R

Cro

Ma

b

Q

1

satmi[t[tnlir

0d

Behavioural Brain Research 192 (2008) 232–247

Contents lists available at ScienceDirect

Behavioural Brain Research

journa l homepage: www.e lsev ier .com/ locate /bbr

esearch report

ineradiographic (video X-ray) analysis of skilled reaching in a single pelleteaching task provides insight into relative contribution of body, head,ral, and forelimb movement in rats

ariam Alaverdashvili a,∗, Hugues Leblondb, Serge Rossignolb, Ian Q. Whishawa

Department of Neuroscience, Canadian Centre for Behavioral Neuroscience, University of Lethbridge, 4401 University Drive, Lethbridge, Alberta T1K 3M4, CanadaGroupe de Recherche Sur le Systeme Nerveux Central (GRSNC), Pavillon Paul-G.-Desmarais, 2960 Chemin de la Tour, University de Montreal, Montreal,uebec H3T 1J4, Canada

(skillode

t, pawin relmovelly obontalachinentatists ob segors ale toelatio

a r t i c l e i n f o

Article history:Received 23 January 2008Received in revised form 8 April 2008Accepted 14 April 2008Available online 22 April 2008

Keywords:Bone and joint motionCineradiography/X-ray analysisMotor systemMovement analysisSkilled reaching and upper extremities

a b s t r a c t

The forelimb movementsmouth have been used to moppositions of head–pelleportion of body segmentsbecause the location andwhich is pliable and partiafrom lateral, dorsal, and frmove during the skilled reenabled by the vertical ori(ii) skilled reaching consnumber of concurrent limpaw using either the inciscinematography is valuabgeneral principles of the r

. Introduction

Use of a forelimb to reach for food, conventionally calledkilled reaching, is an evolutionary old movement, displayed widelymong terrestrial vertebrates [17]. Rodents and primates belongo sister clades and share many similarities in their skilled move-

ents. Therefore, the laboratory rat and mouse are useful fornvestigating basic neural questions related to skilled movements5,7,18,19,21,27,30] and serve as animal models to study motor sys-em injury due to stroke [3,7,34], trauma [36], spinal cord injury12,19], and a number of neurological disorders including Hunting-on’s chorea [9] and Parkinson’s disease [22]. Skilled reaching inormal and experimental model rats has been contrasted using at

east three classes of measurement: end point measures, includ-ng measures of reach successes and reach attempts [29], Cartesianepresentations of the movement of the distal ends of various body

∗ Corresponding author. Tel.: +1 403 394 3998; fax: +1 403 329 2775.E-mail address: mariam.alaverdashvil@uleth.ca (M. Alaverdashvili).

166-4328/$ – see front matter © 2008 Elsevier B.V. All rights reserved.oi:10.1016/j.bbr.2008.04.013

ed reaching) used by rats to reach for a single food pellet to place into thel many neurological conditions. They have been described as a sequence of–pellet and pellet–mouth that can be described as movements of the distal

ation to their fixed proximal joints. Movement scoring is difficult, however,ment of body segments is estimated through the overlying fur and skin,scures movement. Using moderately high-speed cineradiographic filmingperspectives, the present study describes how forelimb and skeletal bonesg act. The analysis indicates that: (i) head movements for orienting to food,ion of the rostral spinal cord, are mainly independent of trunk movement,f a sequence of upper arm and extremity movements each involving ament and joint movements and (iii) food pellets are retrieved from the

nd/or tongue. The findings are discussed in relation to the idea that X-rayol for assisting descriptive analysis and can contribute to understandingns between whole body, head, oral, and upper extremity movement.

© 2008 Elsevier B.V. All rights reserved.

segments [32], and taxonomic measures derived from movementnotation techniques [1,28,35]. The notation techniques describe themovement of the distal portions of limb segments in relation totheir fixed proximal portion and so capture the individual con-tributions of many simultaneously moving body segments. Thesimilarities in movement changes in animal models of human dis-ease attest to the utility of these analytical measures.

In most studies using the rat, the movements of reaching havebeen described from video recordings. A drawback of the methodis that whereas the distal end of the paw can be clearly tracked,representations of the proximal and distal ends of other body seg-ments are more difficult to make. Markers are not easily attachedto the overlying skin and the movement of the skin as a conse-quence of limb or trunk movements makes the location of jointsdifficult. Cineradiographic/video X-ray analysis provides a solutionto this problem. Cineradiography gives an X-ray image of the skele-ton as an animal performs the reaching task and thus provides away of documenting skeletal movements. A number of studies haveused cineradiography to describe locomotion [8,15,23,26] and onestudy has examined reaching in the cat [4]. No previous studieshave used cineradiography as a method to measure movements of

ral Brain Research 192 (2008) 232–247 233

M. Alaverdashvili et al. / Behaviou

the rat or to confirm notational scoring systems used for the rat.The present study uses cineradiography to address a number ofquestions, including whether orienting of the head to locate foodis independent from trunk movement, the extent to which move-ments of different limb segments are sequential and concurrent,and how the rat uses its mouth to remove food from its paw.

The rats were viewed using cineradiography or were filmed withvideo as they performed a single pellet reaching task. Reachingmovements from both methods were compared frame-by-frame.The findings are discussed in relation to: topography of skeletonduring skilled reaching and the utility of cineradiographic meth-ods for understanding general principle of skeleton organizationand movement.

2. Materials and methods

2.1. Subjects

Four female (≈100 days old, with body weight 270–310 g) Long-Evans hoodedrats from the University of Lethbridge and four female rats of the same strain(≈100 days old, with body weight 296–317 g) from the University de Montreal ani-mal colony were used. The experiments were conducted in compliance with theguidelines of the animal care committees of the University of Lethbridge and theUniversity of Montreal as outlined by the Canadian Council for Animal Care, whichcomplies with international standards for animal care. Rats were housed in Plex-iglas cages (36 cm long, 20 cm wide, and 21 cm deep) with sawdust bedding, ingroups of two or three in a colony room maintained on a 12-h light:12-h dark cycle(08:00–20:00 h) with controlled temperature and humidity.

2.2. Feeding and food restriction

Prior to and during training, the rats were gradually food deprived to 90–95% oftheir body weight by once a day feedings of 18–20 g of Purina rat chow. The weekprior to training, each rat received twenty 45 mg dustless precision banana-flavoredpellets (product #F0021, Bioserve Inc., Frenchtown, NJ, USA) 1 h prior to the dailychow ration. The pellets would later serve as reaching targets. Once reach trainingbegan and until the end of the study only rat chow was served in the home cage.

2.3. Single pellet reaching box

The single pellet reaching box [28] is made of clear Plexiglas with the dimensions35 cm long × 14 cm wide × 45 cm high. For X-ray imaging the reaching box has thesame dimensions except the height (30 cm) to position the amplifier close to therat to improve X-ray images. At the center of each front wall of the box there is avertical slot, 1 cm wide extending from 2 cm above the floor to a height of 15 cm. Onthe outside of the wall, in front of the slot, a 2 cm deep shelf is mounted 3 cm abovethe floor. Two indentations on the surface of the shelf are located 2 cm from theinside of the wall in order to hold the food targets. The indentations were alignedwith the edges of the slit. This location prevents a rat from lapping the food withits tongue, and because the paw pronates medially to grasp, requires the use of thecontralateral paw for retrieval.

2.4. Recording apparatus

2.4.1. Video recordingReaching performance was recorded using a Sony 3CCD camcorder (shutter,

speed 1/1000 s) and a cold light source. A Sony videocassette recorder DSR-11 wasused for subsequent frame-by-frame analysis (30 frames/s).

2.4.2. X-ray recordingCineradiographic recording used a high-speed motion analysis camera hooked

to an X-ray machine (Siemens Coroskop Bicor) as illustrated in Fig. 1. Continuous X-rays (100 kV, 8–9 mA) were applied during the reaching movement of the rats whiledigital images were acquired at 120 frames/s either in the lateral, frontal or dorsalperspective with a camera (DALSTAR CA-D6, Dalsa Corp., Ontario, Canada) directlymounted on the 27 cm diameter image intensifier. In order to freeze the movementswithout compromising the quality of the image, the shutter speed was set at 1/500 s.These digital images were directly recorded on a PC using VisionNow Imaging Sys-tem [Boulder Imaging, Louisville, CO, USA; http://www.boulderimaging.com].

2.5. Training

The training consisted of daily 15–20 min sessions for each rat. The first objectivewas to introduce the rat to the reaching cage with pellets on the shelf and to have therat retrieve the pellet by paw or tongue. Once a rat was successfully retrieving thepellet, pellets on the shelf were moved further away in order to encourage the use of

Fig. 1. Experimental setup. The number corresponds to: 1, X-ray source; 2, imageamplifier of X-ray system; 3, camera recording the X-ray sequence from the imageamplifier; 4, reaching cage. Bidirectional arrow indicates to movement range of theimage amplifier for appropriate zooming of the rat or forelimb segments; curvedarrows – range of motion of X-ray system to mount the X-ray source and imageamplifier in desired position for filming the rats either from lateral, dorsal or frontalperspective.

a paw. After a rat demonstrated a preference for one paw, by making more reachingattempts with it, individual pellets were placed into the indentation contralateral tothat paw. During the second week, rats continued to receive daily 20 min trainingsessions, consisting of discrete trials with inter-trial intervals during which the ratswere shaped to leave the slot, walk to the rear wall of the cage, turn and approachthe slot again for the next pellet. In addition, by withholding food on semi-randomlyselected trials, rats were taught to sniff the shelf for a pellet and to reach only if apellet was present. Thus, each rat eventually learned to orient to the food pellet,transport its limb through the slot, grasp the food pellet, and retract its paw throughthe slot to release the food into its mouth [11]. Over days, the rats were trained withdifferent orientation of the cage relative to experimental room. In addition, the ratsto be used for cineradiography were acquainted to the room and the sounds relatedto X-ray recording and equipment prior to the formal tests.

2.6. Timeline

After achieving 50% in total success in the reaching task the rats were recordedby conventional video and X-ray cinematography techniques. Reaching performance(success level) was scored only during training period but not during recording ofthe reaching movements. In separate blocks of trials video camera or X-ray sourceand amplifier was mounted either laterally, dorsoventrally or frontally to obtainimages of the rats’ forelimb and head in the vertical and anteroposterior directions,anteroposterior and mediolateral directions or vertical and mediolateral directions,respectively.

2.7. Behavioral analysis

2.7.1. Quantitative analysisReaching behavior was assessed by the following measures:

(1) Total success. A successful reach was defined as one in which an animal graspeda food pellet, transported it with the paw into the cage, and placed it intoits mouth. To achieve a successful reach, the rat was free to advance the pawtoward the food in as many “tries” as required. Total success was calculated as:success% = number of pellets obtained/20 × 100;

(2) First trial success. First trial successes were those in which a rat obtained afood pellet on the first advance of the limb toward the food as calculated as:success% = number of pellets obtained on first advance/20 × 100.

2.7.2. Movement pattern analysisThe various movements comprising a reach have been identified using a con-

ceptual framework derived from Eshkol–Wachmann Movement Notation (EWMN)[6,28]. In brief, EWMN is designed to express relations and changes of relationbetween the parts of the body. The body is treated as a system of articulated axes,i.e. body and limb segments. A limb is any part of the body that either lies betweentwo joints or has a joint and a free extremity. These are imagined as straight lines(axes), of constant length, which move with one end fixed to the center of a sphere.An important feature of EWMN is that the same movements can be notated in sev-eral polar coordinate systems. The coordinates of each system are determined withreference to the environment, to the animal’s body midline axis, and to the nextproximal or distal limb or body segment. By transforming the description of the

234 M. Alaverdashvili et al. / Behavioural Brain Research 192 (2008) 232–247

th cerital orarmto ho

phaland souldee extelder p

llet. Me andion o

bow, am froso ass

d to mbow a

Table 1Limb/limb segments movement during skilled reaching

Movements X-ray

1 Orient to locate the food pellet The spine has an “S-shape” wivertical plane at atlanto-occip

2 Limb lift with digits to the midline Backward movement of uppershoulder carries the lower armmidline of the body

3 Digits semiflexed Movement of metacarpals and4 Aim with elbow in The rotation of humerus head5 Advance Extension at the elbow and sh6 Digits extend Metacarpals and phalanges ar7 Pronation/arpeggio Abduction of elbow from shou

positions the paw over the pe8 Grasp The pellet is grasped by closur

are accompanied by slight flex9 Supination I Flexion at the shoulder and el

10 Supination II Backward movement of the arrotation at wrist and elbow alcontact the paw

11 Release The phalanges open and exten12 Replacing Flexion at the shoulder and el

same behavior from one coordinate system to the next, invariances in that behaviormay emerge in some coordinate systems but not in others. Thus, the behavior maybe invariant in relation to some or all of the following: the animal’s longitudinal axis,gravity, or bodywise in relation to the next proximal or distal segment. On the basisof descriptions derived from EWMN, 12 of the movements (Table 1) were describedfrom X-ray recordings.

Representative reaching movements from video and X-ray recordings were cap-tured with Final Cut Express HD (V.3.5 http://www.apple.com) and later analyzedframe-by-frame. Only successful reaches on first reach attempts were captured andanalyzed. On representative X-ray frames position/orientation of bones were out-lined at the beginning and end of the each reach movement component printed asschematic diagram of the reach component movement. As X-ray recordings fromfrontal view due to superimposition of bones in antero-posterior space did not suitanalysis of forelimb movement during reaching while this perspective was the bestview for video analysis, here we demonstrate video recordings from frontal and

X-ray recordings from lateral and dorsal views.

2.7.3. Kinematic analysis of movementX-ray images (e.g. Fig. 2) copied on video tapes (Final Cut Express HD;

V.3.5 http://www.apple.com) were further captured and analyzed by software of“Peak Motus” (Version 8; http://www.peakperform.com). The data were acquiredvia manual mode, digitizing moving points by cursor. Displacement of selectedpoints and angles of joints either in horizontal or vertical plane was calculatedduring whole reaching movement and plotted on the graphs. The positions ofdigitized landmarks to calculate movement in the parasagittal plane are illus-trated in Fig. 3A. Angles calculated are the projections of the atlanto-occipitaland cervico-thoracic junctions onto the sagittal plane representing their contri-bution to movements of the head in vertical and forward motion during reaching(Fig. 3A). Vector-based angle calculation method was used at the flexor side of eachjoint.

3. Results

3.1. Quantitative analysis

At the completion of behavioral training on the reaching task,the rats achieved a mean of 61 ± 7% success as an overall reachingscore, with 42 ± 15% on first reach successes.

Fig. 2. Lateral X-ray view of the rat approaching the food. Note the S-shaped curvature ofduring extension of the head to sniff (panel 2).

vical vertebra directed upward; the head moves independent from trunk in theat cervico-thoracic junctions

from scapula and slight closure at the shoulder joint lifts the paw; opening at therizontal; supination the paw around the wrist and elbow brings the palm to the

anges semiflexes and closes the phalangeshoulder extension adducts elbow and then moves forward the limbr advances the forelimbnded and targeted to pellet; the distal portions of the digits remain slightly flexedartially pronates the paw, adduction at the shoulder moves the limb mediallyovement around the wrist assists arpeggio movementflexure of metacarpals and phalanges, and by wrist extension. These movements

f elbow joint and retraction of the shouldernd adduction of the elbow supinate the paw by ∼90◦

m the shoulder and adduction at the elbow supinates the paw toward the mouth;ists supination of paw. The head bends down at atlanto-occipital junction to

ake room for either the incisors or tongue to grasp the foodnd later extension of the shoulder places the paw on the floor/the wall

3.2. Movement element analyses

Forelimb movements during reaching were compared by ana-lyzing the video and X-ray recordings frame by frame for threereaches per rat. The results are described in relation to 12 move-ment elements comprising a successful reach, first as observedwith video recording and following this with X-ray recording[31]. Reaching movements both on video and X-ray images aredemonstrated on the rat reaching with the right paw. A sum-mary of the main findings from X-ray images is documented inTable 1.

3.2.1. Orient to locate the food pellet3.2.1.1. Video. The movement of orienting to the food consists ofwalking from the back of the test box to the front of the box, atwhich point the rat sniffs to determine whether food is locatedon the shelf. In preliminary training, the rat learns to reach if afood pellet is present, and to make another round trip if a foodpellet is not present. Before sniffing, the rat arrests its forwardmovement and makes a number of postural adjustments and step-ping movements to appropriately position itself before the foodslot (Fig. 4, top). The rat then moves its snout until it locates thefood, and sniffs the food, as indicated by opening and closingthe nostrils. After sniffing the food, it raises its head to provideroom for the paw to be directed through the slot to the food.There is ambiguity whether the head movements are indepen-dent of the trunk or carried by trunk movement. This is becausewhen the head moves, so does the skin and fur over the neck andshoulders.

3.2.1.2. X-ray. The X-ray image shows that the head is moved infour different ways during the course of orienting and reachingmovements:

the vertebrae during walking (panel 1) and orienting to the food (panel 3) but not

M. Alaverdashvili et al. / Behavioural Brain Research 192 (2008) 232–247 235

Fig. 3. Stick figure and kinematics of the head and body movement during skilledreaching from lateral perspective in a representative rat. (A) Stick figure shows trunkand head posture during approach (Apr), orient (O), limb lift (L), grasp (Gr) andreplace (Rp). (B) “Peak Motus” displacement in the horizontal plane of differentbody parts (S, snout; AO, atlanto-occipital junction; CT, cervico-thoracic junction;BS, the base of the spine). Note: Displacement of the snout is produced by movementof the forequarters. (C) Displacement in the vertical plane indicates that movementof the snout is produced by movement at the atlanto-occipital and cirvico-thoracicjunction. (D) Angular changes at AO and CT show independent and/or simultaneousangular changes produce head movements.

(1) As the rat walks toward the slot from the back of the box, thehead is carried nose down by movement of the body, with thehead and rostral vertebrae in an “S-shape” (Fig. 2, panel 1). Asthe rat approaches the slot it might extend the head to sniff.This movement is achieved mainly by extension at the tho-racic vertebrae, leaving the rostral spinal cord almost straightbetween the skull and the lumbar vertebrae (Fig. 2, panel 2). As

Fig. 4. Orient. Video and X-ray images of the rat from lateral view during locating thefood pellet. Solid outline corresponds beginning of orient; dotted line – the end of theaction. Note movement of the head independent from trunk at the cervico-thoracicjunction and the vertical orientation of cervical vertebrae.

the rat positions itself before the slot the S-shape is resumed(Fig. 2, panel 3). Thus, the cervical vertebrae are oriented in anupward direction to the atlanto-occipital joint of the head fromtheir apogee at about the 7th cervical vertebrae. There is a sim-ilar upward orientation of the thoracic vertebrae toward thecervico-thoracic junction. The contribution of the trunk in theforward motion of the head is shown in Fig. 3B.

ral Bra

236 M. Alaverdashvili et al. / Behaviou

(2) The head itself moves in two ways: it can be flexed and extendedat the atlanto-occipital joint, and it can move by openingand closing at the cervical–thoracic junction mainly involvingmovement at the cervical vertebrae (Fig. 3). When the snoutis moved to sniff the food pellet and then is moved away, itsmovement is mediated by a combination of these two move-ments. Fig. 3C shows that the raising and bending of the snoutis associated to the displacement of the atlanto-occipital andcervico-thoracic junctions in the vertical plane while the trunkmaintains remains relatively still. Moreover, extension (∼morethan 90◦) and flexion (∼90◦ or less) of the atlanto-occipital andcervico-thoracic junctions (Fig. 3D) during reaching shows thatthe head movement is achieved by independent and/or simul-taneous angular changes at the given junctions independentfrom trunk motion. On the other hand, when the rat approachesthe slot, the forward movement of the head is achieved byboth body displacement and the displacement of the atlanto-occipital and cervico-thoracic junctions (Fig. 3B). The headmovement during “orient to locate the pellet” is illustrated inFig. 4 (bottom) in which the head is moved independent of trunkmovement.

(3) As the rat begins the reach, the head is carried upward by theforequarters with little independent movement relative to therest of the body. The upward movement is achieved by exten-sion of the supporting forelimb and flexion in the supportinghind limbs (Fig. 5, bottom (lateral view)).

(4) Once the rat has obtained the food and withdraws the foodthrough the slot (Fig. 3D), the head flexes at the atlanto-occipitaljunction (angle approximates to 90◦) and there is an light andtransient extension of the cervical vertebrae relative to the tho-racic vertebrae so that the mouth is lowered to contact the pawholding the food pellet. Throughout this movement, the head isalso lowered by flexion of the supporting hind limbs. The weightshift produced by the backward movement of the forequartersallows the supporting forelimb to be lifted to assist in holdingthe reaching paw and the food pellet.

3.2.2. Limb lift with digits to the midline—digits semiflexed3.2.2.1. Video. Video recordings show that when the limb is liftedwith, “the digits to the midline movement”, the movement is mademainly with the upper arm such that the lower arm takes an almosthorizontal position with its distal end oriented toward the midlineof the body. As the limb is lifted, the digits are closed and semiflexedand the paw is supinated so that palm surface is in a vertical plane.

A frontal view shows the tips of the digits (the end of claws) arealigned with the midline of the body. Throughout the movementthe body maintains its position in an antero-posterior plane (Fig. 5,top), the head raises and trunk tilts toward body-supporting pawprobably to make some room for the forelimb to reach the pelletand to maintain balance the other three limbs. The video images areclear with respect to the start and end position of the lift movement,but the relative contributions of the scapula, which is tethered andwrapped by musculature, and the humerus are not obvious.

3.2.2.2. X-ray. X-ray images (Fig. 5, middle and bottom) confirmthat in the lift, the radio-ulnar axis is taken to an almost horizontalposition with the phalanges oriented toward the midline of thebody. This movement is achieved by two separate actions. First,the upper arm (humerus) moves backward, enabled in part by aforward and planar movement of the scapula, and in part by a slightclosure of the humerus–scapula angle. This movement lifts the pawfrom the surface. Second, the scapula–humerus joint opens andthe humerus–radius/ulnar joint closes and adducts. This movementadvances the forelimb to its almost horizontal position with thepaw oriented toward the midline of the body. Thus, the humerus

in Research 192 (2008) 232–247

first moves backward to lift the paw and then forward to positionit toward the midline.

Three separate movements are associated with the paw reachingits “lift position” with the digits semiflexed. First, the digits pushagainst the floor to aid in raising the forequarters. Secondly, oncethey leave the floor surface, the digits close and flex. Third, the pawis supinated in part by a movement around the wrist and in part bya movement around the elbow.

3.2.3. Aim with elbow in3.2.3.1. Video. The movement of aiming involves adducting theelbow to the midline of the body, with the paw remaining on themidline, just under the mouth. The movement positions the paw sothat it can move directly forward through the slot toward the food.The aim (Fig. 6, top) appears to consist of an adducting the elbowthat is secondarily accompanied by forward movement of the limb.The paw fixation in the midline of the body is achieved via abduc-tion of the lower arm. The digits remain semiflexed and no motionis observed around the wrist.

3.2.3.2. X-ray. X-ray images (Fig. 6, middle) confirm the videoimages in that there is first an adduction of the elbow that is thenaccompanied by forward movement of the humerus. Thus, it is amedial and forward movement of the distal end of the humerus(from shoulder joint) that brings the elbow to midline of the body(Fig. 6, middle). The fixation that maintains the paw on the mid-line of the body, while the elbow moves inward, is produced inlarge part by the clockwise rotation (from dorsal perspective), i.e.outward rotation (from lateral perspective) of the humerus head.X-ray images (Fig. 6, bottom) also indicate that there is no indepen-dent extensor movement of the lower arm (at radio-ulnar/humerusjoint) as the angle between the humerus and the radio-ulnaraxis remains constant. Together, these observations suggest thatthroughout most of the aiming movement, the lower arm is carriedby the upper arm and shoulder movement.

3.2.4. Advance and digits extend3.2.4.1. Video. The movement of advance (Fig. 7, top) consists ofmoving the paw directly forward through the slot. Toward the endof the movement the digits begin to extend. The movement ofadvancing the limb appears to be achieved by extension of elbowjoint.

3.2.4.2. X-ray. X-ray images (Fig. 7, middle and bottom) confirm

that the advance is mainly produced by extension at the radio-ulnar/humerus joint but also involves some extension at theshoulder joint with the scapula. There appeared to be minimalmovement of the supporting limbs or the trunk in moving the limbforward. During the last part of the advance the metacarpals andphalanges extend but extension is incomplete, as there remains aslight flexure of the distal portions of the digits.

3.2.5. Pronation/arpeggio3.2.5.1. Video. The movement of pronation (Fig. 8, top) positionsthe paw over the food pellet with the palm down. As pronation takesplace, the digits almost fully extend and open. Pronation takes placewith digit 5 (the outer digit) through digit 2 sequentially contact-ing the shelf, in an arpeggio movement [35]. Pronation moves thepaw palm medially toward the location of the food pellet. Prona-tion appears to be produced by abduction of the elbow while at thesame time rotation at wrist assists arpeggio.

3.2.5.2. X-ray. X-ray images (Fig. 8, middle and bottom) show thatthree movements contribute to pronation/arpeggio. First, the ratboth extends the limb and pronates the paw by abduction of

M. Alaverdashvili et al. / Behavioural Brain Research 192 (2008) 232–247 237

Fig. 5. Lift/digits semiflexed. Video and X-ray images of the start and the end of lift/digits semiflexed. Note: (Frontal view) the paw raises from the floor and the digits alignto midline in a semiflexed posture; (lateral view) the forelimb and scapula move together indicating that the forelimb is carried by movement of the scapula; (dorsal view)the movement of the paw to the midline is produced by rotation at the shoulder and movement aligning the paw with slot is produced by a lateral shift of the body. (Bottom,solid outline: the start of lift; dotted outline: the end of the action.)

238 M. Alaverdashvili et al. / Behavioural Brain Research 192 (2008) 232–247

Fig. 6. Aim. Video and X-ray images of the start and the end of the aim. Note: (Frontal view) elbow adducts to midline of the body and the paw is positioned under the mouth;(lateral view) the upper arm movement carries the lower arm; (dorsal view) the upper arm movement adducts the lower arm. (Bottom, solid outline: the start of lift; dottedoutline: the end of the action.)

M. Alaverdashvili et al. / Behavioural Brain Research 192 (2008) 232–247 239

Fig. 7. Advance/digits extend. Video and X-ray images of the start and the end of the aim. Note: (Frontal view) forelimb advances through the slit with the digits extend;(lateral view) upper arm carries the lower arm and extension at the elbow carries the limb forward while the digits extend; (dorsal view) adduction of the elbow by movementat the shoulder. (Bottom, solid outline: the start of lift; dotted outline: the end of the action.)

240 M. Alaverdashvili et al. / Behavioural Brain Research 192 (2008) 232–247

Fig. 8. Pronation. Video and X-ray images of the start and the end of pronation. Note: (Frontal view) the paw pronates over the pellet by movement from shoulder joint;(lateral view) stability of wrist position; (ventral view) pronation produced by a movement of the upper arm and wrist. (Bottom, solid outline: the start of lift; dotted outline:the end of the action.)

M. Alaverdashvili et al. / Behavioural Brain Research 192 (2008) 232–247 241

Fig. 9. Grasp. Video and X-ray images of the start and the end of grasp. Note: (Frontal view) the digits flex and close around the pellet and wrist extends to take the pellet;(lateral view) flexion of the phalanges and extension of the wrist occur independently; (dorsal view) digit flexion and closing occur independently. (Bottom, solid outline:the start of lift; dotted outline: the end of the action.)

242 M. Alaverdashvili et al. / Behavioural Brain Research 192 (2008) 232–247

Fig. 10. Supination I. Video and X-ray images of the start and the end of supination I. Note: (Frontal view) the paw rotates by ∼90◦ by movement at wrist and prixomal endof upper arm; (lateral view) rotation of the humerus; (dorsal view) rotation of the humerus. (Bottom, solid outline: the start of lift; dotted outline: the end of the action.)

M. Alaverdashvili et al. / Behavioural Brain Research 192 (2008) 232–247 243

Fig. 11. Supination II. Video and X-ray images of the start and the end of Supination II. Note: (Frontal view) the paw supinates by ∼45 toward the mouth; (lateral view)withdrawal is produced by retraction of the upper arm from the shoulder and the head is bent to paw by flexion at the atlanto-occipital junction; (dorsal view) light tilt ofthe trunk. (Bottom, solid outline: the start of lift; dotted outline: the end of the action.)

ral Bra

244 M. Alaverdashvili et al. / Behaviou

the elbow by a movement at the shoulder. Whereas the videoimage suggests that the radio-ulnar/humerus joint opens almostfully so that the limb is straight, the X-ray image shows thatthe radio-ulnar/humerus joint does not fully extend. The ole-cranon at the proximal end of the ulna, limits the extension ofthe elbow to about 135◦ (Fig. 8, middle). Second, pronation ofthe paw does not fully move the paw over the food pellet, butthere is adduction at the shoulder that moves the limb medi-ally so that the paw is positioned over the food pellet. Third,movement about the wrist and movement of the digits appear tocontribute to the arpeggio movement, but neither the video northe X-ray records are sufficiently clear to adequately describe thesemovement.

3.2.6. Grasp3.2.6.1. Video. The grasp (Fig. 9, top) consists of two separate move-ments: the flexion and closure of the digits to embrace the foodpellet and an extension of the wrist. During these movements,the limb is not withdrawn. A rat contacts the pellet using eitherdigit 3 or 4. If the pellet is contacted first by the pad of digit3, then the grasp occurs between the digits 3 and 4. If the rattouches the pellet with the pad of digit 4, the rat grasps the pel-let between digits 4 and 5. In a previous study [33], grasping ofsmall food pellets was observed between digits 4 and 5, but obser-vations using smaller food pellets were not made in the presentstudy.

3.2.6.2. X-ray. X-ray images (Fig. 9, middle and bottom) confirmthat the movement of grasp is achieved by flexion and closure ofthe phalanges and metacarpals over the food and that grasping isaccompanied by wrist extension. The X-ray image also indicatedthat wrist extension is accompanied by a slight flexion of the radio-ulnar/humerus joint and a slight flexion of the shoulder joint. TheX-ray images were not sufficiently clear to indicate whether therewere independent movements of the digits with grasping. An anal-ysis of this aspect of the grasp would require close up filming andhigher speed filming.

3.2.7. Supination I3.2.7.1. Video. The movement of supination I (Fig. 10, top) involvesrotation of the paw by 90◦ around the wrist and by adduction ofthe elbow to position the paw palm in a vertical orientation. Thismovement begins as soon as the grasp is complete and it is complete

before the paw is withdrawn through the slot.

3.2.7.2. X-ray. X-ray images (Fig. 10, middle and bottom) confirmvideo images in that the paw rotates to vertical orientation (90◦)as it is withdrawn to the slot. Supination I appears to be producedby at least two movements of the arm: first there is flexion at theshoulder and at the radio-ulnar/humerus joint and second thereis an adduction of the elbow. Additionally, there is adduction atthe shoulder that moves the paw laterally so that it is positioneddirectly in front of the slot.

3.2.8. Supination II3.2.8.1. Video. The movement of supination II (Fig. 11, top) involvesrotation of the paw so that the paw palm faces the mouth.The movement is mainly achieved by rotation of the pawaround the wrist and by rotation of lower arm at the elbowjoint.

3.2.8.2. X-ray. X-ray images (Fig. 11, middle and bottom) confirmthat the rat supinates the paw toward the mouth using two move-ments: flexion at the shoulder and movement at the elbow joint.

in Research 192 (2008) 232–247

X-ray recordings show that at the end of the movement of supina-tion II, the elbow angle and it’s position are similar to what is seenat the end of the movement of “aim” but now with the paw inventroflexed position rather vertical.

3.2.9. Food release3.2.9.1. Video. The movement of food release consists opening ofthe digits and the releasing the food pellet into the mouth (Fig. 12,top). Because on the video images the paw covers the mouth, theway in which the mouth grasps the food is unclear.

3.2.9.2. X-ray. X-ray images (Fig. 12, middle and bottom) confirmthat during food release, the phalanges are extended and the foodpellet is placed into the mouth. At this time, there appears to bea slight flexion at the wrist accompanying digit extension. The X-ray record also shows that the opening of the mouth accompaniesthe opening of the paw. The pellet is taken either by the incisorsor it is lapped out of the paw using the tongue. After either grasp,the tongue sweeps the food pellet backward for chewing with themolars.

3.2.10. Replacing the limb3.2.10.1. Video. The movement of replacing the paw on the floorinvolves releasing the paw from the mouth and then placing it onfloor surface (or sometimes the rat may rear and place the paw onthe wall). When the paw touches the surface of the table it does sowith digit 5 contacting the floor followed by the other digits in anarpeggio movement.

3.2.10.2. X-ray. The X-ray images indicate that paw removal fromthe mouth and placing it on the floor involves two movements. First,the paw is removed from the mouth by flexion at the shoulder andby a slight flexion at the radio-ulnar/humeris joint. This positionsthe forearm at an angle horizontal to the floor surface. Second, thepaw is then placed on the floor by extension of the shoulder andwith little or no extension at the radio-ulnar/humerus joint. Pawplacing is frequently accompanied by lowering of the body, nev-erthless a clear temporal break between the two movements issometimes also observed.

4. Discussion

Skilled reaching for food by the rat has been extensively used asa model of a number of neurological conditions that affect arm and

hand movements in humans [37]. To this end, scoring systems havebeen developed to describe the movement and its impairments fol-lowing brain injury. A drawback in using descriptive methods is thatthe rat is small, the movement is performed quickly, and the move-ments of limb segments must be estimated through an overlying furand a skin that is pliable. Thus, joint locations are not always clearand they cannot be accurately located using external markers. Thepurpose of the present study was to examine conventional scoringmethods in the light of moderately high-speed cineradiographicrecords of rat reaching for single pellet. The results provide a num-ber of new insights in rat skilled reaching movements and so extentconventional descriptions made from frame-by-frame analysis ofvideo recordings.

Video and cineradiographic recordings gave very similar per-spectives of the rat reaching for food. In addition, the rats adaptedwell to the cineradiographic recording situation so that X-rayviews of skilled reaching were obtained from frontal, lateral, anddorsal perspectives. The high shutter speeds of video recordsin both methods also provided blur-free images of the animal’smovements. Success measures of performance in the two testingsituation indicated that the rats’ performance was comparable so

M. Alaverdashvili et al. / Behavioural Bra

Fig. 12. Release. Video and X-ray images of the start and the end of release. Note:(Frontal view) the digits extends and opens independently; (lateral view) the pelletis grasped with incisors. (Bottom, solid outline: the start of lift; dotted outline: theend of the action.)

that the movements used in the two situation were likely much thesame.

The first and most obvious insight into the rat’s movementsrelate to the movement of the animal’s head during orientingtoward the food pellet. When reaching for food, the rat locates thefood by sniffing and so must orient its head to detect whether food

in Research 192 (2008) 232–247 245

is on the shelf. If food is present, it must raise its head to advance thepaw toward the food. From the conventional video recording, thevertebral column from the shoulders to the head appears horizon-tal and it is not clear whether the trunk assists head movementor whether the head is moved independently of the trunk dur-ing the reaching action. The cineradiographic record revealed thatduring head orienting the vertebral column has S-shaped config-uration, with the head positioned upon a proximal curvature thatapproached the perpendicular, and not horizontal as suggested bythe macroscopic appearance of the neck. Previously, Vidal et al.[24,25] have described S-shaped skeleton topography in rats andmice, and other quadruped animals during the resting posture.Recently, the S-shaped vertebral configuration was observed in therabbit and mouse during exploration and locomotion, respectively[13,14,25]. Here, we extend these previous findings by showing thatthe rats maintain S-shaped vertebral configuration during move-ment and we show that this position allows head movement at theatlanto-occipital and cervico-thoracic junctions independent of thetrunk both when a rat orients to locate food and when head is raisedto advance the limb to food.

A relatively stable trunk position and independent head motionprobably provides energy saving and balanced movement with areduced number of degrees of freedom that the central nervoussystem needs to control [2]. The S-shaped vertebra configurationand relative independence of the head and body movements con-tribute to structuring of movements in the geocentrically orientedspace (parasagittal plane). Willock and Pearson [38] argue that theS-shaped configuration is a favorable configuration for scanningthe horizon. This notion becomes more important for species thatmake orienting movements guided by senses other than vision. Asthe reaching movement in rats is guided by olfaction [31,34], inde-pendent movement of the head and stabilization of body might beimportant for the organization of reaching movement with respectto the mid-sagittal plane.

X-ray images confirmed that midline of the body often servesas reference plane for limb and limb segments during reach-ing as described by Whishaw and Pellis [31] from video-images.When the rat lifts the paw, the digits are aligned to midline;when it aims, the elbow and radio-ulnar axis are brought to theparasagittal plane. Probably information provided from sensorygraviceptors/proprioceptors to central nervous system are essen-tial to guide motor actions in the environment and determines thepath structure of forelimb during reaching in rats. This may be inpart because the rat does not monitor its reach using vision and

so reaches as does a human in the absence of visual feedback [10].Movement of the limb along the parasagittal plane is additionallyuseful for directing the paw through the slow toward the food.

Cineradiographic images confirmed that skilled reaching com-prises a set of segmental movements but they are somewhat morecomplex that has been revealed by EWMN description based onvideo recording. Nevertheless, confirming EWMN, reaching wasachieved via movements emerging sequentially: (1) orient to locatethe food pellet; (2 and 3) limb lift with the digits to midline and dig-its semiflexed; (4) aim with elbow in; (5 and 6) advance with thedigits extend; (7) pronation/arpeggio; (8) grasp; (9) supination I;(10) supination II; (11) release the food into the mouth and (12)replacing the limb on the floor or the wall of the reaching box [32].

Cineradiographic analysis clarifies the intrinsic organiza-tion of these movements. The X-ray images clearly indicate aproximo-distal organization of reaching action reminiscent of theproximo-distal fractionation of the “body scheme” in humans [20].The proximal end of the limb generated almost all movement com-ponents associated with transport and withdrawal. For example,lift, aim and advance were initiated from the shoulder; supina-tion I/II, and replacing the paw was largely dependent on the

ral Bra

[

[

[

246 M. Alaverdashvili et al. / Behaviou

movements at the shoulder joint. The specific role of the shoul-der segment in initiation of reaching action supports descriptionsof muscle activity during reaching action in rats [16]. Hyland andJordan [16] show that lattisumus dorsi activity (a shoulder flexor)is the earliest muscle event in reaching and contributes to raisingthe paw from the ground. The X-ray images also identified how thescapula and humerus head, tethered by muscles, participate in thereaching action. The X-ray findings showing that there are a numberof concurrent movements at joints during each limb action are thusin line with electromyographic findings that most limb muscles areactive throughout a reaching movement [16].

Video images do show that movements do take place in the moredistal portion of the limb and these involved rotation at the wristand digit opening and closing for grasping. This was confirmed fromthe X-ray images. In primates, rotation of the paw is contributed toby rotation of the radius and ulna, but in the rat these bones arefused. The X-ray images also indicate that paw rotation did not relyonly on movement at the wrist but was contributed by movementsof the upper arm.

Because the rat supinates its paw to release food to the mouth,and covers the mouth with the paw in doing so, it is unclear fromvideo records how the food is grasped by the mouth. The X-rayimages show that with the paw adjacent to the mouth, the digitsare partially extended and the rat either grasps the food with theincisors or it laps the food out of the paw using the tongue. Then,it manipulates the food with the tongue toward the molar teethfor chewing and subsequent swallowing. Thus, it is clear that therat does not place the food in the mouth or grasp the food withits lips, as do primates. It will be interesting to examine whetherthis “grasping by the teeth” and “lapping by the tongue” actionsare affected in representative animal models of neurologicalconditions.

5. Conclusion

Cineradiography during forelimb skilled reaching for a food pel-let to place it into the mouth extends the taxonomic analysis ofreaching movement derived from conventional video-image anal-ysis in rats. Cineradiographic analysis provides new findings withrespect to skeleton topography, head movement and the involve-ment of scapula and humerus during skilled reaching. BecauseX-ray analysis indicates that forelimb and paw movements are more

complex than is suggested by video recording, X-ray video-imagingopens new perspectives in understanding general principal ofskeleton movement in rats and other mammals. Thus, it can be avaluable tool for investigating motoric deficits after brain injury andfor understanding the evolution and organization of the motor sys-tem. Because the contribution of motor system injury to changes inmouth and tongue use in taking food from the paw and in eating arenot easily obtained from conventional video recording, X-ray anal-ysis also provides a new vista for exploring oral functions. Finally,X-ray imaging shows that the rodent head can make a numberof movements independently or in conjunction with trunk move-ment, which suggests that the effects of motor system changes onorienting can be studied using X-ray images.

Acknowledgements

The research was supported by grants from the Natural Sciencesand Engineering Research Council of Canada (IQW) and CanadaFoundation for Innovation (SR and HL). Authors wish to thank Mau-rice Tremblay for helpful advises during ceneradiograpic filmingand Maryanne Tran for her assistance in behavioral training andfilming the rats.

[

[

[

[

[

[

[

[

[

[

in Research 192 (2008) 232–247

References

[1] Alaverdashvili M, Foroud A, Lim DH, Whishaw IQ. “Learned baduse” limitsrecovery of skilled reaching for food after forelimb motor cortex stroke inrats: a new analysis of the effect of gestures on success. Behav Brain Res2008;188:281–90.

[2] Bernstein NA. O postroyenii dvizheniy (on the construction of movements).Moscow: Medgiz; 1947 (English translation: the coordination and regulationof movements. Oxford, New York: Pergamon; 1967).

[3] Biernaskie J, Corbett D. Enriched rehabilitative training promotes improvedforelimb motor function and enhanced dendritic growth after focal ischemicinjury. J Neurosci 2001;21:5272–80.

[4] Buczek-Funcke A, Kuhtz-Buschbeck JP, Paschmeyer B, Illert M. X-ray kinematicanalysis of forelimb movements during target reaching and food taking in thecat. Eur J Neurosci 2000;12:1817–26.

[5] Castro AJ. Limb preference after lesions of the cerebral hemisphere in adult andneonatal rats. Physiol Behav 1977;18:605–8.

[6] Eshkol N, Wachmann A. Movement notation. London: Weidenfeld and Nichol-son; 1958.

[7] Farr TD, Whishaw IQ. Quantitative and qualitative impairments in skilled reach-ing in the mouse (Mus musculus) after a focal motor cortex stroke. Stroke2002;33:1869–75.

[8] Fischer MS, Schilling N, Schmidt M, Haarhaus D, Witte H. Basic limb kinematicsof small therian mammals. J Exp Biol 2002;205:1315–38.

[9] Fricker-Gates RA, Smith R, Muhith J, Dunnett SB. The role of pretraining onskilled forelimb use in an animal model of Huntington’s disease. Cell Transplant2003;12:257–64.

[10] Ghafouri M, Lestienne FG. Contribution of reference frames for movementplanning in peripersonal space representation. Exp Brain Res 2006;169:24–36.

[11] Gharbawie OA, Whishaw IQ. Parallel stages of learning and recovery of skilledreaching after motor cortex stroke: “oppositions” organize normal and com-pensatory movements. Behav Brain Res 2006;175:249–62.

12] Girgis J, Merrett D, Kirkland S, Metz GA, Verge V, Fouad K. Reaching trainingin rats with spinal cord injury promotes plasticity and task specific recovery.Brain 2007;130:2993–3003.

[13] Graf W, de Waele C, Vidal PP, Wang DH, Eviger C. The orientation of the cervicalvertebral column in unrestrained awake animals. II. Movement strategies. BrainBehav Evol 1995;45:209–31.

[14] Graf W, Waele DE, Vidal PP. Functional anatomy of the head–neck movementsystem of quadrupedal and bipedal mammals. J Anat 1995;186:55–74.

[15] Herbin M, Hackert R, Gasc JP, Renous S. Gait parameters of treadmill versusoverground locomotion in mouse. Behav Brain Res 2007;181:173–9.

[16] Hyland BI, Jordan VMB. Muscle activity during forelimb reaching movement inrats. Behav Brain Res 1997;85:175–86.

[17] Iwaniuk AN, Whishaw IQ. On the origin of skilled forelimb movements. TrendsNeurosci 2000;23:372–6.

[18] Monfils MH, Teskey GC. Skilled-learning-induced potentiation in rat. Sen-sorimotor cortex: a transient form of behavioral long-term potentiation.Neuroscience 2004;25:329–36.

[19] Muir GD, Webb AA, Kanagal S, Taylor L. Dorsolateral cervical spinal injurydifferentially affects forelimb and hindlimb action in rats. Eur J Neurosci2007;25:1501–10.

20] Paillard J. Motor and representational framing space. In: Paillard J, editor. Brainand space. Oxford: Oxford University Press; 1991. p. 163–82.

21] Peterson GM, Francarol C. The relative influence of the locus and mass of dis-truction upon the control of handedness by the cerebral cortex. J Comp Neurol

1951;68:173–90.

22] Schallert T, Kozlowski DA, Humm JL, Cocke RR. Use-dependent structural eventsin recovery of function. Adv Neurol 1997;73:229–38.

23] Schmidt M. Quadrupedal locomotion in squirrel monkeys (Cibedae: Saimirisciureus): a cineradiographic study of limb kinematics and related substratereaction forces. Am J Phys Anthropol 2005;128:359–70.

24] Vidal PP, Degallaix L, Josset P, Gasc JP, Cullen KE. Postural and locomotor controlin normal and vestibularly deficient mice. J Physiol 2004;559:625–38.

25] Vidal PP, Graf W, Berthoz A. The orientation of the cervical vertebral columnin unrestrained awake animals. I. Resting position. Exp Brain Res 1986;61:549–59.

26] Weijs WA. Mandibular movements of the albino rat during feeding. J Morphol1975;145:107–24.

27] Whishaw IQ. Loss of the innate cortical engram for action patterns used inskilled reaching and the development of behavioral compensation followingmotor cortex lesions in the rat. Neuropharmacology 2000;39:788–805.

28] Whishaw IQ. Prehension. In: Whishaw IQ, Kolb B, editors. The behavior of thelaboratory rat: a handbook with tests. Oxford: Oxford University Press; 2005.p. 62–71.

29] Whishaw IQ, Gorny B. Arpeggio and fractionated digit movements used inprehension by rats. Behav Brain Res 1994;60:15–24.

30] Whishaw IQ, Miklyaeva EI. A rat’s reach should exceed its grasp: analysis ofindependent limb and digit use in the laboratory rat. In: Ossenkop K, Kavaliers P,Sanberg PR, editors. Measuring movement and locomotion: from invertebratesto humans. New York: RG Landes Co.; 1996. p. 135–69.

31] Whishaw IQ, Pellis S. The structure of skilled forelimb reaching in the rat: aproximally driven movement with a single distal rotatory component. BehavBrain Res 1990;41:49–59.

[

[

[

[

M. Alaverdashvili et al. / Behavioural Bra

32] Whishaw IQ, O’Connor WT, Dunnett SB. Contributions of motor cortex, nigros-triatal dopamine and caudate–putamen systems to skilled forepaw use in therat. Brain 1986;109:805–43.

33] Whishaw IQ, Pellis S, Gorny BP, Pellis VC. The impairments in reaching andthe movements of compensation in rats with motor cortex lesions: an end-point, videorecording, and movement notation analysis. Behav Brain Res1991;42:77–91.

34] Whishaw IQ, Pellis S, Gorny BP. Skilled reaching in rats and humans: evidencefor parallel development or homology. Behav Brain Res 1992;47:59–70.

35] Whishaw IQ, Pellis S, Gorny BP, Kolb B, Tetzlaff W. Proximal and distal impair-ments in rat forelimb use in reaching follow unilateral pyramidal tract lesion.Behav Brain Res 1993;56:59–76.

[

[

[

in Research 192 (2008) 232–247 247

36] Whishaw IQ, Gorny B, Foroud A, Kleim JA. Long-Evans and Sprague–Dawleyrats have similar skilled reaching success and limb representations in motorcortex but different movements: some cautionary insights into the selectionof rat strains for neurobiological motor research. Behav Brain Res 2003;145:221–32.

37] Whishaw IQ, Piecharka DM, Zeeb F, Stein DG. Unilateral frontal lobecontusion and forelimb function: chronic quantitative and qualitative impair-ments in reflexive and skilled forelimb movements in rats. J Neurotrauma2004;21:1584–600.

38] Willock C, Pearson J. Predators of the wild, vol. 6. Burbank, CA, USA: WarnerHouse Video; 1992.