Download State-Dependent TMS Reveals a Hierarchical

Cerebral Cortex September 2010;20:2252--2258
doi:10.1093/cercor/bhp291
Advance Access publication January 4, 2010
State-Dependent TMS Reveals a Hierarchical Representation of Observed Acts in the
Temporal, Parietal, and Premotor Cortices
Luigi Cattaneo1, Marco Sandrini1 and Jens Schwarzbach1,2
1
Neuroimaging Laboratories, Center for Mind/Brain Sciences and 2Department of Cognitive Sciences and Education, University of
Trento, 38100 Mattarello (Trento), Italy
Address correspondence to Luigi Cattaneo, Neuroimaging Laboratories, Center for Mind/Brain Sciences, University of Trento, Via delle Regole, 101,
38100 Mattarello (Trento), Italy. Email: [email protected].
A transcranial magnetic stimulation (TMS) adaptation paradigm
was used to investigate the neural representation of observed
motor behavior in the inferior parietal lobule (IPL), ventral premotor
cortex (PMv), and in the cortex around the superior temporal sulcus
(STS). Participants were shown adapting movies of a hand or a foot
acting on different objects and were asked to compare to the
movie, a motor act shown in test pictures. The invariant features
between adapting and test stimuli fitted a 2 3 2 design: same or
different action made by the same or different effector. Neuronavigated TMS pulses were delivered at the onset of each test
picture. TMS over the left and right PMv and over the left IPL
induced a selective shortening of reaction times (RTs) to stimuli
showing a repeated (adapted) action, regardless of the effector
performing it. In a second experiment, TMS applied over the left
STS induced shortening of RTs for adapted actions but only if also
the effector was repeated. The results indicate that observed motor
behavior is encoded with the body part that performs it in the
temporal lobe. A hierarchically higher level of representation is
carried by neural populations in the parietofrontal regions, where
acts are encoded in an abstract way.
Keywords: action observation, adaptation, hierarchy, magnetic stimulation,
premotor
Introduction
It is a well-developed capacity in humans to extract effortlessly
the meaning of an observed motor act from the visual scene
that is presented to them. For example, others’ behavior is
immediately categorized as ‘‘grasping’’ regardless either of the
numerous possible kinematic variants of this movement or
even of the body part (e.g., a right hand, a left hand, or a foot)
that is used to perform the act. Neuroimaging studies have
consistently shown that the observation of an individual acting
upon an object activates in the observer an interconnected
system of cortical nodes represented by extrastriate visual
areas, the superior temporal sulcus (STS), and a parietofrontal
system consisting of the intraparietal sulcus (IPS) and inferior
parietal lobule (IPL) plus the ventral premotor cortex (PMv)
and caudal part of inferior frontal gyrus (IFG). In some
instances also, the superior parietal lobule (SPL) and the dorsal
premotor cortex have been shown to be activated by
observation of motor behavior (for a review, see Cattaneo
and Rizzolatti 2009).
It has been proposed that the representation of observed
actions is hierarchically organized (Grafton and Hamilton 2007).
According to this view at some point of the action observation
system, a transformation between the infinite physical variants of
a movement into its meaning must occur, that is, a transformation
Ó The Author 2010. Published by Oxford University Press. All rights reserved.
For permissions, please e-mail: [email protected]
into a hierarchically higher and more abstract representation of
observed acts. Some studies have shown a capacity of abstraction
of observed acts by neural populations in the inferior parietal
and frontal cortices, demonstrating in some instances also
a lower hierarchical level of processing by extrastriate visual
areas (Hamilton and Grafton 2006, 2008; Grafton and Hamilton
2007; Urgesi et al. 2007; Candidi et al. 2008). On the other hand,
other studies have reported representation of low-level features
of observed actions within the frontoparietal system. Among
others is, for example, the presence of spatially segregated
representations of different effectors (Buccino et al. 2001) or the
evidence of the processing of kinematics of the observed motor
acts (Pobric and Hamilton 2006; Suchan et al. 2008). In summary,
there is no agreement on what respective functional roles the
different nodes of the action observation system play in
recognizing motor behavior. Whereas some authors hypothesize
that action understanding takes place in the ventral part of the
dorsal stream (Rizzolatti and Matelli 2003), others claim that
actions are fully recognized and categorized outside the motor
system, in the ventral stream (Mahon and Caramazza 2008).
In order to further investigate the relative contribution of
the distinct ventral and dorsal nodes of the action observation
system, we applied a novel approach to TMS, the transcranial
magnetic stimulation adaptation (TMSA) paradigm, (Silvanto
et al. 2008) to action observation. TMSA is able to provide
information on the cortical topography of brain functions and
the causal relation of neural activity in the targeted areas to
behavior. The TMSA paradigm is based on the well-established
idea that the effects of TMS are state dependent. Specifically,
TMS behaviorally facilitates the attributes encoded by
adapted neural populations, compared with nonadapted,
within the stimulated brain area (Silvanto et al. 2008). In
the TMSA paradigm, the state of the cortex prior to the
stimulus is manipulated in a controlled way by means of
perceptual adaptation. The adapting stimulus induces habituation in a subset of cells that code particular stimulus
features, making them a selective target for TMS. Stimulation
time locked to the cognitive task delivered over the cortical
area containing the adapted neurons should selectively
improve the performance in processing the adapted stimulus.
This TMS paradigm therefore allows targeting functionally
distinct neuronal pools in spite of their spatial overlap with
respect to each other or with other neural populations. The
TMSA paradigm has been tested with stimulation of visual
areas such as V1/V2 and V5/middle temporal whose
properties are relatively well known (Silvanto and Muggleton
2008). Also, a recent study applied this paradigm to the
language domain revealing abstract letter selectivity in the
left posterior parietal cortex (Cattaneo, Rota, et al. 2009).
In the present experiments, we used the repeated visual
exposure to the same motor act repeated in time as adapting
stimulus, which participants had to match to test stimuli. We
considered as an index of abstractness of action representation
the presence of cross-adaptation between vision of the same
motor behavior but performed by different effectors.
Materials and Methods
Participants
Fourteen healthy participants (7 females, mean age 25 years) with no
history of neurological or psychiatric illness took part in Experiment 1
and 9 (5 females, mean age 28 years) took part in Experiment 2. All
participants were right handed according to the Edinburgh Inventory.
The experiments were undertaken with the understanding and written
consent of each subject, with the approval of the appropriate local
ethics committee, and in compliance with national legislation and the
Code of Ethical Principles for Medical Research Involving Human
Subjects of the World Medical Association (Declaration of Helsinki).
Setting and Protocol
Participants were sitting in an armchair with their head on a chin-rest
in front of a 17-inch computer screen at viewing distance of 57 cm. All
participants wore earplugs. The width of stimuli corresponded to 20°
of visual angle. Trials consisted in the presentation of an adapting movie
followed by the presentation of a series of 8 static test pictures (Fig. 1).
Each trial began with a red fixation cross appearing in the middle of the
screen for 15 s on a gray background. Participants were required to
watch carefully the adapting movie that showed repeatedly a single
motor act performed by a hand or a foot. They were then required to
respond as fast and as accurately as possible to static test pictures
whether the depicted motor act was same or different from the one
seen in the adaptation movie. Responses were made with the index and
middle finger of the right hand on a keyboard. Response mapping (same
Figure 1. Schematization of a single trial: a 15-s fixation cross preceded the
adapting movie in which the same motor act made by the same effector was shown
for 40 times at a 1-Hz rate. Eight test pictures were then presented in random
sequence. When applied, TMS was delivered at the onset of each test picture.
or different) to each finger was randomized between participants. They
were explicitly told to ignore the effector and make a judgment on the
type of act only. The visual presentation of the test picture was
terminated by a manual response of the participant or, in case of no
response, ended by default after 1500 ms. Each trial (adapting movie
plus 8 test pictures) was repeated 16 times within each block, and
therefore, a total of 8 3 16 = 128 single responses were collected in
each block. A total of 6 blocks were carried out for each participant.
The first one was always the baseline block, without TMS. In each of the
remaining 5 blocks, one site of TMS was applied. The order of these 5
blocks was pseudorandomized between subjects. We used the E-Prime
version 2.0 (Psychology Software Tools, Inc.) software for stimulus
presentation and reaction time (RT) recordings. The experiment was
preceded by a brief training session.
Adapting and Test Stimuli
Adapting movies contained a series of 40 consecutive clips of 1 s in
duration, showing either a right hand or right foot that were either
grasping or pushing an object. Half of the 40 clips showed a male actor
and the other half a female actor. The manipulated objects were of 2
possible shapes (a cube and a ball, both 3 cm in diameter). All features
were ordered pseudorandomly such that the only invariant features
between 1-s clips belonging to the same adaptation movie were 1) the
type of effector (a hand or a foot) and 2) the type of motor act
performed upon the object. This resulted in 4 adapting movies (hand/
grasp; hand/push; foot/grasp; foot/push), which were repeated 4 times
in every block. The images in test pictures were single frames extracted
from the adaptation movies. The frames corresponded to the moment
of contact with the object. The test pictures represented, compared
with the adapting movie, a different act performed by a different
effector, the same act performed by a different effector, a different act
performed by the same effector, or the same act performed by the same
effector (Fig. 1). Pictures of each of these 4 types were presented twice
following an adapting movie, for a total of 8 test pictures every trial.
Transcranial Magnetic Stimulation
Biphasic TMS pulses were applied with a figure-of-eight coil (MC-B70)
and a MagPro 3100 stimulator (MagVenture A/S, Denmark). Before the
experiment, the individual ‘‘visible’’ resting motor excitability threshold
of stimulation was established as the lowest stimulation intensity
applied over the primary motor cortex capable of evoking a visible
contraction in the relaxed right first dorsal interosseous muscle on at
least 4 out of 8 consecutive stimulations. The stimulation intensity used
during the experiment was set at 110% of the individual threshold. The
magnetic stimulator was triggered by Eprime through the parallel port.
Single stimuli were delivered at the onset of each test picture. The coil
was attached to a mechanical arm fixed to a tripod and placed
tangentially to the skull. The coil position was monitored online with
the BrainVoyager (Brain Innovation BV, The Netherlands) neuronavigation system and adjusted to the target location based on
reconstructions of individual brain anatomies. Prior to the experiment,
a high-resolution T1-weighted magnetization-prepared rapid gradient
echo sequence (176 axial slices, in-plane resolution 256 3 224, 1-mm
isotropic voxels, generalized autocalibrating partially parallel acquisition
with acceleration factor = 2, time repetition = 2700 ms, time echo =
4.180 ms, time to inversion = 1020 ms, flip angle = 7°) scan of the
brain of each subject was obtained using a MedSpec 4-T head scanner
(Bruker BioSpin, Ettlingen, Germany) with an 8-channel array head coil.
Experiment 1
Six blocks were carried out in every participant, each with a different
TMS condition: 1) baseline—no TMS; 2) TMS over midline SPL; 3)
TMS over left supramarginal gyrus (SMG); 4) TMS over left PMv; 5) TMS
over right SMG; and 6) TMS over right PMv. The no-TMS block was
always run first. The order of the blocks with TMS was randomized
between subjects. Stimulation sites were identified on individual brain
reconstructions on the basis of macroanatomical landmarks. PMv
was defined as the portion of the precentral gyrus posterior to the
point where the inferior frontal sulcus meets the precentral sulcus
Cerebral Cortex September 2010, V 20 N 9 2253
(Tomassini et al. 2007). The SMG was targeted over its dorsalmost
portion, defined as the part of the SMG immediately ventral to the IPS
and dorsal to the caudal end of the posterior branch of the sylvian
fissure. SPL was targeted on a point over the midline halfway between
the postcentral sulcus and the caudal end of the IPS. The locations on
a representative brain are shown in Figure 2.
Coil orientation was anteroposterior with the handle pointing
backward for PMv, perpendicular to the midline with the handle
pointing outward for the SMG, and anteroposterior with the handle
pointing backward for the SPL position. To test for current spread to
adjacent motor areas, participants were monitored off-line before the
experiment for any muscular twitch in the hand, lips, jaw, or tongue
time locked with the stimulus.
Experiment 2
Experiment 2 was designed to test the action observation node of the
temporal lobe, that is, the cortex associated with the posterior part of
the STS. It was constructed after the results of Experiment 1 had been
analyzed, and therefore, a no-TMS condition was not carried out.
Instead, the midline SPL stimulation was used as control. Three blocks
were carried out in every participant, each with a different TMS
condition: 1) TMS over the left STS, 2) TMS over midline SPL, and 3)
TMS over left PMv; the order of the TMS conditions was randomized
between subjects. The cortex around STS was targeted over its middle
portion by dividing in 2 parts the STS from the temporal pole to the end
of its posterior ascending branch in the parietal lobe and positioning
the coil over the midpoint in a site roughly corresponding to where
most activation foci are found for observation of transitive distal
biological movements (Puce and Perrett 2003). An example of coil
location is shown in Figure 2. Coil orientation was initially vertical with
the handle pointing upward; however, due to the proximity of the ear,
the coil orientation was changed in some participants for the purpose
of minimizing discomfort and improving the contact of the TMS coil
with the scalp. The midline SPL and the left PMv positions were
localized as in Experiment 1.
Data Processing and Analysis
Analysis was conducted on RTs and error rates. Data from single trials
with incorrect responses were discarded. Subsequently, RTs below 300
ms were excluded from analysis. As stated above, responses longer than
1500 ms were not logged by the system. In Experiment 1, the number
of trials discarded because falling outside the 300- to 1500-ms time
interval corresponded to 0.5% of all trials, ranging in all cells from 0% to
1.1%, and in Experiment 2, they corresponded to 0.6% of all trials,
ranging in all cells from 0% to 1.2%.
RTs were grouped and averaged according to 3 different factors: 1)
stimulation modality, 2) adaptation of action, and 3) adaptation of
effector. Each measure was obtained by averaging of a pool of 32 single
RTs (minus number of excluded error trials). In each of the 2
experiments, an analysis of variance (ANOVA) was performed with the
3 within-subjects variables: ‘‘stimulation modality’’ (6 levels in Experiment 1: no TMS, SPL, left SMG, left PMv, right SMG, and right PMv and 3
levels in Experiment 2: left STS, SPL, and left PMv); ‘‘action adaptation’’ (2
levels in both experiments: adapted or nonadapted); and ‘‘effector
adaptation’’ (2 levels in both experiments: adapted or nonadapted). Post
hoc comparisons were made with 2-tailed t-tests for paired data, and the
P value was Bonferroni corrected for the number of comparisons. The
same analyses were performed on raw error rates for both experiments.
Results
Experiment 1
None of the subjects experienced significant side effects from
TMS aside from transient low-intensity neck pain in one case
with an onset after the experiment and a moderate--severe
frontal headache in another case that was reported during
stimulation of the right PMv position.
A significant main effect of ‘‘TMS condition’’ was found
(F5,65 = 23.000, P < 0.00001). RTs in all the conditions with TMS
were faster than RTs in the no-TMS condition. These RT
enhancements are probably due to nonspecific factors such as
auditory intersensory facilitation of the click produced by the
magnetic stimulator at the onset of the test pictures
(Hershenson 1962; Marzi et al. 1998; Bien et al. 2009). A
significant main effect of action adaptation was also found
(F1,13 = 12.575, P = 0.004) with RTs to adapted actions being
overall faster than those to nonadapted actions. However, the
most important finding was a significant interaction of the 2
factors TMS condition and action adaptation (F5,65 = 3.7055, P =
0.005). No other effect or interaction was found (all P values >
0.42). In particular, no significant 3-way interaction was
observed (F5,65 = 0.99, P = 0.43).
Significance level for post hoc comparisons was set to
0.0083. The post hoc comparisons investigating the TMS
condition 3 action adaptation showed a significant difference
between nonaction-adapted and action-adapted stimuli only in
3 TMS conditions: left SMG (P = 0.006), left PMv (P = 0.0005),
and right PMv (P = 0.008). The mean values of the interaction
are given in Table 1, whereas Figure 3 shows the mean
differences between adapted action and nonadapted action
conditions. No significant effect of any factor was found in the
ANOVA on error rates (all P values > 0.33). The grand mean of
the error rate across all blocks and participants was of 5.7%.
Table 1
RTs (ms) and error rates (%) from Experiment 1
Figure 2. Depiction on a representative brain surface of the macroanatomical
landmarks used for coil positioning in the 2 experiments. The sulci are represented as
thick white lines. IFS, inferior frontal sulcus; PrCS, precentral sulcus; CS, central
sulcus; PoCS, postcentral sulcus; PBSF, posterior branch of the sylvian fissure. The
putative stimulation spots are indicated as white circles.
2254 TMS Adaptation in Action observation
d
Cattaneo et al.
No TMS
SPL
Left SMG
Left PMv
Right SMG
Right PMv
Same action
Different action
P value
672
569
541
540
563
534
679
575
576
589
581
573
0.35
0.69
0.0062
0.0005
0.12
0.0080
(78)/5.2
(64)/6.0
(55)/6.4
(60)/5.5
(80)/7.0
(69)/5.7
(80)/3.9
(78)/6.3
(66)/6.5
(69)/4.9
(77)/5.8
(77)/5.4
Note: Values represent mean RTs (standard deviation)/error rate. P values refer to t-tests
between RTs to same actions and RTs to different actions. Significance level is set to P 5
0.0083. Significant P values are highlighted in bold.
Experiment 2
Three subjects reported moderate pain associated with
stimulation of the left STS position, which was solved by
varying the orientation of the coil. Overall, the STS stimulation
site was surprisingly painless for such a lateral scalp position.
A significant interaction of the 2 factors TMS condition and
action adaptation (F2,16 = 4.2215, P = 0.034) was found.
However, the most relevant finding was a significant 3-way
interaction of the factors TMS condition, action adaptation, and
effector adaptation (F2,16 = 3.9755, P = 0.039). No other
significant effect was found (all P values > 0.24). Significance
level for post hoc t-tests was set to 0.0083. The post hoc
analysis of this interaction showed the following pattern: with
left PMv TMS, the RTs to adapted actions were faster than to
nonadapted actions regardless of the effector being the same or
different (P = 0.0080 and P = 0.0009, respectively). With SPL
stimulation, no difference was observed between adapted and
nonadapted actions. For TMS over the left STS, responses to
adapted actions were significantly faster than to nonadapted
actions only if also the effector was repeated (P = 0.0051). The
mean values of the interaction are shown in Table 2. Figure 4
shows the mean differences between adapted action and
nonadapted action conditions. No significant effect of any
factor was found in the ANOVA on error rates (all P values >
0.24). The grand mean of the error rate across all blocks and
participants was of 5.2%.
Discussion
The results of the present experiments demonstrate that TMS
applied over a cortical area produces different effects, depending on the quality of visual experience to which the subjects
have been previously exposed. In particular, we observed that
TMS induced an improved performance in tasks involving the
analysis of features to which participants have been previously
adapted. Such pattern of findings suggests that TMS has
specifically enhanced performance of the neural subpopulations that respond to the invariant features between the
adapting stimulus and the test stimulus.
The concept that repeated visual exposure to motor
behavior can produce perceptual aftereffects in the observer
is not surprising. This has been shown in one study in both
humans and in single cells of monkeys using grasping or
placing acts (Barraclough et al. 2009). Another study showed
the possibility of inducing an effect compatible with priming
to the meaning of the act with visual presentation of motor
acts (Costantini et al. 2008). Imaging studies have also shown
that several brain structures have adaptation effects in the
form of repetition suppression for observed motor behavior
(Chong et al. 2008; Kilner et al. 2009; Lingnau et al. 2009). In
the present experiment, a clear correlate of adaptation was
not evident behaviorally as an increase in reaction times to
repeated (adapted) stimuli in the no-TMS session. However,
the clear state dependency of the effects induced by TMS
shows that we did induce a ‘‘state change’’ in the subjects’
perceptual system by means of visual exposure to acts even
in the absence of clear behavioral effects without TMS. It has
been already proposed that state-dependent effects of TMS
can be obtained without a behavioral effect of adaptation in
the no-TMS condition. TMS state-dependent effects have
been obtained in spite of a weak behavioral effect not
present in all the subjects or not present at all (Cattaneo
et al. 2008; Cattaneo, Rota, et al. 2009; Cohen Kadosh and
Walsh 2009). Also, in the case of adaptation to color, the
effects of TMS have been shown to be present several
Figure 3. Experiment 1: mean values of the individual differences between adapted
action trials and nonadapted action trials for all TMS conditions. The values represent
the behavioral effects of adaptation. Positive values indicate a relative cost of
adaptation in terms of reaction time, and negative values indicate a relative
advantage of adaptation. Error bars indicate 95% confidence intervals. Conditions with
significant P values of post hoc comparisons are labeled with an asterisk.
Table 2
RTs (ms) and error rates (%) from Experiment 2
TMS condition Same effector
Same action
STS
SPL
Left PMv
Different effector
Different action P value Same action Different action P value
573 (100)/5.1 612 (101)/5.9
588 (89)/4.6 580 (56)/5.7
531 (64)/4.1 574 (61)/5.3
0.0051 615 (91)/6.1 594 (83)/4.8
0.63
569 (60)/5.4 554 (59)/4.8
0.0081 540 (67)/5.7 574 (60)/5.1
0.080
0.012
0.0009
Note: Values represent mean RTs (standard deviation)/error rate. P values refer to t-tests
between RTs to same actions and RTs to different actions. Significance level is set to P 5
0.0083. Significant P values are highlighted in bold.
Figure 4. Experiment 2: mean values of the individual differences between adapted
action trials and nonadapted action trials for all TMS conditions. The values represent
the behavioral effects of adaptation. Positive values indicate a relative cost of
adaptation in terms of reaction time, and negative values indicate a relative
advantage of adaptation. Error bars indicate 95% confidence intervals. Conditions with
significant P values of post hoc comparisons are labeled with an asterisk. D-Eff,
different effector; S-Eff, same effector.
Cerebral Cortex September 2010, V 20 N 9 2255
minutes after the perceptual aftereffect has faded out
(Silvanto et al. 2007).
In Experiment 1, we clearly establish a causal role of the
parietofrontal nodes in the task that we tested, that is,
categorizing the semantics of 2 well-distinct types of distal
goal-directed motor behavior. In particular, we identified as
active nodes the left PMv, left SMG, and right PMv. Neuronal
populations in these 3 areas are able to code the type of act
(push vs. grasp), irrespective of physical features such as the
effector that performs them. Indeed, the only invariant feature
between adapting stimuli and test stimuli was the type of act.
The fact that TMS facilitated responses to the adapted action
regardless of the effector indicates that neuronal populations in
this region perform abstract coding of the motor act.
A role of the left PMv--IFG complex in the encoding of
observed actions has been hypothesized in previous works using
either a ‘‘virtual lesion’’ TMS approach (Pobric and Hamilton
2006; Avenanti et al. 2007; Urgesi et al. 2007; Candidi et al. 2008)
or describing real lesions in patients (Saygin, Wilson, Dronkers,
and Bates 2004; Moro et al. 2008; Pazzaglia et al. 2008). Also the
left SMG has been found to be active in action coding in imaging
experiments (Hamilton and Grafton 2006). It is not surprising
that we did not find an effect of right SMG stimulation. In the
literature, right-lateralized effects of action observation have
been found in tasks that contained features of prediction in time
of the outcome of the action or of the final intention of the agent
(Iacoboni et al. 2005; Hamilton and Grafton 2008; Liepelt et al.
2008). These tasks are clearly different from the one of the
present experiment, which consists in recognizing ‘‘what’’ an
observed agent is doing in that exact moment as opposed to
‘‘why’’ the agent is doing it (Boria et al. 2009). Aside from
imaging experiments, however, prior to the present study,
a causal role of the inferior parietal cortex in action coding has
been suggested only by human lesional studies (Heilman et al.
1982; Rothi et al. 1985). In these descriptions, apraxic patients
with parietal lesions failed to distinguish well-performed from
poorly performed gestures. They also failed to recognize the
meaning of pantomimes in both verbal and nonverbal tasks.
These descriptions in patients confirm the role of the inferior
parietal cortex in generalizing the type of action as suggested by
our data.
It has been suggested that the parietofrontal coding of motor
acts happens through a process of internal simulation of the
same observed acts (Gallese et al. 1996; Grafton 2009). In this
context, the results from Experiment 1 support the hypothesis
that motor representations within PMv and IPL are coded in
terms of goals rather than movements of body parts made to
achieve them. This capacity of the motor system to code action
meanings has been shown in animals and humans for both
execution of movements (Alexander and Crutcher 1990a,
1990b; Kakei et al. 2001; Fogassi et al. 2005; Umilta et al. 2008)
and their observation (Umilta et al. 2001; Gangitano et al. 2004;
Cattaneo, Caruana, et al. 2009). To this respect, our findings are
clearly supported by previous experiments investigating motor
resonance to observed actions by applying TMS over the motor
cortex, which have shown that the simulation process
occurring with action observation can be tuned to action goals
rather than to the physical features of the behavior (Gangitano
et al. 2004; Cattaneo, Caruana, et al. 2009).
Given the partial left lateralization of the effect that we
found, it could be argued that what we see is the effect of
habituation to verbal material, provided that subjects employed
2256 TMS Adaptation in Action observation
d
Cattaneo et al.
a strategy based on subvocal rehearsal for this task rather than
a real adaptation to the visual features of the stimulus.
However, no subject reported any such strategy. Some
semantic aspects of verbal processing have been found in the
SMG, but in reading tasks (Stoeckel et al. 2009). Previous TMS
studies on action or object naming would localize such
processes in distant areas such as dorsolateral prefrontal cortex
(Cappa et al. 2002) or posteriorly in Wernicke’s area (Mottaghy
et al. 2006).
The results of Experiment 2 on the other hand indicate that
coding of observed motor acts is performed by neural
populations within the STS-associated cortex with distinct
properties than in the ventral parietofrontal network. Neurons
in STS can code distal goal-directed motor acts, but they cannot
categorize motor acts disjoint from the effector that performs
them. In other words, neurons in STS have the capacity to
recognize motor behavior, but they cannot generalize its
semantic valence across all its possible physical variants. We
have observed evidence for such a capacity of abstraction only in
the parietofrontal system among the locations that we tested.
Our results on STS stimulation are also of interest considering
that the active task required discriminating the observed actions
while ignoring the effector. However, we find an interactive
effect of stimulation and effector, that is, we are acting with TMS
on an implicit or passively adapted feature. This type of implicit
effect has never been reported with TMSA paradigms (Silvanto
et al. 2008) and shows that TMS can act on passively induced
aftereffects in a complex manner.
The data on body part specificity in STS are controversial;
though most papers show tuning of different neural structures
to different body parts (Puce and Perrett 2003; Pelphrey et al.
2005), other work seems to contradict this (Thompson et al.
2007). At this point, one methodological remark needs to be
made. For the sake of simplicity, we have been attributing our
functional results to entire areas that we have targeted.
However, the main advantage of the TMSA technique is that
we can define functionally distinct neural populations embedded within other topographically overlapping population with
different properties. Therefore, all our findings, including that
of effector-dependent representation of motor behavior in STS,
are not mutually exclusive with other findings, especially when
investigating complex polimodal areas.
Experiment 2 was designed to test the action observation
node of the temporal lobe, the cortex associated with the
posterior part of the STS. The choice of stimulation sites, aside
from STS, was limited to one site where TMS had proven to be
ineffective in Experiment 1 (SPL, serving as control) and one
site where TMS had proven effective (left PMv) in order to
show replicability of the data. STS was stimulated on one side
only (the left side) for 2 reasons. First, though previous TMS
work applying off-line repetitive TMS over right posterior STS
disrupted the processing of point-light displays of biological
motion (Grossman et al. 2005), there is no clear evidence for
a lateralization of STS activity in biological motion processing
(Saygin 2007), so we chose the side where we found TMS
effects in both parietal and frontal stimulation sites. Second, we
wanted to minimize the number of stimulation sites due to
discomfort of TMS applied to a spot near to the ear and to the
‘‘temporalis’’ muscle fascia. Such discomfort, however, was not
subsequently reported by participants.
Our work combines 2 advantages: we have employed the
innovative method of state-dependent TMS and second we have
investigated simultaneously all the putative cortical stations of
the action observation system. Taken together, these 2 experiments demonstrate that the task of recognizing a motor act is
a 2-step process occurring at first in the STS-associated cortex in
a low-level modality that actually describes the physical features
of the observed scene. The second step occurs in the ventral
parietofrontal network where the perceptive features of the
observed scene are transformed into the abstract coding of the
motor act. Our data show that the parietofrontal action
observation system is ‘‘downstream’’ to the point where the
transformation between the infinite variants of a motor act are
transformed into the abstract coding of the motor valence of the
act independent of physical instantiation. On the contrary, the
STS node of the action observation system seems to ‘‘upstream’’
of such transformation.
The idea of a 2-step processing is indirectly suggested but
not clearly demonstrated by previous imaging studies.
However, there is general agreement that a preliminary
coding of observed acts linked to the physical features of
the observed behavior is carried out in STS, with a subsequent
addition of motor meaning in the parietofrontal system.
Evidence for this is found in one paper where STS did not
show the capacity for visual learning when observing other’s
acts (Cross et al. 2009). Others have shown that the premotor
system probably ‘‘fills’’ point-light biological motion displays
with motor significance (Saygin, Wilson, Hagler, et al. 2004).
Aside from human data, the model of the 2-step processing of
action cognition has been also proposed on the basis of
extensive physiological and anatomical data in monkeys
(Rizzolatti and Matelli 2003).
Notes
Conflict of Interest : None declared.
References
Alexander GE, Crutcher MD. 1990a. Neural representations of the
target (goal) of visually guided arm movements in three motor areas
of the monkey. J Neurophysiol. 64:164--178.
Alexander GE, Crutcher MD. 1990b. Preparation for movement: neural
representations of intended direction in three motor areas of the
monkey. J Neurophysiol. 64:133--150.
Avenanti A, Bolognini N, Maravita A, Aglioti SM. 2007. Somatic and
motor components of action simulation. Curr Biol. 17:2129--2135.
Barraclough NE, Keith RH, Xiao D, Oram MW, Perrett DI. 2009. Visual
adaptation to goal-directed hand actions. J Cogn Neurosci. 21:1806--1820.
Bien N, Roebroeck A, Goebel R, Sack AT. 2009. The brain’s intention to
imitate: the neurobiology of intentional versus automatic imitation.
Cereb Cortex. 19:2338--2351.
Boria S, Fabbri-Destro M, Cattaneo L, Sparaci L, Sinigaglia C, Santelli E,
Cossu G, Rizzolatti G. 2009. Intention understanding in autism. PLoS
One. 4:e5596.
Buccino G, Binkofski F, Fink GR, Fadiga L, Fogassi L, Gallese V, Seitz RJ,
Zilles K, Rizzolatti G, Freund HJ. 2001. Action observation activates
premotor and parietal areas in a somatotopic manner: an fMRI study.
Eur J Neurosci. 13:400--404.
Candidi M, Urgesi C, Ionta S, Aglioti SM. 2008. Virtual lesion of ventral
premotor cortex impairs visual perception of biomechanically
possible but not impossible actions. Soc Neurosci. 3:388--400.
Cappa SF, Sandrini M, Rossini PM, Sosta K, Miniussi C. 2002. The role of
the left frontal lobe in action naming: rTMS evidence. Neurology.
59:720--723.
Cattaneo L, Caruana F, Jezzini A, Rizzolatti G. 2009. Representation of
goal and movements without overt motor behavior in the human
motor cortex: a TMS study. J Neurosci. 29:11134--11138.
Cattaneo L, Rizzolatti G. 2009. The mirror neuron system. Arch Neurol.
66:557--560.
Cattaneo Z, Rota F, Vecchi T, Silvanto J. 2008. Using state-dependency of
transcranial magnetic stimulation (TMS) to investigate letter selectivity in the left posterior parietal cortex: a comparison of TMS-priming
and TMS-adaptation paradigms. Eur J Neurosci. 28:1924--1929.
Cattaneo Z, Rota F, Walsh V, Vecchi T, Silvanto J. 2009. TMS-adaptation
reveals abstract letter selectivity in the left posterior parietal cortex.
Cereb Cortex. 19:2321--2325.
Chong TT, Cunnington R, Williams MA, Kanwisher N, Mattingley JB.
2008. fMRI adaptation reveals mirror neurons in human inferior
parietal cortex. Curr Biol. 18:1576--1580.
Cohen Kadosh R, Walsh V. 2009. Numerical representation in the parietal
lobes: abstract or not abstract? Behav Brain Sci. 32:313--328.
Costantini M, Committeri G, Galati G. 2008. Effector- and targetindependent representation of observed actions: evidence from
incidental repetition priming. Exp Brain Res. 188:341--351.
Cross ES, Kraemer DJ, Hamilton AF, Kelley WM, Grafton ST. 2009.
Sensitivity of the action observation network to physical and
observational learning. Cereb Cortex. 19:315--326.
Fogassi L, Ferrari PF, Gesierich B, Rozzi S, Chersi F, Rizzolatti G. 2005.
Parietal lobe: from action organization to intention understanding.
Science. 308:662--667.
Gallese V, Fadiga L, Fogassi L, Rizzolatti G. 1996. Action recognition in
the premotor cortex. Brain. 119(Pt 2):593--609.
Gangitano M, Mottaghy FM, Pascual-Leone A. 2004. Modulation of
premotor mirror neuron activity during observation of unpredictable grasping movements. Eur J Neurosci. 20:2193--2202.
Grafton ST. 2009. Embodied cognition and the simulation of action to
understand others. Ann N Y Acad Sci. 1156:97--117.
Grafton ST, Hamilton AF. 2007. Evidence for a distributed hierarchy of
action representation in the brain. Hum Mov Sci. 26:590--616.
Grossman ED, Battelli L, Pascual-Leone A. 2005. Repetitive TMS over
posterior STS disrupts perception of biological motion. Vision Res.
45:2847--2853.
Hamilton AF, Grafton ST. 2006. Goal representation in human anterior
intraparietal sulcus. J Neurosci. 26:1133--1137.
Hamilton AF, Grafton ST. 2008. Action outcomes are represented in
human inferior frontoparietal cortex. Cereb Cortex. 18:1160--1168.
Heilman KM, Rothi LJ, Valenstein E. 1982. Two forms of ideomotor
apraxia. Neurology. 32:342--346.
Hershenson M. 1962. Reaction time as a measure of intersensory
facilitation. J Exp Psychol. 63:289--293.
Iacoboni M, Molnar-Szakacs I, Gallese V, Buccino G, Mazziotta JC,
Rizzolatti G. 2005. Grasping the intentions of others with one’s own
mirror neuron system. PLoS Biol. 3:e79.
Kakei S, Hoffman DS, Strick PL. 2001. Direction of action is represented
in the ventral premotor cortex. Nat Neurosci. 4:1020--1025.
Kilner JM, Neal A, Weiskopf N, Friston KJ, Frith CD. 2009. Evidence of
mirror neurons in human inferior frontal gyrus. J Neurosci.
29:10153--10159.
Liepelt R, Von Cramon DY, Brass M. 2008. How do we infer others’
goals from non-stereotypic actions? The outcome of contextsensitive inferential processing in right inferior parietal and
posterior temporal cortex. Neuroimage. 43:784--792.
Lingnau A, Gesierich B, Caramazza A. 2009. Asymmetric fMRI
adaptation reveals no evidence for mirror neurons in humans. Proc
Natl Acad Sci USA. 106:9925--9930.
Mahon BZ, Caramazza A. 2008. A critical look at the embodied
cognition hypothesis and a new proposal for grounding conceptual
content. J Physiol Paris. 102:59--70.
Marzi CA, Miniussi C, Maravita A, Bertolasi L, Zanette G, Rothwell JC,
Sanes JN. 1998. Transcranial magnetic stimulation selectively
impairs interhemispheric transfer of visuo-motor information in
humans. Exp Brain Res. 118:435--438.
Moro V, Urgesi C, Pernigo S, Lanteri P, Pazzaglia M, Aglioti SM. 2008.
The neural basis of body form and body action agnosia. Neuron.
60:235--246.
Mottaghy FM, Sparing R, Topper R. 2006. Enhancing picture
naming with transcranial magnetic stimulation. Behav Neurol.
17:177--186.
Cerebral Cortex September 2010, V 20 N 9 2257
Pazzaglia M, Smania N, Corato E, Aglioti SM. 2008. Neural underpinnings
of gesture discrimination in patients with limb apraxia. J Neurosci.
28:3030--3041.
Pelphrey KA, Morris JP, Michelich CR, Allison T, McCarthy G. 2005.
Functional anatomy of biological motion perception in posterior
temporal cortex: an FMRI study of eye, mouth and hand movements.
Cereb Cortex. 15:1866--1876.
Pobric G, Hamilton AF. 2006. Action understanding requires the left
inferior frontal cortex. Curr Biol. 16:524--529.
Puce A, Perrett D. 2003. Electrophysiology and brain imaging of
biological motion. Philos Trans R Soc Lond B Biol Sci. 358:435--445.
Rizzolatti G, Matelli M. 2003. Two different streams form the dorsal
visual system: anatomy and functions. Exp Brain Res. 153:146--157.
Rothi LJ, Heilman KM, Watson RT. 1985. Pantomime comprehension
and ideomotor apraxia. J Neurol Neurosurg Psychiatr. 48:207--210.
Saygin AP. 2007. Superior temporal and premotor brain areas necessary
for biological motion perception. Brain. 130:2452--2461.
Saygin AP, Wilson SM, Dronkers NF, Bates E. 2004. Action comprehension in aphasia: linguistic and non-linguistic deficits and their lesion
correlates. Neuropsychologia. 42:1788--1804.
Saygin AP, Wilson SM, Hagler DJ, Jr, Bates E, Sereno MI. 2004. Point-light
biological motion perception activates human premotor cortex. J
Neurosci. 24:6181--6188.
Silvanto J, Muggleton N, Walsh V. 2008. State-dependency in brain
stimulation studies of perception and cognition. Trends Cogn Sci.
12:447--454.
Silvanto J, Muggleton NG. 2008. Testing the validity of the TMS statedependency approach: targeting functionally distinct motion-
2258 TMS Adaptation in Action observation
d
Cattaneo et al.
selective neural populations in visual areas V1/V2 and V5/MT+.
Neuroimage. 40:1841--1848.
Silvanto J, Muggleton NG, Cowey A, Walsh V. 2007. Neural adaptation
reveals state-dependent effects of transcranial magnetic stimulation.
Eur J Neurosci. 25:1874--1881.
Stoeckel C, Gough PM, Watkins KE, Devlin JT. 2009. Supramarginal gyrus
involvement in visual word recognition. Cortex. 45:1091--1096.
Suchan B, Melde C, Herzog H, Homberg V, Seitz RJ. 2008. Activation
differences in observation of hand movements for imitation or
velocity judgement. Behav Brain Res. 188:78--83.
Thompson JC, Hardee JE, Panayiotou A, Crewther D, Puce A. 2007.
Common and distinct brain activation to viewing dynamic
sequences of face and hand movements. Neuroimage. 37:966--973.
Tomassini V, Jbabdi S, Klein JC, Behrens TE, Pozzilli C, Matthews PM,
Rushworth MF, Johansen-Berg H. 2007. Diffusion-weighted imaging
tractography-based parcellation of the human lateral premotor
cortex identifies dorsal and ventral subregions with anatomical and
functional specializations. J Neurosci. 27:10259--10269.
Umilta MA, Escola L, Intskirveli I, Grammont F, Rochat M, Caruana F,
Jezzini A, Gallese V, Rizzolatti G. 2008. When pliers become fingers
in the monkey motor system. Proc Natl Acad Sci USA.
105:2209--2213.
Umilta MA, Kohler E, Gallese V, Fogassi L, Fadiga L, Keysers C,
Rizzolatti G. 2001. I know what you are doing. A neurophysiological
study. Neuron. 31:155--165.
Urgesi C, Candidi M, Ionta S, Aglioti SM. 2007. Representation of body
identity and body actions in extrastriate body area and ventral
premotor cortex. Nat Neurosci. 10:30--31.