to lead to reduced motoneuronfiring and decreased locomotor-related mo-toneuron output (85). In both rats and mice,group I mGluRs also modulate synaptic trans-mission within the spinal cord. In rats, group ImGluRs appear to depress inhibitory transmissioninvolved in sensorimotor pathways (115) and re-current inhibitory transmission to motoneurons(114). In mice, activation of group I mGluRs de-presses excitatory locomotor-related input (85).

Acetylcholine

All cholinergic inputs to mammalian spinal neu-rons are thought to originate from within the spi-nal cord (119, 147, 166). Motoneurons representthe major acetylcholine-producing cells of the spi-nal cord. The acetylcholine they produce is mostlydestined for the periphery where it mediates trans-mission at the neuromuscular junction, althoughmotoneuron axon collaterals also activate recur-rent inhibitory pathways mediated by Renshawcells (38, 98). Other intraspinal sources of acetyl-choline include small cholinergic neurons scat-tered in the dorsal horn, central canal cluster cellssurrounding the central canal, and “partition cells”that lie between the dorsal and ventral horns in aregion extending from lamina X to the lateral edgeof the gray matter (8, 83, 124, 135, 182). Along withneuronal sources of acetylcholine, there is consid-erable expression of muscarinic metabotropic acetyl-choline receptors in the ventral horn of themammalian spinal cord (75, 125, 171, 173, 179).Taken together with the fact that cholinergic interneu-rons are activated during locomotion (83), these datahighlight the potential importance of acetylcholineas a modulator of locomotor behavior.

In the early embryonic mouse spinal cord, cho-linergic transmission, arising from the axon collat-erals of motoneurons and involving nicotinicreceptors, drives early locomotor-related rhythmand appears to contribute to the appropriate as-sembly of spinal locomotor circuitry (72, 126).Later, at neonatal stages in the rat, locomotor-like activity can be initiated by activation ofmetabotropic receptors following the application

of acetylcholine or cholinesterase inhibitors (36,151, 152). It should, however, be noted that thisacetylcholine-driven activity often lacks appro-priate alternation between ipsilateral extensorsand flexors (36).

Although there have been no reports of acetylcho-line modulating the frequency of ongoing locomotoractivity in mammals, cholinergic transmission at C-bouton synapses, which densely cover motoneuronalsomata and arise from spinal interneurons (V0cinterneurons), has recently been shown to modu-late the strength of locomotor-related motoneuronoutput in a task-dependent manner (124, 182)(FIGURE 3). Spinal V0c interneurons, which are thesource of C-bouton inputs to motoneurons, appearto be tonically active throughout locomotion with thefrequency of their activity tightly phase locked tomotoneuron output. Genetic inactivation of theoutput of V0c interneurons results in an impairedability to increase the activation of hindlimb mus-cles during motor tasks that require greater force.From these data, it is predicted that task-specificmodulation of the intensity of motoneuron outputis in part controlled by V0c interneurons and theirC-bouton contacts with motoneurons.

At the cellular level, acetylcholine-mediatedmodulation generally has a net excitatory effect onspinal neurons. Studies in mice have shown thatacetylcholine depolarizes the resting membranepotential and decreases input resistance in bothcommissural interneurons and interneurons thatare activated during locomotion (which may in-clude some commissural interneurons) (24, 40).Additional effects of acetylcholine observed in lo-comotor activity-related interneurons include ahyperpolarization of the action potential thresh-old, an increase in the magnitude of the AHP, anda decrease in Ih (40, 41). The net excitatory effect ofacetylcholine on mouse interneurons is evidencedby a leftward shift in frequency current relation-ships (24), although in locomotor activity-relatedinterneurons this is accompanied by a decrease inslope such that excitability is only increased atlower stimulus intensities (40).

In mouse motoneurons, activation of metabo-tropic muscarinic receptors also has a net excit-atory effect on firing output (124). Althoughmuscarinic receptor activation can either hyperpo-larize or depolarize the resting membrane potentialof motoneurons, it always leads to a reductionin AHP amplitude via the activation of m2-typemuscarinic receptors (124), which are found at C-bouton synapses (75, 125, 179) (FIGURE 3). Thisreduction in the AHP leads to increased motoneu-ron excitability as evidenced by an increase in theslope of frequency current plots (124) and isthought to underlie the task-dependent regulation

REVIEWS




ownloaded from


of the strength of motoneuron output during loco-motion discussed above (182).

For lower vertebrates, much less is currentlyknown about the role of metabotropic muscarinicreceptors for acetylcholine in the modulation oflocomotor activity. However, there is ample evi-dence that motoneurons express muscarinic re-ceptors. In the salamander, for example, M2-typereceptor activation alters the integrative electricalproperties of motoneurons by modulating threeionic currents: a hyperpolarization-activated cationic

current (Ih), a calcium-dependent potassium cur-rent (KCa), and an inwardly rectifying potassiumcurrent (IKir) (32).

Conventional Fast InhibitoryTransmittersGABA/Glycine

There is an abundance of inhibitory interneuronswithin the mammalian spinal cord, which, in ad-dition to controlling key aspects of the locomotorpattern such as the alternation of activity between

FIGURE 3. Intrinsic cholinergic modulation of mouse locomotionA: a small cluster of cholinergic interneurons (V0C interneurons) located near the central canal of the spinal cord represents the sole source of C-bouton synaptic inputs to motoneurons (MN). Both motoneurons and V0C interneurons receive rhythmic input from the locomotor CPG. B: acetyl-choline (Ach), released at C-bouton synapses, activates postsynaptic m2-type muscarinic receptors on motoneurons, which in turn leads to areduction in currents mediated by SK-type Ca2�-dependent K� channels. C-bouton activation therefore leads to a decrease in the action potentialafterhyperpolarisation (AHP) and an increase in motoneuron firing frequency. C: recordings of EMG activity from gastrocnemius (Gs) and tibialisanterior (TA) hindlimb muscles during walking and swimming in control mice and mice in which C-bouton signaling has been silenced demonstratea task-dependent and muscle-specific role for the C-bouton system in the modulation of motoneuron output. Note reduced modulation of EMGamplitude from walking to swimming in the Gs muscle in mutant animals. D: these findings suggest that V0C interneurons and their C-bouton con-tacts with motoneurons form a feed-forward facilitatory system for the control of motoneuron output during locomotion. Inputs from the locomotor CPG,sensory afferents, and descending systems appear to control the level of activity of V0C interneurons and hence motoneuron output and muscle activa-tion in a task-dependent manner. The question mark indicates the uncertain nature of the descending and/or sensory inputs that mediate task-dependentregulation of the activity of V0C interneurons. Images and data are modified from Refs. 124 and 182 and used with permission.

REVIEWS




ownloaded from


the left and right sides of the spinal cord and be-tween flexor and extensor motoneurons, may alsomodulate on-going locomotor activity. Commonlystudied examples of inhibitory ventral interneu-rons include Ia inhibitory interneurons (53, 84),Renshaw cells (141), and inhibitory commissuralinterneurons (e.g., Refs. 22, 101). Along with thepresence of many sources of inhibitory transmis-sion within the spinal cord, there is also widespreadexpression of metabotropic GABAB receptors (30, 164),supporting a potential role for neuromodulation of lo-comotion via inhibitory transmission.

In the mammalian spinal cord, a number ofstudies in cats (e.g., Refs. 39, 48, 87, 105, 155) androdents (17, 170) have shown that the main mod-ulatory actions of GABAB receptors involve thepresynaptic inhibition of transmitter release withlittle or no effect on postsynaptic properties ofneurons. In isolated neonatal rat spinal cordpreparations, the consequences of GABAB recep-tor activation on locomotor output include a re-duction in the amplitude of ventral root bursts,probably due to reduced locomotor-related syn-aptic drive to motoneurons, and a reduction inlocomotor frequency due to undefined actionson CPG interneurons (17).

With respect to GABAB receptors in lower verte-brates, it has been shown that their activation inXenopus embryos also causes presynaptic inhibi-tion of transmitter release, specifically from theterminals of glycinergic interneurons that mediatereciprocal inhibition during swimming (169). Theeffect involves a direct presynaptic action onomega-conotoxin-sensitive (N-type) calciumchannels, leading to a reduction in the probabilityof quantal release. In addition, a GABAB receptoragonist reduces the firing threshold of motoneu-rons, indicating parallel pre- and postsynaptic ef-fects that unite to reduce the duration andfrequency of fictive swimming. Similar effects ofGABAB receptor activation have been reported inthe lamprey where tonic activation of these recep-tors during fictive swimming contributes to thesetting of the locomotor burst frequency. In addi-tion, GABAB receptors play an important role incontrolling the intersegmental propagation of ac-tivity along the spinal cord during fictive locomo-tion (115a, 161a). Activation of GABAB receptors byendogenously released GABA during locomotornetwork activity reduces the locomotor frequencyand modifies the intersegmental phase lag as theactivity propagates rostrocaudally along the spinalcord (161a). These effects are mediated in largepart by the GABAB-mediated depression of volt-age-activated calcium currents and Ca2�-depen-dent K� channels, which affects post-inhibitoryrebound and the post-spike after hyperpolariza-tion (115a).

Purinergic Transmitters

In Xenopus embryos, ATP and adenosine controlthe duration of swimming episodes (43). Thesources of ATP are not identified, but it is pre-sumed to be co-released by neurons that are partof the spinal CPG. At the onset of a swimming bout,extracellular ATP levels rise as a result of the activ-ity. Initially, near the onset of the bout when swimfrequency is at its highest, ATP blocks K� channelsto help maintain a high level of excitability withinthe network so the swimming frequency is at itshighest. However, as the episode progresses, theATP is gradually broken down in the extracellularspace by a 5’-ectonucleotidase to adenosine. Adeno-sine is able to block Ca2� channels, which impairssynaptic transmission and reduces excitability. Ev-idence suggests that adenosine mediates its effectsvia A1 receptors and that both an N-type calciumcurrent and a further unidentified HVA calciumcurrent are inhibited by adenosine (18b). The ef-fect of adenosine’s inhibition of calcium channelsis to lower the swim frequency, reducing the activ-ity-dependent production of ATP by the swimCPG. With time, the inhibitory effects of adenosineovercome the excitatory effects of ATP, and theswim frequency eventually reaches a critical lowlevel that can no longer be sustained, and thus theswimming bout ceases. This is a very elegant ex-ample of a role for an intrinsic neuromodulatorthat may be important in ensuring that swimmingdoes not continue ad infinitum. There are moretraditional mechanisms for terminating swimmingsuch as inhibitory reticulospinal neurons that pro-duce rapid GABAergic inhibition whenever the ros-tral cement gland is contacted by objects in theenvironment (18a). In behavioral terms, this de-scending inhibition from the brain stem is likely tobe involved before the purinergic biochemicalclock mechanism is engaged.

Recent work utilizing isolated neonatal mousespinal cord preparations has also revealed a rolefor purines in the intrinsic modulation of the mam-malian locomotor CPG (180). In the mouse, spinalcord adenosine, derived from the breakdown ofATP following its release from glia, appears to playan endogenous role in setting the excitability andfrequency of the locomotor CPG. This is revealedby an increase in locomotor frequency upon block-ade of adenosine receptors that is absent inthe presence of glial toxins or ectonucleotidaseinhibitors.

Peptide Neuromodulators

The best understood peptide with respect to rolesin motor control is the tackykinin, substance P. In

REVIEWS




ownloaded from


lower vertebrates, it has been extensively studiedin the isolated spinal cord of the lamprey where abrief (10 min) application of substance P leads toboth short-term (�2 h) and long-term (�24 h)increases in the frequency of NMDA-induced fic-tive locomotion (131a, 131b). The short-term ef-fects are protein kinase C-dependent and appearto involve a potentiation of cellular responses toNMDA and the associated induction of membranepotential oscillations. The long-term effects, inwhich a 10-min exposure to substance P triggers anincrease in the baseline swimming frequency thatlasts for up to 24 h, require changes in proteinsynthesis. The tachykinin receptor antagonistspantide II leads to a lowering of the baselineNMDA-induced locomotor frequency, stronglysuggesting that the release of substance P or arelated peptide during swimming modulates thelocomotor CPG. Substance P also has a presynapticfacilitatory effect on the release of glutamate ontospinal neurons via a calcium-independent mecha-nism (Parker & Grillner, 1998). There is an intrinsicsource of tackykinins in the spinal cord: substanceP immunoreactivity is found in cells of the ventralplexus. However, there is also evidence for twotackykinins (substance P and neurokinin B) in nu-merous cell groups in the brain, including regionsthat project to the spinal cord and may play a rolein the initiation and/or modulation of swimming.It is not known under what circumstances the lam-prey might engage this positive feedback system tomaintain high-frequency swimming, but sincethese animals undergo long migrations to theirbreeding grounds it has been hypothesized thatthe tackykinin system may be well suited to con-tributing to such a behavioral role.

In mammals, a large number of peptides andtheir associated receptors are present in the ven-tral aspect of the spinal cord and are therefore ina position to modulate motor activity (12, 133).In one study (9), 11 such peptides were appliedto the isolated rat spinal cord, either on theirown or in the presence of NMA or NMA plus5-HT, a cocktail of drugs that activates the loco-motor CPG, while recordings were made fromlumbar ventral roots. Only four of the peptides(oxytocin, vasopressin, bombesin, and TRH)were able to trigger tonic or loosely coordinatedrhythmic activity when bath applied alone. How-ever, all of the peptides were able to modulateongoing locomotor cocktail-induced CPG activ-ity, although the effects were different for eachpeptide. For example, proctolin and neurotensinincreased the frequency of locomotion, whereasmet-enkephalin and oxytocin decreased it. Somepeptides (bombesin, somatostatin) acted purelyon the frequency of locomotion, whereas mostaffected both the rhythm frequency and the lo-

comotor burst amplitudes. Presumably, the pep-tides acting solely on amplitude can directlyinfluence motoneuron firing, whereas the othersalso act at the level of premotor interneuronalcircuitry.

Gaseous Neuromodulation

The free radical nitric oxide (NO) is an unusual butpotent modulator of locomotor activity in lowervertebrates such as the Xenopus tadpole (120, 122)and the lamprey (96, 97). NO is manufactured bycells that possess the enzyme NO synthase, whichbreaks down L-arginine to L-citrulline, liberating amolecule of NO in the process. In Xenopus tad-poles, the neuronal sources of NO, as revealed byNADPH-diaphorase histochemistry, are located inthe brain stem in distinct and bilaterally symmet-rical neuronal clusters (120, 123). NOS expressionin these neurons appears in a well defined tempo-ral sequence such that all three clusters are labeledaround the time of hatching from the egg mem-brane when the young larva is only �3 days old.These three clusters remain throughout develop-ment (139). NO has a net inhibitory effectand acts as a break on swimming: NO donorsreduce episode durations and slow swimming;NOS inhibitors or NO scavengers have the oppositeeffect, indicating that NOS is constitutively activeand produces an intrinsic NO tone that regulatesswimming output (120). The primary mechanismsthrough which these effects on locomotion occurinclude NO’s presynaptic facilitation of 1) glycin-ergic transmission (which slows the swimmingrhythm) and 2) GABAergic transmission (whichleads to premature termination of activity by acti-vating a descending “stopping” pathway mediatedby reticulospinal inhibitory interneurons) (121,122) (FIGURE 4). At these early stages of develop-ment, before the onset of free swimming, NO isinhibitory. However, later in larval development,around the time when the limbs are developing,NOS-positive neurons appear in the spinal cord(139), and NO donors have the opposite effect offacilitating rhythmic activity (148).

Although there is extensive anatomical evi-dence that NO synthesis is widespread in thenervous system of most vertebrates studied, itspotential role in the modulation of locomotion inother model systems has received considerablyless attention than other more conventional sig-naling molecules. Recent studies have shownthat NO plays a significant role in modulatinglocomotion in both the lamprey (96, 97) andmouse (1, 47).

In the lamprey system, NOS is located in avariety of spinal neuron types including propri-oceptive edge cells and gray matter neurons. NO

REVIEWS




ownloaded from


has an excitatory effect, increasing the frequencyof NMDA-induced locomotor activity. NOS inhi-bition or NO scavenging has the opposite effect,showing that there is an intrinsic release of NOduring locomotor rhythm generation. The syn-aptic effects of NO in the lamprey are complex.NO simultaneously reduces mid-cycle (glyciner-gic) inhibition and potentiates on-cycle (glutamater-gic) excitation. Analysis of miniature synapticcurrents reveals a purely presynaptic effect on gly-cine release but a parallel pre- and postsynaptic ef-fect on glutamatergic transmission.

Spinal interneurons expressing NOS are also foundthroughout the mouse spinal cord. Although NOS-positive interneurons are most abundant in the dor-sal horn, they are also present at lower levels in areassurrounding the central canal and parts of the ventralhorn, particularly lamina VII (47, 124, 153). The ad-dition of NO donors to isolated mouse spinal cordpreparations modulates both the amplitude andfrequency of on-going locomotor activity (47).Interestingly, the amplitude of locomotor activitycan either be increased or decreased, dependingon the dose of the NO donor, whereas frequencyis consistently reduced. NO scavengers also af-fect the amplitude and frequency of locomotoroutput, indicating an endogenous role for NO inmodulating the activity of both CPG interneurons andmotoneurons.

Metamodulation and InteractionsBetween Modulators

In some notable cases, the effects of 5-HT and NAare opposing and can switch during development(118, 140). 5-HT, in general, facilitates the loco-motor drive, increasing the duration and inten-sity of ventral root bursts. In Xenopus tadpoles,NA exerts effectively the opposite effects bylengthening the cycle period without any obvi-ous effect on the duration of motor bursts. Howare these differential effects produced? The twoamines have a common target in modulating thestrength of glycinergic inhibition (118). Bothamines act presynaptically to regulate glycinerelease from commissural interneurons, butwhereas 5-HT depresses release, NA enhances it.In the case of 5-HT, the effect is to reduce theamplitude of the inhibition at mid-cycle, allowingneurons to escape from inhibition sooner and firefor longer. In the case of NA, the effect is to in-crease the amplitude of mid-cycle inhibition anddelay the onset of the next cycle.

Later in development, during the metamorphicperiod when the limbs and associated central cir-cuits are being assembled, the roles of 5-HT andNA, although still acting in opposition, switch.Rauscent and colleagues studied the effects of two

biogenic amines, 5-HT and NA, on the spontane-ously generated fictive locomotor patterns re-corded from ventral roots at this later stage (140). Itis first shown that these isolated preparations dis-play one of two forms of circuit coordination: ei-ther the limb and tail circuits generate rhythmswith independent frequencies or they generate aconjoint rhythm with 1:1 coordination. Followingthe addition of 5-HT to preparations displaying thefrequency-independent rhythms, the slower limbrhythm accelerates while the faster axial rhythmslows until the two are united at a single commonfrequency. Conversely, in preparations already dis-

FIGURE 4. Modulation and metamodulation of the Xenopus tadpole loco-motor networkSummary diagram illustrating the metamodulation of spinal inhibition by NO in Xeno-pus tadpoles. At larval stage 42 (top), the major subdivisions of the CNS (middle) arerecognizable, including the forebrain (fb), midbrain (mb), hindbrain (hb), and spinalcord (sc). Brain regions are separated by dotted lines. The noradrenergic locus coer-uleus (in orange) is located just caudal to the midbrain-forebrain boundary. The mid-hindbrain reticulospinal (Mhr) region (purple) is located dorsally in the caudalhindbrain, with axons that cross the midline (dashed line) and descend to the spinalcord, providing a major source of spinal GABAergic inhibition. Commissural interneu-rons (blue) cross the midline and provide the major source of spinal glycinergic inhibi-tion. NO modulates swimming frequency by altering endogenous NA release (orange),which in turn modulates glycine release (blue) to motoneurons (MN). NO modulatesepisode durations by directly modulating GABA release (purple) to motoneurons.However, the nitrergic modulation of GABA could also affect swimming frequency.Basal levels of NO are indicated in middle circle, reduced NO levels in left circle (palecolors), raised NO levels in right circle (dark colors). Figure adapted from Ref. 121,and used with permission from Journal of Neuroscience.

REVIEWS




ownloaded from


playing the frequency-coupled rhythm, bath appli-cations of NA exerts the opposite effect ofdecelerating the limb and accelerating the axialrhythm until the two display independent frequen-cies. These effects are similar to those exerted bythe amines at stages both before and after meta-morphosis; 5-HT decelerates the axial rhythm ofearlier stages before limb development, whereasNA slows (while 5-HT accelerates) the limb rhythmgenerated by froglets after the tail has disappeared.At the intermediate stages when both circuits arepresent, it is presumably important for the organ-ism to be able to functionally couple and decouplethe two locomotor systems, and Rauscent and col-leagues suggest that two aminergic systems withopposing behavioral effects may be involved in thistactical switching between two locomotorynetworks.

Metamodulation defines a second-order form ofmodulation in which the effects of one modulatorysystem are influenced by the effects of anothermodulatory system (89). Metamodulation can occurin series or in parallel. A clear example of serial meta-modulation in vertebrate locomotor control involv-ing NO signaling is to be found in the swimmingsystem of Xenopus tadpoles (121) (FIGURE 4). Thedifferential effects of the amines 5-HT and NA oninhibitory transmission have already been described(see above). In the case of NA, there is a presynapticfacilitation of glycine release from commissural in-hibitory interneurons and a parallel presynapticfacilitation of GABA release from descending mid-hindbrain reticulospinal interneurons. The effectsof NO bear a remarkable similarity, raising thequestion of whether these two modulators haveconvergent targets at the terminals of GABAergicand glycinergic interneurons or whether theremight be serial interactions between the two mod-ulatory systems. Considering first the modulationof GABA synapses, there is an apparent conver-gence along two parallel routes: one a nitrergiceffect and the other noradrenergic. The evidence isthat the effects of a given neuromodulator onGABA transmission is unaffected by blockage ofthe other modulator and vice versa. For example,the shortening of swim episode durations (due toenhanced GABA release) by NA persists when NOis scavenged or NOS is inhibited, and the effects ofNO donors on episode durations remain when al-pha-adrenoreceptors are blocked.

In the case of glycinergic transmission, however,the situation is very different because the effects ofNO are occluded when alpha-adrenoreceptors areblocked by phentolamine, but the effects of nor-adrenaline are unaffected by NO scavenging orNOS inhibition. This leads to the conclusion thatNO’s effects on glycinergic transmission involve afacilitation of the noradrenergic system, which in

turn mediates the direct presynaptic facilitation ofglycine release.

Conclusions and Future Prospects

Vertebrate locomotor control networks are richlyendowed with neuromodulatory inputs that pro-vide the locomotor output with the degree of flex-ibility that is necessary to maneuver efficientlythrough the environment. These inputs derivefrom multiple locations both within the spinal cordand also from extrinsic modulatory centers in thebrain. At the cellular level, we know most aboutmodulators deriving from neuronal sources, but itis clear that, even though we appreciate them less,glia are able to synthesize and release neuromodu-lators with very potent effects on central circuitry,including ATP, D-serine, and neurosteroids.

In aquatic vertebrates, such as the lamprey and frogtadpole, the links between cellular mechanisms of ac-tion and their consequences for locomotor net-work output are well described for a number ofneuromodulators. In comparison, the relation-ships between cellular mechanisms and networkoutput are less clear for the neuromodulation ofcomplex mammalian systems. This largely reflectsour lack of knowledge regarding the componentsof the mammalian CPG and their interconnectiv-ity. However, with an increasing number of mam-malian interneurons being defined and describedusing molecular genetic techniques (63), we arenow much better placed to go on to define specificsources of neuromodulation, their cellular targets,and their roles in the regulation of mammalianlocomotor output (e.g., Ref. 182).

Although knowledge of locomotor circuitry is crit-ical toward understanding the actions of neuro-modulators, conversely, studies of neuromodulationmay aid in deciphering locomotor networks. This hasindeed proved to be the case for brain stem networkscontrolling mammalian respiration where knowl-edge of substance P and opioid-mediated modula-tion of the respiratory rhythm led to the discoverythat NK-1 and mu-opioid receptors serve as markersof critical respiratory rhythm-generating neuronswithin the pre-Botzinger complex (64). More re-cently, the specific neuromodulatory effects of opi-oids on pre-Botzinger complex neurons have alsorevealed the existence of a second respiratory oscil-lator in the brain stem, the parafacial respiratorygroup (reviewed by Ref. 54).

From the perspective of the individual compo-nents comprising locomotor control networks (i.e.,neurons and synapses), it is abundantly clear thatthey are constantly exposed to an ever changingneuromodulatory environment. At any moment intime, the extent to which the plethora of modula-tory receptors embedded in the neuronal mem-

REVIEWS




ownloaded from


brane are activated will determine the levels of theintracellular second messenger molecules that ul-timately determine the computational ability of theneuron to integrate fast synaptic inputs and gen-erate output. In this sense, the neurons can bethought of as biochemical integrators, a role thatought to be given equal credence with their role oftransmitting electrical information to their ownsynaptic targets. This idea is by no means a newone (89), and yet in the last decade there has beenvery little progress in advancing the hypothesis.This is surely an area in which future researcheffort should advance our understanding of howneuromodulation sculpts an infinite variety of out-puts from a single anatomically defined network. �

No conflicts of interest, financial or otherwise, are de-clared by the author(s).

References1. Agboh K, Sillar KT, Miles GB. Physiological characterisation

of nitrergic neurons in the mouse lumbar spinal cord In:Society for Neuroscience, 2009. Chicago, IL, 2009, programno. 177.172.

2. Al-Mosawie A, Wilson JM, Brownstone RM. Heterogeneity ofV2-derived interneurons in the adult mouse spinal cord. Eur JNeurosci 26: 3003–3015, 2007.

3. Alvarez FJ, Pearson JC, Harrington D, Dewey D, Torbeck L, FyffeRE. Distribution of 5-hydroxytryptamine-immunoreactive bou-tons on alpha-motoneurons in the lumbar spinal cord of adultcats. J Comp Neurol 393: 69–83, 1998.

4. Alvarez FJ, Villalba RM, Carr PA, Grandes P, Somohano PM.Differential distribution of metabotropic glutamate receptors1a, 1b, and 5 in the rat spinal cord. J Comp Neurol 422:464–487, 2000.

5. Anwyl R. Metabotropic glutamate receptors: electrophysio-logical properties and role in plasticity. Brain Res Rev 29:83–120, 1999.

6. Barbeau H, Rossignol S. The effects of serotonergic drugs onthe locomotor pattern and on cutaneous reflexes of the adultchronic spinal cat. Brain Res 514: 55–67, 1990.

7. Barbeau H, Rossignol S. Initiation and modulation of thelocomotor pattern in the adult chronic spinal cat by norad-renergic, serotonergic and dopaminergic drugs. Brain Res546: 250–260, 1991.

8. Barber RP, Phelps PE, Houser CR, Crawford GD, SalvaterraPM, Vaughn JE. The morphology and distribution of neuronscontaining choline acetyltransferase in the adult rat spinalcord: an immunocytochemical study. J Comp Neurol 229:329–346, 1984.

9. Barriere G, Bertrand S, Cazalets JR. Peptidergic neuromodu-lation of the lumbar locomotor network in the neonatal ratspinal cord. Peptides 26: 277–286, 2005.

10. Barriere G, Mellen N, Cazalets JR. Neuromodulation of thelocomotor network by dopamine in the isolated spinal cordof newborn rat. Eur J Neurosci 19: 1325–1335, 2004.

11. Barriere G, Tartas M, Cazalets JR, Bertrand SS. Interplaybetween neuromodulator-induced switching of short-termplasticity at sensorimotor synapses in the neonatal rat spinalcord. J Physiol 586: 1903–1920, 2008.

12. Barthe JY, Clarac F. Modulation of the spinal network forlocomotion by substance P in the neonatal rat. Exp Brain Res115: 485–492, 1997.

13. Beato M, Bracci E, Nistri A. Contribution of NMDA andnon-NMDA glutamate receptors to locomotor pattern gen-eration in the neonatal rat spinal cord. Proc Biol Sci 264:877–884, 1997.

14. Beato M, Nistri A. Serotonin-induced inhibition of locomotorrhythm of the rat isolated spinal cord is mediated by the5-HT1 receptor class. Proc Biol Sci 265: 2073–2080, 1998.

15. Bennett DJ, Li Y, Siu M. Plateau potentials in sacrocaudalmotoneurons of chronic spinal rats, recorded in vitro. J Neu-rophysiol 86: 1955–1971, 2001.

16. Berger AJ, Takahashi T. Serotonin enhances a low-voltage-activated calcium current in rat spinal motoneurons. J Neu-rosci 10: 1922–1928, 1990.

17. Bertrand S, Cazalets JR. Presynaptic GABAergic control ofthe locomotor drive in the isolated spinal cord of neonatalrats. Eur J Neurosci 11: 583–592, 1999.

18. Bjorklund A, Skagerberg G. Evidence for a major spinal cordprojection from the diencephalic A11 dopamine cell group inthe rat using transmitter-specific fluorescent retrograde trac-ing. Brain Res 177: 170–175, 1979.

18a. Boothby KM, Roberts A. The stopping response of Xenopuslaevis embryos: pharmacology and intracellular physiology ofrhythmic spinal neurones and hindbrain neurones. J Exp Biol169: 65–86, 1992.

18b. Brown P, Dale N. Adenosine A1 receptors modulate highvoltage-activated Ca2� currents and motor pattern genera-tion in the Xenopus embryo. J Physiol 525: 655–667, 2000.

19. Brustein E, Chong M, Holmqvist B, Drapeau P. Serotoninpatterns locomotor network activity in the developing ze-brafish by modulating quiescent periods. J Neurobiol 57:303–322, 2003.

20. Brustein E, Drapeau P. Serotoninergic modulation of chloridehomeostasis during maturation of the locomotor network inzebrafish. J Neurosci 25: 10607–10616, 2005.

21. Brustein E, Rossignol S. Recovery of locomotion after ventraland ventrolateral spinal lesions in the cat. II. Effects of nor-adrenergic and serotoninergic drugs. J Neurophysiol 81:1513–1530, 1999.

22. Butt SJ, Kiehn O. Functional identification of interneuronsresponsible for left-right coordination of hindlimbs in mam-mals. Neuron 38: 953–963, 2003.

23. Candiani S, Augello A, Oliveri D, Passalacqua M, Pennati R, DeBernardi F, Pestarino M. Immunocytochemical localization of se-rotonin in embryos, larvae and adults of the lancelet, Branchios-toma floridae. Histochem J 33: 413–420, 2001.

24. Carlin KP, Dai Y, Jordan LM. Cholinergic and serotonergicexcitation of ascending commissural neurons in the thoraco-lumbar spinal cord of the neonatal mouse. J Neurophysiol 95:1278–1284, 2006.

25. Carp JS, Anderson RJ. Dopamine receptor-mediated depression of spinal monosynaptic transmission.Brain Res 242: 247–254, 1982.

26. Carr PA, Pearson JC, Fyffe RE. Distribution of 5-hydroxytryp-tamine-immunoreactive boutons on immunohistochemically-identified Renshaw cells in cat and rat lumbar spinal cord.Brain Res 823: 198–201, 1999.

27. Cazalets JR, Sqalli-Houssaini Y, Clarac F. Activation of thecentral pattern generators for locomotion by serotonin andexcitatory amino acids in neonatal rat. J Physiol 455: 187–204, 1992.

28. Chapman RJ, Issberner JP, Sillar KT. Group I mGluRs increaselocomotor network excitability in Xenopus tadpoles via pre-synaptic inhibition of glycinergic neurotransmission. Eur JNeurosci 28: 903–913, 2008.

29. Chapman RJ, Sillar KT. Modulation of a spinal locomotornetwork by metabotropic glutamate receptors. Eur J Neuro-sci 26: 2257–2268, 2007.

30. Charles KJ, Calver AR, Jourdain S, Pangalos MN. Distributionof a GABAB-like receptor protein in the rat central nervoussystem. Brain Res 989: 135–146, 2003.

31. Chau C, Barbeau H, Rossignol S. Effects of intrathecal alpha1-and alpha2-noradrenergic agonists and norepinephrine onlocomotion in chronic spinal cats. J Neurophysiol 79: 2941–2963, 1998.

32. Chevallier S, Nagy F, Cabelguen JM. Cholinergic control ofexcitability of spinal motoneurones in the salamander. JPhysiol 570: 525–540, 2006.

33. Clemens S, Hochman S. Conversion of the modulatory ac-tions of dopamine on spinal reflexes from depression tofacilitation in D3 receptor knock-out mice. J Neurosci 24:11337–11345, 2004.

REVIEWS




ownloaded from


34. Commissiong JW, Hellstrom SO, Neff NH. A newprojection from locus coeruleus to the spinal ven-tral columns: histochemical and biochemical evi-dence. Brain Res 148: 207–213, 1978.

35. Conway BA, Hultborn H, Kiehn O, Mintz I. Pla-teau potentials in alpha-motoneurones inducedby intravenous injection of L-dopa and clonidinein the spinal cat. J Physiol 405: 369–384, 1988.

36. Cowley KC, Schmidt BJ. A comparison of motorpatterns induced by N-methyl-D-aspartate, ace-tylcholine and serotonin in the in vitro neonatalrat spinal cord. Neurosci Lett 171: 147–150,1994.

37. Crone C, Hultborn H, Kiehn O, Mazieres L, Wig-strom H. Maintained changes in motoneuronalexcitability by short-lasting synaptic inputs in thedecerebrate cat. J Physiol 405: 321–343, 1988.

38. Cullheim S, Kellerth JO. A morphological studyof the axons and recurrent axon collaterals of catalpha-motoneurones supplying different func-tional types of muscle unit. J Physiol 281: 301–313, 1978.

39. Curtis DR, Lacey G. GABA-B receptor-mediatedspinal inhibition. Neuroreport 5: 540–542, 1994.

40. Dai Y, Carlin KP, Li Z, McMahon DG, BrownstoneRM, Jordan LM. Electrophysiological and phar-macological properties of locomotor activity-related neurons in cfos-EGFP mice. J Neuro-physiol 102: 3365–3383, 2009.

41. Dai Y, Jordan LM. Multiple effects of serotoninand acetylcholine on hyperpolarization-activatedinward current in locomotor activity-related neu-rons in Cfos-EGFP mice. J Neurophysiol 104:366–381, 2010.

42. Dai Y, Jordan LM. Multiple patterns and compo-nents of persistent inward current with seroton-ergic modulation in locomotor activity-relatedneurons in cfos-EGFP mice. J Neurophysiol 103:1712–1727, 2010.

43. Dale N, Gilday D. Regulation of rhythmic move-ments by purinergic neurotransmitters in frogembryos. Nature 383: 259–263, 1996.

44. Delgado-Lezama R, Perrier JF, Nedergaard S,Svirskis G, Hounsgaard J. Metabotropic synapticregulation of intrinsic response properties of tur-tle spinal motoneurones. J Physiol 504: 97–102,1997.

45. Diaz-Rios M, Dombeck DA, Webb WW, Harris-Warrick RM. Serotonin modulates dendritic cal-cium influx in commissural interneurons in themouse spinal locomotor network. J Neurophysiol98: 2157–2167, 2007.

46. Dunbar MJ, Tran MA, Whelan PJ. Endogenousextracellular serotonin modulates the spinal loco-motor network of the neonatal mouse. J Physiol588: 139–156, 2010.

47. Dunford C, Sillar KT, Miles GB. Regulation ofmammalian spinal locomotor networks by nitricoxide In: Society for Neuroscience, 2009. Chi-cago, IL, 2009, program no. 178.174.

48. Edwards FR, Harrison PJ, Jack JJ, Kullmann DM.Reduction by baclofen of monosynaptic EPSPs inlumbosacral motoneurones of the anaesthetizedcat. J Physiol 416: 539–556, 1989.

49. El Manira A, Kettunen P, Hess D, Krieger P.Metabotropic glutamate receptors provide in-trinsic modulation of the lamprey locomotor net-work. Brain Res Rev 40: 9–18, 2002.

50. El Manira A, Kyriakatos A. The role of endocan-nabinoid signaling in motor control. Physiology25: 230–238, 2010.

51. Elliott P, Wallis DI. Serotonin and L-norepinephrineas mediators of altered excitability in neonatal ratmotoneurons studied in vitro. Neuroscience 47:533–544, 1992.

52. Fedirchuk B, Dai Y. Monoamines increase theexcitability of spinal neurones in the neonatal ratby hyperpolarizing the threshold for action po-tential production. J Physiol 557: 355–361, 2004.

53. Feldman AG, Orlovsky GN. Activity of interneu-rons mediating reciprocal 1a inhibition during lo-comotion. Brain Res 84: 181–194, 1975.

54. Feldman JL,Del Negro CA. Looking for inspira-tion: new perspectives on respiratory rhythm.Nat Rev Neurosci 7: 232–242, 2006.

55. Feraboli-Lohnherr D, Barthe JY, Orsal D. Serotonin-induced activation of the network for locomotion inadult spinal rats. J Neurosci Res 55: 87–98, 1999.

56. Forssberg H, Grillner S. The locomotion of theacute spinal cat injected with clonidine iv. BrainRes 50: 184–186, 1973.

57. Gabbay H, Lev-Tov A. Alpha-1 adrenoceptoragonists generate a “fast” NMDA receptor-independent motor rhythm in the neonatal ratspinal cord. J Neurophysiol 92: 997–1010,2004.

58. Gabriel JP, Mahmood R, Kyriakatos A, Soll I,Hauptmann G, Calabrese RL, El Manira A. Sero-tonergic modulation of locomotion in zebrafish:endogenous release and synaptic mechanisms. JNeurosci 29: 10387–10395, 2009.

59. Gerin C, Becquet D, Privat A. Direct evidence forthe link between monoaminergic descendingpathways and motor activity. I. A study with mi-crodialysis probes implanted in the ventral funic-ulus of the spinal cord. Brain Res 704: 191–201,1995.

60. Giroux N, Rossignol S, Reader TA. Autoradiographicstudy of alpha1- and alpha2-noradrenergic andserotonin1A receptors in the spinal cord of normaland chronically transected cats. J Comp Neurol406: 402–414, 1999.

61. Gladden MH, Maxwell DJ, Sahal A, Jankowska E.Coupling between serotoninergic and noradren-ergic neurones and gamma-motoneurones in thecat. J Physiol 527: 213–223, 2000.

62. Gordon IT, Whelan PJ. Monoaminergic control ofcauda-equina-evoked locomotion in the neonatalmouse spinal cord. J Neurophysiol 96: 3122–3129, 2006.

63. Goulding M. Circuits controlling vertebrate loco-motion: moving in a new direction. Nat Rev Neu-rosci 10: 507–518, 2009.

64. Gray PA, Rekling JC, Bocchiaro CM, Feldman JL.Modulation of respiratory frequency by peptider-gic input to rhythmogenic neurons in the pre-Botzinger complex. Science 286: 1566–1568,1999.

65. Grillner S. Biological pattern generation: the cel-lular and computational logic of networks in mo-tion. Neuron 52: 751–766, 2006.

65a. Grillner S, Wallen P, Hill R, Cangiano L, El ManiraA. Ion channels of importance for the locomotorpattern generation in the lamprey brainstem-spi-nal cord. J Physiol 533: 23–30, 2001.

66. Hagglund M, Borgius L, Dougherty KJ, Kiehn O.Activation of groups of excitatory neurons in themammalian spinal cord or hindbrain evokes loco-motion. Nat Neurosci 13: 246–252, 2010.

67. Hammar I, Bannatyne BA, Maxwell DJ, EdgleySA, Jankowska E. The actions of monoaminesand distribution of noradrenergic and serotonin-ergic contacts on different subpopulations ofcommissural interneurons in the cat spinal cord.Eur J Neurosci 19: 1305–1316, 2004.

68. Hammar I, Chojnicka B, Jankowska E. Modulationof responses of feline ventral spinocerebellartract neurons by monoamines. J Comp Neurol443: 298–309, 2002.

69. Hammar I, Stecina K, Jankowska E. Differentialmodulation by monoamine membrane receptoragonists of reticulospinal input to lamina VIII fe-line spinal commissural interneurons. Eur J Neu-rosci 26: 1205–1212, 2007.

70. Han P, Nakanishi ST, Tran MA, Whelan PJ. Dopa-minergic modulation of spinal neuronal excitabil-ity. J Neurosci 27: 13192–13204, 2007.

71. Han P, Whelan PJ. Modulation of AMPA currentsby D(1)-like but not D(2)-like receptors in spinalmotoneurons. Neuroscience 158: 1699–1707,2009.

72. Hanson MG, Landmesser LT. Characterization ofthe circuits that generate spontaneous episodesof activity in the early embryonic mouse spinalcord. J Neurosci 23: 587–600, 2003.

73. Harvey PJ, Li X, Li Y, Bennett DJ. 5-Ht2 receptoractivation facilitates a persistent sodium currentand repetitive firing in spinal motoneurons of ratswith and without chronic spinal cord injury. JNeurophysiol 96: 1158–1170, 2006.

74. Heckman CJ, Lee RH, Brownstone RM. Hyperex-citable dendrites in motoneurons and their neu-romodulatory control during motor behavior.Trends Neurosci 26: 688–695, 2003.

75. Hellstrom J, Oliveira AL, Meister B, Cullheim S.Large cholinergic nerve terminals on subsets ofmotoneurons and their relation to muscarinic re-ceptor type 2. J Comp Neurol 460: 476–486,2003.

76. Hill RH, Svensson E, Dewael Y, Grillner S. 5-HTinhibits N-type but not L-type Ca2� channels via5-HT1A receptors in lamprey spinal neurons. EurJ Neurosci 18: 2919–2924, 2003.

77. Hinckley CA, Hartley R, Wu L, Todd A, Ziskind-Conhaim L. Locomotor-like rhythms in a geneti-cally distinct cluster of interneurons in themammalian spinal cord. J Neurophysiol 93:1439–1449, 2005.

78. Hokfelt T, Phillipson O, Goldstein M. Evidencefor a dopaminergic pathway in the rat descend-ing from the A11 cell group to the spinal cord.Acta Physiol Scand 107: 393–395, 1979.

79. Holstege JC, Bongers CM. Ultrastructural as-pects of the coeruleo-spinal projection. ProgBrain Res 88: 143–156, 1991.

80. Holstege JC, Van Dijken H, Buijs RM, GoedknegtH, Gosens T, Bongers CM. Distribution of dopa-mine immunoreactivity in the rat, cat and monkeyspinal cord. J Comp Neurol 376: 631–652, 1996.

81. Hounsgaard J, Hultborn H, Jespersen B, Kiehn O.Bistability of alpha-motoneurones in the decere-brate cat and in the acute spinal cat after intra-venous 5-hydroxytryptophan. J Physiol 405: 345–367, 1988.

82. Hounsgaard J, Kiehn O. Serotonin-induced bistabil-ity of turtle motoneurones caused by a nifedipine-sensitive calcium plateau potential. J Physiol 414:265–282, 1989.

83. Huang A, Noga BR, Carr PA, Fedirchuk B, JordanLM. Spinal cholinergic neurons activated duringlocomotion: localization and electrophysiologicalcharacterization. J Neurophysiol 83: 3537–3547,2000.

84. Hultborn H, Jankowska E, Lindstrom S. Recurrentinhibition of interneurones monosynaptically ac-tivated from group Ia afferents. J Physiol 215:613–636, 1971.

85. Iwagaki N, Miles GB. Activation of group Imetabotropic glutamate receptors modulateslocomotor-related motoneuron output in mice. JNeurophysiol 105: 2108–2120, 2011.

86. Jankowska E, Hammar I, Chojnicka B, Heden CH.Effects of monoamines on interneurons in fourspinal reflex pathways from group I and/or groupII muscle afferents. Eur J Neurosci 12: 701–714,2000.

REVIEWS




ownloaded from


86a. Jiang Z, Carlin KP, Brownstone RM. An in vitrofunctionally mature mouse spinal cord prepara-tion for the study of spinal motor networks. BrainRes 816: 493–439, 1999.

87. Jimenez I, Rudomin P, Enriquez M. Differentialeffects of (�)-baclofen on Ia and descendingmonosynaptic EPSPs. Exp Brain Res 85: 103–113,1991.

88. Jordan LM, Liu J, Hedlund PB, Akay T, PearsonKG. Descending command systems for the initi-ation of locomotion in mammals. Brain Res Rev57: 183–191, 2008.

89. Katz PS (editor). Beyond Neurotransmission:Neuromodulation and Its Importance for Infor-mation Processing. Oxford, UK: Oxford Univ.Press, 1999.

90. Kettunen P, Krieger P, Hess D, El Manira A. Sig-naling mechanisms of metabotropic glutamatereceptor 5 subtype and its endogenous role in alocomotor network. J Neurosci 22: 1868–1873,2002.

91. Kiehn O, Hultborn H, Conway BA. Spinal locomo-tor activity in acutely spinalized cats induced byintrathecal application of noradrenaline. Neuro-sci Lett 143: 243–246, 1992.

92. Kiehn O, Kjaerulff O. Spatiotemporal character-istics of 5-HT and dopamine-induced rhythmichindlimb activity in the in vitro neonatal rat. JNeurophysiol 75: 1472–1482, 1996.

93. Kiehn O, Sillar KT, Kjaerulff O, McDearmid JR.Effects of noradrenaline on locomotor rhythm-generating networks in the isolated neonatal ratspinal cord. J Neurophysiol 82: 741–746, 1999.

94. Kjaerulff O, Kiehn O. 5-HT modulation of multipleinward rectifiers in motoneurons in intact prepa-rations of the neonatal rat spinal cord. J Neuro-physiol 85: 580–593, 2001.

95. Krieger P, Grillner S, El Manira A. Endogenousactivation of metabotropic glutamate receptorscontributes to burst frequency regulation in thelamprey locomotor network. Eur J Neurosci 10:3333–3342, 1998.

96. Kyriakatos A, El Manira A. Long-term plasticity ofthe spinal locomotor circuitry mediated by endo-cannabinoid and nitric oxide signaling. J Neuro-sci 27: 12664–12674, 2007.

97. Kyriakatos A, Molinari M, Mahmood R, Grillner S,Sillar KT, El Manira A. Nitric oxide potentiation oflocomotor activity in the spinal cord of the lam-prey. J Neurosci 29: 13283–13291, 2009.

98. Lagerback PA, Ronnevi LO, Cullheim S, KellerthJO. An ultrastructural study of the synaptic con-tacts of alpha-motoneurone axon collaterals. I.Contacts in lamina IX and with identified alpha-motoneurone dendrites in lamina VII. Brain Res207: 247–266, 1981.

99. Lakke EA. The projections to the spinal cord ofthe rat during development: a timetable of de-scent. Adv Anat Embryol Cell Biol 135: I–XIV,1–143, 1997.

100. Landry ES, Lapointe NP, Rouillard C, Levesque D,Hedlund PB, Guertin PA. Contribution of spinal5-HT1A and 5-HT7 receptors to locomotor-likemovement induced by 8-OH-DPAT in spinal cord-transected mice. Eur J Neurosci 24: 535–546,2006.

101. Lanuza GM, Gosgnach S, Pierani A, Jessell TM,Goulding M. Genetic identification of spinal in-terneurons that coordinate left-right locomotoractivity necessary for walking movements. Neu-ron 42: 375–386, 2004.

102. Lapointe NP, Rouleau P, Ung RV, Guertin PA.Specific role of dopamine D1 receptors in spinalnetwork activation and rhythmic movement in-duction in vertebrates. J Physiol 587: 1499–1511,2009.

103. Laslo P, Lipski J, Funk GD. Differential expressionof Group I metabotropic glutamate receptors inmotoneurons at low and high risk for degenera-tion in ALS. Neuroreport 12: 1903–1908, 2001.

104. Lee RH, Heckman CJ. Enhancement of bistabilityin spinal motoneurons in vivo by the noradrener-gic alpha1 agonist methoxamine. J Neurophysiol81: 2164–2174, 1999.

105. Lev-Tov A, Meyers DE, Burke RE. Activation oftype B gamma-aminobutyric acid receptors in theintact mammalian spinal cord mimics the effectsof reduced presynaptic Ca2� influx. Proc NatlAcad Sci USA 85: 5330–5334, 1988.

106. Li X, Murray K, Harvey PJ, Ballou EW, BennettDJ. Serotonin facilitates a persistent calcium cur-rent in motoneurons of rats with and withoutchronic spinal cord injury. J Neurophysiol 97:1236–1246, 2007.

107. Liu J, Akay T, Hedlund PB, Pearson KG, JordanLM. Spinal 5-HT7 receptors are critical for alter-nating activity during locomotion: in vitro neona-tal and in vivo adult studies using 5-HT7 receptorknockout mice. J Neurophysiol 102: 337–348,2009.

108. Liu J, Jordan LM. Stimulation of the parapyrami-dal region of the neonatal rat brain stem pro-duces locomotor-like activity involving spinal5-HT7 and 5-HT2A receptors. J Neurophysiol 94:1392–1404, 2005.

109. Machacek DW, Hochman S. Noradrenaline un-masks novel self-reinforcing motor circuits withinthe mammalian spinal cord. J Neurosci 26: 5920–5928, 2006.

110. MacLean JN, Cowley KC, Schmidt BJ. NMDAreceptor-mediated oscillatory activity in the neo-natal rat spinal cord is serotonin dependent. JNeurophysiol 79: 2804–2808, 1998.

111. MacLean JN, Schmidt BJ. Voltage-sensitivity ofmotoneuron NMDA receptor channels is modu-lated by serotonin in the neonatal rat spinal cord.J Neurophysiol 86: 1131–1138, 2001.

112. Madriaga MA, McPhee LC, Chersa T, Christie KJ,Whelan PJ. Modulation of locomotor activity bymultiple 5-HT and dopaminergic receptor sub-types in the neonatal mouse spinal cord. J Neu-rophysiol 92: 1566–1576, 2004.

113. Maitra KK, Seth P, Thewissen M, Ross HG, Gan-guly DK. Dopaminergic influence on the excit-ability of antidromically activated Renshaw cellsin the lumbar spinal cord of the rat. Acta PhysiolScand 148: 101–107, 1993.

114. Marchetti C, Taccola G, Nistri A. Activation ofgroup I metabotropic glutamate receptors de-presses recurrent inhibition of motoneurons inthe neonatal rat spinal cord in vitro. Exp BrainRes 164: 406–410, 2005.

115. Marchetti C, Taccola G, Nistri A. Distinct sub-types of group I metabotropic glutamate recep-tors on rat spinal neurons mediate complexfacilitatory and inhibitory effects. Eur J Neurosci18: 1873–1883, 2003.

115a.Matsushima T, Tegner J, Hill RH, Grillner S.GABAB receptor activation causes a depressionof low- and high-voltage-activated Ca2� cur-rents, postinhibitory rebound, and postspike af-terhyperpolarization in lamprey neurons. JNeurophysiol 70: 2606–2619, 1993.

116. Maxwell DJ, Riddell JS, Jankowska E. Serotonin-ergic and noradrenergic axonal contacts associ-ated with premotor interneurons in spinalpathways from group II muscle afferents. Eur JNeurosci 12: 1271–1280, 2000.

117. McCrea DA. Spinal circuitry of sensorimotor con-trol of locomotion. J Physiol 533: 41–50, 2001.

118. McDearmid JR, Scrymgeour-Wedderburn JF, Sil-lar KT. Aminergic modulation of glycine releasein a spinal network controlling swimming in Xe-nopus laevis. J Physiol 503: 111–117, 1997.

119. McLaughlin BJ. Propriospinal and supraspinalprojections to the motor nuclei in the cat spinalcord. J Comp Neurol 144: 475–500, 1972.

120. McLean DL, Sillar KT. The distribution of NADPH-diaphorase-labelled interneurons and the role ofnitric oxide in the swimming system of Xenopuslaevis larvae. J Exp Biol 203: 705–713, 2000.

121. McLean DL, Sillar KT. Metamodulation of a spinallocomotor network by nitric oxide. J Neurosci 24:9561–9571, 2004.

122. McLean DL, Sillar KT. Nitric oxide selectivelytunes inhibitory synapses to modulate vertebratelocomotion. J Neurosci 22: 4175–4184, 2002.

123. McLean DL, Sillar KT. Spatiotemporal pattern ofnicotinamide adenine dinucleotide phosphate-diaphorase reactivity in the developing centralnervous system of premetamorphic Xenopus lae-vis tadpoles. J Comp Neurol 437: 350–362, 2001.

123a.Meer DP, Buchanan JT. Apamin reduces the lateafterhyperpolarization of lamprey spinal neurons,with little effect on fictive swimming. NeurosciLett 143: 1–4, 1992.

124. Miles GB, Hartley R, Todd AJ, Brownstone RM.Spinal cholinergic interneurons regulate the ex-citability of motoneurons during locomotion.Proc Natl Acad Sci USA 104: 2448–2453, 2007.

125. Muennich EA, Fyffe RE. Focal aggregation ofvoltage-gated, Kv2.1 subunit-containing, potas-sium channels at synaptic sites in rat spinal mo-toneurones. J Physiol 554: 673–685, 2004.

126. Myers CP, Lewcock JW, Hanson MG, GosgnachS, Aimone JB, Gage FH, Lee KF, Landmesser LT,Pfaff SL. Cholinergic input is required during em-bryonic development to mediate proper assem-bly of spinal locomotor circuits. Neuron 46: 37–49, 2005.

127. Nishimaru H, Takizawa H, Kudo N. 5-Hydroxytryptamine-induced locomotor rhythm inthe neonatal mouse spinal cord in vitro. NeurosciLett 280: 187–190, 2000.

128. Noga BR, Johnson DM, Riesgo MI, Pinzon A.Locomotor-activated neurons of the cat. I. Sero-tonergic innervation and co-localization of 5-HT7,5-HT2A, and 5-HT1A receptors in the thoraco-lumbar spinal cord. J Neurophysiol 102: 1560–1576, 2009.

129. Noga BR, Johnson DM, Riesgo MI, Pinzon A.Locomotor-activated neurons of the cat. II. Nor-adrenergic innervation and colocalization withNE�1a or NE�2b receptors in the thoraco-lumbar spinal cord. J Neurophysiol 105: 1835–1849, 2011.

130. Nygren LG, Olson L. A new major projectionfrom locus coeruleus: the main source of norad-renergic nerve terminals in the ventral and dorsalcolumns of the spinal cord. Brain Res 132: 85–93,1977.

130a.Okado N, Sako H, Homma S, Ishikawa K. Devel-opment of serotoninergic system in the brain andspinal cord of the chick. Prog Neurobiol 38: 93–123, 1992.

131. Ono H, Fukuda H. Pharmacology of descendingnoradrenergic systems in relation to motor func-tion. Pharmacol Ther 68: 105–112, 1995.

131a.Parker D, Zhang W, Grillner S. Substance P mod-ulates NMDA responses and causes long-termprotein synthesis-dependent modulation of thelamprey locomotor network. J Neurosci 18:4800–4813, 1998.

131b.Parker D, Grillner S. Cellular and synaptic modu-lation underlying substance P-mediated plasticityof the lamprey locomotor network. J Neurosci18: 8095–8110, 1998.

132. Pearlstein E, Ben Mabrouk F, Pflieger JF, Vinay L.Serotonin refines the locomotor-related alterna-tions in the in vitro neonatal rat spinal cord. Eur JNeurosci 21: 1338–1346, 2005.

REVIEWS




ownloaded from


133. Pearson SA, Mouihate A, Pittman QJ, Whelan PJ.Peptidergic activation of locomotor pattern gen-erators in the neonatal spinal cord. J Neurosci 23:10154–10163, 2003.

134. Perrier JF, Hounsgaard J. 5-HT2 receptors pro-mote plateau potentials in turtle spinal motoneu-rons by facilitating an L-type calcium current. JNeurophysiol 89: 954–959, 2003.

135. Phelps PE, Barber RP, Houser CR, Crawford GD,Salvaterra PM, Vaughn JE. Postnatal develop-ment of neurons containing choline acetyltrans-ferase in rat spinal cord: an immunocytochemicalstudy. J Comp Neurol 229: 347–361, 1984.

136. Pin JP, Duvoisin R. The metabotropic glutamatereceptors: structure and functions. Neurophar-macology 34: 1–26, 1995.

137. Qu S, Ondo WG, Zhang X, Xie WJ, Pan TH, LeWD. Projections of diencephalic dopamine neu-rons into the spinal cord in mice. Exp Brain Res168: 152–156, 2006.

138. Rajaofetra N, Ridet JL, Poulat P, Marlier L, Sand-illon F, Geffard M, Privat A. Immunocytochemicalmapping of noradrenergic projections to the ratspinal cord with an antiserum against noradren-aline. J Neurocytol 21: 481–494, 1992.

139. Ramanathan S, Combes D, Molinari M, SimmersJ, Sillar KT. Developmental and regional expres-sion of NADPH-diaphorase/nitric oxide synthasein spinal cord neurons correlates with the emer-gence of limb motor networks in metamorphos-ing Xenopus laevis. Eur J Neurosci 24: 1907–1922, 2006.

140. Rauscent A, Einum J, Le Ray D, Simmers J,Combes D. Opposing aminergic modulation ofdistinct spinal locomotor circuits and their func-tional coupling during amphibian metamorpho-sis. J Neurosci 29: 1163–1174, 2009.

141. Renshaw B. Central effects of centripetal im-pulses in axons of spinal ventral roots. J Neuro-physiol 9: 191–204, 1946.

142. Roudet C, Mouchet P, Feuerstein C, Savasta M.Normal distribution of alpha 2-adrenoceptors inthe rat spinal cord and its modification after nor-adrenergic denervation: a quantitative autora-diographic study. J Neurosci Res 39: 319–329,1994.

143. Roudet C, Savasta M, Feuerstein C. Normal dis-tribution of alpha-1-adrenoceptors in the rat spi-nal cord and its modification after noradrenergicdenervation: a quantitative autoradiographicstudy. J Neurosci Res 34: 44–53, 1993.

144. Schmidt BJ, Jordan LM. The role of serotonin inreflex modulation and locomotor rhythm produc-tion in the mammalian spinal cord. Brain Res Bull53: 689–710, 2000.

145. Schwartz EJ, Gerachshenko T, Alford S. 5-HTprolongs ventral root bursting via presynapticinhibition of synaptic activity during fictive loco-motion in lamprey. J Neurophysiol 93: 980–988,2005.

146. Seth P, Gajendiran M, Maitra KK, Ross HG, GangulyDK. Evidence for D1 dopamine receptor-mediatedmodulation of the synaptic transmission from mo-tor axon collaterals to Renshaw cells in the ratspinal cord. Neurosci Lett 158: 217–220, 1993.

147. Sherriff FE, Henderson Z, Morrison JF. Furtherevidence for the absence of a descending cholin-ergic projection from the brainstem to the spinalcord in the rat. Neurosci Lett 128: 52–56, 1991.

148. Sillar KT, Combes D, Ramanathan S, Molinari M,Simmers J. Neuromodulation and developmentalplasticity in the locomotor system of anuran am-phibians during metamorphosis. Brain Res Rev57: 94–102, 2008.

149. Sillar KT, Woolston AM, Wedderburn JF. Involve-ment of brainstem serotonergic interneurons inthe development of a vertebrate spinal locomo-tor circuit. Proc Biol Sci 259: 65–70, 1995.

150. Skagerberg G, Lindvall O. Organization of dien-cephalic dopamine neurones projecting to thespinal cord in the rat. Brain Res 342: 340–351,1985.

151. Smith JC, Feldman JL. In vitro brainstem-spinalcord preparations for study of motor systems formammalian respiration and locomotion. J Neuro-sci Methods 21: 321–333, 1987.

152. Smith JC, Feldman JL, Schmidt BJ. Neural mech-anisms generating locomotion studied in mam-malian brain stem-spinal cord in vitro. FASEB J 2:2283–2288, 1988.

153. Spike RC, Todd AJ, Johnston HM. Coexistenceof NADPH diaphorase with GABA, glycine, andacetylcholine in rat spinal cord. J Comp Neurol335: 320–333, 1993.

154. Sqalli-Houssaini Y, Cazalets JR. Noradrenergiccontrol of locomotor networks in the in vitrospinal cord of the neonatal rat. Brain Res 852:100–109, 2000.

155. Stuart GJ, Redman SJ. The role of GABAA andGABAB receptors in presynaptic inhibition of IaEPSPs in cat spinal motoneurones. J Physiol 447:675–692, 1992.

156. Svensson E, Woolley J, Wikstrom M, Grillner S.Endogenous dopaminergic modulation of thelamprey spinal locomotor network. Brain Res970: 1–8, 2003.

157. Taccola G, Marchetti C, Nistri A. Effect ofmetabotropic glutamate receptor activity onrhythmic discharges of the neonatal rat spinalcord in vitro. Exp Brain Res 153: 388–393, 2003.

158. Taccola G, Marchetti C, Nistri A. Modulation ofrhythmic patterns and cumulative depolarizationby group I metabotropic glutamate receptors inthe neonatal rat spinal cord in vitro. Eur J Neu-rosci 19: 533–541, 2004.

159. Taccola G, Marchetti C, Nistri A. Role of group IIand III metabotropic glutamate receptors inrhythmic patterns of the neonatal rat spinal cordin vitro. Exp Brain Res 156: 495–504, 2004.

160. Talpalar AE, Kiehn O. Glutamatergic mechanismsfor speed control and network operation in therodent locomotor CpG. Front Neural Circuits 4:2010.

161. Tartas M, Morin F, Barriere G, Goillandeau M,Lacaille JC, Cazalets JR, Bertrand SS. Noradren-ergic modulation of intrinsic and synaptic prop-erties of lumbar motoneurons in the neonatal ratspinal cord. Front Neural Circuits 4: 4, 2010.

162. Theiss RD, Heckman CJ. Systematic variation ineffects of serotonin and norepinephrine on re-petitive firing properties of ventral horn neurons.Neuroscience 134: 803–815, 2005.

163. Thirumalai V, Cline HT. Endogenous dopaminesuppresses initiation of swimming in prefeedingzebrafish larvae. J Neurophysiol 100: 1635–1648,2008.

164. Towers S, Princivalle A, Billinton A, Edmunds M,Bettler B, Urban L, Castro-Lopes J, Bowery NG.GABAB receptor protein and mRNA distributionin rat spinal cord and dorsal root ganglia. Eur JNeurosci 12: 3201–3210, 2000.

165. Ung RV, Landry ES, Rouleau P, Lapointe NP,Rouillard C, Guertin PA. Role of spinal 5-HT2receptor subtypes in quipazine-induced hindlimbmovements after a low-thoracic spinal cord tran-section. Eur J Neurosci 28: 2231–2242, 2008.

166. VanderHorst VG, Ulfhake B. The organization ofthe brainstem and spinal cord of the mouse: re-lationships between monoaminergic, cholinergic,and spinal projection systems. J Chem Neuro-anat 31: 2–36, 2006.

167. Viala D, Buser P. Methods of obtaining locomo-tor rhythms in the spinal rabbit by pharmacolog-ical treatments (DOPA, 5-HTP, D-amphetamine).Brain Res 35: 151–165, 1971.

168. Vinay L, Brocard F, Pflieger JF, Simeoni-Alias J,Clarac F. Perinatal development of lumbar mo-toneurons and their inputs in the rat. Brain ResBull 53: 635–647, 2000.

169. Wall MJ, Dale N. GABAB receptors modulateglycinergic inhibition and spike threshold in Xe-nopus embryo spinal neurones. J Physiol 469:275–290, 1993.

169a.Wallen P, Grillner S. N-methyl-D-aspartate recep-tor-induced, inherent oscillatory activity in neu-rons active during fictive locomotion in thelamprey. J Neurosci 7: 2745–2755, 1987.

170. Wang MY, Dun NJ. Phaclofen-insensitive presyn-aptic inhibitory action of (�/�)-baclofen in neo-natal rat motoneurones in vitro. Br J Pharmacol99: 413–421, 1990.

171. Wei J, Walton EA, Milici A, Buccafusco JJ. M1-m5muscarinic receptor distribution in rat CNS byRT-PCR and HPLC. J Neurochem 63: 815–821,1994.

172. Weil-Fugazza J, Godefroy F. Dorsal and ventraldopaminergic innervation of the spinal cord:functional implications. Brain Res Bull 30: 319–324, 1993.

173. Welton J, Stewart W, Kerr R, Maxwell DJ. Differ-ential expression of the muscarinic m2 acetylcho-line receptor by small and large motoneurons ofthe rat spinal cord. Brain Res 817: 215–219, 1999.

174. Westlund KN, Bowker RM, Ziegler MG, CoulterJD. Noradrenergic projections to the spinal cordof the rat. Brain Res 263: 15–31, 1983.

175. Whelan P, Bonnot A, O’Donovan MJ. Propertiesof rhythmic activity generated by the isolatedspinal cord of the neonatal mouse. J Neuro-physiol 84: 2821–2833, 2000.

176. White SR, Fung SJ, Barnes CD. Norepinephrineeffects on spinal motoneurons. Prog Brain Res88: 343–350, 1991.

177. Wikstrom M, Hill R, Hellgren J, Grillner S. Theaction of 5-HT on calcium-dependent potassiumchannels and on the spinal locomotor network inlamprey is mediated by 5-HT1A-like receptors.Brain Res 678: 191–199, 1995.

178. Wilson JM, Hartley R, Maxwell DJ, Todd AJ,Lieberam I, Kaltschmidt JA, Yoshida Y, JessellTM, Brownstone RM. Conditional rhythmicity ofventral spinal interneurons defined by expressionof the Hb9 homeodomain protein. J Neurosci 25:5710–5719, 2005.

179. Wilson JM, Rempel J, Brownstone RM. Postnataldevelopment of cholinergic synapses on mousespinal motoneurons. J Comp Neurol 474: 13–23,2004.

180. Witts E, Panetta K, Miles GB. Purinergic receptoractivation modulates locomotor-related activityin the isolated mouse spinal cord. In: Society forNeuroscience, 2010. San Diego, CA: Soc. Neuro-sci., 2010, program no. 288.212.

181. Yoshida M, Tanaka M. Existence of new dopami-nergic terminal plexus in the rat spinal cord: as-sessment by immunohistochemistry using anti-dopamine serum. Neurosci Lett 94: 5–9, 1988.

182. Zagoraiou L, Akay T, Martin JF, Brownstone RM,Jessell TM, Miles GB. A cluster of cholinergicpremotor interneurons modulates mouse loco-motor activity. Neuron 64: 645–662, 2009.

183. Zhang W, Grillner S. The spinal 5-HT system con-tributes to the generation of fictive locomotion inlamprey. Brain Res 879: 188–192, 2000.

184. Zhong G, Diaz-Rios M, Harris-Warrick RM. Intrin-sic and functional differences among commis-sural interneurons during fictive locomotion andserotonergic modulation in the neonatal mouse.J Neurosci 26: 6509–6517, 2006.

REVIEWS




ownloaded from


185. Zhong G, Diaz-Rios M, Harris-Warrick RM. Sero-tonin modulates the properties of ascendingcommissural interneurons in the neonatal mousespinal cord. J Neurophysiol 95: 1545–1555, 2006.

186. Zhong G, Droho S, Crone SA, Dietz S, Kwan AC,Webb WW, Sharma K, Harris-Warrick RM. Elec-trophysiological characterization of V2a interneu-rons and their locomotor-related activity in theneonatal mouse spinal cord. J Neurosci 30: 170–182, 2010.

187. Zhu H, Clemens S, Sawchuk M, Hochman S. Ex-pression and distribution of all dopamine recep-tor subtypes [D(1)-D(5)] in the mouse lumbarspinal cord: a real-time polymerase chain reac-tion and non-autoradiographic in situ hybridiza-tion study. Neuroscience 149: 885–897, 2007.

188. Zhu H, Clemens S, Sawchuk M, Hochman S. Un-altered D1, D2, D4, and D5 dopamine receptormRNA expression and distribution in the spinalcord of the D3 receptor knockout mouse. JComp Physiol A 194: 957–962, 2008.

REVIEWS




ownloaded from


Neuromodulation of Vertebrate

Documents

Transcript of Neuromodulation of Vertebrate