B R A I N R E S E A R C H R E V I E W S 5 7 ( 2 0 0 8 ) 1 3 – 2 1

ava i l ab l e a t www.sc i enced i r ec t . com

www.e l sev i e r. com/ loca te /b ra in res rev

Review

Some historical reflections on the neural control of locomotion

François Clarac⁎

Laboratoire Plasticité et Physio-Pathologie de la Motricité, CNRS Université de la Méditerranée, 31 chemin Joseph Aiguier,13402 Marseille Cedex 20, France

A R T I C L E I N F O

⁎ Fax: +33 491775084.E-mail address: clarac@dpm.cnrs-mrs-fr.

0165-0173/$ – see front matter © 2007 Elsevidoi:10.1016/j.brainresrev.2007.07.015

A B S T R A C T

Article history:Accepted 1 July 2007Available online 22 August 2007

Thought on the neural control of locomotion dates back to antiquity. In this article, however,the focus is more recent by starting with some major 17th century concepts, which weredeveloped by René Descartes, a French philosopher; Thomas Willis, an English anatomist;and Giovanni Borelli, an Italian physiologist and physicist. Each relied on his personalexpertise to theorize on the organization and control ofmovements. The 18th and early 19thcenturies saw work on both the central and peripheral control of movement: the formermost notably by Johann Unzer, Marie Jean-Pierre Flourens and Julien-Jean-César Legallois,and the latter by Unzer, Jirí Procháska and many others. Next in the 19th century,neurologists used human locomotion as a precise tool for characterizing motor pathologies:e.g., Guillaume Duchenne de Boulogne's description of locomotor ataxia. Jean-Martin Charcotconsidered motor control to be organized at two levels of the central nervous system: thecerebral cortex and the spinal cord. Maurice Philippson's defined the dog's step cycle andconsidered that locomotion used both central and reflex mechanisms. Charles Sherringtonexplained that locomotor control was usually thought to consist of a succession ofperipheral reflexes (e.g., the stepping reflexes). Thomas Graham Brown's thencontemporary evidence for the spinal origin of locomotor rhythmicity languished inobscurity until the early 1960s. By then the stage was set for an international assault on theneural control of locomotion, which featured research conducted on both invertebrate andvertebrate animal models. These contributions have progressively became more integratedand interactive, with current work emphasizing that locomotor control involves a seamlessintegration between central locomotor networks and peripheral feedback.

© 2007 Elsevier B.V. All rights reserved.

Keywords:Movement neuroscienceLocomotionCentral pattern generationInvertebrate motor controlVertebrate motor control

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142. Locomotion: is it voluntary or automatic? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143. Locomotor pathologies described by neurologists during the 19th century . . . . . . . . . . . . . . . . . . . . . . . . 154. Central versus peripheral control of locomotion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165. Identification of locomotor neural circuitry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

er B.V. All rights reserved.

14 B R A I N R E S E A R C H R E V I E W S 5 7 ( 2 0 0 8 ) 1 3 – 2 1

6. Concluding thoughts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

1. Introduction

Following some earlier noteworthy and precedent symposiaheld in 1966 (Wiersma, 1967), 1967 (Brazier, 1969) and 1973(Stein et al., 1973), a symposium of particular significance forwork on the neural control of locomotion was held inPhiladelphia, USA in 1975. A primary goal of this meetingwas to optimize interactions between investigators workingon a large spectrum of invertebrate and vertebrate prepara-tions, literally from arthropods and molluscs to the human(Herman et al., 1976). The same group of gradually evolvingorganizers then arranged subsequent similar meetings. Thesewere held in Stockholm, Sweden in 1985 (Grillner et al., 1986)and Tucson, Arizona, USA in 1995 (Stein et al., 1997). Now, the4th meeting in this series has just been held in 2006, with thisonce again in Stockholm. It has been fascinating to observethat each of these meetings has featured somewhat differentissues as dictated by the state-of-the-art in the field. None-theless, some enduring problems have been continuouslystudied these past 30 years and these have been progressivelyanalyzed to a deeper degree. Remarkably, many of the“pioneers” of the first meeting are still active in locomotionresearch and a host of relatively young investigators havejoined in this effort, thereby attesting to the significance ofsuch research in neuroscience, in general, and movementneuroscience, in particular.

The main subject of discussion on locomotion at the 1975meeting was the neuronal basis of selected locomotorbehaviors as studied across phyla. After a preceding 60-yearperiod during which a reflex control of locomotion was thepredominant concept, the recent (1960s–early 1970s) discoveryof central pattern generating networks was emerging on theforefront and many presentations focused on the potentialneuronal circuitry for such control. At the 2006meeting, muchemphasis was placed on the plasticity of interactions betweencentrally scored locomotor networks and peripheral sensoryinput. This brief historical review highlights some of thesuccessive concepts that have brought the field to this point.

1 Twelve of Borelli's creative and unique illustrations of bodymechanics appear in Enoka (1988). The originals are in theHuntington Library, San Marino, CA, USA.

2. Locomotion: is it voluntary or automatic?

This question was fundamental after the dualist positiontaken by the French philosopher, René Descartes (1596–1650).He proposed that biological activities had two components:one controlled by a “machine” and the other by psychicalactivities due to the “spirit” and hence, the “soul” (Descartes,1664). Despite his then sophisticated understanding of neuro-anatomy, he argued that interfaces between both componentswere regulated by the pineal gland (epiphysis). He classifiedlocomotion as an automaticmovement that was controlled bythe biological machine (Clarac, 2005a). This idea of automa-tism was emphasized a century later with the development of

automatons like the flute player and swimming duck, whichwere demonstrated to the French Royal Academy of Sciencesby Jacques de Vaucanson (1709–1782). In the 17th century,Thomas Willis (1621–1675), an English professor of naturalphilosophy at the University of Oxford, UK, provided relativelyaccurate descriptions of the different parts of the centralnervous system (CNS) and the neuronal organization of move-ments (Willis, 1664). He argued that involuntary “permanent”movements like circulation and respiration were initiated andcontrolled by the cerebellum and the brainstem. In contrast,other involuntary movements with an adaptive (learned)component, like locomotion, were controlled by the striatum(Molnar, 2004).

Shortly thereafter, a detailed, quantitative mechanicalstudy of locomotion appeared (Borelli, 1680). This was thebest known work of Giovanni Borelli (1608–1679), an Italianphysiologist and physicist, who was the first to use the lawsof mechanics to explain body movements like locomotion(Ashley-Ross and Gillis, 2002). Borelli was the professor ofmathematics at the Universities of Messina (1649–1655; 1667–1673) and Pisa (1656–1673). From 1674 until his demise, helived in Rome, under the patronage of ex-Queen Christina ofSweden, who has been exiled to Rome for converting toCatholicism. Borelli is considered the founder of iatrophysics(the combination of physics and medicine). He was stronglyinfluenced by the work of Galileo Galilei (1564–1642), JohannesKepler (1571–1630) and Isaac Newton (1642–1727). Borelli wasalso a practical experimentalist. For example, he proved thatmuscle contraction was not due to the action of gaseousspirits. This was shown by submerging a struggling animal inwater, slitting some of its muscles, and showing that no bub-bles appeared!

Borelli compared the bodies of animals and humans andanalyzed the different types of movement that they could ac-complish, including, for example, those of horses and insects.He was fascinated, in particular, by insects walking upsidedown on the ceiling of a room. Borelli reasoned that animalmovements had to be considered as a legitimate aspect ofmathematics and physics. His 1680 book contained the firsttruly scientific description of human locomotion,with analysisof the action of different joints, the various forces exerted,including the propulsive force during the stance phase of thestep, and the particular role of pelvic rotation.1

Throughout the 18th century, there was an extensiveanalysis of different types of reflexes, many using the earlier-developed nerve–muscle preparation of the Dutch microsco-pist and naturalist, Jan Swammerdam (1637–1680), as docu-mentedbyCobb (2002). For example, Jirí Procháska (1749–1820),a professor of anatomy, physiology and ophthalmology in

15B R A I N R E S E A R C H R E V I E W S 5 7 ( 2 0 0 8 ) 1 3 – 2 1

Vienna and Prague (Schiller, 1999), delineated many types ofreflex action throughout the body and proposed laws for theiroperation (Procháska, 1780–84). Johann Unzer (1727–1799), amedical graduate of the University of Halle and Procháska'sresearch mentor, was also prominent in developing the con-cept of the reflex (Laycock, 1851; Liddell, 1960). Importantly, hedemonstrated that spinalized birds and mammals couldlocomote in a near-normal fashion. This was convincingevidence for a predominant spinal control of locomotion albeitit with peripheral feedback also a key controller.

The first truly accurate description of an automatic motorcenter in the CNS followed shortly thereafter, at the beginningof the 19th century. This was the work of the French phys-iologist, Julien-Jean-César Legallois (1770–1840), who localizedthe primary center of respiration to a tiny part of the medullaoblongata. His seminal work on young decorticate rabbits withtransections at various levels throughout the midbrain, pons,andmedulla led to his conclusion (p. 37 in 1815 English editionof Legallois, 1812) that “ … It is not on the entire brain that therespiration depends, but only on a very circumscribed area ofthe medulla oblongata situated a short distance from theoccipital opening toward the origin of the 8th pair of nerves”(also cited on p. 27 in Finger, 1994). This finding was sub-sequently confirmed by the French physiologist, Marie Jean-Pierre Flourens (1794–1867), who coined the term Le noeud vitalfor Legallois's medullary respiratory center, which Flourensthoughtwas no greater in size than the head of a pin. Note alsothat Flourens confirmed and extended upon Unzer's observa-tions on the predominant spinal control of locomotion in birdsand mammals (Flourens, 1824).

Interestingly, the debate on the voluntary versus automaticcontrol of locomotion has continued up to the present day, asexemplified in the report of Prochazka et al. (2000). Perhaps theissue will be laid to rest if consensus builds for the idea of theSwedish workers, Sten Grillner and Peter Wallén, that diversemotor patterns across different vertebrate species, includingthe human “ … should be labeled as voluntary, because theycan be recruited at will. Moreover, most, if not all, of the motorpatterns available at birth are subject to maturation and aremodified substantially through learning” (p. 3 in Grillner andWallén, 2004).

3. Locomotor pathologies described byneurologists during the 19th century

Modern medicine began in this century with the advent ofnosography, the systematic description of diseases. Neurolo-gists contributed substantially to this effort with detailedanalysis of different motor pathologies. Locomotion wasconsidered an excellent tool for characterizing such deficits,and this led to much speculation on how locomotion wascontrolled by the CNS. For example, Guillaume Duchenne(1806–1875), a French doctor from Boulogne-sur-Mer, becameinterested in the muscular responses to “localized electrica-lisation,” using an induction coil to produce a faradic stim-ulating current (Parent, 2005). His results led him to believethat movements resulted from groups of muscles acting insynergy (Duchenne, 1867). Like Jean-Martin Charcot (1825–1893), he worked at the famous Paris hospital, la Salpêtrière

(Goetz et al., 1995). He studied in great detail what he termedlocomotor ataxia, in which leg coordinationwas severely altered(Schiller, 1995). Similar pathologywas studied a little earlier bya German neurologist at the University of Berlin, MoritzRomberg (1795–1873), who called it tabes dorsalis in his classic,first-ever systematic textbook in neurology (Romberg, 1846).Both Duchenne de Boulogne and Romberg demonstrated thatthemotor deficit was even greater when the patient walked inthe dark. Anatomical studies soon confirmed that thepathology was due to a sclerosis of the spinal dorsal column.Romberg defined his “sign” with the presence or exaggerationof postural disequilibrium when standing with eyes closed(Pearce, 2005).

Duchenne de Boulogne concluded that movements werecentrally organized on the basis of his study of several motorpathologies. He first thought that the cerebellum was thecenter for coordinating locomotor actions, but later heproposed that the primary center was the medulla oblongata.He did not deny the significance of the findings on peripheralmuscular control of the Scottish surgeon–anatomist, CharlesBell (1774–1842), who called sensory input from the muscula-ture the sixth sense (Bell, 1830). Rather, Duchenne de Boulognedissociated a central sensation, a sort of muscular conscious-ness associated with the elaboration of a voluntary action,from the afferent control, a muscle sense due to the ongoingmovement. By the late 19th century, muscle sense was largelydiscussed as being due to both centrally and peripherallyderived sensations (see pp.171–194 in Scheerer, 1985). Theperipheral component was predominant in the work of theEnglish biologist and physician, Henry Bastian (1837–1915; seeBastian, 1887–1888) and, shortly thereafter, it was supportedby a much clearer definition of proprioception, which wasprovided by Charles Sherrington (1856–1952), an English pro-fessor of physiology at the Universities of Liverpool (1895–1912) and Oxford (1913–1935; see Sherrington, 1906). Thenature of the centrally derived component of muscle senselay dormant until it was re-formulated in themid 20th centuryby the German zoologist, Eric von Holst (1908–1962) with workundertaken on fish and on insects with absolute and relativecoordination (von Holst, 1943; von Holst and Mittelstaedt,1950) and by the American neuropsychologist, Roger Sperry(1913–1994), who used frogs and fish (Sperry, 1950). Theircombined work led to the present concepts of kinesthesia,including the presence of comparators between internalfeedback to higher centers of the descending central motorcommand (corollary discharge) and external feedback from theperiphery (for the most recent full review of kinesthesia, seeGandevia, 1996).

The 19th century neurologists considered that, at least inhumans, movements were centrally organized. Followingpioneering and soon confirmed work on the motor cortex bytwo German neurosurgeons, Gustav Fritsch (1838–1891), asurgeon and anthropologist, and Eduard Hitzig (1838–1907) apsychiatrist, which included direct electrical stimulation ofdiscrete cortical areas (Fritsch and Hitzig, 1870), two theoriesof higher motor control became popular. The first was themore widely spread by such notable scientists as Charcot,the English neurologist, John Hughlings Jackson (1835–1911),the English neurosurgeon, David Ferrier (1843–1928), and theAustrian psychiatrist and all-round neuroscientist, Theodor

16 B R A I N R E S E A R C H R E V I E W S 5 7 ( 2 0 0 8 ) 1 3 – 2 1

Meynert (1833–1892). They argued individually that a mentalrepresentation (like a motor intuition) resided in the cerebralcortex, which was able to induce voluntary movement(Gasser, 1995). The alternative theory was championed largelyby Bastian, who considered that the afferents (both internaland external feedback) played a prominent role by stimulatingthe motor centers to produce movement.

Charcot never worked on the neural control of locomotion,but in one of his famous Tuesday lessons (the 5th one inMarch1889; see Fig. 7 in Sanders, 2002), he presented a particularlyfascinating theory about such control. He argued that humanwalking was a learned movement that soon became automat-ic. He argued that someCNS centerswith “partialmemories” ofwhat is required for locomotion were able to finalize thecomplexity of the motor behavior. He believed these centerswere located at two levels, in the spinal cord and in the cerebralcortex, both linked by commissural fibers. He thought thespinal center was the more important of the two, organizingthe coordination of locomotor movements and adapting themto different environmental conditions. The cortexwas deemedneeded only to trigger the behavior, and increase or decreasethe speed and force of the locomotor rhythm. Once the cortexdelivered the voluntary command, the spinal cellular groupspossessed the partial memories necessary for the execution ofthe complexity of the movements. This vision was subse-quently expanded; first, spontaneous locomotion was de-scribed in precollicular or premammillary cats (Hinsey et al.,1930); second, under the influence of the Russian movementneuroscientist, Nicolai Bernstein (1896–1966) in his prescienttext on the construction of movements (Bernstein, 1947),Orlovsky and Shik, members of the Moscow, USSR MotorControl School, together with their student Severin, showedthat locomotion could be induced by neurons in the midbrain.If this command area for locomotion that now is known toexist in all vertebrates (e.g., Shik et al., 1966) is facilitated byelectrical stimulation in high decerebrate cats, a locomotorcontrol is revealed that involve interactions between descend-ing command signals, central locomotor networks, and sen-sory feedback (for details, see Stuart and McDonagh, 1998;Orlovsky et al., 1999). Charcot's lecture is also partially pre-cedent to the modern-day ideas of Grillner and Wallén (2004)that were presented above.

4. Central versus peripheral control oflocomotion

In the first decade of the 20th century, physiological interest inrhythmic motor behavior was on the rise. This interest wasgreatly enhanced by the 19th century work of two ingeniousphotographers who had the same lifespan (1830–1904):Eadweard Muybridge, an Englishman who worked largely inthe USA, and Etienne-Jules Marey, a professor of physiology atthe Collège de France in Paris. They invented different types ofapparatus for measuring a variety of kinematic and kineticparameters (Tosi, 1992; Bouisset, 1992). Examples includeMuybridge's classical recording of the horse's trot (He was toexplore the question whether there could be a phase duringtrot in which all forelimbs were not in contact with theground) and Marey's high speed (96 frames/s) filming of dog

locomotion during the walk, trot and gallop (Fig. 1A). Mareygave such recordings to Maurice Philippson (1877–1938), aBelgian professor of physiology who analyzed the kinematicsof hind limb hip, knee and ankle displacements and providedthe still-used division of the phases of the step cycle as shownin Fig. 1B (Philippson, 1905).

Philippson learned to maintain chronic spinal dogs in thelaboratory of FriedrichGoltz (1834–1902), aGermanprofessor ofphysiology at the University of Strasbourg. When Philippson'sspinal animals were held off the ground, they often exhibitedrhythmic hind limb sequences ofmovements, whichhe filmedand analyzed. These and other measurements led Philippsonto conclude that spinal cord controlled locomotion using bothcentral and reflex mechanisms (Fig. 1B).

In a particularly influential and long (93-pg.) paper,Sherrington (1910) considered different types of reflex action(flexion, nociceptive flexion, crossed extension), which mayassist locomotion when elicited by different types of stimula-tion (e.g., cutaneous stimulation by activation of the perineum,back, tail, or pinna). For these reflexes, he delineated the in-volved hind limb muscles and the relative intensity of thereflex responses. He compared these reflexes in differentpreparations (intact, decerebrate, spinal animal) and demon-strated their similarities and the differences (for furtherdetails, see Clarac, 2005b). Sherrington's views on the neuralcontrol of rhythmic movements obtained when the animalwas off the ground differently somewhat from those ofPhilippson. For example, with the preparation's legs hangingdown freely, Sherrington argued thatmuscular proprioceptorswere excited and their input to the spinal cord could inducereflex stepping. By contrast, Philippson, in this experimentalsituation, emphasized the role of the spinal cord. Sherringtoninsisted more on the periphery where he detailed the role ofmuscular and cutaneous afferents. Centrally, he argued for analternating depression of flexor and extensor centers. In his1910 paper, he conceded a rhythm-generating role for thespinal cordwith the statement that “… The seat of the rhythmis obviously not peripheral. It is not in the muscles or theirmotornerve... nor can it lie in the receptiveorgansof the skinortheir afferent nerve trunks for direct stimulation of the crosssection of the spinal axis itself provokes the rhythmicity reply.The rhythm is therefore central in its seat” (p. 87). Puzzlingly,however, he waxed and waned on this latter possibilitythroughout his subsequent career, such that his more consis-tent conclusion seems tohavebeen that locomotionwasundera predominant peripheral control (for further details, seeStuart et al., 2001; Stuart, 2005).

Thomas Graham Brown (1882–1965), a Scottish-trainedclinician, worked independently on locomotion throughout1910–1913 in Sherrington's Liverpool laboratory. He alsocollaborated with Sherrington on the pilomotor system, anon-locomotor aspect of spinal reflexes, and the motor cortexof a non-human primate (Adrian, 1966). It would appearthat Graham Brown received informal mentoring from Sher-rington and there is evidence that they remained close friendsthroughout their lifetimes (see pp. 333–334 in Stuart et al.,2001). In later years, Graham Brown was the professor ofphysiology at the University of Cardiff (1920–1947) where heretained an office until 1961. His work on the neural control oflocomotion is described in 16 articles published between

Fig. 1 – The locomotor step cycle analysis of Philippson (1905). (A) Photographs of dog locomotion which were given toPhilippson by Marey. (B) Different phases of the step cycle, which Philippson defined as F (foot lift-off in the swing phase),E1 (foot descent in the swing phase), E2 (limb “give” in the stance phase) and E3 (limb thrust in the stance phase. Beneath thesephases, Philippson drew the timings of reflexes which he believed to assist and/or bring about the successive step cyclemovements.

17B R A I N R E S E A R C H R E V I E W S 5 7 ( 2 0 0 8 ) 1 3 – 2 1

1911 and 1922 (for review, see pp. 44, 48–49 and 91 in Wetzeland Stuart, 1976) and an ingenious film on the treadmilllocomotion of the decerebrate cat. He presented this at anunpublished 1941 meeting of the Physiological Society inLondon. This film had no captions and its availabilityremained relatively unknown until Lundberg and Phillips

(1973) wrote a short account about it. For the present purposes,it should be noted that Graham Brown's approach to the studyof locomotion differed substantially from that of Philippsonand Sherrington. For example, Fig. 2A shows an experiment inwhich he deafferented the two hind limbs of the decerebratecat, cut most of the dorsal and lumbar roots, and observed

Fig. 2 – Central activation of the spinal cord to produce rhythmic locomotor movements. (A) From Graham Brown (1911).(B) From Viala and Buser (1971). (C) From Grillner and Zangger (1979). For details, consult the original articles.

18 B R A I N R E S E A R C H R E V I E W S 5 7 ( 2 0 0 8 ) 1 3 – 2 1

spontaneous rhythmic bursts of activity in nerves to a flexor(tibialis anterior) and extensor (gastrocnemius) muscle (Gra-ham Brown, 1911).

Graham Brown confirmed his Fig. 2A finding in severalsubsequent reports, including one in which he proposed theconcept of “half centers” (Graham Brown, 1914): “The experi-ments seem to show that the fundamental unit of activity inthe nervous system is not that which we term the spinalreflex. They show the independence of the efferent neurone,and suggest that the functional unit is …the mutuallyconditioned activity of the linked antagonistic efferent neu-rons (“half centres”) which together form the “ centre” …”(p. 45). This idea received no support, however, and itlanguished in obscurity until it was revitalized by AndersLundberg, a Swedish professor of physiology at GöteborgUniversity, in the 1960s (for review, see Lundberg, 1969).Rather, for the first 60 years of the 20th century, a reflexcontrol of the stepping rhythm held sway (e.g., Gray, 1950)even though there was much evidence available by 1900

that the control of locomotion featured a substantial CNScomponent.

5. Identification of locomotor neural circuitry

In opposition to the reflex theory, evidence began to emergefrom the 1930s onwards that motor behavior could be induced“spontaneously.” Edgar Adrian (1889–1977), an English profes-sor of physiology at Cambridge University, reported that in theisolated insect ventral cord of Dytiscus, it was possible torecord spontaneous activity in selected nerve bundles thatwas phase related to respiratory movements (Adrian, 1931). C.Ladd Prosser (1907–2002), the “father” of American compara-tive physiology and a professor of physiology and neurosci-ence at the University of Illinois, argued in a review on thephysiology of invertebrate nervous systems that while loco-motion was “reflex in nature,” there was some spontaneousactivity when part of the nervous system was isolated

19B R A I N R E S E A R C H R E V I E W S 5 7 ( 2 0 0 8 ) 1 3 – 2 1

(Prosser, 1946). These two reports preceded the concept ofcentral pattern generators (CPGs) that emerged in the early1960s. Theodore Bullock (1915–2005), the father of Americaninvertebrate neurobiology, professor of zoology and neurosci-ence at the Los Angeles and later San Diego campuses of theUniversity of California, stated that “ … central patterning isthe necessary and often the sufficient condition for determin-ing the main characteristic features of almost all actions”(p. 56 in Bullock, 1961). Two of the best seminal examples ofthis viewpoint appeared in invertebrate research of the veryearly 1960s (Clarac and Pearlstein, 2007). First was work un-dertaken in the laboratory of a Dutch-born professor of biologyat the California Institute of Technology in Los Angeles, USA,Cornelius (“Cag”) Wiersma (1905–1979). He demonstrated aswimmeret beating at a near-normal frequency (1 Hz) in acompletely isolated crayfish abdominal nervous cord (Hughesand Wiersma, 1960). Shortly thereafter, in 1961, the AmericanDonaldWilson (1933–1970), while a postdoctoral trainee in theCopenhagen, Denmark laboratory of Torkel Weis-Fogh, dem-onstrated a slow rhythmic flight pattern at 10 Hz in adeafferented and surgically reduced locust preparation (vs.18 Hz in the intact locust) (Wilson, 1961). Following these twotruly seminal reports, there was an explosion of CPG researchundertaken on a wide variety of invertebrate preparations (forreview, Marder et al., 2005; Clarac and Pearlstein, 2007).

For mammals, the 1960s and early 1970s work of theMoscow, USSR workers Maurice Shik and Grigori Orlovskyand their colleagues, as we mentioned previously, providedcompelling indirect evidence for descending command signalsand sensory feedback operating on spinal CPGs to initiate andcontrol locomotion (for review, see Orlovsky et al., 1999). Forintact cats, Engberg and Lundberg (1969) proposed a compro-mise between the ideas of Sherrington and Graham-Brown.

They recorded legmovements and theEMGsof selectedhindlimb extensor muscles in freely walking cats and demonstratedthat the hind limb touches the ground at the onset of the stancephase, 5 to 10 ms after the activation of the extensor muscles.This confirmed that the beginning of that extensor phase wasdue to a central command. They proposed that the spinalcomponent of locomotor control featured a half-center “CPG”activation that was tuned by sensory input. In concomitantintracellular recording work on the acute spinal cat, Lundbergand his Polish collaborator, Elzbieta Jankowska, demonstratedthat after L-DOPA injection, the short latency reflex induced byhigh threshold muscle afferents was depressed, whereas verystrong response was evident at a much longer latency. Theysuggested that the latter indicated the first flexor responseoccurring at the onset of locomotion (Jankowska et al., 1967). Aclear rhythmic alternating activity was presented by Grillner(1969) (see his Fig. 1). The French workers, Denise Viala andPierre Buser, also demonstrated centrally controlled locomotorrhythmicity in both decerebrate and spinal rabbits (Viala andBuser, 1971; see Fig. 2B) and Perret and Cabelguen (1976)dissociated central and peripheral participation in decorticatecat locomotor pattern. Sten Grillner and his group reported onthe hind limb motor activity before and after a completedeafferentation in the decerebrate walking cats (Grillner andZangger, 1975). They showed that a complex varied motorpattern remainedafter deafferentation, thus demonstrating thepresence of a centrally generated output pattern. Finally, it

should be noted that in the 1970s and early 1980s, Sten Grillnerand his Canadian collaborator, Serge Rossignol, extended onLundberg's concepts of alternative spinal reflex pathways (seeStuart, 2002) and the above-mentioned DOPA-inducedmodula-tion of spinal circuitry. They demonstrated that phase-depen-dent reflex reversals occurred in both moving animalpreparations (e.g., decerebrate, spinal) and when such prepara-tions were paralyzed and primed to generate a locomotorrhythm (fictive locomotion; see Grillner, 1975; Rossignol, 1996).They thereby provided seminal examples of the reconfigurationof spinal neuronal circuitry on a very rapid time scale and theinterplay between CPG activity and sensory input for a smoothcontrol of locomotion.

6. Concluding thoughts

Since 1970 an extensive analysis of CNS circuitry has beenundertaken on largely invertebrates and non-mammalianvertebrates (Orlovsky et al., 1999). Thiswork has demonstratedthe complexity and the plasticity of locomotor networks in theCNS and their modulation by sensory input. Clearly, argu-ments about a predominantly central versus peripheral con-trol of locomotion are now obsolete!

Locomotor CNS network analysis has played a key role inadvancing general understanding of circuit modules and theirproperties, with applicability to even the cerebral cortical level(Yuste et al., 2005). Although microcircuits in the CNSpresumably have their own idiosyncratic features, it seemalso likely that they share some features of locomotor CPGs,thereby suggesting that “ …wewill most likely be able to iden-tify a much richer repertoire of the tools for precise circuitdesign in different part of the nervous system” (p. 532 inGrillner et al., 2005).

Acknowledgments

This work was aided by support from the CNRS. I would like tothank, in particular, Douglas Stuart, who provided somecomplementary historical information and edited the penul-timate version of themanuscript. Many thanks also to Abdel ElManira, who has made some helpful comments on an earlierversion. I would also like to thankNgaNguyen (Arizona HealthSciences Library, University of Arizona), who undertook somelibrary and Internet research.

R E F E R E N C E S

Adrian, E.D., 1931. Potential changes in the isolated nervoussystem of Dytiscus marginalis. J. Physiol. (London) 72, 132–151.

Adrian, E.D., 1966. Thomas Graham Brown, 1882–1965. Biogr. Mem.Fellows R. Soc. 12, 13–33.

Ashley-Ross, M.A., Gillis, G.B., 2002. A brief history of vertebratefunctional morphology. Integr. Comp. Biol. 42, 183–189.

Bastian, H.C., 1887–1888. The “muscular sense”, its nature andcortical localisation. Brain 10, 1–137.

Bell, C., 1830. The Nervous System of the Human Body. Longman,London.

20 B R A I N R E S E A R C H R E V I E W S 5 7 ( 2 0 0 8 ) 1 3 – 2 1

Bernstein, N.A., 1947. O Postroenii Dvizhenii [On the Constructionof Movements]. Medzig, Moscow.

Borelli, G.A., 1680. De Motu Animalium [On the Movement ofAnimals]. Superiorum Permissu, Rome.

Bouisset, S., 1992. Etienne-Jules Marey, or when motionbiomechanics emerged as a science. In: Cappozo, A., Marchetti,M., Tosi, V. (Eds.), Biolocomotion: A Century of Research UsingMoving Pictures. Promograph, Rome, pp. 71–88.

Brazier, M.A.B. (Ed.), 1969. The Interneurone. University ofCalifornia Press, Los Angeles.

Bullock, T.H., 1961. The origins of patterned nervous discharge.Behaviour 17, 48–59.

Clarac, F., 2005a. The History of Reflexes: Part I. From Descartes toPavlov. IBRO Neuro History, http://www.ibro.info/Pub_Main_Display.asp?Main_ID=421.

Clarac, F., 2005b. The History of Reflexes: Part 2. From Sherringtonto 2004. IBRO Neuro History, http://www.ibro.info/Pub_Main_Display.asp?Main_ID=439.

Clarac, F., Pearlstein, E., 2007. Invertebrate preparations and theircontribution to neurobiology in the second half of the 20thcentury. Brain Res. Rev. 54 (1), 113–161.

Cobb, M., 2002. Timeline: exorcizing the animal spirits: JanSwammerdam on nerve function. Nat. Rev., Neurosci. 3,395–400.

Descartes, R., 1664. Traité de l'homme [Treatise on Man]. HarvardUniversity Press, Cambridge, Mass. Paris. Angot. French textwith English translation in Thomas Steele Hall (1972).

Duchenne, G., 1867. Physiologie des mouvements démontrée àl'aide de l'expérimentation électrique et de l'observationclinique et applicable à l'étude des paralysies et desdéformations [Physiology of Movements Demonstrated withElectrical Experimentation and with Clinical Observation andApplied to the Study of Paralysis and Deformations]. Baillère,Paris.

Engberg, I., Lundberg, A., 1969. An electromyographic analysis ofmuscular activity in the hindlimb of the cat duringunrestrained locomotion. Acta Physiol. Scand. 75, 614–630.

Enoka, R.M., 1988. Neuromechanical Basis of Kinesiology. HumanKinetics, Champaign (IL, USA).

Finger, S., 1994. Origins of Neuroscience. Oxford University Press,Oxford.

Flourens, M.J.-P., 1824. Recherches Expérimentales sur lesPropriétés et les Fonctions du Système Nerveux dans lesAnimaux Vertébrés [Experimental Research on the Propertiesand Functions of the Nervous System in Vertebrate Animals].Paris, Crevot.

Fritsch, G., Hitzig, E., 1870. Über die elektrische Erregbarkeit desGrosshirns. Archiv für Anatomie und Physiologie, 300–332. InG. von Bonin, Some Papers on the Cerebral Cortex. Translatedas “On the electrical excitability of the cerebrum.” Springfield,IL: Charles C Thomas 1960, pp. 73–96.

Gandevia, S.C., 1996. Kinesthesia: roles for afferent signals andmotor commands. In: Rowell, L.B., Shepherd, J.T. (Eds.),Handbook of Physiology: Sec. 12. Exercise: Regulation andIntegration of Multiple Systems. Oxford University Press, NewYork.

Gasser, J., 1995. Aux Origines du Cerveau Moderne: Localisations,Langage et Mémoire dans L'œuvre de Charcot [On the Originsof the Modern Brain. Charcot's Work on Localizations,Language and Memory]. Fayard, Paris, p. 335.

Goetz, C.G., Bonduelle, M., Gelfand, T., 1995. Charcot: ConstructingNeurology. Oxford University Press, New York.

Graham Brown, T., 1911. The intrinsic factors in the act ofprogression in the mammal. Proc. R. Soc. Lond. 84B, 308–319.

Graham Brown, T., 1914. On the nature of the fundamental activityof the nervous centres: together with an analysis of theconditioning of rhythmic activity in progression, and a theoryof the evolution of function in the nervous system. J. Physiol.(London) 48, 18–46.

Gray, J., 1950. The role of peripheral sense organs duringlocomotion in vertebrates. Symp. Soc. Exp. Biol. 4, 112–126.

Grillner, S., 1969. The influence of DOPA on the static and thedynamic fusimotor activity to the triceps surae of the spinalcat. Acta Physiol. Scand. 77 (4), 490–509.

Grillner, S., 1975. Locomotion in vertebrates: central mechanismsand reflex interaction. Physiol. Rev. 55, 247–304.

Grillner, S., Wallén, P., 2004. Innate versus learned movements—afalse dichotomy? Prog. Brain Res. 143, 3–12.

Grillner, S., Zangger, P., 1975. How detailed is the central patterngeneration for locomotion? Brain Res. 88 (2), 367–371.

Grillner, S., Zangger, P., 1979. On the central generation in the lowspinal cat. Exp. Brain Res. 34, 241–261.

Grillner, S., Stein, P.S.G., Stuart, D.G., Forrsberg, H., Herman, R.M.(Eds.), 1986. Neurobiology of Vertebrate Locomotion.Macmillan Press, London.

Grillner, S., Markram, H., De Shutter, E., Silberg, G., LeBeau, F.E.N.,2005. Microcircuits in action—from CPGs to neocortex. TrendsNeurosci. 28, 525–533.

Herman, R.M., Grillner, S., Stein, P.S.G, Stuart, D.G., 1976. NeuralControl of Locomotion. Plenum, New York.

Hinsey, J.C., Ranson, S.W., McNattin, R.F., 1930. The role of thehypothalamus and mesencephalon in locomotion. Arch.Neurol. Psych. 23, 1–43.

Hughes, G.M., Wiersma, C.A.G., 1960. The coordination ofswimmeret movements in the crayfish, Procambarus clarkii(Girard). J. Exp. Biol. 37, 657–670.

Jankowska, E., Jukes, M.G.M., Lund, S., Lundberg, A., 1967. Theeffect of DOPA on the spinal cord: VI. Half-centre organizationof interneurones transmitting effects from flexor reflexafferents. Acta Physiol. Scand. 70, 389–402.

Laycock, T., Ed./Translator, 1851. The Principles of Physiology byJohn August Unzer, and A Dissertation on the Functions of theNervous System by George Procháska. London: SydenhamSociety.

Liddell, E.G.T., 1960. The Discovery of Reflexes. Clarendon Press,Oxford.

Legallois, J.-J.-C., 1812. Expériences sur le Principe de la Vie,Notamment sur Celui des Mouvements du Cœur, et sur le Siègede ce Principe [Experiments on the Vital Principle in particularon that of Heart Movements and the Site of this Principle].Hautel, Paris.

Lundberg, A., 1969. Reflex Control of Stepping. Nansen memoriallecture to Norwegian Academy of Sciences.Universitetsforlarget, Oslo.

Lundberg, A., Phillips, C.G., 1973. T Graham Brown's film onlocomotion in the decerebrate cat. J. Physiol. (London) 231,90–91.

Marder, E., Bucher, D., Schulz, D.J., Taylor, A.L, 2005. Invertebratecentral pattern generation moves along. Curr. Biol. 15,R685–R699.

Molnar, Z., 2004. ThomasWillis (1621–1675), the founder of clinicalneuroscience. Nat. Rev., Neurosci. 5, 329–333.

Orlovsky, G.N., Deliagina, T.G., Grillner, S., 1999. Neuronal Controlof Locomotion—FromMollusc to Man. Oxford University Press,Oxford.

Parent, A., 2005. Duchenne De Boulogne: a pioneer in neurologyand medical photography. Can. J. Neurol. Sci. 35, 369–377.

Pearce, J.M.S., 2005. Romberg and his sign. Eur. Neurol. 53,210–213.

Perret, C., Cabelguen, J.M., 1976. Central and reflex participation inthe timing of a bifunctionalmuscle, the semi-tendinosus in thecat. Brain Res. 106, 390–395.

Philippson, M., 1905. L'autonomie et la centralisation dans lesystème nerveux des animaux [Autonomy and centralizationin the animal nervous system]. Trav. Lab. Physiol. Inst. Solvay(Bruxelles) 7, 1–208.

Procháska, J., 1780–84. De Structura Nervorum [On NervousStructures]. Prague, Gerle. 3 volumes.

21B R A I N R E S E A R C H R E V I E W S 5 7 ( 2 0 0 8 ) 1 3 – 2 1

Prochazka, A., Clarac, F., Loeb, G.E., Rothwell, J.C., Wolpaw, J.R.,2000. What do reflex and voluntary mean. Modern views on anancient debate. Exp. Brain Res. 130, 417–432.

Prosser, C.L., 1946. The physiology of nervous systems ofinvertebrate animals. Physiol. Rev. 26, 337–382.

Romberg, M.H., 1846. Lehrbuch der Nervenkrankheiten desMenschen [Textbook of Nervous Diseases]. Duncker, Berlin.

Rossignol, S., 1996. Neural control of stereotypic limb movements.In: Rowell, L.B., Shepherd, J.T. (Eds.), Handbook of Physiology:Sec. 12. Exercise: Regulation and Integration of MultipleSystems. Oxford University Press, New York, pp. 173–216.

Sanders, L.J., 2002. Jean-Martin Charcot (1825–1893). The manbehind joint disease. J. Am. Podiatr. Med. Assoc. 92, 375–380.

Scheerer, E., 1985. Muscle Sense and Innervation Feelings: aChapter in the History of Perception and Action. UniversitätZentrum für interdisziplinäre Forschung, Bielefeld (GER).

Schiller, F., 1995. Staggering gait in medical history. Ann. Neurol.37, 127–135.

Schiller, F., 1999. Jirí Procháska: on the 250th anniversary of hisbirth. J. Hist. Neurosci. 8, 70–72.

Sherrington, C.S., 1906. On the proprioceptive system especially inits reflex aspect. Brain 29, 467–482.

Sherrington, C.S., 1910. Flexion-reflex of the limb, crossedextension-reflex, and reflex stepping and standing. J. Physiol.(London) 40, 28–121.

Shik, M.L., Severin, F.V., Orlovsky, G.N., 1966. Control of walkingand running by means of electrical stimulation of themid-brain. Biophysics 11, 756–765.

Sperry, R.W., 1950. Neural basis of the spontaneous optokineticresponse produced by visual inversion. J. Comp. Physiol.Psychol. 43, 482–489.

Stein, R.B., Pearson, K.G., Smith, R.S., Redford, J.B. (Eds.), 1973.Control of Posture and Locomotion. Plenum Press, New York.

Stein, P.S.G., Grillner, S., Selverston, A.I., Stuart, D.G. (Eds.), 1997.Neurons, Networks, and Motor Behavior. MIT Press, Boston.

Stuart, D.G., 2002. Reflections on spinal reflexes. Adv. Exp. Med.Biol. 508, 249–257.

Stuart,D.G., 2005. Integrationofpostureandmovement: contributionsof Sherrington, Hess and Bernstein. Hum. Mov. Sci. 24, 621–643.

Stuart, D.G., McDonagh, J.C., 1998. Reflections on a Bernsteinianapproach to systems neuroscience: the controlled locomotionof high-decerebrate cats. In: Latash, M.L. (Ed.), Progress inMotor Control—Bernstein's Traditions in Movement Studies.Human Kinetics, Champaign (IL, USA), pp. 21–49.

Stuart, D.G., Pierce, P.A., Callister, R.J., Brichta, A., McDonagh, J.C.,2001. Sir Charles Sherrington: humanist, mentor, andmovement neuroscientist. In: Latash, M.L., Zatsiorsky, V. (Eds.),Classical Papers in Movement Science. Human Kinetics,Champaign (IL, USA), p. 397–374.

Tosi, V., 1992. Marey and Muybridge; how modern biolocomotionanalysis started? In: Cappozo, A., Marchetti, M., Tosi, V. (Eds.),Biolocomotion: A Century of Research Using Moving Pictures.Promograph, Rome, pp. 51–71.

Viala, D., Buser, P., 1971. Modalités d'obtention de rythmeslocomoteurs chez le lapin spinal par traitementspharmacologiques (Dopa, 5-HTP, d'amphétamine) [Effects ofDOPA, 5-HTP and D-amphetamine on locomotor rhythms inthe spinal rabbit]. Brain Res. 35, 151–165.

von Holst, E., 1943. Uber relative Koordination bei Arthropoden.Pflügers Arch. Gesamte Physiol. 246, 846–865.

von Holst, E., Mittelstaedt, H., 1950. Das reaffenrenzprinzip:Wechselwirkungen zwischen Zentralnervensystem undPeripherie [The re-afference principle: mutual effects betweenthe central nervous system and the periphery].Naturwissenschaften 37, 464–476.

Wetzel, M.C., Stuart, D.G., 1976. Cat locomotion and its neuralcontrol. Prog. Neurobiol. 7, 1–98.

Wiersma, C.A.G. (Ed.), 1967. Invertebrate Nervous Systems: TheirSignificance for Mammalian Neurophysiology. University ofChicago Press, Chicago.

Willis, T., 1664. Cerebri Anatome: Cui Accessit NervorumDescriptio et Usus [The Anatomy of the Brain and Nerves].Flesher, London.

Wilson, D.M., 1961. The central nervous control of flight in a locust.J. Exp. Biol. 38, 471–490.

Yuste, R., MacLean, J.N., Smith, J., Lansner, A., 2005. The cortexas a central pattern generator. Nat. Rev., Neurosci. 6,477–483.