Circuits for Skilled Reaching and Grasping

NE35CH26-Alstermark ARI 2 April 2012 17:46

RE V I E W

S

IN

AD V A

NC

E

Circuits for Skilled Reachingand GraspingBror Alstermark,1 and Tadashi Isa2

1Department of Integrative Medical Biology, Section of Physiology, Umea University,S-901 87 Umea, Sweden; email: [email protected] of Developmental Physiology, National Institute for Physiological Sciences,Myodaiji, Okazaki 444-8585, Japan; email: [email protected]

Annu. Rev. Neurosci. 2012. 35:559–78

The Annual Review of Neuroscience is online atneuro.annualreviews.org

This article’s doi:10.1146/annurev-neuro-062111-150527

Copyright c© 2012 by Annual Reviews.All rights reserved

0147-006X/12/0721-0559$20.00

Keywords

corticomotoneuronal pathways, motoneuron, propriospinal neuron,segmental interneuron

Abstract

From an evolutionary perspective, it is clear that basic motor func-tions such as locomotion and posture are largely controlled by neuralcircuitries residing in the spinal cord and brain-stem. The control ofvoluntary movements such as skillful reaching and grasping is gener-ally considered to be governed by neural circuitries in the motor cor-tex that connect directly to motoneurons via the corticomotoneuronal(CM) pathway. The CM pathway may act together with several brain-stem systems that also act directly with motoneurons. This simple viewwas challenged by work in the cat, which lacks the direct CM system,showing that the motor commands for reaching and grasping couldbe mediated via spinal interneurons with input from the motor-cortexand brain-stem systems. It was further demonstrated that the spinalinterneurons mediating the descending commands for reaching andgrasping constitute separate and distinct populations from those in-volved in locomotion and posture. The aim of this review is to describepopulations of spinal interneurons that are involved in the control ofskilled reaching and grasping in the cat, monkey, and human.

559

Review in Advance first posted online on April 9, 2012. (Changes may still occur before final publication online and in print.)

Changes may still occur before final publication online and in print

Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


Contents

INTRODUCTION . . . . . . . . . . . . . . . . . . 560REACHING AND GRASPING IN

THE CAT AND MACAQUEMONKEY . . . . . . . . . . . . . . . . . . . . . . . . 565

SUBCORTICAL UPDATING OFCORTICAL COMMAND TOTHE C3-C4 PROPRIOSPINALNEURONS . . . . . . . . . . . . . . . . . . . . . . . 569

FEEDBACK INHIBITORYCONTROL OF REACHING VIAC3-C4 PROPRIOSPINALNEURONS . . . . . . . . . . . . . . . . . . . . . . . 570

PROPRIOSPINAL PATHWAYSIN HUMANS . . . . . . . . . . . . . . . . . . . . . 571

NEW TECHNIQUES FORINVESTIGATING THEFUNCTION OF NEURALCIRCUITS . . . . . . . . . . . . . . . . . . . . . . . 571

FUNCTIONAL RECOVERYFOLLOWING SPINALCORD INJURY . . . . . . . . . . . . . . . . . . . 572

INTRODUCTION

Neural circuits for skilled reaching and grasp-ing are considered to be mainly located inthe motor cortex and exert their effects inmotoneurons (MNs) via the “direct” cortio-comotoneuronal (CM) pathway (Figure 1a,b).This 60-year-old idea has been firmly estab-lished because the direct CM pathway is lackingin lower species and becomes more prominentin primates together with a parallel increasein manual dexterity (Bernhard & Bohm 1954).However, most of the descending controlfrom the motor cortex to MNs is mediatedvia “indirect” CM pathways with intercalatedneurons in the brain stem and spinal cord. Thegeneral understanding of interneuronal (IN)circuits in the brain stem and spinal cord is thatthey mainly control automatic movements suchas locomotion and scratching in addition toposture, reflexes, and tonus (Figure 1a)and are of less importance for skilledmovements. In classical behavioral studies in

the macaque monkey using pyramidal lesionsat the brain-stem level (Lawrence & Kuypers1968a, 1968b), it was concluded that highlyfractionated finger movements in primatesprimarily depend on the direct CM pathway.From these studies, researchers proposed thata major advantage of the direct CM systemwas its ability to bypass the INs, which wereconsidered to have too widespread connectionswith MNs innervating distal hand muscles tobe able to mediate the command for indepen-dent digit movements (Kuypers 1982). As aresult, researchers recently further suggestedthat the direct CM pathway has replacedone of the indirect CM pathways (Lemon2008). However, in this review, we describespecialized neural circuits of spinal pre-MNcenters controlling reaching and grasping.The focus is on the functional organization ofthese circuits, with less attention to the detailsregarding synaptic effects. For the latter, seereviews by Alstermark & Lundberg (1992),Alstermark et al. (2007), and Isa et al. (2007).

Sherrington (1906) together with hisstudents laid the foundation for circuit analysisin the spinal cord with classical work on theorganization of the stretch reflex, reciprocalinhibition, scratch reflex, flexor and crossedextensor reflexes (Creed et al. 1932), and loco-motion (Brown 1911) as shown in Figure 1a.Although such circuits included the MNs asthe final common path (Figure 1b), Sherring-ton also included INs in the common path(Figure 1c) in the sense that it represented the“highest degree of communism” (Sherrington1906). The great advantage of INs is that amajor integration of sensory and descendingsignals can be made at a stage just priorto the MNs (Lundberg 1999). Descendingsupraspinal systems may facilitate or inhibit thesignals from the primary afferents to the MNs.By using either excitatory or inhibitory INs, thesign of an effect can be reversed from excitationto inhibition or the reverse (disinhibition).We refer to INs with direct monosynapticprojection to MNs as last-order INs.

Modern investigations of the neuronalorganization of descending pathways began

560 Alstermark · Isa


Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


a b

ReachingGrasping

New functionsdiscovered:ReachingGrasping

Old functionsdemonstrated:LocomotionScratchingReflexesPostureTonus

Motor cortexHigher motor centers

Primaryafferent

MuscleMusclespindlespindleMusclespindle

Muscle

Spinal cord

Final common path

Motoneuron

cHigher motor centers

Common path

Motoneuron

Interneuron

JointMuscleSkin

Figure 1Motor principles. (a) Skilled voluntary movements such as reaching and grasping are commonly considered to be dependent on neuralcircuits mainly located in the motor cortex, whereas stereotyped motor behavior such as locomotion, scratching, reflexes, posture, andtonus are controlled by spinal circuits. However, as described in this review, the descending command for skilled reaching and graspingcan also be controlled by spinal circuits. (b) Schematic neural diagram showing monosynaptic input from muscle spindle Ia afferents anddescending pathways from higher motor centers in the cortex and brain stem. Sherrington (1906) considered motoneurons to be the“final common path.” (c) Schematic neural diagram showing disynaptic input, via an interneuron, from afferents and higher motorcenters to motoneurons. Sherrington (1906) also included the interneurons together with the motoneurons in the “common path.”

with Lloyd’s (1941) study of pyramidal effectsin the cat spinal cord. He used the techniqueof conditioning monosynaptic test reflexes(Renshaw 1940) and showed that temporalfacilitation was required to evoke pyramidalexcitation in MNs, suggesting that interca-lated neurons also mediated the effect fordescending systems (illustrated in Figure 2a).A similar approach was used by Lundberg &Voorhoeve (1962), who recorded intracellu-larly from MNs and found facilitation fromthe sensorimotor cortex of excitatory andinhibitory segmental reflexes. The final proofof a disynaptic excitatory CM pathway wasprovided in experiments by Lundberg and col-leagues recording from forelimb MNs in the cat(Illert et al. 1977). They showed that disynapticexcitation could be mediated by a group of

INs located in the upper cervical spinal cordsegments C3-C4 and termed them C3-C4propriospinal neurons (PNs) as shown inFigure 2b. The C3-C4 PNs are characterizedby an extensive convergence from severaldescending pathways: the cortico-, rubro-,reticulo-, and tectospinal tracts. These de-scending systems may also give feed-forwardinhibition of subpopulations of PNs via feedfor-ward inhibitory INs (Figure 2b). Whereas thedominating descending input is excitatory, theinput from forelimb afferents is predominantlyinhibitory via feedback inhibitory INs. In ad-dition to the MN projection, the C3-C4 PNshave an ascending projection to the lateral retic-ular nucleus (LRN). The LRN mediates themajor mossy-fiber input to the cerebellumfrom the spinal cord. The ascending LRN

www.annualreviews.org • Circuits for Reaching and Grasping 561


Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


projection may thus provide the cerebellumwith an efferent copy of the command to theMNs (see below; Figure 6). On the basis ofthis information, the cerebellum may then in-teract quickly with the cortical and brain-stemsystems to regulate activity in the C3-C4 PNs(see below). Some inhibitory C3-C4 PNs canalso mediate disynaptic inhibition in MNs(Figure 2b). These inhibitory PNs have thesame extensive descending convergence as theexcitatory ones, and they also have an ascend-ing projection to the LRN (see references inAlstermark et al. 2007).

Disynaptic CM excitation can also be me-diated via INs located in the same segments asthe MNs (Illert & Wiedeman 1984, Alstermark& Sasaki 1985, Kitazawa et al. 1993). Such INsare termed segmental interneurons (sINs) andare shown in Figure 3. To visualize the relativecontributions of PNs and sINs to reaching andgrasping, two groups of cats were comparedusing activity-dependent transneuronal up-take of wheat-germ agglutinated horseradishperoxidase (WGA-HRP) into last-order INsfollowing injection into the motor axons ofthe deltoid muscle (Alstermark & Kummel

Stimulus 2

Stimulus 2 Stimuli 2 and 3

Stimulus 1

Stimulus 3

Stimulus 1 or 3

Record

Monosynaptic EPSP

Muscle

Motoneuron

1×

1×

1×32×

2×

3×2 + 3

2

Electrophysiological technique for analyzing circuitry

The C3–C4 propriospinal system

LRN LRN LRN

PN PN

MN MN MN

IN IN

a

b

Supraspinalneuron

Spinalinterneuron

Group Iamuscle afferents

Temporal facilitation ofdisynaptic EPSP

Spatial facilitation ofdisynaptic EPSP

CS

Forelimbafferents

MuscleMusclespindlespindleMusclespindle

Te Re Ru CS



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


1990). Investigators pursued this study becausethe deltoid MNs are active in the reaching,grasping, as well as locomoting groups, butthey may also be differentially controlled. Onegroup performed unrestrained locomotion,and the other combined target reaching andgrasping to remove a morsel of food froma small tube with the forelimb. As shown inFigure 3, reaching consists of rapid elbow flex-ion followed by combined shoulder and elbowextension. During this lifting and protractionphase of reaching, the forepaw is dorsi-flexedand the toes adducted just prior to the insertioninto the tube. Inside the tube, the digits areabducted and flexed to grasp the morsel offood. After grasping, the forepaw is supinatedduring retraction out from the tube and thenbrought to the mouth (Alstermark et al. 1981b).

Figure 3 also shows the transverse andlongitudinal distributions of last-order labeledspinal INs. In the locomoting group, only afew labeled cells were found in the C2-C4 seg-ments, whereas many labeled cells were locatedin C5-Th1 with two peaks in the C6 and C7segments. In contrast, labeled cells were foundfrom C2 to Th1 in the cats performing reach-ing and grasping with a peak in the C6 segment.These results suggest that spinal INs devoted tothe control of reaching and grasping exist all the

way from C2 to Th1, whereas INs involved inthe generation of locomotion are mainly con-fined to the C5-Th1 segments. Why the spinalINs involved in the control of reaching andgrasping are also distributed rostral to the fore-limb segments is still not known. The sINs arenot as well characterized with respect to the de-scending systems; also unclear is whether theyproject to the LRN.

Although transneuronal labeling of last-order INs with WGA-HRP revealed a greaternumber of sINs than C3-C4 PNs (Alstermark& Kummel 1990), the synaptic strength issimilar for the CM pathway via the C3-C4PNs (Alstermark & Sasaki 1985). Apparently,the sINs constitute heterogenous groups,with only a part receiving strong corticospinalinput. Disynaptic CM pathways via C3-C4PNs have also been shown in the macaquemonkey (Alstermark et al. 1999) and in humans(Malmgren & Pierrot-Deseilligny 1988), butthey are absent in the rat (Alstermark et al.2004) and mouse (Alstermark & Ogawa 2004).As illustrated in Figure 3, corticoreticu-lospinal (Peterson et al. 1979, Alstermark &Sasaki 1986, Riddle et al. 2009) and corti-corubrospinal (Fujito & Aoki 1995) pathwayscan also mediate disynaptic CM excitation;these pathways constitute phylogenetically old

←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−Figure 2Electrophysiological techniques for analyzing circuitry. The excitability of motoneurons (MNs) can betested by using the monosynaptic Ia reflex pathway as described by Renshaw (1940) and Lloyd (1941).Although this is usually technically easier than stimulating last-order interneurons (INs) directly, researcherswere able to stimulate last-order INs in case of the C3-C4 propriospinal neurons (PNs) because of theirascending projection to the lateral reticular nucleus (LRN) (Alstermark et al. 1981a) (not shown in figure).However, last-order INs can also be activated synaptically, as in the case of the C3-C4 PN, by stimulatingone of their descending inputs, e.g., the corticospinal tract (indicated here as a supraspinal neuron). Such adisynaptic pathway usually requires temporal summation of excitatory postsynaptic potentials (EPSPs) in theintercalated neurons (in this case, the C3-C4 PNs) to make them fire. When firing, a disynaptic EPSP can berecorded in the MNs. To investigate the convergence onto common intercalated spinal INs, two differentinputs to the last-order INs can be stimulated together, producing spatial facilitation indicated by the largeEPSP in the MNs. Within the C3-C4 propriospinal system, PNs receive monosynaptic excitation from thecortico-, rubro-, reticulo-, and tectospinal tracts. Via activation from these descending systems, the PNs mayevoke disynaptic excitation in forelimb MNs. In addition, the PNs have ascending collaterals to the LRN.All the descending systems may evoke disynaptic inhibition in the PNs, denoted as feed-forward inhibition.There is only a weak and sparse monosynaptic excitation from the forelimb afferents. The major effect fromthe forelimb afferents is disynaptic inhibition, denoted as feedback inhibition. In addition to the excitatoryPNs, a population of inhibitory PNs has a descending convergence similar to that of excitatory PNs.Inhibitory PNs, as well as the inhibitory INs mediating feedback inhibition, also project to the LRN.



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


CST

ReST

RuST

MCx

C2

C3– C4

C5C2

100

200

0

N

C4C3 C5 C6 C7 C8 Th1

Locomotion

Reaching + grasping

LocomotionReaching +grasping

Differential activation of last-order interneurons during reaching and grasping versus locomotion

PN

C6 –Th1

MN

sIN

LocomotionReaching +grasping

g p g oco ot o

Figure 3Longitudinal and transverse location of last-order spinal interneurons (INs) active during combined reaching and grasping orlocomotion. (Center diagonal image) The organization of descending pathways controlling propriospinal and segmental interneurons(PNs and sINs, respectively); (upper right corner) the reaching and grasping movement sequences (from Alstermark et al. 1981b) and thelongitudinal distribution of labeled INs in two groups of cats performing either combined reaching and grasping (light blue line) orunrestrained overground locomotion (dark yellow line); (lower left corner) the transverse distribution of labeled INs for the two groups ofcats. The detailed images of the paw are courtesy of L.G. Pettersson. Wheat-germ agglutinated horseradish peroxidase was injectedinto the motor nerve of the deltoid muscle and was then taken up transneuronally from the motoneurons (MN) into last-order INs inan activity-dependent manner. The direct cortiocomotoneuronal (CM) pathway and monosynaptic MN connections from therubrospinal (RuST) and reticulospinal (ReST) tracts are not shown. Other abbreviation: CST, corticospinal tract. Adapted withpermission from Alstermark & Kummel (1990).



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


systems (Shapovalov 1975). In primates, thereis also a direct CM pathway (Bernhard & Bohm1954, Landgren et al. 1962). It appears thatan evolution of the CM pathways has takenplace via indirect corticobulbospinal and cor-ticopropriospinal pathways as well as a directCM pathway. We describe evidence showinghow these direct and indirect CM pathwaysare used to control reaching and grasping.

REACHING AND GRASPINGIN THE CAT ANDMACAQUE MONKEY

The above experiments using WGA-HRPsuggest that spinal INs may be active duringboth reaching and grasping. To investigate therelative contribution of C3-C4 PNs and sINsto reaching and grasping, the corticospinal andrubrospinal tracts were lesioned in C5 in the cat(Alstermark et al. 1981a, Alstermark & Sasaki1985) and the macaque monkey (Alstermarket al. 1999, Sasaki et al. 2004, Alstermark et al.2011) (Figure 4). This lesion eliminates thecortico- and rubrospinal inputs to the sINs butspares them to the C3-C4 PNs. The lesionswere confirmed electrophysiologically. In thecat, intracellular recording from MNs showedthat the disynaptic pyramidal excitatory post-synaptic potential (EPSP) could still be evokedafter the C5 lesion, although the amplitudewas decreased owing to the interruption of theinput to the sINs. In the cat, fast and accuratereaching of the forelimb could be observedduring the first postoperative week (Figure 4).However, after the forepaw was inserted intothe tube, the cats could not use their digits tograsp the morsel of food and instead made araking movement. In the monkey, the lesioninterrupted the inputs to the sINs as well asthe direct CM pathways. In this species, aftera complete lesion to the corticospinal tractthat spared the majority of the propriospinalaxons, both reaching to the tube and the pre-cision grip formed by flexion and oppositionof the index finger and thumb, which wereobserved preoperatively, could be performedduring the first postoperative week (see

Supplemental Movie 1. Follow the Supple-mental Material link from the Annual Reviewshome page at http://www.annualreviews.org). Transient deficits during the first twoweeks were seen in the preshaping of the digitsbefore insertion and coflexion of the otherdigits. These findings suggest that, in the cat,C3-C4 PNs can mediate the command forreaching, whereas the command for graspingrequires sINs. In contrast, in the macaquemonkey, the C3-C4 PNs seem capable ofmediating the commands for reaching andgrasping. The early deficits in digit controlindicate that the sINs and the direct CMpathway play a critical role in preshaping andalso in fractionated digit movements.

To test whether the reaching that remainedin the cat and the reaching and grasping in themacaque monkey could be ascribed to trans-mission via the C3-C4 PNs or via the ven-tral corticoreticulospinal system, a dorsal lesionwas made in C2 as shown in Figure 5. Thislesion interrupts the cortico- and rubrospinaltracts to the C3-C4 PNs as well as to the sINs,but it spares the corticoreticular input. This le-sion abolished the disynaptic pyramidal EPSPin forelimb MNs in both the cat and mon-key; it also abolished the monosynaptic pyra-midal EPSP in the monkey. Both reaching andgrasping were severely impaired in the cat andmonkey. In both species, reaching toward thetube could still be initiated, but precision wasimpaired in aiming toward the tube openingand in the coordination between joints. Al-though reaching recovered after approximatelyone month, fine-digit grasping was still abol-ished (see Supplemental Movie 2).

Taken together, the results following thedorsal C5 and C2 lesions show that C3-C4PNs can mediate the command for reachingin the cat and for reaching and grasping in themacaque monkey. A lesion of the propriospinalaxons made by transecting the ventral partof the lateral funiculus in C5 in the cat(Alstermark et al. 1981b, Pettersson et al. 1997)resulted in ataxia during reaching near thetarget (see section below) (Figure 6d ), eventhough grasping was still intact. On the basis



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.

http://www.annualreviews.org

http://www.annualreviews.org


StimulateCST

ReST

RuST

MCx

C2

C3– C4

C5

Lesion

PN

C6 –Th1

MN

Effects on reaching and grasping after lesion of CST and RuST in C5

Post-lesion

Pre-lesion

Post-lesion

Monkey

Cat

Intracellularrecordingfrom MN

sIN

Figure 4Behavioral and electrophysiological testing of propriospinal versus segmental interneurons (PNs and sINs, respectively) in the controlof reaching and grasping after lesion of the cortico- and rubrospinal tracts in the C5 segment. The transection of these pathways (centerdiagonal image; red ) eliminated the input to the sINs but spared the input to the C3-C4 PNs. In the cat (top panels), reaching wasnormal, whereas grasping with the digits was abolished. In the monkey (middle panels), reaching was normal and precision grip could beperformed with some deficits in preshaping and conjoint extension in the other digits: (middle upper row) preoperative condition and(middle lower row) postoperative day 1. Abbreviations: CST, corticospinal tract; MCx, motor cortex; MN, motoneuron;ReST, reticulospinal tract; RuST, rubrospinal tract. Adapted from Alstermark et al. (1981b) and Sasaki et al. (2004) with permission.



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


StimulateCST

ReST

RuST

MCx

C2

C3– C4

C5

PN

C6 –Th1

MN

Effects on reaching and grasping after lesion of CST and RuST in C2

Monkey

Cat


Post-lesion

Pre-lesion

Post-lesion

Lesion

sIN

Figure 5Behavioral and electrophysiological testing of propriospinal and segmental interneurons (PNs and sINs, respectively) versuscorticoreticulospinal pathway in the control of reaching and grasping after lesion of the cortico- and rubrospinal tracts in C2. Thetransection of these pathways (bottom diagonal image; red ) eliminated the input both to the C3-C4 PNs and sINs but spared the input tothe corticoreticulospinal pathway. In the cat (top panels), reaching and grasping were defective. In the monkey (center panels), reachingwas impaired and precision grip could not be performed: (center upper row) preoperative condition and (center lower row) postoperativeday 1. Abbreviations: CST, corticospinal tract; MCx, motor cortex; MN, motoneuron; ReST, reticulospinal tract; RuST, rubrospinaltract. Adapted from Alstermark et al. (1981b, 2011) with permission.



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


a b

c d

ReST

RuST CST

MCx

C2

C3– C4

C5

Cortical visualpathways

Subcortical visualpathways

Movingtarget

Cerebellum

Spinal cord

Motor cortex

Bulbospinalsystems

C6 –Th1

MN

C6–Th1

MN

C3–C4

Efferent copy

Lateral reticular nucleus

Deep CN

Granule cell Purkinje cell

PN

Bulbospinalsystems

Cerebellum

Motor cortex

PN

Postlesion

sIN

Lesion

Cat


Figure 6Subcortical visual updating via C3-C4 propriospinal neurons. (a) New visual information about changing target position can bemediated via both cortical and subcortical pathways: (blue boxes) reaching to the left target and switching from the left to the righttarget. (b) Schematic circuit consisting of the C3-C4 propriospinal system, efferent loop via the cerebellar cortex, and bulbospinalsystems and the motor cortex. This circuit may be used to update the cortical command for reaching. (c) Ventral lesion in C5 of thereticulospinal and propriospinal pathways sparing the cortico- and rubrospinal pathways to segmental interneurons. Intracellularrecordings from motorneurons (MNs) revealed reduced disynaptic corticospinal tract (CST) excitation. (d ) Reaching behavior after theventral C5 lesion, showing ataxia near the tube opening. Other abbreviations: CST, corticospinal tract; MCx, motor cortex;ReST, reticulospinal tract; RuST, rubrospinal tract.



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


of this observation, researchers proposed thatsINs can mediate the command for graspingin the cat (Alstermark et al. 1981b). However,these results do not exclude the possibility thatsINs can also help control reaching, becausereaching, although defective, could be executedeven when the axons of the PNs had beentransected. Thus, the dual control of reachingand grasping by the PNs in the monkey mayhave developed because precision grip requiresfine control of the upper arm (Alstermark et al.2011).

In addition to its important role in control-ling skilled reaching and grasping via the directCM pathway in primates, the spinal cord con-tains separate neural circuits that are capableof conveying the descending motor commandsfrom higher motor centers to the MNs. Thesespinal centers are presumably complementaryto the higher centers in the motor cortex andbrain stem. If so, what advantage could such in-tegration at a stage just prior to the “final com-mon path” provide? Illert et al. (1977) proposedthat the descending cortical command could beupdated on its way to the MNs via subcorti-cal systems that have convergent input to theC3-C4 PNs and sINs.

SUBCORTICAL UPDATING OFCORTICAL COMMAND TO THEC3-C4 PROPRIOSPINALNEURONS

During reaching, if the target suddenly changesits position in space, the movement trajectoryhas to be corrected using visual information.There are visual pathways to both the cortexand subcortical systems including the cere-bellum and bulbospinal systems (Figure 6a).Cortical visual pathways play an importantrole in the control of reaching (Goodale &Westwood 2004), usually with latencies longerthan 200 msec for correction of an ongoingmovement. However, shorter latencies in therange of 120-160 msec have been proposed tobe subcortically mediated in humans (Day &Lyon 2000) as well as in subjects with completecorpus callosum agenesis (Day & Brown 2001)

or whose primary visual area was temporarilyblocked (Christensen et al. 2008). Recently, us-ing electromyography recordings, researchersobserved latencies around 100 msec in humans(Pruszynki et al. 2010).

To test the hypothesis of subcortical updat-ing via the C3-C4 propriospinal system, catswere trained to perform reaching to one of twoor three tubes. The correct tube was illumi-nated, and during reaching, the illuminationcould be switched to another tube with variabledelay. The cats (Figure 6a) showed remarkablyshort latencies for the correction of the move-ment trajectory, down to 50-60 msec after thelight shift (Alstermark et al. 1990, Petterssonet al. 1997, Pettersson & Perfiliev 2002). Laten-cies longer than 100 msec were also observed.If transmission in the bulbospinal systems wasinterrupted by lesions in C2 of the rubro- andreticulospinal tracts, the cats could still correctthe trajectory but only with latencies longerthan 100 msec (Pettersson et al. 1997). In thosecases, the new visual information was mostlikely processed at a cortical level.

These results are compatible with theneural circuitry known to control the C3-C4propriospinal system (Figure 6b), becausebulbospinal systems receive very fast visualinformation (Werner 1993). However, a majorproblem is how to correct the ongoing corticaldescending command via the C3-C4 PNs.As described above, the PNs, in addition tothe MN projection, have ascending collateralsto the LRN in the lower brain stem (Illert& Lundberg 1978, Alstermark et al. 1981a,Alstermark & Ekerot 1992, Isa et al. 2006). TheLRN is a major mossy-fiber input from thespinal cord to the cerebellum (Clendenin et al.1974, Ekerot 1990). The cortical input to theLRN is separate from the corticospinal tract(Alstermark & Lundberg 1982, Alstermark& Ekerot 1992, Matsuyama & Drew 1997),which may control transmission via specificascending spinocerebellar pathways such asthe C3-C4 propriospinal system. The rubro-and reticulospinal neurons with projection tothe C3-C4 PNs receive input from the motorcortex and the deep cerebellar nuclei (Illert



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


Inhibitory feedback control of reaching

0 ms 75 125 150

250 275 350 400

Rest

IntracellularExtracellular

Rest

Stretch

Stretch

Skin afferent

C5 DC

C3–C4PN IN

Stretch

MNC6–Th1

a b Postlesion

PrelesionPrelesion LesionLesionLesion

Figure 7Feedback inhibition of the C3-C4 propriospinal neurons (PNs) during reaching. (a) Recording from PNs and feedback inhibitoryinterneurons (INs) in the C3-C4 segments during skin stretching revealed strong inhibition in the PNs. (b) Lesion of the dorsal columnin C5 eliminates the feedback inhibition and results in the hypermetria in reaching. Abbreviation: MN, motorneuron.

et al. 1977). The ascending projection from theC3-C4 PNs may provide the cerebellum withan efferent copy of the descending commandfor reaching to the forelimb MNs (Alstermarket al. 1981b, Alstermark 1983). Using the newvisual information, the cerebellum may thencalculate a correction of the ongoing movementtrajectory and issue new additional inputs to theC3-C4 PNs via the bulbospinal systems. Thecorrective signal will be matched by the newefferent copy until the activity level of the PNshas changed to correct for the new trajectory.

This hypothesis has received indirectsupport from behavioral testing of reachingfollowing transection of the propriospinalaxons in C5 (Figure 6c). After this lesion, thecats had no problem initiating reaching towardthe tube opening, but near the tube opening,they displayed severe ataxia during decelera-tion (Figure 6d ) (Alstermark et al. 1981b). Inthis case, transmission via the C3-C4 PNs tothe MNs was interrupted, but because of theremaining ascending projection to the LRN,the cerebellum may be misinformed, resultingin the ataxia (Alstermark 1983).

FEEDBACK INHIBITORYCONTROL OF REACHING VIAC3-C4 PROPRIOSPINALNEURONS

In addition to reaching, aiming and final posi-tioning of the paw are critical for a successfulgrasp. Such control may depend heavily onafferent information. The C3-C4 PNs receiveexcitation from forelimb muscle and skin af-ferents, but the dominating effect is inhibition(Alstermark et al. 1984a, 1984b, 1986b). Asshown in Figure 7a, stretching the skin of theforelimb evokes firing of feedback inhibitoryINs and concomitant inhibition in the PNs inthe C3-C4 segments. To investigate the func-tion of inhibitory feedback control of the C3-C4 PNs, researchers compared cats performingtarget reaching after dorsal column transectionin C5 or C2. The C5 lesion interrupted thefeedback inhibition to the PNs, although itwas spared after the C2 lesion. Both lesionsabolished the ascending information to highercenters. Interestingly, severe hypermetria inreaching was observed after the C5, but notafter the C2, dorsal column lesion; the C2



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


lesion resulted only in dysmetria near the tubeopening (Alstermark et al. 1986a) (Figure 7b).The hypermetria was not compensated untilafter approximately one week of testing, pre-sumably with the help of vision. These resultssuggest that feedback inhibition of PNs may beused conjointly with the descending commandcontrolling reaching. Deceleration and correcttermination of the forepaw may be difficultto control with only vision and may requiremultimodal control (Alstermark et al. 1986a).The INs mediating the inhibition of the PNsreceive cortical input. Thus, the motor andsensory cortices could activate them as well asthe PNs during reaching. The INs may thenbecome activated by spatial facilitation fromforelimb afferents when reaching close to thetarget. Similar to the PNs, the medial INsproviding feedback inhibition of the PNs haveascending collaterals to the LRN and to thecuneate nucleus (Alstermark et al. 1984b). Inanalogy with the discussion above on the role ofthe efferent copy from the PNs to the LRN, theC5 dorsal column lesion may cause erroneoussignaling from the feedback INs that could alsocontribute to the hypermetria. The dorsal col-umn transections in both C5 and C2 eliminateafferent information to higher centers, whichmay be the cause of the dysmetria mainlyobserved after the C2 lesion. The additionalhypermetria after the C5 dorsal column lesionsuggest that the inhibitory INs are essential forinhibition of the C3-C4 PNs during reachingin its terminal phase close to the target.

PROPRIOSPINAL PATHWAYSIN HUMANS

Using indirect electrophysiological techniques,researchers found propriospinal pathways inthe cervical and lumbar spinal cord segments ofhumans (for a review, see Pierrot-Deseilligny &Burke 2005). The cervical propriospinal path-way shares several properties with those in thecat and monkey, e.g., descending inhibition andstrong feedback inhibition evoked by skin affer-ents. A difficulty in the study of humans whenusing indirect techniques is to determine the

segments in which the PNs are located. How-ever, in one patient with a spinal cord injury atthe C6-C7 segmental junction restricted to theventral and lateral funicles on one side, the pro-priospinally mediated excitation was abolishedon the side of the lesion but remained intacton the nonlesioned side (Marchand-Pauvertet al. 2001). This finding, together with thesystematically longer latency of propriospinallymediated excitation to more caudally locatedmotor nuclei, suggests that the cervical PNs inhumans are located rostral to C6. In the patientwith a restricted lesion of the propriospinalpathway on one side, reaching and precisiongrip were normal on the nonlesioned side. Onthe lesioned side, researchers observed deficitsin reaching close to the target with transientataxia as well as difficulties in holding smallobjects with the thumb and index finger (B.Alstermark & L.G. Pettersson, unpublished ob-servations). Surprisingly, transcranial magneticstimulation revealed a similar short-latency-evoked cortical electromyography responseon both the lesioned and the nonlesionedsides, indicating that transmission in the directCM pathway was not significantly impaired(Marchand-Pauvert et al. 2001). These findingsagree well with the behavioral observationsdescribed above for the monkey. Transmissionvia sINs in humans was not accessible withina similar investigation, but it likely plays anessential role as does the direct CM pathway.

NEW TECHNIQUES FORINVESTIGATING THEFUNCTION OF NEURALCIRCUITS

As described in Reaching and Grasping in theCat and Macaque Monkey (section above),most of the behavioral work was performedusing selective spinal cord lesions (but seeabove for a description of the technique usingactivity-dependent transneuronal labeling)(Alstermark & Kummel 1990). The lesioningtechnique has several limitations such as caus-ing changes in excitability in both the acuteand chronic states and with compensatory



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


mechanisms in the nonlesioned systems (seesection below). With this technique, it is alsodifficult for researchers to make exactly similarlesions from one experiment to another. Thelesion resulting from this technique will alsounavoidably affect other axons with a locationsimilar to that of the targets. To demonstratedirectly the function of the C3-C4 PNs in theintact state, a novel technique to manipulatethe activity of the C3-C4 PNs without a lesionwould be desirable. For this purpose, we estab-lished a novel genetic tool of pathway-specificand reversible transmission block and applied itto macaque monkeys (M. Kinoshita, R. Matsui,S. Kato, T. Hasegawa, H. Kasahara, K. Isa,A. Watakabe, T. Yamamori, Y. Nishimura,B. Alstermark, D. Watanabe, D. Kobayashi,and T. Isa, unpublished observations) asshown in Figure 8. First, we injected a highlyefficient retrogradely transported (HiRet)lentivirus vector carrying the enhanced GFPand enhanced tetanus neurotoxin (eTeNT) inthe downstream of the tetracycline responsiveelement into the motor nuclei of the C6-Th1segments. This caused infection through theaxonal terminals of the fibers terminating inthe motor nuclei.

Seven to ten days later, the second vector,the adenoassociated vector carrying a highlyefficient Tet-ON sequence, rtTAV16, wasinjected in a section ranging from the caudalC2 to the rostral C5 segments. These injectionscaused infection via the cell somata, but itwas not transported retrogradely to the motorcortex. Together, the lentivirus (via axons)and adenovirus (somata) would specificallyinfect INs with somata located in C2-C5with projection to forelimb motor nuclei inthe C6-Th1 segments. In this condition, theenhanced GFP and eTeNT were transcribedunder the presence of doxycycline (Dox), atetracycline derivative, in the C3-C4 PNs withdouble infection. One to two months after thevirus injections, Dox was orally administeredto the monkeys. Two to five days after Doxapplication, allowing time for the expression ofeTeNT, the monkeys showed deficits in reach-ing and/or grasping movements. The monkeys

exhibited weakness in the index finger andthumb precision grip, grasping movements be-came slower, and, in some monkeys, reachingshowed dysmetria. Terminal electrophysiolog-ical experiments under anesthesia showed thatthe synaptic transmission through the C3-C4PNs was markedly reduced, presumably owingto the eTeNT that reduced synaptic release(Yamamoto et al. 2003). As shown in Figure 8c,stimulation of the corticospinal tract evokedmainly the monosynaptic field but not thedisynaptic field mediated via the C3–C4 PNs.Also, stimulation of the ascending axons ofthe PNs in the LRN revealed a large block intransmission of the monosynaptic field. Theseelectrophysiological findings also showed thatthe direct CM pathway was not affected by thevirus injections. In addition, with anti-GFPimmunohistochemistry, researchers were ableto visualize the location of the somata of theC3-C4 PNs whose transmission had beenblocked. These somata were located mainlyin the lateral part of the intermediate zoneof the C3-C4 segments, which transneu-ronal labeling with WGA-HRP revealed wasalso where the C3-C4 PNs were located(Alstermark & Kummel 1990). Another ad-vantage of using this technique is its abilityto visualize the axons all the way through thesomata to the terminals. We found that theaxons terminated on presumed MNs in theC6-Th1 segments and on neurons in the LRN.These results confirmed that the neuronsaffected by eTeNT during Dox administrationbelong to the group of C3-C4 PNs. Takentogether with the previous behavioral findings,these new results demonstrate that C3-C4 PNsplay a critical role in the control of reachingand grasping movements in macaque monkeysboth under normal conditions and after spinalcord lesions.

FUNCTIONAL RECOVERYFOLLOWING SPINALCORD INJURY

Although the macaque monkeys could performthe precision-grip exercise following lesion



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


MNs in C6–Th1

VAMP2

Dox

eGFP

PEST

TRE eTeNT

eTeNT

rtTAV16

C3–C4

PN

d

cCMV WPREITR ITR

psi RRE cPPT WPREeGFP

PEST

AAV2-CMV-rtTAV16a

b

HiRet-TRE-EGFP/eTeNT

LTR TRETight LTReTeNT

Sv40pArtTAV16

Stimulus LRN 1×

Monosynaptic

Disynaptic

Motor nucleiextracellular fields

Stimulus CST 3×

StimulateCST

StimulatePN axons

MCx

AAV2-CMV-rtTAV16No retrograde transport

HiRet-TRE-EGFP/eTeNT

C6–Th1

MN

Retrogradetransport

LRN

Cerebellum

C3–C4 PNs

sIN

Extracellularrecordingfrom motornuclei

Following PN blockade

Figure 8Selective and reversible blockade of synaptic transmission through the C3-C4 PNs via a double viral vector system. (a) Schematicdiagram of the vector injections, showing how HiRet-TRE-eGFP/eTeNT (red dots) and AAV2-CMV-rtTAV16 ( green dots) interact inthe double-infected cells. The former vector could infect PNs, sINs, and corticospinal neurons in the MCx; the double infectionoccurred only in the C3-C4 PNs. The AAV2-CMV-rtTAV16 vector is not transported retrogradely. (b) Injection of vectors in themotor nuclei in C6-Th1 (red ) and close to the PNs in C3-C4 ( green). Schematic diagram showing that the eGFP and eTeNT that weretranscribed in the presence of Dox and eTeNT cleave VAMP2, leading to the blockade of the synaptic transmission. (c) Extracellularrecordings from the motor nuclei in C6-Th1, obtained while stimulating the axons of the PNs directly in the LRN or by stimulatingthe corticospinal tract in the contralateral pyramid. Single and double asterisks indicate the monosynaptic and disynaptic fieldpotentials, respectively. (d ) Impairment of the precision grip on the second day after initiation of Dox administration in a monkey withdouble vector infection. Abbreviations: Dox, doxycycline; CST, corticospinal tract; eGFP, enhanced GFP; eTeNT, enhanced tetanusneurotoxin (eTeNT); HiRet, highly efficient retrogradely transported; LRN, lateral reticular nucleus; MCx, motor cortex;MN, motorneuron; PN, propriospinal neuron; sIN, segmental interneuron.



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


of the cortico- and rubrospinal tracts in C5(as described above), they showed deficits intheir grasping movements: Independent fingercontrol and their ability to preshape their digitsaccurately were both impaired. With dailytraining, lesioned monkeys showed a markedrecovery, and all showed near-full recovery ofprecision grip within two weeks to two monthspost lesion (Sasaki et al. 2004; Nishimuraet al. 2007, 2009). However, after a similarlesion in C2, monkeys displayed only partial,at best, recovery of precision grip. Theseresults suggest that the C3-C4 PNs are, inaddition to their normal function, involvedin the postlesion recovery of dexterous handmovements within the direct CM pathway. Inthe cat, which lacks direct CM connections,the postlesion recovery of digit grasping withinthe cortico-and rubrospinal tracts in C5-C6depends on takeover via C3-C4 PNs and thecorticoreticulospinal pathway (Alstermarket al. 1987, Blagovechtchenski et al. 2000,Pettersson et al. 2007).

Researchers have used brain imaging withpositron emission tomography to study cere-bral activation during functional recovery after

lesion of the cortico- and rubrospinal tracts inC5 in the macaque monkey (Nishimura et al.2007). During the early phase of recovery (1month), increased activation was present bilat-erally in the primary motor cortex (M1), andduring the late stage of recovery (3-4 months af-ter lesion), activation was present in an extendedarea of the contralateral M1 and bilaterally inthe ventral premotor cortex. These results sug-gest that a number of different pathways may beinvolved during the recovery process. Studiesof humans who have had a stroke suggest thatthe corticoreticulospinal with projection tothe propriospinal system may contribute torecovery (Pierrot-Deseilligny & Burke 2005).

To summarize, the experimental models re-viewed here offer good opportunities to extendthe research on motor recovery of voluntarymovements after spinal cord injury. Recently, avast literature (see Flynn et al. 2011, Nishimura& Isa 2011) argues for the involvement of PNsin functional recovery after spinal cord injury.Researchers may be able to take advantage ofthe ability of PNs to control object-orientedreaching and dexterous hand movements to de-velop future therapeutic strategies.

SUMMARY POINTS AND FUTURE ISSUES

1. The control of skilled reaching and grasping involves not only cortical circuits with directCM connection, but also indirect CM pathways via PNs and sINs in the spinal cord. Inthe cat, the C3-C4 PNs are specialized to control reaching, whereas grasping is moredependent on the sINs. In the macaque monkey, the C3-C4 PNs can also convey thecommand for precision grip involving the thumb and index fingers. These spinal circuitsare different from other spinal networks involved in locomotion.

2. The motor cortex yields parallel control of the PNs and sINs via the corticospinal andbulbospinal pathways. Sending visual information to the bulbospinal pathway, the spinalcircuits can quickly update the descending cortical command. An efferent copy fromthe PNs to the cerebellum via the LRN may play a critical role during reaching inupdating the cortical command to a moving target. In addition to visual control, feed-back inhibition of the C3-C4 PNs sent from the cutaneous receptors in the forelimbis critical for correct deceleration and accurate termination of reaching at the targetlocation.



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


3. Future work remains to be performed on the sINs to investigate the descending con-vergence pattern. Research needs to determine whether sINs also have an ascendingprojection to the LRN. Molecular work must also contend with the challenges of char-acterizing and possibly finding differential markers for the C3-C4 PNs and sINs. Froman evolutionary point of view, it is also interesting to understand why the PNs are locatedrostral to the forelimb segments.

4. In humans and rodents, a cervical propriospinal system also exists, but its normal functionremains to be clarified.

5. So far, most behavioral work has been conducted using lesions. A new technique based ona specific infection of PNs using double viral vectors has enabled pathway-selective andreversible inactivation of the C3-C4 propriospinal system. In monkeys, this inactivationhappened within days, resulting in deficits in reaching and grasping. It is also necessaryto develop new techniques in which reversible inactivation of discrete circuits can beperformed at a millisecond time resolution. The newly established optogenetic techniquemay be most important in this respect. Genetic techniques based on unique cell identitymay also offer a new possibility for manipulating specific neural circuits.

6. The behavioral work using lesions has given new interesting results with respect to plasticchanges during the early phase of recovery. These findings may be used to improvefunctional recovery after spinal cord injury in humans.

DISCLOSURE STATEMENT

The authors are not aware of any affiliations, memberships, funding, financial holdings, or anyother conflicts of interest that might be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS

B.A. is supported by the Swedish Research Council and Human Frontier Science Program. T.I.is supported by the Strategic Research Program of Brain Sciences by the Ministry of Educa-tion, Culture, Sports, Science, and Technology of Japan. The authors thank Dr. Pettersson forconstructive criticism on an earlier version of this article.

LITERATURE CITED

Alstermark B. 1983. Functional role of propriospinal neurons in the control of forelimb movements; a behavioural andelectrophysiological study. PhD thesis. Dep. Physiol., Univ. Goteborg. 32 pp.

Alstermark B, Ekerot CF. 1992. Organization of the ascending projection from C3-C4 propriospinal neuronesto cerebellum via the lateral reticular nucleus. Acta Physiol. Scand. 146(Suppl. 608):P2.35, 151

Alstermark B, Gorska T, Johannisson T, Lundberg A. 1986a. Hypermetria in forelimb target-reaching afterinterruption of the inhibitory pathway from forelimb afferents to C3-C4 propriospinal neurones. Neurosci.Res. 5:457–61

Alstermark B, Gorska T, Lundberg A, Pettersson LG. 1990. Integration in descending motor pathwayscontrolling the forelimb in the cat. 16. Visually guided switching of target-reaching. Exp. Brain Res.80:1–11

Alstermark B, Isa T, Ohki Y, Saito Y. 1999. Disynaptic pyramidal excitation in forelimb motoneurons mediatedvia C3-C4 propriospinal neurons in the Macaca fuscata. J. Neurophysiol. 82:3580–95



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


Alstermark B, Isa T, Pettersson LG, Sasaki S. 2007. The C3-C4 propriospinal system in the cat and monkey:a spinal pre-motoneuronal centre for voluntary motor control. Acta Physiol. Oxf. 189:123–40

Alstermark B, Johannisson T, Lundberg A. 1986b. The inhibitory feedback pathway from the forelimb toC3-C4 propriospinal neurones investigated with natural stimulation. Neurosci. Res. 5:451–56

Alstermark B, Kummel H. 1990. Transneuronal transport of wheat germ agglutinin conjugated horseradishperoxidase into last order spinal interneurones projecting to acromio- and spinodeltoideus motoneuronesin the cat. 2. Differential labelling of interneurones depending on movement type. Exp. Brain Res. 80:96–103

Alstermark B, Lindstrom S, Lundberg A, Sybirska E. 1981a. Integration in descending motor pathways con-trolling the forelimb in the cat. 8. Ascending projection to the lateral reticular nucleus from C3-C4propriospinal also projecting to forelimb motoneurones. Exp. Brain Res. 42:282–98

Alstermark B, Lundberg A. 1982. Electrophysiological evidence against the hypothesis that corticospinal fibressend collaterals to the lateral reticular nucleus. Exp. Brain Res. 47:148–50

Alstermark B, Lundberg A. 1992. The C3-C4 propriospinal system: target-reaching and food-taking. In MuscleAfferents and Spinal Control of Movement, ed. L Jami, E Pierrot-Deseilligny, D Zytnicki, pp. 327–54.Oxford/New York/Seoul/Tokyo: Pergamon

Alstermark B, Lundberg A, Norrsell U, Sybirska E. 1981b. Integration in descending motor pathways con-trolling the forelimb in the cat. 9. Differential behavioural defects after spinal cord lesions interruptingdefined pathways from higher centres to motoneurones. Exp. Brain Res. 42:299–318

Alstermark B, Lundberg A, Pettersson LG, Tantisira B, Walkowska M. 1987. Motor recovery after serialspinal cord lesions of defined descending pathways in cats. Neurosci. Res. 5:68–73

Alstermark B, Lundberg A, Sasaki S. 1984a. Integration in descending motor pathways controlling the forelimbin the cat. 10. Inhibitory pathways to forelimb motoneurones via C3-C4 propriospinal neurones. Exp.Brain Res. 56:279–92

Alstermark B, Lundberg A, Sasaki S. 1984b. Integration in descending motor pathways controlling the fore-limb in the cat. 11. Inhibitory pathways from higher motor centres and forelimb afferents to C3-C4propriospinal neurones. Exp. Brain Res. 56:293–307

Alstermark B, Ogawa J. 2004. In vivo recordings of bulbospinal excitation in adult mouse forelimb motoneu-rones. J. Neurophysiol. 92:1958–62

Alstermark B, Ogawa J, Isa T. 2004. Lack of monosynaptic corticomotoneuronal EPSPs in rats: disynap-tic EPSPs mediated via reticulospinal neurones and polysynaptic EPSPs via segmental interneurones.J. Neurophysiol. 91:1832–39

Alstermark B, Pettersson LG, Nishimura Y, Yoshino-Saito K, Tsuboi F, et al. 2011. Motor command forprecision grip in the macaque monkey can be mediated by spinal interneurons. J. Neurophysiol. 106:122–26

Alstermark B, Sasaki S. 1985. Integration in descending motor pathways controlling the forelimb in the cat.13. Corticospinal effects in shoulder, elbow, wrist, and digit motoneurones. Exp. Brain Res. 59:353–64

Alstermark B, Sasaki S. 1986. Integration in descending motor pathways controlling the forelimb in the cat.15. Comparison of the projection from excitatory C3-C4 propriospinal neurones to different species offorelimb motoneurones. Exp. Brain Res. 63:543–56

Bernhard CG, Bohm E. 1954. Cortical representation and functional significance of the corticomotoneuronalsystem. Arch. Neurol. Psychiat. 72:473–502

Blagovechtchenski E, Pettersson LG, Perfiliev S, Krasnochokova E, Lundberg A. 2000. Control of digits viaC3-C4 propriospinal neurones in cats; recovery after lesions. Neurosci. Res. 38:103–7

Brown TG. 1911. The intrinsic factors in the act of progression in the mammal. Proc. R. Soc. Lond. Ser. B84:309–19

Christensen MS, Kristiansen L, Rowe JB, Nielsen JB. 2008. Action-blindsight in healthy subjects after tran-scranial magnetic stimulation. Proc. Natl. Acad. Sci. USA 105:1353–57

Clendenin M, Ekerot CF, Oscarsson O. 1974. The lateral reticular nucleus in the cat. III. Organization ofcomponent activated from the ipsilateral forelimb tract. Exp. Brain Res. 21:501–13

Creed RS, Denny-Brown DE, Eccles JC, Liddell EGT, Sherrington CS. 1932. Reflex Activity of the SpinalCord. London: Oxford Univ. Press



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


Day BL, Brown P. 2001. Evidence for subcortical involvement in the visual control of human reaching. Brain124:1832–40

Day BL, Lyon IN. 2000. Voluntary modification of automatic arm movements evoked by motion of a visualtarget. Exp. Brain Res. 130:159–68

Ekerot CF. 1990. The lateral reticular nucleus in the cat. VI. Excitatory and inhibitory paths. Exp. Brain Res.79:109–19

Flynn JR, Graham BA, Galea MP, Callister RJ. 2011. The role of propriospinal interneurons in recovery fromspinal cord injury. Neuropharmacology 60:809–22

Fujito Y, Aoki M. 1995. Monosynaptic rubrospinal projections to distal forelimb motoneurons in the cat. Exp.Brain Res. 105:181–90

Goodale MA, Westwood DA. 2004. An evolving view of duplex vision: separate but interacting corticalpathways for perception and action. Curr. Opin. Neurobiol. 2:203–11

Illert M, Lundberg A. 1978. Collateral connections to the lateral reticular nucleus from cervical propriospinalneurones projecting to forelimb motoneurones in the cat. Neurosci. Lett. 7:167–72

Illert M, Lundberg A, Tanaka R. 1977. Integration in descending motor pathways controlling the forelimbin the cat. 3. Convergence on propriospinal neurones transmitting disynaptic excitation from the corti-cospinal tract and other descending tracts. Exp. Brain Res. 29:323–46

Illert M, Wiedemann E. 1984. Pyramidal actions in identified radial motor nuclei of the cat. Pflugers Arch.41:132–42

Isa T, Ohki Y, Alstermark B, Pettersson LG, Sasaki S. 2007. Direct and indirect cortico-motoneuronal path-ways and control of hand/arm movements. Physiology 22:145–52

Isa T, Ohki Y, Seki K, Alstermark B. 2006. Properties of propriospinal neurons in the C3-C4 segments medi-ating disynaptic pyramidal excitation to forelimb motoneurones in the Macaque monkey. J. Neurophysiol.95:3674–85

Kitazawa S, Ohki Y, Sasaki M, Xi M, Hongo T. 1993. Candidate premotor neurons of skin reflex pathwaysto T1 forelimb motoneurons of the cat. Exp. Brain Res. 95:291–307

Kuypers HG. 1982. A new look at the organization of the motor system. Prog. Brain Res. 57:381–403Landgren S, Phillips CG, Porter R. 1962. Cortical fields of origin of the monosynaptic pyramidal pathways

to some alpha motoneurones of the baboon’s hand and forearm. J. Physiol. 161:112–25Lawrence DG, Kuypers HGJM. 1968a. The functional organization of the motor system in the monkey. I.

The effects of bilateral pyramidal lesions. Brain 91:1–14Lawrence DG, Kuypers HGJM. 1968b. The functional organization of the motor system in the monkey. II.

The effects of lesions of the descending brain-stem pathways. Brain 91:15–36Lemon RN. 2008. Descending pathways in motor control. Annu. Rev. Neurosci. 31:195–218Lloyd DPC. 1941. The spinal mechanism of the pyramidal system in cats. J. Neurophysiol. 4:525–46Lundberg A. 1999. Descending control of forelimb movements in the cat. Brain Res. Bull. 50:323–24Lundberg A, Voorhoeve P. 1962. Effects from the pyramidal tract on spinal reflex arcs. Acta Physiol. Scand.

56:201–19Malmgren K, Pierrot-Deseilligny E. 1988. Evidence for non-monosynaptic Ia excitation of human wrist flexor

motoneurones, possibly via propriospinal neurones. J. Physiol. 405:747–64Marchand-Pauvert V, Mazevet D, Pradat-Diehl P, Alstermark B, Pierrot-Deseilligny E. 2001. Interruption

of a relay of corticospinal excitation by a spinal lesion at C6-C7. Muscle Nerve 24:1554–61Matsuyama K, Drew T. 1997. Organization of the projections from the pericruciate cortex to the pon-

tomedullary brainstem of the cat: a study using the anterograde tracer Phaseolus vulgaris-leucoagglutin.J. Comp. Neurol. 389:617–41

Nishimura Y, Isa T. 2011. Cortical and subcortical compensatory mechanisms after spinal cord injury inmonkeys. Exp. Neurol. Epub ahead of print

Nishimura Y, Morichika Y, Isa T. 2009. A subcortical oscillatory network contributes to recovery of handdexterity after spinal cord injury. Brain 132:709–21

Nishimura Y, Onoe H, Morichika Y, Perfiliev S, Tsukada H, Isa T. 2007. Time-dependent central compen-satory mechanisms of finger dexterity after spinal cord injury. Science 318:1150–55

Peterson B, Pitts NG, Fukushima K. 1979. Reticulospinal connections with limb and axial motoneurons. Exp.Bran Res. 36:1–20



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.


Pettersson LG, Alstermark B, Blagovechtchenski E, Isa T, Sasaki S. 2007. Skilled digit movements in felineand primate-recovery after selective spinal cord lesions. Acta Physiol. 189:141–54

Pettersson LG, Lundberg A, Alstermark B, Isa T, Tantisira B. 1997. Effect of spinal cord lesions on forelimbtarget-reaching and on visually guided switching of target-reaching in the cat. Neurosci. Res. 29:241–56

Pettersson LG, Perfiliev S. 2002. Descending pathways controlling visually guided updating of reaching incats. Eur. J. Neurosci. 16:1349–60

Pierrot-Deseilligny E, Burke D. 2005. The Circuitry of the Human Spinal Cord. Cambridge, UK: CambridgeUniv. Press

Pruszynki JA, King GL, Boisse L, Scott S, Flanagan JR, Munoz DP. 2010. Stimulus-locked responses onhuman arm muscles reveal rapid neural pathway linking visual input to arm motor output. Eur. J. Neurosci.32:1049–57

Renshaw B. 1940. Activity in the simplest spinal reflex pathways. J. Neurophysiol. 3:373–87Riddle CN, Edgley SA, Baker SN. 2009. Direct and indirect connections with upper limb motoneurons from

the primate reticulospinal tract. J. Neurosci. 29:4993–99Sasaki S, Isa T, Pettersson LG, Alstermark B, Naito K, et al. 2004. Dexterous finger movements in primate

without monosynaptic corticomotoneuronal excitation. J. Neurophysiol. 92:3142–47Shapovalov AI. 1975. Neuronal organization and synaptic mechanisms of supraspinal motor control in verte-

brates. Rev. Physiol. Biochem. Pharmacol. 72:1–54Sherrington CS. 1906. The Integrative Action of the Nervous System. New Haven, CT/London: Yale Univ. PressWerner W. 1993. Neurons in the primate superior colliculus are active before and during arm movements to

visual targets. Eur. J. Neurosci. 5:335–40Yamamoto M, Wada N, Kitabatake Y, Watanabe D, Anzai M, et al. 2003. Reversible suppression of gluta-

matergic neurotransmission of cerebellar granule cells in vivo by genetically manipulated expession oftetanus neurotoxin light chain. J. Neurosci. 23:6759–67



Ann

u. R

ev. N

euro

sci.

2012

.35.

Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

Kor

ea A

dvan

ced

Inst

itute

of

Scie

nce

and

Tec

hnol

ogy

(KA

IST

) on

04/

15/1

2. F

or p

erso

nal u

se o

nly.

Circuits for Skilled Reaching and Grasping

Documents

motor cortex

motor commands

skillful reaching

grasping inthe cat

motor cortex

online atneuro

ne35ch26alstermark ari

basic motor functions