Investigating the role of the medial prefrontal cortex in the consolidation of schematic associative memories using Transcranial Magnetic Stimulation.

New incoming information is not encoded on a blank slate, but instead is
reconciled with and incorporated into the existing body of knowledge. An
extensive body of literature suggests that congruency of new information with
this prior knowledge, also called schemas, can enhance later memory
performance. A series of recent neuroscientific studies on this so-called
schema-congruency effect in animals and humans suggest an important role for
the interplay between the medial temporal lobe (MTL) and medial prefrontal
cortex (mPFC). A recent proposal suggests that the mPFC drives cortical
plasticity in the rest of the cortex, facilitating the integration of new
information within pre-existing schemas, both upon encoding and subsequent
reactivation. This is reflected in neural signals at later retrieval, such as
stronger activity in the mPFC as well as connectivity between the mPFC and
other neocortical associative storage areas, such that it suggests that the
retrieval of schema-congruent mnemonic traces become more dependent upon
representations mediated by the mPFC
In addition to correlational evidence of the relation between the mPFC and the
schema-congruency effect in human memory, it would be important to manipulate
the mPFC directly to further investigate the role of this region in selective
processing of schema-congruent associative memories. A window of opportunity to
modify the memory trace is given after reactivating associative information by
partial cueing, which temporarily increases cortical plasticity of the memory
trace. Several studies have shown that reactivation can strengthen the memory
trace as evidenced in improved later memory performance, or that the
presentation of new information after reactivation can modify or distort the
memory trace. Two recent studies have applied brain stimulation after
reactivation of memory traces. One found that applying electroconvulsive shocks
after reactivation of a memory trace erases this memory trace,whereas another
found that applying stimulation with TMS after reactivation can improve later
memory. Here, we combine the TMS-induced modification of memory traces with a
schema-manipulation to investigate the selective role of the mPFC in processing
schema-congruent memory traces.

Primary Objective:
We aim to investigate the possibility of erasing multimodal memories by
applying repetitive TMS to a somatosensory region that represents part of the
memory trace. Specifically, we expect that application of the stimulation
protocol to the somatosensory hand representational area in the critical window
of cortical plasticity after reactivation leads to a selective impairment for
later retrieval of specifically those associative memories that are reactivated
(compared to sham stimulation). We also aim to investigate how temporarily
impairing processing in the mPFC with TMS after reactivation selectively
influences the consolidation and eventual retrieval of those associative
memories that are schema-congruent. Specifically, we expect that application of
the stimulation protocol to the medial prefrontal cortex region after
reactivation leads to an impairment of subsequent recall of reactivated
schema-congruent associations.

In this study, we regard multisensory congruency with prior knowledge as a
pre-existing schema. We manipulate this congruency with prior knowledge to
investigate its influence on the retrieval of associative memories. Associative
information will be learned (day 1) in a visuotactile learning paradigm where
participants study visual motifs that are itself randomly associated with
object word-fabric combinations that are either schema-congruent (a leather
fabric paired with the word 'jacket') and schema-incongruent (a lace fabric
paired with the word 'umbrella') (as in van Kesteren et al., 2010, 2013). This
paradigm is extended by including a reactivation procedure on the day after
encoding (day 2) as a method to selectively manipulate cortical plasticity of
the trace well into the consolidation process. Half of the associative triplets
will be reactivated by means of tactile exploration of fabrics and probing for
retrieval of object words, which will temporarily render the corresponding
associative memory trace labile (or modifiable). We then use this as a window
of opportunity to selectively interfere with the consolidation process by
applying continuous theta-burst stimulation (40 sec) with a MC-B70 butterfly
coil (Magventure, stimulator is MagVenture, MagPro X100 with Magoption) for
stimulation of deeper brain regions. We use three separate groups here
(between-subject manipulation), as the memory paradigm cannot be repeated
multiple times across the same participants. Participants will be randomly
assigned to these three groups. The stimulation protocol will be applied to the
mPFC in one group. Alternatively, the stimulation protocol with the same coil
and stimulator will be applied in another group to the somatosensory cortex as
a task control. The stimulation control will be formed in a third group by
stimulating the primary motor cortex on the cortical midline (the area that
represents leg motor movements), which is presumed not to have any effect on
the task. We use this form of stimulation control instead of the usual
stimulation of a medial control site over the vertex, as the underlying cortex
of the vertex has a presumed function in declarative memory.
One day after the experimental intervention (day 3), we test memory by
performing a recognition memory test on the visual motifs, as well as an
associative memory task probing participants to select the correct
corresponding object word associated with the visual motif. We therefore
employ a 3 by 2 by 2 factorial design with a between group factor (stimulation
site), and two within group factors (respectively reactivation and
schema-congruency) as independent factors. The outcome (dependent) measures
will be recognition memory (d-prime) and associative memory (% correct).

We will employ an off-line rTMS-protocol with the intention to produce
short-term effects on the functioning of the target regions. The intervention
will consist of a standard continuous theta-burst stimulation (cTBS) procedure,
consisting of a total of 600 pulses administered across 40 sec. The stimulation
protocol is patterned, and consists of bursts of 3 pulses at 50 Hz, and each
burst itself is repeated at a frequency of 5Hz. The stimulation intensity will
be anchored in all three experimental groups relative to 80 % of the measured
active motor threshold from the tibialis anterior, following previous studies.
The reason for using aMT as measured from the tibialis anterior muscle instead
of a distal hand muscle, is that the representational motor area for the
tibialis anterior effector is located at a similar depth in the interhemisperic
fissure to our target location in the medial prefrontal cortex. During the
intake, we will determine the active motor-evoked threshold (aMT) of the
tibialis anterior as well as the first dorsal interosseous, as measured by
electromyographic recordings in response to single TMS-pulses delivered to the
appropriate motor hotspots in the primary motor cortex following the method of
limits. Specifically, during moderate contraction of the tibialis anterior
muscle respectively the first dorsal interosseous, the aMT will be determined
as the minimal stimulation intensity at which 5 out of 10 pulses evokes a
visible motor-evoked potential (MEP) on the electromyographic recordings. The
intensity at which the protocol of cTBS is applied is defined at 80% of the aMT
of the tibialis anterior, unless the intensity defined in this manner falls
above 120 % of aMT of the first dorsal interosseous. In the latter case, we
will employ a stimulation intensity for the cTBS-protocol that is anchored
below 120% of the aMT of the first dorsal interosseous. We use this
precautionary upper threshold of stimulation intensity, as the single case that
is known (out of 4,500 total cases) where the application of cTBS induced a
tonic-clonic seizure used a threshold of approximately 120 % of the aMT as
measured by electromyographic recordings on the first dorsal interosseous.
The cTBS-intervention will be delivered twice, once at the intake session to
determine tolerability, and later in the reactivation session (day 2) as an
experimental intervention (see study design). Depending on the experimental
group, the stimulation will be delivered to the medial prefrontal cortex
(navigated by anatomical midline landmarks moving the coil two-third the
direction from vertex to the nasion, somatosensory cortex (by neuronavigating
to anatomical somatosensory region corresponding to the finger representations,
or primary motor cortex (by neuronavigating to the midline motor cortex
representing the leg area).

Participants will not directly benefit from their participation in the study.
Transcranial magnetic stimulation (TMS) is widely used as a
non-invasive brain stimulation technique, based on principles of
electromagnetic induction. During stimulation the participant
will likely hear the clicks of the TMS pulses and experience a slight
stimulation of nerves and muscles of the head and face.

Theta-burst stimulation
A recent meta-analysis of the published literature on continuous theta-burst
repetitive (TBS) stimulation concluded that both the reported symptoms and
general risk of adverse events during TBS is comparable or less than other high
frequency rTMS protocols. Seizure is a reported severe adverse event, but still
only occurred once in over 4,500 sessions resulting in a crude risk of 0.02%.
We will take precautions with regard to the one report of a cTBS-induced
seizure, by using a more conservative threshold for anchoring stimulation
intensity. The overall crude risk of any adverse event from occurring was
estimated at 1.1%. The most commonly reported adverse event during TBS are
transient headaches and neck pains, which is similar to other rTMS protocols.
This adverse event was reported by less than 3% of all subjects receiving TBS.
Stimulating deeper brain regions with the butterfly coil
For this study, we are employing a butterfly coil in order to stimulate deeper
regions of the neocortex. According to a recent modeling paper, stimulation of
targets located approximately 4 cm or more below the skull surface is likely
unsafe, due to increased intensity of stimulation at the surface of the skull.
The current study will aim to stimulate target cortex within this 4cm margin,
to minimize this increased risk when stimulating deeper brain regions.
Stimulation intensity will be defined with reference to a motor threshold,
determined by stimulating the toe/leg representation area in the motor cortex.
We will be relying on a measured motor-evoked potential by electromyography of
the tibialis anterior to define the motor threshold (in contrast with other
studies that required visible motor twitches) during active muscle contraction
(in contrast with resting muscles), which potentially would lower the
stimulation intensities used in our study. Specifically, the active motor
threshold is defined as the minimal stimulation intensity where 50% (5 out of
10) pulses delivered to the central sulcus in the primary motor cortex evokes a
visible motor-evoked potential as measured with electromyography of the
tibialis anterior muscle. The stimulation intensity of the theta-burst protocol
will be put at 80% of this active motor threshold. If the intensity defined in
this manner falls above 120 % of aMT of the first dorsal interosseous, we will
employ a stimulation intensity for the cTBS-protocol that is anchored at 115 %
of the aMT of the first dorsal interosseous.
All subjects are screened for their relevant medical history and other safety
aspects (e.g. presence of metal parts in the head). Overall, the risk and
burden associated with participation can be considered low and acceptable based
on the previous literature, and we do not expect serious adverse events during
the project.