Attenuation processes in positive social emotion upregulation: Disentangling functional role of ventrolateral prefrontal cortex

Full text

Article
iScience
Attenuation processes in positive social emotion
upregulation: Disentangling functional role of
ventrolateral prefrontal cortex
Graphical abstract
Highlights
dHealthy women performed upregulation of positive social
emotions
dDistinct brain activity and connectivity patterns in vlPFC
revealed
dAttenuation of arvlPFC activity under control influences of
SFG modulated by vmPFC
dIndividual perception of sociality mediates affective and
social appraisals
Authors
Dmitriy D. Bezmaternykh,
Mikhail Ye. Melnikov,
Evgeny D. Petrovskiy, ..., Mark B. Shtark,
Patrik Vuilleumier, Yury Koush
Correspondence
yurykoush@gmail.com
In brief
Neuroscience; Behavioral neuroscience
Bezmaternykh et al., 2025, iScience 28, 111909
February 21, 2025 ª2025 The Author(s). Published by Elsevier Inc.
https://doi.org/10.1016/j.isci.2025.111909 ll

iScience
Article
Attenuation processes in positive social emotion
upregulation: Disentangling functional role of
ventrolateral prefrontal cortex
Dmitriy D. Bezmaternykh,
1
Mikhail Ye. Melnikov,
1,2
Evgeny D. Petrovskiy,
3
Ksenia G. Mazhirina,
1
Andrey A. Savelov,
3
Mark B. Shtark,
1
Patrik Vuilleumier,
4,5
and Yury Koush
6,7,8,
*
1
Institute of Molecular Biology and Biophysics, Federal Research Center of Fundamental and Translational Medicine, Novosibirsk, Russia
2
Department of Biophysics, Biomedicine, and Neuroscience, Faculty of Biology and Biotechnology, Al-Farabi Kazakh National University,
Almaty, Kazakhstan
3
International Tomography Center SB RAS, Novosibirsk, Russia
4
Department of Neuroscience, Medical School, University of Geneva, Geneva, Switzerland
5
Swiss Center of Affective Sciences, University of Geneva, Campus Biotech, Geneva, Switzerland
6
Vladimir Zelman Center for Neurobiology and Brain Rehabilitation, Skolkovo Institute of Technology, Moscow, Russia
7
Department of Biomedical Engineering, Yale University, New Haven, CT, USA
8
Lead contact
*Correspondence: yurykoush@gmail.com
https://doi.org/10.1016/j.isci.2025.111909
SUMMARY
Positive emotions determine individual well-being and sustainable social relationships. Here, we examined
the neural processes mediating upregulation of positive social emotions using functional magnetic reso-
nance imaging in healthy female volunteers. We identified brain regions engaged in upregulation of positive
social emotions and applied a parametric empirical Bayes approach to isolate modulated network connec-
tivity patterns and assess how these effects relate to individual measures of social perception. Our findings
indicate that upregulation of positive social emotions shapes the functional interplay between affective valu-
ation and cognitive control functions. We revealed a selective increase of bilateral posterior ventrolateral pre-
frontal cortex (vlPFC) activity and attenuation of activity in right anterior vlPFC under control influences from
more superior prefrontal regions. We also found that individual perception of sociality modulates connectivity
between affective and social networks. This study expands our understanding of neural circuits required to
balance positive emotions in social situations and their rehabilitative potential.
INTRODUCTION
Positive emotions support psychological resilience, individual
and societal well-being, and sustainable relationships.
1–4
Conversely, a lack of positive attitude to social events, known
as social anhedonia, affects social functioning and may
contribute to neuropsychiatric conditions, including schizo-
phrenia, post-traumatic stress, and depressive disorders.
4–6
The recent COVID-19 pandemic and associated restrictions
led to higher social anxiety
7
and radically altered interpersonal
interactions that typically support social-emotional well-being.
8,9
Previous research on emotion regulation suggests that
decreasing response to negative emotions requires inhibiting
prepotent appraisal of a stimulus in favor of an alternative reap-
praisal, and that these processes are associated with activation
of ventrolateral prefrontal cortex (vlPFC),
10–13
which is thought to
play an inhibitory role across several other tasks.
14–16
The pre-
frontal control system may also suppress overactive pleasure,
a process that might lead to depression in people with mood dis-
orders due to inappropriate activation. Accordingly, depressed
patients who demonstrate less downregulation of positive emo-
tions show quicker and better recovery from anhedonia in
response to antidepressant treatment, associated with lower
activity in rvlPFC, relative to patients with more persistent anhe-
donia who may exert stronger inhibition of positivity.
17
On the
other hand, transcranial magnetic stimulation (TMS) studies re-
ported that suppression of rvlPFC activity can lead to more
negative appraisal, whereas its activation can lead to less nega-
tive appraisal during regulation of negative emotions.
18
Likewise,
activation of rvlPFC is associated with more positive social eval-
uation,
18
relief of social pain, and increased reward.
19
However,
the engagement of vlPFC in successful generation and regula-
tion of negative emotions is also dependent on positive media-
tors, including nucleus accumbens/ventral striatum and subge-
nual anterior cingulate cortex (sgACC) that is associated with
greater reappraisal success, and negative mediators such as
amygdala, associated with lower reappraisal success.
12
In
sum, the exact role of rvlPFC in emotion regulation remains un-
resolved. It is unclear whether this region primarily mediates
regulation processes by suppressing negative affect, enhancing
iScience 28, 111909, February 21, 2025 ª2025 The Author(s). Published by Elsevier Inc. 1
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
ll
OPEN ACCESS

positive affect, or dampening instead both positive and negative
emotions. More generally, the involvement of vlPFC (in tandem
with homologous regions in left hemisphere) during reappraisal
of negative emotions might be related to more demanding con-
trol goals compared to upregulation of positive emotions, as well
as to differences between approach and avoidance processes
modulated by regulation strategies.
10
Unlike emotion self-regulation where the regulator and the
target are the same person, social regulation of emotions refers
to process where an individual attempts to regulate the
emotional response of another individual(s).
20
Nonetheless,
both social regulation and self-regulation of emotions in social
situations may recruit similar brain structures,
20–23
including in-
teractions between higher-order regions for cognitive control
and affective networks.
20
It is well-established that a key neural
mechanism of emotion regulation is the top-down influence from
prefrontal cortices onto emotion generating system, primarily
onto amygdala.
24,25
Amygdala plays a central role in conscious
and unconscious emotion processing and essentially acts to
extract the meaning of social signals.
26–28
In addition, sgACC
transfers information from limbic to cognitive control sys-
tem,
29–31
supports social processing,
32,33
and promotes positive
social emotion regulation.
21,31
Other crucial brain regions for the
regulation of positive social emotions include ventromedial pre-
frontal cortex (vmPFC),
21
a key part of affective, social, and
reward processing systems
34–39
; whereas dorsomedial (dmPFC)
and lateral prefrontal cortices are implicated in more general
cognitive control systems,
10,37
and engaged in positive social
emotion upregulation along with superior frontal gyrus (SFG)
adjacent to dmPFC.
21,31
However, the regulation of positive and positive social emotions
remains much less studied as compared to negative emotion
regulation,
10,40,41
which limits our mechanistic understanding of
emotion regulation processes. Moreover, growing evidence sug-
gests that different emotion regulation goals and different levelsof
stimulus valence may involve both common and distinct brain
structures.
22,40
Recent research also indicates that cognitive up-
and downregulation of emotions involves activation of partly
similar brain regions including, among others, bilateral vlPFC
and dorsomedial PFC (dmPFC).
21,40–42
Activation of vlPFC was
primarily observed during reappraisal of negative emotions,
11–13
and most recently, during reappraisal regardless of valence,
40,43
also known as anterior inferior frontal gyrus involved in inhibitory
control.
44,45
However, it remains unclear whether vlPFC modu-
lates positive social emotion regulation network.
To address these open questions, we used functional MRI
(fMRI) experiments requiring active self-engagement in positive
social situations and upregulation. By focusing on vlPFC as a
key region, we studied activation in positive social situations
compared to viewing neutral social situations and applied dy-
namic causal modeling (DCM) to probe for the interplay between
this region and other brain areas implicated in emotion regula-
tion. Distinct alternative models were examined to unveil how
vlPFC mediates affective and control processes during positive
social upregulation. To enable robust behavioral and fMRI ana-
lyses of individual data, we acquired two fMRI runs per partici-
pant and introduced rest periods between regulatory trials.
Unlike structural and resting state functional connectivity, our
effective connectivity analysis allowed us to investigate causal
influences between interconnected brain areas together with
their modulation by contextual factors and stimuli, through
group-level DCM analysis and parametric empirical Bayes
(PEB).
46,47
We hypothesized that vlPFC would be more active
during positive social upregulation if it directly acts to increase
positive affect. Alternatively, vlPFC should be either less active
or show no change if it mediates a more general suppression
of affect regardless of valence, as suggested by previous
research on reappraisal of negative emotions.
12
In addition, pos-
itive emotion upregulation in social situations should recruit other
regions implicated in affective processing, social cognition, and
cognitive control, which might interact with vlPFC through either
attenuating or enhancing modulatory influences according to the
exact role of this region in affective and control processes. We
therefore hypothesized selective connectivity changes in DCM
results. Thus, we expected that vlPFC might exert direct influ-
ences onto limbic and affect valuation regions, receive modula-
tory inputs from other cognitive control regions, and show
context-dependent modulations of these connections during
the positive social emotion upregulation task. Given that vlPFC
(de)activation was not explicitly disambiguated during upregula-
tion of positive emotions in the past literature, we had no hypoth-
esis about the sign of these interactions. In addition, we hypoth-
esized that connectivity strength between vlPFC and other brain
regions could be predicted by individual perception of sociality in
emotion eliciting situations. Since positive social upregulation is
associated with prosocial behavior,
18,19,48–50
understanding
neural processes controlling positive social emotion upregula-
tion may help better assessing and preventing dysregulations
that can lead social exclusion on mood and well-being.
RESULTS
We investigated attenuation processes in positive social emotion
upregulation and the role of rvlPFC in this regulation in a group of
healthy female volunteers using two whole-brain fMRI experi-
ments. We modeled PEB variations of endogenous connectivity
and contextual modulations of positive social emotion regulation
network with individual sociality scores as covariates, and specif-
ically investigated the functional interplay of activated bilateral
posterior ventrolateral prefrontal cortex (pvlPFC) and deactivated
right anterior vlPFC (arvlPFC) with positive social emotion regula-
tion network and the influence of individual social perception.
Behavioral ratings and questionnaire scores
For the first experiment, behavioral results were reported else-
where
21
but did not distinguish valence from (non-)sociality rat-
ings. For the second experiment (Figure 1), participants reported
good adherence to the experimental protocol (ability to focus:
3.7 ±1.9) and moderate vividness of imagery (2.1 ±2.7). Partic-
ipants did not increase positive affect after compared to before
the experiment (Figure 2A, Table S1; PANAS-P, before: 56.4 ±
12.4, after: 56.7 ±13.4, paired one-tailed t-test, t
19
= 0.13,
p= 0.89), but showed a trend decrease in negative affect
(PANAS-N, before: 20.5 ±3.1, after: 19.4 ±2.3, paired one-tailed
t-test, t
19
= 1.72, p= 0.051). This lack of increase in positive affect
and trend reduction of negative affect is consistent with past
2iScience 28, 111909, February 21, 2025
iScienc
e
Article
ll
OPEN ACCESS

research using similar duration and intensity of positive social
stimuli.
21
Participants consistently rated positive social pictures
with higher valence, arousal, and sociality scores, confirming
that positive social situations were perceived as more pleasant
and socially engaging interactions than the neutral social situa-
tions in our dataset (Figure 2A; valence: positive, 7.33 ±0.53;
neutral, 5.49 ±0.39; paired t-test, t
19
= 19.55, p< 0.001; arousal:
positive, 6.46 ±1.03; neutral, 4.93 ±1.15, paired t-test, t
19
=
17.69, p< 0.001; sociality: positive, 6.31 ±0.98; neutral, 4.86 ±
1.00; paired t-test, t
19
= 9.80, p< 0.001). We also found signifi-
cant correlations between scores of positive and neutral pictures
(Figure 2B; valence, r = 0.62, p= 0.004; arousal, r = 0.94,
p< 0.001; sociality, r = 0.78, p< 0.001). There were no correla-
tions between sociality and valence (positive, r = 0.21,
p= 0.37; neutral, r = 0.28, p= 0.23) or arousal (positive,
r = 0.33, p= 0.16; neutral, r = 0.35, p= 0.13) scores. In line
with prior work, we did not find associations between avoidance
attachment style scores and valence ratings of social situa-
tions.
51
Interestingly, we observed a significant positive correla-
tion between avoidance attachment style (ECR-AVS) scores and
sociality scores of social neutral images (Figure 2C, r = 0.71,
adjusted p= 0.037, FDR correction across all questionnaires
and behavioral ratings applied, n= 300, q < 0.05). Higher avoi-
dant attachment scores were associated with higher individual
sociality ratings of neutral social situations (not positive), sug-
gesting that participants who tend to feel uncomfortable with in-
timacy treated neutral social scenes as more socially engaging
relative to non-avoidant participants.
Brain activity associated with positive social emotion
upregulation
For the first experiment, the main effect of upregulation (upregula-
tion vs. viewing) revealed activations in SFG, and the main effect
of positive social stimuli (positive social vs. neutral nonsocial) re-
vealed, among other brain areas, activations in dmPFC, vmPFC,
sgACC, and bilateral amygdala.The specific effect of upregulating
positive social emotions compared to passive viewing of positive
social pictures was associated with increased activity of SFG and
dmPFC (Table S2, details reported elsewhere
21
). This experiment
did not reveal arvlPFC (de)activation when comparing other upre-
gulation and passive viewing conditions.
For the second experiment, blocks with upregulation of posi-
tive social emotion (in comparison to passive viewing of neutral
social pictures) activated a widespread network including bilat-
eral SFG, dmPFC, bilateral amygdala, vmPFC, left anterior
insula, inferior frontal gyrus (IFG/pvlPFC), thalamus/caudate,
hippocampus, putamen, superior temporal sulcus, and superior
parietal lobule (Figure 3A, Table S2, upregulate positive
social > view neutral social). Conversely, this comparison was
also associated with significant decrease in arvlPFC, as well as
inferior parietal lobule, superior and middle temporal gyri, and
medial regions overlapping with default mode network (Fig-
ure 3A, Table S2, view neutral social > upregulate positive so-
cial). The upregulation of positive social emotion in comparison
to fixation also shared similar effects than those in the first exper-
iment, with activation in dmPFC, vmPFC, sgACC, pvlPFC, and
left temporoparietal junction (TPJ), among other areas (Figure 3B,
Table S2, upregulate positive social > fixation). These results
therefore reject the hypothesis that arvlPFC would be activated
stronger during positive social upregulation as compared to
neutral social situations. Instead, bilateral pvlPFC was activated
and arvlPFC was deactivated. Consistently with previous find-
ings on selective processing of social stimuli,
27
we did not find
any functional lateralization of amygdala activity (comparing indi-
vidual contrast images with their flipped counterparts) but
observed a specialization of the superficial part of amygdala
Figure 1. Design of a functional run in experiment 2
During two functional runs, alternating blocks of neutral and positive social pictures were presented via MR-compatible monitor (17.6 min run duration). Each run
included five trials that comprised four upregulation blocks interleaved with five passive viewing blocks of 18 s each, followed by a 46 s fixation period (3 social
pictures per block with 6 s display duration). Participants were instructed to passively look at neutral social situations (blue frame; passive viewing) or imagine
experiencing positive social situations from a first-person perspective and feel positive about it (green frame; emotion upregulation). For illustrative purposes, we
used social pictures found via Google search engine under the Creative Common Zero license.
iScience 28, 111909, February 21, 2025 3
iScienc
e
Article
ll
OPEN ACCESS

activity during upregulation of positive social emotions
(Table S3). In addition to superficial part of amygdala, elicited
activation in basolateral complex in upregulation condition
related to fixation could be associated with social perception.
52
Regression analyses revealed that individual sociality scores of
neutral social pictures correlated negatively with upregulation
vs. neutral viewing activity in bilateral amygdala (Figure 3C,
r=0.69, adjusted p= 0.005; FDR-corrected for multiple compar-
isons across ROIs, n= 6, q < 0.05). The cognitive reappraisal
(ERQ-R) scores were also negatively correlated with upregulation
activity vs. neutral viewing in bilateral amygdala and sgACC (Fig-
ure 3D, amygdala: r = 0.61, adjusted p=0.014,sgACC:r=
0.63, adjusted p= 0.014; FDR-corrected for multiple compari-
sons across ROIs, n= 6, q < 0.05). Despite the BDI scores being
relatively low (mean ±SD, 3.9 ±3.3), they correlated negatively
with upregulation activity in pvlPFC (Figure S1,r=0.58, adjusted
p= 0.043, FDR-corrected for multiple comparisons across ROIs,
n=6,q<0.05).
Effective connectivity in positive social emotion
upregulation
To further investigate regulation processes modulated by rvlPFC
and identify network parameters associated with subjective soci-
ality ratings, we performed an effective connectivity analysis using
hierarchical PEB framework. We specifically focused on disentan-
gling the functional roles of activated pvlPFC and deactivated
arvlPFC during positive social emotion upregulation (Table S2),
given the known reappraisal role of bilateral pvlPFC regardless of
valence,
40,43
its engagement in reappraisal of negative emo-
tions,
10–13
regulation of social emotions,
19
and inhibitory con-
trol.
44,45
Thus, we defined key network nodes including bilateral
pvlPFC, arvlPFC, vmPFC, sgACC, bilateral SFG that included
dmPFC, and bilateral amygdala. Positive and negative DCM PEB
connectivity strengths indicate excitatory and inhibitory influ-
ences, respectively. Negative self-connectivity of nodes indicates
decreased self-inhibition (i.e., disinhibition), expressed in terms of
the log-scaled self-inhibitory prior 0.5*exp(A
i,i
).
When considering PEB variations of endogenous connectiv-
ity, we found a highly interconnected network (Figure 4A, model
with sociality covariate of positive pictures). The strongest
excitatory influence was exerted from sgACC onto pvlPFC
and from pvlPFC onto SFG. Interestingly, pvlPFC had positive
outgoing endogenous connections with all other nodes of the
model except arvlPFC. Conversely, arvlPFC received a single
selective attenuating influence from pvlPFC and exerted an
excitatory influence onto pvlPFC and SFG. We also examined
the PEB variations of endogenous connectivity with sociality
covariate of neutral pictures, which revealed a network archi-
tecture generally similar to model variations with the sociality
covariate of positive pictures (Figure S2A). They differed in
additional inhibitory connection of vmPFC onto arvlPFC and
lack of sgACC onto AMY endogenous positive connectivity.
An inhibitory influence of vmPFC onto arvlPFC suggests that
in presence of deactivation of arvlPFC, more active regulatory
processes were required to evaluate neutral social situations,
assign affective values to these situations, and assess their
personal self-relevance.
We then determined contextual PEB variations of this network
related to positive social upregulation demands. The effect of
emotion upregulation task was such that AMY and sgACC posi-
tively modulated affective valuation in vmPFC, which in turn
positively modulated pvlPFC and SFG (Figure 4B). Furthermore,
SFG strongly attenuated arvlPFC, which in turn positively modu-
lated sgACC and pvlPFC. Thus, we confirmed our hypothesis
that rvlPFC areas exerted direct influences onto limbic and affec-
tive systems and were contextually modulated by emotion upre-
gulation demands via attenuating connectivity from cognitive
control system (i.e., SFG). However, we also observed distinct
interactions between pvlPFC and arvlPFC with the rest of the
network. When considering PEB variations of contextual con-
nectivity models with sociality covariate of neutral pictures,
emotion upregulation task modulated not only SFG onto arvlPFC
inhibitory influence, but also outgoing arvlPFC onto SFG con-
nectivity (Figure S2B).
Figure 2. Psychometric scores and correlations between behavioral characteristics
(A) Emotion regulation (during experiment 2) did not increase positive mood but showed a trend-level decrease in negative mood. Positive social pictures
were rated significantly higher than neutral social pictures in valence, arousal, and sociality. For average values, we reported the mean and standard deviation
(SD). * indicate statistical significance (p< 0.05).
(B) Higher individual scores of positive social images were associated with higher scores of social neutral images, indicating high consistency of social scene
ratings within a participant. * denote significant Pearson correlations (p< 0.05).
(C) Higher attachment avoidance scores were associated with higher sociality scores of neutral social pictures, suggesting that participants who seek inde-
pendence and tend to experience discomfort from intimacy perceived neutral social scenes as more convenient for social interaction (while this relation was not
significant for positive social scenes, r = 0.58, p= 0.15). * Survived FDR correction for multiple comparisons (across all questionnaires and behavioral ratings,
n= 300, q < 0.05).
4iScience 28, 111909, February 21, 2025
iScienc
e
Article
ll
OPEN ACCESS

We also examined how these functional networks were modu-
lated by the individual social perception of positive scenes.
Although individual ratings of positive social and neutral social
pictures were correlated, we estimated two separate PEB varia-
tions of endogenous connectivity and their contextual modula-
tions using sociality ratings of neutral and positive social pictures
as distinct covariates, respectively. When considering PEB var-
iations of endogenous connectivity, sociality ratings of positive
pictures were positively associated with endogenous connectiv-
ity from SFG onto vmPFC, vmPFC onto pvlPFC, and from AMY
onto pvlPFC and arvlPFC (Figure 4C; for neutral pictures, see
Figure S2). In parallel, they were negatively associated with the
endogenous connectivity from sgACC onto AMY and pvlPFC,
pvlPFC onto vmPFC, and from SFG onto pvlPFC. Sociality rat-
ings of positive and neutral pictures also modulated disinhibition
self-connectivity of AMY and arvlPFC nodes. These data confirm
our hypothesis that connectivity strength between arvlPFC and
AMY engaged in socio-affective processes could be predicted
by individual traits making sociality effects complementary to
the effects of regulation. When considering contextual PEB var-
iations related to upregulation demands, we found significant
negative association between sociality ratings of positive pic-
tures and contextual connectivity from vmPFC onto AMY during
emotion upregulation (Figure 4B, dashed line). In comparison to
positive pictures, models of endogenous connectivity with
sociality covariate of neutral pictures were characterized with
additional positive associations between sociality scores and
connectivity from SFG onto sgACC, AMY onto vmPFC, and
negative associations between sociality scores and connectivity
from sgACC onto vmPFC (Figure S2C). Distinctively from social-
ity perception of positive scenes, this network had no associa-
tions for connectivity from SFG and vmPFC onto pvlPFC, and
from sgACC onto AMY. The negative association between
contextual modulation of connectivity from vmPFC onto AMY
and sociality ratings was not found for neutral social pictures.
Finally, for fully connected DCM models, we post-hoc illus-
trated a positive association between individual sociality ratings
of positive pictures and arvlPFC disinhibition self-connectivity
(Figure 4D, one-tailed Pearson r = 0.49, p= 0.013 unc.), and a
positive association with endogenous connectivity strength
from AMY onto arvlPFC (Figure 4E, one-tailed Pearson
r = 0.38, p= 0.047 unc.). We also plotted the only significantly
negative contextual modulation by social perception, namely
connectivity from vmPFC onto AMY (Figure 4F, one-tailed Pear-
son r = 0.53, p= 0.008 unc.).
DISCUSSION
Regulation of positive and positive social emotions remains
much less studied as compared to negative emotion regulation.
Figure 3. Brain activations and their associations with sociality and cognitive reappraisal scores in experiment 2
(A) Upregulation of positive social emotions compared to passive viewing of neutral social situations was associated with increased (red scale) activity in SFG,
vmPFC, pvlPFC and bilateral amygdala, as well as decreased (blue scale) activity in arvlPFC and DMN (FWE, p< 0.05). For illustration purposes, activation maps
were threshold at FWE p< 0.05.
(B) In upregulation contrasted to fixation we additionally observed significant activa tions in sgACC (FWE, p< 0.05). For illustration purposes, activation maps were
threshold at p< 0.005 unc.
(C) Higher sociality scores of neutral social pictures predicted lower and negative changes in bilateral amygdala activity during positive social emotion upre-
gulation, indicating that individuals with stronger ability to socialize in neutral social content exhibited lower emotional responses.
(D) Higher cognitive reappraisal scores predicted lower and negative activity changes in bilateral amygdala and sgACC during upregulation, indicating that
individuals with higher reappraisal preferences showed lower emotional responses.
(C and D) * survived FDR correction for multiple comparisons (across ROIs, n= 6, q < 0.05).
iScience 28, 111909, February 21, 2025 5
iScienc
e
Article
ll
OPEN ACCESS

Despite suggested similarity between negative and positive
emotion regulation processes, e.g., on generic reappraisal role
of activated bilateral pvlPFC,
40,43
there is growing evidence
that different emotion regulation goals and emotional valence
may involve distinct brain structures.
22,40
The present study pro-
vides new insights into the neural mechanisms underlying posi-
tive social emotion upregulation from a first-person perspective
and the role of rvlPFC in this regulation in a group of healthy fe-
male volunteers using whole-brain fMRI.
Brain (de)activation related to positive social emotion
upregulation
Upregulation of positive emotions in positive social situations as
compared to viewing neutral social scenes confirmed an
engagement of several brain regions typically implicated in
cognitive reappraisal,
10,12,40
social processing,
20
and positive
social emotion upregulation.
10,21,31
These involved limbic sys-
tem (e.g., AMY, sgACC), higher-order prefrontal regions impli-
cated in executive control and social cognition (e.g., SFG,
dmPFC), and affect generation (e.g., vmPFC). As expected in
both experiments, upregulation of positive social emotions was
associated with significant activations of dmPFC and adjacent
bilateral SFG regions implicated in cognitive control of emotions
and social processing. Expectedly, SFG, dmPFC, vmPFC, TPJ,
sgACC and bilateral amygdala were more active during upregu-
lation of positive social emotions as compared to fixation base-
line. Higher cognitive reappraisal scores predicted lower and
negative activity changes in bilateral amygdala and sgACC dur-
ing upregulation, suggesting that participants with larger
emotion reappraisal capability also show lower recruitment of
Figure 4. Effective connectivity underlying positive social emotion upregulation
To evaluate directional connections between nodes, we varied PEB models of endogenous and contextual modulation connectivity with sociality scoresof
positive social pictures as covariates. The numbers and thickness of the arrows indicate BMA values (cyan arrows for negative and black arrows for positive
connectivity strengths).
(A) The group average endogenous connectivity revealed a highly interconnected network. The pvlPFC showed positive connections with all the other network
nodes except arvlPFC, and arvlPFC received a single attenuating influence from pvlPFC and exerted an excitatory influence onto pvlPFC and SFG.
(B) The emotion upregulation task positively modulated connectivity from AMY and sgACC onto vmPFC, which positively modulated pvlPFC and SFG. Moreover,
SFG attenuated arvlPFC, which in turn positively modulated sgACC and pvlPFC, indicating highly selective changes in the interplay between control processes
mediated by pvlPFC and arvlPFC. Illustrative brain activity waveforms were scaled proportionally to the contrast-to-noise-ratio of the corresponding nodes
(average CNR estimated for upregulate positive vs. view neutral contrast; 0.19 for bilateral AMY, 0.45 for bilateral pvlPFC, 0.36 for bilateral SFG and dmPFC,
0.24 for arvlPFC, and 0.31 for vmPFC). Dashed line denotes significant negative association between sociality ratings of positive pictures and contextual
connectivity from vmPFC onto AMY during emotion upregulation, illustrated on this panel for simplicity.
(C and D) Higher sociality scores of positive social pictures were associated with significantly increased disinhibition of arvlPFC and AMY, indicating increased
sensitivity of these nodes to social content. In addition, sociality scores were associated with several endogenous connectivity strengths, including modulation of
connectivity between pvlPFC and arvlPFC and the rest of the network, indicating that socio-affective processes could be predicted by individual traits. We
showcased that sociality scores of positive pictures correlated positively with (D) magnitude of disinhibition of arvlPFC,
(E) connectivity strength from AMY onto arvlPFC, and negatively with (F) context ual connectivity from vmPFC onto AMY during positive social upregulation.
(A–C) * denote parameters with posterior probability (Pp) > 0.95 (strong evidence). (D–F) * denote significant post-hoc one-tailed Pearson correlations (p< 0.05
unc.).
6iScience 28, 111909, February 21, 2025
iScienc
e
Article
ll
OPEN ACCESS

emotion generation regions during positive social emotion
upregulation.
21,31
We also identified activation in bilateral pvlPFC, in line with
prior studies on reappraisal of positive and negative emo-
tions
40,43
and inhibitory control.
44,45
Interestingly, participants
who scored lower on BDI showed greater bilateral pvlPFC acti-
vation, which is consistent with patients in depression compared
to healthy controls during reappraisal of negative emotions.
13
Identified deactivation of arvlPFC suggests distinct functional
rvlPFC subdivisions and more general role in attenuating affect
as reflected by increased recruitment for negative downregula-
tion and additionally reduced recruitment for positive social up-
regulation. Functional segregation of rvlPFC was recently sug-
gested in meta-analytic study,
43
where arvlPFC and pvlPFC
were assigned to different co-activation groups implicated in
response inhibition or executive control and appraisal or lan-
guage processing during emotion regulation, respectively.
Although it was suggested that these two networks support
mainly the regulatory processes specific to reappraisal of nega-
tive emotions, perhaps due to the permanent literature shift to-
ward negative emotion regulation.
10,40,41
Implications of arvlPFC attenuation for mediation of
affective appraisal
Self-referential positive social emotion upregulation revealed
strong interconnections between prefrontal and limbic systems,
which agrees with previous studies on positive social emotion
regulation
21,31
and reappraisal of negative emotions.
10,12
Criti-
cally, contextual modulation of endogenous connectivity by
emotion upregulation highlighted the key functional interplay of
arvlPFC with affective valuation, cognitive control, and social
emotion regulation network. This was evidenced by strongly
attenuated contextual connectivity from SFG onto arvlPFC and
positively modulated contextual connectivity from arvlPFC
onto pvlPFC and sgACC. While functional role of SFG (dmPFC
and adjacent SFG) is generally thought to mediate cognitive con-
trol during social emotional appraisal and introspection,
53,54
to
evaluate social information,
55,56
to situate oneself in social con-
texts,
55,57
and to maintain a goal-relevant regulation strategy,
57
it
also appears to be crucially engaged during positive social
emotion upregulation.
21
Notably, we revealed direct interactions
between SFG and pvlPFC but contextual modulation form SFG
onto arvlPFC. Since rvlPFC is positively engaged in reappraisal
of negative emotions and response inhibition,
14,44
deactivation
of arvlPFC in positive upregulation of emotions in positive social
situations in comparison to passive viewing of neutral social sit-
uations and its selective connectivity pattern with pvlPFC and
SFG could be explained by distinctive roles of rvlPFC subdivi-
sions in effect attenuation processes.
Excitatory influence of arvlPFC onto sgACC suggests its medi-
atory role in gatekeeping processes between limbic and prefron-
tal cortices in social cognition
32,33,35
and positive social emotion
upregulation.
21,31
Moreover, higher cognitive reappraisal scores
predicted lower activity changes in bilateral amygdala and
sgACC, indicating that individuals with stronger preference for
reappraisal as an emotion regulation strategy exhibit lower pos-
itive emotional response and weaker connectivity changes be-
tween limbic and prefrontal cortices.
31
This is well in line with
findings on negative stimuli
58,59
and downregulation of negative
emotions,
60,61
suggesting a coupling between preference to re-
appraise and amygdala responsivity irrespective of valence.
Observed contextual modulation during positive social emotion
upregulation is largely consistent with prefrontal-subcortical
interactions modulated by reappraisal success during negative
emotion upregulation.
62
Specifically, authors observed positive
correlation of negative emotion upregulation success with
coupling of IFG (like pvlPFC) and amygdala, dmPFC and sgACC
and amygdala, vmPFC and IFG, and negative correlation of upre-
gulation success with couplingbetween IFG and dmPFC. Interest-
ingly, theyalso found reversedcorrelation pattern between vmPFC
and IFG during negative emotion downregulation (via distancing
from negative situations). Therefore, for emotion upregulation
regardless of valence, vmPFC was found positively associated
with pvlPFC, while reappraisal of negative emotions reversed
this dependency, confirming a key role of pvlPFC in regulation of
positive social emotions as in regulation of negative emotions
62
and in regulation of social emotions on a regulator side.
20
Notably, we did not reveal direct interactions between vmPFC
and arvlPFC regions but between vmPFC and pvlPFC, as well as
contextual connectivity from vmPFC onto pvlPFC. Widespread
endogenous connectivity pattern of pvlPFC and contextual con-
nectivity from vmPFC and arvlPFC onto pvlPFC could reflect its
mediatory role in positive social emotion regulation processes.
This finding is consistent with positive effective connectivity be-
tween vmPFC and left IFG (i.e., pvlPFC) in upregulation of aver-
sive stimuli.
62
Additionally, positive social upregulation task
modulated affective appraisal through the excitatory influence
of sgACC and AMY onto vmPFC, which was further translated
to SFG and pvlPFC. This is consistent with the known mediatory
role of vmPFC in prefrontal-subcortical connectivity.
12,21,63,64
The role of vmPFC is also central in regulation of social emotions,
connecting social engagement with affective and reward sys-
tems,
34,35,38
as it computes mainly affective valuations and is
modulated by both sociality and valence dimensions,
65,66
as-
signing and updating the subjective value of stimuli.
37,39
Implications of arvlPFC attenuation and sociality
perception for mediation of social appraisal
Bilateral amygdala plays a prominent role in evaluating social
cues and facial expressions.
23,26,28
Here, lower individual social-
ity scores of neutral social pictures predicted higher activation in
bilateral amygdala during emotion upregulation. This suggests
that participants with larger emotion upregulation capability
and greater recruitment of emotion generation regions might
perceive neutral social situations as less suitable for social inter-
action, which fits with the notion that amygdala activity is modu-
lated by sociality dimension.
22,65,67
Our endogenous connectivity models revealed significant as-
sociations between connectivity strength and perceived sociality
of positive and neutral social pictures. Specifically, higher indi-
vidual sociality scores were associated with reduction in self-in-
hibition inputs for bilateral AMY and arvlPFC, and with higher
positive connectivity from AMY onto arvlPFC and pvlPFC. This
suggests increased sensitivity of these regions to activity in the
rest of the network mediated by the perception of sociality in
positive social situations.
68
iScience 28, 111909, February 21, 2025 7
iScienc
e
Article
ll
OPEN ACCESS

For sociality covariates of neutral pictures, we identified nega-
tive associations for connectivity from SFG onto arvlPFC as
compared to connectivity from SFG onto pvlPFC for positive pic-
tures. This indicates specific top-down regulatory processes in
social situations reflected in inverse coupling between putative
attenuation processes mediated by arvlPFC and/or pvlPFC
and individual perception of sociality of social situations. The
coupling between sociality perception and arvlPFC disinhibition
complements prior findings that increases or decreases of
rvlPFC activity by TMS distinctively modulate emotions and pro-
social behavior,
18
while TMS-induced activation of rvlPFC im-
proves positive memory regarding social feedback.
19
Individual perception of sociality mediated connectivity from
SFG onto pvlPFC and vmPFC and reciprocal connectivity be-
tween pvlPFC and vmPFC in positive social situations as well
as connectivity from SFG onto arvlPFC, vmPFC and sgACC
and from pvlPFC onto vmPFC in neutral social situations, which
indicates that attenuation processes are integrated in both the
affective valuation of social situations and cognitive control pro-
cesses. Accordingly, a mediatory role of vmPFC and rvlPFC was
also suggested by TMS-induced activation of rvlPFC which
demonstrated attenuated activity in amygdala and insula but
enhanced coupling of prefrontal-subcortical areas via vmPFC
during the reappraisal of negative social exclusion.
63
Conclusions
Contrary to common observations of rvlPFC activation during re-
appraisal of negative emotions, we identified an increase of bilat-
eral pvlPFC activity and specific decrease of arvlPFC activity
during the upregulation of positive social emotions, as compared
to passive viewing of neutral social situations. Our findings sug-
gest that arvlPFC function might be attenuated during the effort-
ful self-engaged positive social emotion upregulation, and that
perception of sociality of positive and neutral social situations
also modulates these processes. We show evidence that upre-
gulation of positive social emotions involves attenuation pro-
cesses that modulate the network interplay between several
brain regions implicated in affective and social appraisal, and
cognitive control processes. This knowledge advances our un-
derstanding of neural circuits related to balancing upregulation
of positive emotions in social situations and explores the rehabil-
itative potential of modulating different parts of this network.
Limitations of the study
There were several limitations to the current study design. The
all-female sample limits our ability to generalize the results and
examine potential sex differences. However, female participants
may show stronger emotional responses and larger regulation
effects in both PFC and subcortical areas, as compared to
males, particularly for positive emotions and reappraisal.
69,70
This gender selection served to optimize the sensitivity of our
measures. In addition, we focused on a realistic scenario of ther-
apeutic relevance engaging oneself in positive social emo-
tions.
21,31,71,72
However, our data were collected across two
separate experiments. Because the first experiment allows con-
trasting upregulation and passive viewing of positive social situ-
ations, the second experiment included only the upregulation of
positive social pictures versus passive viewing of neutral social
pictures,
12
allowing to most effectively capture upregulation ef-
fects for social stimuli.
73
Nevertheless, the current design does
not allow for determining whether the observed effects are due
to emotional valence, context type, or their interaction, which
has been done elsewhere.
22,40
Our findings motivate future
research revisiting balanced positive, neutral, and negative so-
cial study designs to shed more light on specificity of emotion
regulation processes.
RESOURCE AVAILABILITY
Lead contact
Further information and requests for resources should be directed to the lead
contact, Yury Koush, (yurykoush@gmail.com).
Materials availability
This study did not generate plasmids, mouse lines, or unique reagents.
Data and code availability
dCode: the code generated during conventional data analyses is based
on SPM12 and MATLAB functions and is available from the lead contact
upon request.
dData: group-level fMRI and DCM results reported in the present study
are available from OSF (OSF: https://osf.io/8tvec/). The single subjects’
raw MRI data reported in this study cannot be deposited in a public re-
pository due to restrictions imposed by the local ethics committee.
dAdditional information: all data needed to evaluate the conclusions in
this paper are available within the paper and the Supplemental Informa-
tion. Any additional information required to reanalyze the data reported
in this paper is available from the lead contact upon request.
ACKNOWLEDGMENTS
The study was supported by the Russian Science Foundation (project no. 21-
15-00209).
AUTHOR CONTRIBUTIONS
The authors confirm contribution to the study as follows: conceptualization:
M.Y.M. and Y.K.; investigation and methodology: D.D.B., M.Y.M., E.D.P.,
K.G.M., A.A.S., M.B.S., and Y.K.; data analyses: D.D.B. and Y.K.; supervision:
Y.K.; writing and edits: D.D.B., M.Y.M., P.V., and Y.K. All authors approved the
final version of the manuscript.
DECLARATION OF INTERESTS
The authors declare no competing interests.
STAR+METHODS
Detailed methods are provided in the online version of this paper and include
the following:
dKEY RESOURCES TABLE
dEXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS
BParticipants
BInformed consent and ethics approval
dMETHOD DETAILS
BStimuli
BExperimental paradigm
BMRI data acquisition
BfMRI data preprocessing
dQUANTIFICATION AND STATISTICAL ANALYSIS
BGLM analysis
BDynamic causal modeling
BAnalyses of the behavioral and psychometric data
8iScience 28, 111909, February 21, 2025
iScienc
e
Article
ll
OPEN ACCESS

SUPPLEMENTAL INFORMATION
Supplemental information can be found online at https://doi.org/10.1016/j.isci.
2025.111909.
Received: February 29, 2024
Revised: September 27, 2024
Accepted: January 23, 2025
Published: January 27, 2025
REFERENCES
1. Diener, E., Napa Scollon, C., and Lucas, R.E. (2009). The Evolving
Concept of Subjective Well-Being: The Multifaceted Nature of
Happiness. In Assessing Well-Being: The Collected Works of Ed Diener,
E. Diener, ed. (Springer Netherlands), pp. 67–100. https://doi.org/10.
1007/978-90-481-2354-4_4.
2. Fredrickson, B.L. (2001). The role of positive emotions in positive psy-
chology. The broaden-and-build theory of positive emotions. Am. Psy-
chol. 56, 218–226. https://doi.org/10.1037//0003-066x.56.3.218.
3. Kok, B.E., Coffey, K.A., Cohn, M.A., Catalino, L.I., Vacharkulksemsuk, T.,
Algoe, S.B., Brantley, M., and Fredrickson, B.L. (2013). How positive emo-
tions build physical health: perceived positive social connections account
for the upward spiral between positive emotions and vagal tone. Psychol.
Sci. 24, 1123–1132. https://doi.org/10.1177/0956797612470827.
4. Mauss, I.B., Shallcross, A.J., Troy, A.S., John, O.P., Ferrer, E., Wilhelm,
F.H., and Gross, J.J. (2011). Don’t hide your happiness! Positive emotion
dissociation, social connectedness, and psychological functioning.
J. Pers. Soc. Psychol. 100, 738–748. https://doi.org/10.1037/a0022410.
5. Pelizza, L., and Ferrari, A. (2009). Anhedonia in schizophrenia and major
depression: state or trait? Ann. Gen. Psychiatry 8, 22. https://doi.org/10.
1186/1744-859X-8-22.
6. Barkus, E., and Badcock, J.C. (2019). A Transdiagnostic Perspective on
Social Anhedonia. Front. Psychiatry 10, 216. https://doi.org/10.3389/
fpsyt.2019.00216.
7. Kindred, R., and Bates, G.W. (2023). The Influence of the COVID-19
Pandemic on Social Anxiety: A Systematic Review. Int. J. Environ. Res.
Public Health 20, 2362. https://doi.org/10.3390/ijerph20032362.
8. Long, E., Patterson, S., Maxwell, K., Blake, C., Boso
´Pe
´rez, R., Lewis, R.,
McCann, M., Riddell, J., Skivington, K., Wilson-Lowe, R., and Mitchell,
K.R. (2022). COVID-19 pandemic and its impact on social relationships
and health. J. Epidemiol. Community. Health. 76, 128–132. https://doi.
org/10.1136/jech-2021-216690.
9. Forbes, P.A.G., Pronizius, E., Feneberg, A.C., Nater, U.M., Piperno, G.,
Silani, G., Stijovic, A., and Lamm, C. (2023). The effects of social interac-
tions on momentary stress and mood during COVID-19 lockdowns. Br. J.
Health Psychol. 28, 306–319. https://doi.org/10.1111/bjhp.12626.
10. Ochsner, K.N., Silvers, J.A., and Buhle, J.T. (2012). Functional imaging
studies of emotion regulation: a synthetic review and evolving model of
the cognitive control of emotion. Ann. N. Y. Acad. Sci. 1251, E1–E24.
11. Kim, S.H., and Hamann, S. (2007). Neural correlates of positive and nega-
tive emotion regulation. J. Cogn. Neurosci. 19, 776–798. https://doi.org/
10.1162/jocn.2007.19.5.776.
12. Wager, T.D., Davidson, M.L., Hughes, B.L., Lindquist, M.A., and Ochs-
ner, K.N. (2008). Prefrontal-subcortical pathways mediating successful
emotion regulation. Neuron 59, 1037–1050. https://doi.org/10.1016/j.
neuron.2008.09.006.
13. Keller, M., Mendoza-Quin
˜ones, R., Cabrera Mun
˜oz, A., Iglesias-Fuster,
J., Virue
´s, A.V., Zvyagintsev, M., Edgar, J.C., Zweerings, J., and Mathiak,
K. (2022). Transdiagnostic alterations in neural emotion regulation cir-
cuits - neural substrates of cognitive reappraisal in patients with depres-
sion and post-traumatic stress disorder. BMC. Psychiatry. 22, 173.
https://doi.org/10.1186/s12888-022-03780-y.
14. Aron, A.R. (2007). The neural basis of inhibition in cognitive control. The
Neuroscientist : a review journal bringing neurobiology. Neuroscientist
13, 214–228. https://doi.org/10.1177/1073858407299288.
15. Tabibnia, G., Monterosso, J.R., Baicy, K., Aron, A.R., Poldrack, R.A.,
Chakrapani, S., Lee, B., and London, E.D. (2011). Different forms of
self-control share a neurocognitive substrate. J. Neurosci. 31, 4805–
4810. https://doi.org/10.1523/JNEUROSCI.2859-10.2011.
16. Aron, A.R., Robbins, T.W., and Poldrack, R.A. (2004). Inhibition and the
right inferior frontal cortex. Trends. Cogn. Sci. 8, 170–177. https://doi.
org/10.1016/j.tics.2004.02.010.
17. Light, S.N., Heller, A.S., Johnstone, T., Kolden, G.G., Peterson, M.J., Ka-
lin, N.H., and Davidson, R.J. (2011). Reduced right ventrolateral prefron-
tal cortex activity while inhibiting positive affect is associated with
improvement in hedonic capacity after 8 weeks of antidepressant treat-
ment in major depressive disorder. Biol. Psychiatry. 70, 962–968. https://
doi.org/10.1016/j.biopsych.2011.06.031.
18. Yu, W., Li, Y., Cao, X., Mo, L., Chen, Y., and Zhang, D. (2023). The role of
ventrolateral prefrontal cortex on voluntary emotion regulation of social
pain. Hum. Brain Mapp. 44, 4710–4721. https://doi.org/10.1002/hbm.
26411.
19. Li, S., Xie, H., Zheng, Z., Chen, W., Xu, F., Hu, X., and Zhang, D. (2022).
The causal role of the bilateral ventrolateral prefrontal cortices on
emotion regulation of social feedback. Hum. Brain Mapp. 43, 2898–
2910. https://doi.org/10.1002/hbm.25824.
20. Reeck, C., Ames, D.R., and Ochsner, K.N. (2016). The Social Regulation
of Emotion: An Integrative, Cross-Disciplinary Model. Trends Cogn. Sci.
20, 47–63. https://doi.org/10.1016/j.tics.2015.09.003.
21. Koush, Y., Pichon, S., Eickhoff, S.B., Van De Ville, D., Vuilleumier, P., and
Scharnowski, F. (2019). Brain networks for engaging oneself in positive-
social emotion regulation. Neuroimage 189, 106–115. https://doi.org/10.
1016/j.neuroimage.2018.12.049.
22. Vrticka, P., Sander, D., and Vuilleumier, P. (2011). Effects of emotion
regulation strategy on brain responses to the valence and social content
of visual scenes. Neuropsychologia 49, 1067–1082.
23. Vrticka, P., Sander, D., and Vuilleumier, P. (2012). Lateralized interactive
social content and valence processing within the human amygdala. Fron-
tiers 6, 358. https://doi.org/10.3389/fnhum.2012.00358.
24. Banks, S.J., Eddy, K.T., Angstadt, M., Nathan, P.J., and Phan, K.L.
(2007). Amygdala - frontal connectivity during emotion regulation. Soc.
Cogn. Affect. Neurosci. 2, 303–312. https://doi.org/10.1093/Scan/
Nsm029.
25. Berboth, S., and Morawetz, C. (2021). Amygdala -prefrontal connectivity
during emotion regulation: A meta-analysis of psychophysiological inter-
actions. Neuropsychologia 153, 107767. https://doi.org/10.1016/j.neu-
ropsychologia.2021.107767.
26. Phelps, E.A., and LeDoux, J.E. (2005). Contributions of the amygdala to
emotion processing: from animal models to human behavior. Neuron 48,
175–187. https://doi.org/10.1016/j.neuron.2005.09.025.
27. Goossens, L., Kukolja, J., Onur, O.A., Fink, G.R., Maier, W., Griez, E.,
Schruers, K., and Hurlemann, R. (2009). Selective processing of social
stimuli in the superficial amygdala. Hum. Brain. Mapp. 30, 3332–3338.
https://doi.org/10.1002/hbm.20755.
28. Fusar-Poli, P., Placentino, A., Carletti, F., Landi, P., Allen, P., Surguladze,
S., Benedetti, F., Abbamonte, M., Gasparotti, R., Barale, F., et al. (2009).
Functional atlas of emotional faces processing: a voxel-based meta-
analysis of 105 functional magnetic resonance imaging studies.
J. Psychiatry. Neurosci. 34, 418–432.
29. Disner, S.G., Beevers, C.G., Haigh, E.A.P., and Beck, A.T. (2011). Neural
mechanisms of the cognitive model of depression. Nat. Rev. Neurosci.
12, 467–477. https://doi.org/10.1038/nrn3027.
30. Benschop, L., Vanhollebeke, G., Li, J., Leahy, R.M., Vande rhasselt, M.A.,
and Baeken, C. (2022). Reduced subgenual cingulate-dorsolateral
iScience 28, 111909, February 21, 2025 9
iScienc
e
Article
ll
OPEN ACCESS

prefrontal connectivity as an electrophysiological marker for depression.
Sci. Rep. 12, 16903. https://doi.org/10.1038/s41598-022-20274-9.
31. Scharnowski, F., Nicholson, A.A., Pichon, S., Rosa, M.J., Rey, G., Eickh-
off, S.B., Van De Ville, D., Vuilleumier, P., and Koush, Y. (2020). The role of
the subgenual anterior cingulate cortex in dorsomedial prefrontal-amyg-
dala neural circuitry during positive-social emotion regulation. Hum.
Brain Mapp. 41, 3100–3118. https://doi.org/10.1002/hbm.25001.
32. Premkumar, P. (2012). Are you being rejected or excluded? Insights from
neuroimaging studies using different rejection paradigms. Clin. Psycho-
pharmacol. Neurosci. 10, 144–154. https://doi.org/10.9758/cpn.2012.
10.3.144.
33. Lockwood, P.L., and Wittmann, M.K. (2018). Ventral anterior cingulate
cortex and social decision-making. Neurosci. Biobehav. Rev. 92,
187–191. https://doi.org/10.1016/j.neubiorev.2018.05.030.
34. Vetter, N.C., Weigelt, S., Do
¨hnel, K., Smolka, M.N., and Kliegel, M.
(2014). Ongoing neural development of affective theory of mind in adoles-
cence. Soc. Cogn. Affect. Neurosci. 9, 1022–1029. https://doi.org/10.
1093/scan/nst081.
35. Klimecki, O.M. (2015). The plasticity of social emotions. Soc. Neurosci.
10, 466–473. https://doi.org/10.1080/17470919.2015.1087427.
36. Roy, M., Shohamy, D., and Wager, T.D. (2012). Ventromedial prefrontal-
subcortical systems and the generation of affective meaning. Trends
Cogn. Sci. 16, 147–156. https://doi.org/10.1016/j.tics.2012.01.005.
37. Braunstein, L.M., Gross, J.J., and Ochsner, K.N. (2017). Explicit and im-
plicit emotion regulation: a multi-level framework. Soc. Cogn. Affect.
Neurosci. 12, 1545–1557. https://doi.org/10.1093/scan/nsx096.
38. Hiser, J., and Koenigs, M. (2018). The Multifaceted Role of the Ventrome-
dial Prefrontal Cortex in Emotion, Decision Making, Social Cognition, and
Psychopathology. Biol. Psychiatry 83, 638–647. https://doi.org/10.1016/
j.biopsych.2017.10.030.
39. Winecoff, A., Clithero, J.A., Carter, R.M., Bergman, S.R., Wang, L., and
Huettel, S.A. (2013). Ventromedial prefrontal cortex encodes emotional
value. J. Neurosci. 33, 11032–11039. https://doi.org/10.1523/JNEURO-
SCI.4317-12.2013.
40. Sokolowski, A., Morawetz, C., Folkierska-Zukowska, M., and Lukasz
Dragan, W. (2022). Brain activation during cognitive reappraisal depend-
ing on regulation goals and stimulus valence. Soc. Cognit. Affect Neuro-
sci. 17, 559–570. https://doi.org/10.1093/scan/nsab117.
41. Morawetz, C., Bode, S., Derntl, B., and Heekeren, H.R. (2017). The effect
of strategies, goals and stimulus material on the neural mechanisms of
emotion regulation: A meta-analysis of fMRI studies. Neurosci. Biobehav.
Rev. 72, 111–128. https://doi.org/10.1016/j.neubiorev.2016.11.014.
42. Li, F., Yin, S., Feng, P., Hu, N., Ding, C., and Chen, A. (2018). The cogni-
tive up- and down-regulation of positive emotion: Evidence from
behavior, electrophysiology, and neuroimaging. Biol. Psychol. 136,
57–66. https://doi.org/10.1016/j.biopsycho.2018.05.013.
43. Morawetz, C., Riedel, M.C., Salo, T., Berboth, S., Eickhoff, S.B., Laird,
A.R., and Kohn, N. (2020). Multiple large-scale neural networks underly-
ing emotion regulation. Neurosci. Biobehav. Rev. 116, 382–395. https://
doi.org/10.1016/j.neubiorev.2020.07.001.
44. Dillon, D.G., and Pizzagalli, D.A. (2007). Inhibition of Action, Thought, and
Emotion: A Selective Neurobiological Review. Appl. Prev. Psychol. 12,
99–114. https://doi.org/10.1016/j.appsy.2007.09.004.
45. Aziz-Safaie, T., M€
uller, V.I., Langner, R., Eickhoff, S.B., and Cieslik, E.C.
(2024). The effect of task complexity on the neural network for response
inhibition: An ALE meta-analysis. Neurosci. Biobehav. Rev. 158, 105544.
https://doi.org/10.1016/j.neubiorev.2024.105544.
46. Zeidman, P., Jafarian, A., Corbin, N., Seghier, M.L., Razi, A., Price, C.J.,
and Friston, K.J. (2019). A guide to group effective connectivity analysis,
part 1: First level analysis with DCM for fMRI. Neuroimage 200, 174–190.
47. Zeidman, P., Jafarian, A., Seghier, M.L., Litvak, V., Cagnan, H., Price,
C.J., and Friston, K.J. (2019). A guide to group effective connectivity
analysis, part 2: Second level analysis with PEB. Neuroimage 200, 12–25.
48. Snippe, E., Jeronimus, B.F., Aan Het Rot, M., Bos, E.H., de Jonge, P.,
and Wichers, M. (2018). The Reciprocity of Prosocial Behavior and Pos-
itive Affect in Daily Life. J. Pers. 86, 139–146. https://doi.org/10.1111/
jopy.12299.
49. Aknin, L.B., Van de Vondervoort, J.W., and Hamlin, J.K. (2018). Positive
feelings reward and promote prosocial behavior. Curr. Opin. Psychol. 20,
55–59. https://doi.org/10.1016/j.copsyc.2017.08.017.
50. Alexander, R., Arago
´n, O.R., Bookwala, J., Cherbuin, N., Gatt, J.M., Kah-
rilas, I.J., Ka
¨stner, N., Lawrence, A., Lowe, L., Morrison, R.G., et al.
(2021). The neuroscience of positive emotions and affect: Implications
for cultivating happiness and wellbeing. Neurosci. Biobehav. Rev. 121,
220–249. https://doi.org/10.1016/j.neubiorev.2020.12.002.
51. Vrticka, P., Bondolfi, G., Sander, D., and Vuilleumier,P. (2012). Theneural
substrates of social emotion perception and regulation are modulated by
adult attachment style. Soc. Neurosci. 7, 473–493. https://doi.org/10.
1080/17470919.2011.647410.
52. Bickart, K.C., Dickerson, B.C., and Barrett, L.F. (2014). The amygdala as
a hub in brain networks that support social life. Neuropsychologia 63,
235–248. https://doi.org/10.1016/j.neuropsychologia.2014.08.013.
53. Goldberg, I.I., Harel, M., and Malach, R. (2006). When the brain loses its
self: prefrontal inactivation during sensorimotor processing. Neuron 50,
329–339. https://doi.org/10.1016/j.neuron.2006.03.015.
54. Ochsner, K.N., and Gross, J.J. (2008). Cognitive emotion regulation:
Insights from social cognitive and affective neuroscience. Curr. Dir. Psy-
chol. Sci. 17, 153–158. https://doi.org/10.1111/j.1467-8721.2008.00566.x.
55. Amodio, D.M., and Frith, C.D. (2006). Meeting of minds: themedial frontal
cortex and social cognition. Nat. Rev. Neurosci. 7, 268–277. https://doi.
org/10.1038/nrn1884.
56. Bzdok, D., Langner, R., Schilbach, L., Engemann, D.A., Laird, A.R., Fox,
P.T., and Eickhoff, S.B. (2013). Segregation of the human medial prefron-
tal cortex in social cognition. Front. Hum. Neurosci. 7, 232. https://doi.
org/10.3389/fnhum.2013.00232.
57. Gusnard, D.A., Akbudak, E., Shulman, G.L., and Raichle, M.E. (2001).
Medial prefrontal cortex and self-referential mental activity: relation to a
default mode of brain function. Proc. Natl. Acad. Sci. USA 98, 4259–
4264. https://doi.org/10.1073/pnas.071043098.
58. Drabant, E.M., McRae, K., Manuck, S.B., Hariri, A.R., and Gross, J.J.
(2009). Individual differences in typical reappraisal use predict amygdala
and prefrontal responses. Biol. Psychiatry 65, 367–373. https://doi.org/
10.1016/j.biopsych.2008.09.007.
59. Gao, W., Biswal, B., Chen, S., Wu, X., and Yuan, J. (2021). Functional
coupling of the orbitofrontal cortex and the basolateral amygdala medi-
ates the association between spontaneous reappraisal and emotional
response. Neuroimage 232, 117918. https://doi.org/10.1016/j.neuro-
image.2021.117918.
60. Kanske, P., Heissler, J., Scho
¨nfelder, S., and Wessa, M. (2012). Neural
correlates of emotion regulation deficits in remitted depression: the influ-
ence of regulation strategy, habitual regulation use, and emotional
valence. Neuroimage 61, 686–693. https://doi.org/10.1016/j.neuro-
image.2012.03.089.
61. Fitzgerald, J.M., MacNamara, A., Kennedy, A.E., Rabina k, C.A., Rauch,
S.A.M., Liberzon, I., and Phan, K.L. (2017). Individual differences in
cognitive reappraisal use and emotion regulatory brain function in com-
bat-exposed veterans with and without PTSD. Depress. Anxiety. 34,
79–88. https://doi.org/10.1002/da.22551.
62. Morawetz, C., Bode, S., Baudewig, J., and Heekeren, H.R. (2017). Effec-
tive amygdala-prefrontal connectivity predicts individual differences in
successful emotion regulation. Soc. Cogn. Affect. Neurosci. 12,
569–585. https://doi.org/10.1093/scan/nsw169.
63. He, Z., Li, S., Mo, L., Zheng, Z., Li, Y., Li, H., and Zhang, D. (2023). The
VLPFC-Engaged Voluntary Emotion Regulation: Combined TMS-fMRI
Evidence for the Neural Circuit of Cognitive Reappraisal. J. Neurosci.
43, 6046–6060. https://doi.org/10.1523/JNEUROSCI.1337-22.2023.
10 iScience 28, 111909, February 21, 2025
iScienc
e
Article
ll
OPEN ACCESS

64. Steward, T., Davey, C.G., Jamieson, A.J., Stephanou, K., Soriano-Mas,
C., Felmingham, K.L., and Harrison, B.J. (2021). Dynamic Neural Interac-
tions Supporting the Cognitive Reappraisal of Emotion. Cereb. Cortex.
31, 961–973. https://doi.org/10.1093/cercor/bhaa268.
65. Britton, J.C., Phan, K.L., Taylor, S.F., Welsh, R.C., Berridge, K.C., and
Liberzon, I. (2006). Neural correlates of social and nonsocial emotions:
An fMRI study. Neuroimage 31, 397–409. https://doi.org/10.1016/j.neu-
roimage.2005.11.027.
66. Hare, T.A., Camerer, C.F., Knoepfle, D.T., and Rangel, A. (2010). Value
computations in ventral medial prefrontal cortex during charitable deci-
sion making incorporate input from regions involved in social cognition.
J. Neurosci. 30, 583–590. https://doi.org/10.1523/JNEUROSCI.4089-
09.2010.
67. Adolphs, R., Baron-Cohen, S., and Tranel, D. (2002). Impaired recogni-
tion of social emotions following amygdala damage. J. Cogn. Neurosci.
14, 1264–1274. https://doi.org/10.1162/089892902760807258.
68. Padilla-Coreano, N., Tye, K.M., and Zelikowsky, M. (2022). Dynamic in-
fluences on the neural encoding of social valence. Nat. Rev. Neurosci.
23, 535–550. https://doi.org/10.1038/s41583-022-00609-1.
69. McRae, K., Ochsner, K.N., Mauss, I.B., Gabrieli, J.J.D., and Gross, J.J.
(2008). Gender Differences in Emotion Regulation: An fMRI Study of
Cognitive Reappraisal. Group Process. Intergroup Relat. 11, 143–162.
https://doi.org/10.1177/1368430207088035.
70. Gaviria, J., Rey, G., Bolton, T., Ville, D.V.D., and Vuilleumier, P. (2021).
Dynamic functional brain networks underlying the temporal inertia of
negative emotions. Neuroimage 240, 118377. https://doi.org/10.1016/j.
neuroimage.2021.118377.
71. Koush, Y., Meskaldji, D.E., Pichon, S., Rey, G., Rieger, S.W., Linden,
D.E.J., Van De Ville, D., Vuilleumier, P., and Scharnowski, F. (2017).
Learning Control Over Emotion Networks Through Connectivity-Based
Neurofeedback. Cereb. Cortex 27, 1193–1202. https://doi.org/10.1093/
cercor/bhv311.
72. Young, K.D., Siegle, G.J., Zotev, V., Phillips, R., Misaki, M., Yuan, H., Dre-
vets, W.C., and Bodurka, J. (2017). Randomized Clinical Trial of Real-
Time fMRI Amygdala Neurofeedback for Major Depressive Disorder:
Effects on Symptoms and Autobiographical Memory Recall. Am. J. Psy-
chiatry. 174, 748–755. https://doi.org/10.1176/appi.ajp.2017.16060637.
73. Grosse Rueschkamp, J.M., Brose, A., Villringer, A., and Gaebler, M.
(2019). Neural correlates of up-regulating positive emotions in fMRI and
their link to affect in daily life. Soc. Cogn. Affect. Neurosci. 14, 1049–
1059. https://doi.org/10.1093/scan/nsz079.
74. Lang, P.J., Greenwald, M.K., Bradley, M.M., and Hamm, A.O. (1993).
Looking at pictures: affective, facial, visceral, and behavioral reactions.
Psychophysiology 30, 261–273. https://doi.org/10.1111/j.1469-8986.
1993.tb03352.x.
75. Marchewka, A., Zurawski, q., Jednoro
´g, K., and Grabowska, A. (2014).
The Nencki Affective Picture System (NAPS): introduction to a novel,
standardized, wide-range, high-quality, realistic picture database. Be-
hav. Res. Methods 46, 596–610. https://doi.org/10.3758/s13428-013-
0379-1.
76. Kurdi, B., Lozano, S., and Banaji, M.R. (2017). Introducing the Open Af-
fective Standardized Image Set (OASIS). Behav. Res. Methods 49,
457–470. https://doi.org/10.3758/s13428-016-0715-3.
77. Weierich, M.R., Kleshchova, O., Rieder, J.K., and Reilly, D.M. (2019). The
Complex Affective Scene Set (COMPASS): Solving the Social Content
Problem. Affective Visual Stimulus Sets 5, 53. https://doi.org/10.1525/
collabra.256.
78. Crone, D.L., Bode, S., Murawski, C., and Laham, S.M. (2018). The Socio -
Moral Image Database (SMID): A novel stimulus set for the study of so-
cial, moral and affective processes. PLoS One 13, e0190954. https://
doi.org/10.1371/journal.pone.0190954.
79. Dan-Glauser, E.S., and Scherer, K.R. (2011). The Geneva affective pic-
ture database (GAPED): a new 730-picture database focusing on valence
and normative significance. Behav. Res. Methods 43, 468–477. https://
doi.org/10.3758/s13428-011-0064-1.
80. Carretie
´, L., Tapia, M., Lo
´pez-Martı
´n, S., and Albert, J. (2019). EmoMa-
drid: An emotional pictures database for affect research. Motiv. Emot.
43, 929–939. https://doi.org/10.1007/s11031-019-09780-y.
81. Bradley, M.M., and Lang, P.J. (1994). Measuring emotion: the Self-
Assessment Manikin and the Semantic Differential. J. Behav. Ther.
Exp. Psychiatry. 25, 49–59. https://doi.org/10.1016/0005-7916(94)
90063-9.
82. Gross, J.J., and John, O.P. (2003). Individual differences in two emotion
regulation processes: implications for affect, relationships, and well-be-
ing. J. Pers. Soc. Psychol. 85, 348–362. https://doi.org/10.1037/0022-
3514.85.2.348.
83. Beck, A.T.,Steer, R.A., Ball,R., and Ranieri, W. (1996).Comparison of Beck
Depression Inventories -IA and -II in psychiatric outpatients. J. Pers.
Assess. 67, 588–597. https://doi.org/10.1207/s15327752jpa6703_13.
84. Zigmond, A.S., and Snaith, R.P. (1983). The hospital anxiety and depres-
sion scale. Acta. Psychiatr. Scand. 67, 361–370. https://doi.org/10.1111/
j.1600-0447.1983.tb09716.x.
85. Spielberger, C., Gorsuch, R., Lushene, R., Vagg, P.R., and Jacobs, G.
(1983). Manual for the State-Trait Anxiety Inventory (Form Y1 –
Y2);Consulting Psychologists (Palo Alto, CA, USA: Press).
86. Mishlove, M., and Chapman, L.J. (1985). Social anhedonia in the predic-
tion of psychosis proneness. J. Abnorm. Psychol. 94, 384–396. https://
doi.org/10.1037//0021-843x.94.3.384.
87. Snaith, R.P., Hamilton, M., Morley, S., Humayan, A., Hargreaves, D., and
Trigwell, P. (1995). A scale for the assessment of hedonic tone the Snaith-
Hamilton Pleasure Scale. Br. J. Psychiatry. 167, 99–103. https://doi.org/
10.1192/bjp.167.1.99.
88. Treynor, W., Gonzalez, R., and Nolen-Hoeksema, S. (2003). Rumination
Reconsidered: A Psychometric Analysis. Cognit. Ther. Res. 27,
247–259. https://doi.org/10.1023/A:1023910315561.
89. Brennan, K.A., Clark, C.L., and Shaver, P. (1998). Self-report measure-
ment of adult attachment: An integrative overview. In Attachment theory
and close relationships, J. A. Simpson and W. S. Rholes, eds. (The Guil-
ford Press), pp. 46–76.
90. Carver, C.S., and White, T.L. (1994). Behavioral inhibition, behavioral
activation, and affective responses to impending reward and punish-
ment: The BIS/BAS Scales. J. Pers. Soc. Psychol. 67, 319–333. https://
doi.org/10.1037/0022-3514.67.2.319.
91. Rotter, J.B. (1966). Generalized expectancies for internal versus external
control of reinforcement. Psychol. Monogr. 80, 1–28. https://doi.org/10.
1037/h0092976.
92. Watson, D., Clark, L.A., and Tellegen, A. (1988). Development and valida-
tion of brief measures of positive and negative affect: the PANAS scales.
J. Pers. Soc. Psychol. 54, 1063–1070. https://doi.org/10.1037//0022-
3514.54.6.1063.
93. Jenkinson, M. (2003). Fast, automated, N-dimensional phase-unwrap-
ping algorithm. Magn. Reson. Med. 49, 193–197. https://doi.org/10.
1002/mrm.10354.
94. Ashburner, J. (2007). A fast diffeomorphic image registration algorithm.
Neuroimage 38, 95–113. https://doi.org/10.1016/j.neuroimage.2007.
07.007.
95. Power, J.D., Barnes, K.A., Snyder, A.Z., Schlaggar, B.L., and Petersen,
S.E. (2012). Spurious but systematic correlations in functional connectiv-
ity MRI networks arise from subject motion. Neuroimage 59, 2142–2154.
https://doi.org/10.1016/j.neuroimage.2011.10.018.
96. Krylova, M., Skouras, S., Razi, A., Nicholson, A.A., Karner, A., Steyrl, D.,
Boukrina, O., Rees, G., Scharnowski, F., and Koush, Y. (2021). Progres-
sive modulation of resting-state brain activity during neurofeedback of
positive-social emotion regulation networks. Sci. Rep. 11, 23363.
https://doi.org/10.1038/s41598-021-02079-4.
iScience 28, 111909, February 21, 2025 11
iScienc
e
Article
ll
OPEN ACCESS

97. Rolls, E.T., Huang, C.C., Lin, C.P., Feng, J., and Joliot, M. (2020). Auto-
mated anatomical labelling atlas 3. Neuroimage 206, 116189. https://
doi.org/10.1016/j.neuroimage.2019.116189.
98. Palomero-Gallagher, N., Eickhoff, S.B., Hoffstaedter, F., Schleicher, A.,
Mohlberg, H., Vogt, B.A., Amunts, K., and Zilles, K. (2015). Functional or-
ganization of human subgenual cortical areas: Relationship between ar-
chitectonical segregation and connectional heterogeneity. Neuroimage
115, 177–190. https://doi.org/10.1016/j.neuroimage.2015.04.053.
99. Friston, K.J., Harrison, L., and Penny, W. (2003). Dynamic causal model-
ling. Neuroimage 19, 1273–1302.
100. Friston, K., Mattout, J., Trujillo-Barreto, N., Ashburner, J., and Penny, W.
(2007). Variational free energy and the Laplace approximation. Neuro-
image 34, 220–234. https://doi.org/10.1016/j.neuroimage.2006.08.035.
101. Guedj, C., and Vuilleumier, P. (2023). Modulation of pulvinar connectivity
with cortical areas in the control of selective visual attention. Neuroimage
266, 119832. https://doi.org/10.1016/j.neuroimage.2022.119832.
12 iScience 28, 111909, February 21, 2025
iScienc
e
Article
ll
OPEN ACCESS

STAR+METHODS
KEY RESOURCES TABLE
EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS
Participants
We used previously published fMRI data
21
from twenty-three healthy participants (11 female, mean ±SD age, 27.7 ±6.2 years, White
cohort), denoted as the first experiment, and newly collected data of twenty-one healthy females aged between 18 and 55 who were
recruited for a single day study, denoted as the second experiment. One subject was excluded from new data due to excessive head
movements in fMRI session, leaving a sample of 20 participants (mean ±SD age, 37.1 ±8.4 years, White cohort). For the second
experiment, we recruited only female participants, because prior work suggests they show stronger emotional responses and larger
regulation effects in both PFC and subcortical areas, as compared to males, particularly for positive emotions and reappraisal.
69,70
They reported no history of neurological, psychiatric, or addictive disorders as well as no contraindications to MR imaging, were free
of any psychotropic medication, and had normal or corrected-to-normal vision.
Informed consent and ethics approval
The first experiment was approved by the ethics committee at the University of Geneva, Geneva, Switzerland.
21
The second exper-
iment was approved by the ethics review board of the Federal research center of fundamental and translational medicine, Novosi-
birsk, Russian Federation (approval number 28/1). All participants gave written informed consent prior to the study and received a
monetary compensation.
METHOD DETAILS
Stimuli
For both experiments, a set of social pictures was collected from the International Affective Pictures System (IAPS),
74
Nencki
Affective Picture System (NAPS),
75
Open Affective Standardized Image Set (OASIS),
76
Complex Affective Scene Set
(COMPASS),
77
Socio-Moral Image Database (SMID),
78
Geneva Affective Picture Database (GAPED),
79
and EMOMadrid data-
base
80
(901 in total). Pictures were classified as social based on presence of people and faces as a part of the scene.
65
Pictures
were resized to fit the screen resolution of display monitor; no additional cropping and color correction was performed. To ac-
count for relative biases in valence and arousal scores between different image databases, we matched their scores given pre-
defined categories with uniform emotional response, e.g., weddings, happy babies, team sports, smiling adults, etc. Using
linear regression model, we examined the association between median valence and arousal scores of each category and re-
coded original scores to those in IAPS. We did not include extreme positive content to avoid ceiling effects. The order of image
presentation for fMRI scans and behavioral ratings was pseudo-randomized per participant. For average values, we reported
the mean and standard deviation (SD).
For the first experiment, stimuli details are listed elsewhere (positive social pictures, n= 112, normative valence 6.97 ±0.68, arousal
4.97 ±0.82; neutral nonsocial pictures, n= 112, normative valence 5.21 ±0.60, arousal 3.61 ±0.96).
21
For the second experiment,
stimuli were assigned to positive social pictures (n= 369, normative valence 7.13 ±0.34, arousal 4.97 ±0.65) or neutral social pictures
(n= 420, normative valence 5.32 ±0.78, arousal 4.35 ±0.48). Social pictures were used for fMRI sessions and behavioral ratings of
valence, arousal, and sociality after fMRI. Pictures used for fMRI and behavioral sets did not overlap.
REAGENT or RESOURCE SOURCE IDENTIFIER
Deposited data
Group level fMRI contrasts OSF repository https://osf.io/8tvec/
Single subject DCM OSF repository https://osf.io/8tvec/
Individual ROI masks OSF repository https://osf.io/8tvec/
Single subject raw data Federal Research Center of Fundamental
and Translational Medicine
N/A
Software and algorithms
MATLAB R2021b MathWorks, Natick, MA https://www.mathworks.com/products/
matlab.html
Statistical Parametric Mapping (SPM12,
v7771), including DCM 12.5
Wellcome Trust Center for Neuroimaging https://www.fil.ion.ucl.ac.uk/spm/
iScience 28, 111909, February 21, 2025 e1
iScienc
e
Article
ll
OPEN ACCESS

Stimuli selected for the fMRI set in the second experiment comprised a series of different pseudo-randomized 120 positive social
(group average; normative valence = 7.11 ±0.01, arousal = 4.94 ±0.03) and 150 neutral social pictures (group average; normative
valence = 5.35 ±0.01, arousal = 4.28 ±0.03). Positive and neutral social pictures were significantly different on the subject level (two-
sample one-tailed t-test, p-values <0.001), which was also confirmed on the trial level (two-sample one-tailed t-test; valence p-values
<0.001; arousal p-values <0.015). For positive and neutral social between-subject stimuli randomizations, we minimized the differ-
ence in valence (two-sample two-tailed t-test; positive p-values >0.76, neutral p-values >0.72) and arousal (two-sample two-tailed
t-test; positive p-values >0.50, neutral p-values >0.23).
Stimuli selected for the behavioral rating sets (30 pictures each; group average; positive social normative valence = 7.15 ±0.03,
arousal = 5.06 ±0.04, neutral social normative valence = 5.27 ±0.03, arousal = 4.46 ±0.04) were also pseudo-randomized and had
significant difference in valence between sets (two-sample one-tailed t-test; p-values <0.001) and in arousal between neutral and
positive pictures (two-sample one-tailed t-test, p-values <0.002). For both categories, we minimized between-subject difference
in valence (two-sample two-tailed t-test; positive social p-values >0.38; neutral social p-values >0.66) and arousal (two-sample
t-test; positive social p-values >0.62; neutral social p-values >0.50). Social pictures from behavioral and fMRI picture sets did
not differ in valence (two-sample t-test; positive p-values >0.37; neutral p-values >0.34) and arousal (two-sample t-test; positive
p-values >0.13; neutral p-values >0.13).
Each picture was repeated a limited number of times per group (median [quartiles] = 6 [3 10]). For individual picture sets, the mean
normative valence and arousal was significantly greater for positive than neutral social pictures (two-sample one-tailed t-test;
valence p-values <0.001, arousal p-values <0.002). There was no significant difference in valence and arousal between fMRI and
behavioral picture sets (p-values >0.13).
Experimental paradigm
The first experiment was set as a 2 32 factorial design based on the factors stimuli (positive-social vs. neutral nonsocial scenes) and
task (passive viewing vs. effortful emotion upregulation).
21
For design efficiency and stronger signal to noise, we did not manipulate
valence and content of pictures separately,
22,65
because this study did not aim at dissecting these two factors but focused on pos-
itive aspects of social interactions only.
21
We acquired two functional runs (11.3 min run duration) consisting of two emotion upre-
gulation or passive viewing epochs randomized across subjects (alternating seven blocks of positive social and neutral nonsocial
pictures per epoch, 4 pictures per block, 6s picture display duration).
Because thefirst experiment allows contrasting upregulation andpassive viewing of positive social pictures, in the secondexperiment
we included only the critical experimental condition with upregulation of positive social pictures, and a reference baseline condition with
passive viewing of neutral social pictures,
12
allowing our analysis to mosteffectively capture upregulation effects for social stimuli.
73
Our
experimental design is particularly different from previous factorial studies
22,40
by balanced passive viewing of neutral social scenes and
fixation baseline. We contrasted blocks of positive social upregulation and neutral social viewing in two subsequent fMRI runs (Figure 1;
17.6min runduration). Both runs consisted of fivetrials alternating fouremotion regulation blocks with five passive viewing (3 pictures per
block with 18s display duration) plus a fixation cross presented for 48s. For both experiments during positive social upregulation blocks,
subjects wereasked to imagine activeand enjoyable interactions with peopledepicted in social pictures from a first-person perspective.
During neutral social viewing blocks, subjects were asked to passively look at neutral social pictures. Stimuliand tasks were indicatedby
differentpicture frame colors.The stimuli were presented on the screenof an MR-compatible monitor visible through a mirror attached to
the head coil. All participants were instructed to breathe steadily and remain as still as possible.
After fMRI runs of the second experiment, participants performed a behavioral rating task with two picture sets (60 pictures in total),
using valence and arousal scores on a continuous scale of self-assessment manikins.
81
For sociality ratings, they evaluated how
much socially engaging or interacting they felt about depicted scenes using customized manikins representing incremental increases
in sociality. Care was taken to ensure that participants did not make their rating depending on the number of people in the scene, for
example if there is only one person, they could base their rating on how easily they could interact with that person in this scene. Un-
derstanding of the task was ensured and response time was not limited.
To assess personality traits that may influence changes in brain activity and connectivity, participants were asked to complete the
Emotion Regulation Questionnaire (ERQ),
82
Beck Depression Inventory-II (BDI),
83
Hospital Anxiety and Depression Scale (HADS),
84
State-Trait Anxiety Inventory (STAI),
85
Revised Social Anhedonia Scale (RSAS),
86
Snaith-Hamilton Pleasure Scale (SHAPS),
87
Ruminative Response Scale (RRS),
88
Experiences in Close Relationships questionnaire (ECR),
89
Behavioral Inhibition and Activation
Systems Scales (BIS/BAS),
90
and Rotter’s Internal-External Locus of Control Scale (LCS).
91
To test whether emotional state changes after upregulation fMRI session, we measured the Positive and Negative Affect Schedule
(PANAS)
92
before and after this session. To assess compliance with instructions and strategy details, we asked participants to rate
whether they were focused or absentminded, and whether their imagery was vivid or not (Likert scale from 5 to +5).
MRI data acquisition
The first experiment was performed on a 3T MRI scanner (Trio Tim, Siemens Medical Solutions, Erlangen, Germany) equipped with a
32-channel head coil at the Brain and Behavior Laboratory (University of Geneva). Functional images were acquired with a whole-
brain single-shot gradient-echo T2*-weighted EPI sequence (TR/TE = 2050/35ms, flip angle = 75, matrix 120 3120, 32 slices, voxel
size 2 3232mm
3
, GRAPPA, iPAT = 3).
e2 iScience 28, 111909, February 21, 2025
iScienc
e
Article
ll
OPEN ACCESS

For the second experiment, MRI recordings were performed on a 3T MRI scanner (Ingenia, Philips, Best, the Netherlands) equip-
ped with a 16-channel head coil at the International Tomography Center SB RAS. A T1-weighted structural image was acquired at the
beginning of the scanning session using 3D turbo field echo sequence with TR/TE = 7.7/3.8ms, flip angle = 8, matrix 288 3288, 181
slices, and voxel size = 0.87 30.87 31mm
3
. Functional T2*-weighted images (528 scans per run) were obtained with a single-shot
gradient-echo EPI sequence with TR/TE = 2000/35ms, flip angle = 90, matrix 112 3112, 31 slices, voxel size 2 3233mm
3
,9
dummy scans, and parallel imaging (SENSE) factor = 3. The EPI protocol was configured to maximize brain coverage and to ensure
optimal signal quality for sgACC and bilateral amygdala.
31
We also acquired a double-echo gradient-echo static magnetic field map
(TE
1
= 7ms, TE
2
= 10ms, flip angle = 70, voxel size = 2.5 32.5 32.5mm
3
).
fMRI data preprocessing
For both experiments, the conventional fMRI data processing was performed using SPM12 (Wellcome Trust Center for Neuroimag-
ing, Queen Square, London, UK) and MATLAB (Mathworks, Inc.). The functional images were spatially realigned to the first scan of
each run, corrected for slice-timing and geometric distortions,
93
co-registered to the individual structural image, normalized to the
standard MNI structural template with an isotropic 2mm
3
voxel size and smoothed with an isotropic Gaussian kernel with 6mm full-
width-at-half-maximum (FWHM) using DARTEL.
94
QUANTIFICATION AND STATISTICAL ANALYSIS
GLM analysis
First-level fMRI analysis for the first experiment is described in detail elsewhere.
21
Similarly for the second experiment at the single-
subject level, we specified a general linear model (GLM) with a separate regressor for each trial of passive viewing and emotion up-
regulation conditions, implemented in SPM12. For two functional runs the fixed-effect model was applied. We modeled regressors as
boxcar functions convolved with the canonical hemodynamic response function (HRF), 6 head movement covariates to capture re-
sidual motion artifacts, and removed scans with excessive head motion (framewise displacement >0.9).
95,96
The data were high-pass
filtered with a conventional 0.008Hz cut-off.
For the second experiment whole-brain group-level analysis, a flexible factorial ANOVA was performed with a random factor
‘subject’ and fixed factor ‘condition’ (trial-based individual contrasts). As contrasts of interest, we computed upregulate positive
vs. view neutral and upregulate positive vs. fixation. Statistical maps were corrected for multiple comparisons using whole brain fam-
ily-wise error correction (FWE, p< 0.05).
Dynamic causal modeling
Definition of individual regions of interest
For DCM estimations, we defined group and individual ROIs. Group ROIs were defined within corresponding anatomical structures
and served as reference areas within which individual functional peaks of activity were located.
46
Group ROIs did not overlap. Group
bilateral amygdala ROIs were defined anatomically based on the AAL3 atlas
97
because it is a small region for which a spherical ROI
would have likely included non-amygdala voxels in proximity. Group sgACC ROI was defined as an anatomically referenced 16 3
27 312mm
3
box centered at [0,14,-12] with the addition of a Brodman Area 25.
98
To localize other key network nodes, we used
8mm spheres centered on ROI peak coordinates at the group level (FWE, p< 0.05). Group vmPFC ([-2, 56, 18]), SFG (left, [-16,
54, 38]; right, [16, 50, 44]; central dmPFC: [-2, 58, 36]), and pvlPFC (left, [-56, 24, 12]; right, [58, 28, 4]) maxima were defined using
upregulate positive > view neutral contrast. For group arvlPFC, we used peak coordinates [46, 41, 4] defined as the intersection
between view neutral > upregulate positive contrast and bilateral vlPFC maps extracted from Neurosynth database (association
tests, entry ‘‘inhibitory control’’ and ‘‘ventrolateral prefrontal’’). The group arvlPFC ROI is in good agreement with emotion regulation
studies
12
and in relatively close proximity yet different from more posterior inferior frontal gyrus areas involved in reappraisal of pos-
itive and negative emotions
40,43
and inhibitory control.
44,45
For DCM node time-series extraction, individual ROIs were defined as spheres of 6mm radius centered on the peaks located within
corresponding group ROIs using individual contrast maps exceeded a liberal statistical threshold (p< 0.05 unc.). Individual sgACC
ROI was defined as a small rectangle around individual peak using upregulate positive > fixation contrast. In case of absence of in-
dividual maximum, we used a group ROI mask. Individual sgACC and bilateral amygdala ROIs were restricted by corresponding
group masks.
PEB DCM analysis
For the second experiment, to model functional networks engaged during the task, we applied bilinear Dynamic Causal Modeling
(DCM). DCM is a Bayesian framework that evaluates directional interactions between different brain areas (i.e., effective connectivity)
and the impact of experimental tasks on these interactions. DCM combines biologically plausible neurovascular model with neuronal
model to predict observed fMRI time-series and estimate model parameters using Variational Bayes under a fixed-form Laplace
scheme.
99,100
Functional brain network was modeled as endogenous bidirectional connectivity between the DCM model nodes
and self-connectivity of nodes (matrix A), plus modulatory inputs onto the endogenous connections related to contextual factors
(i.e., engaging actively in the depicted situations, matrix B) and external inputs (i.e., presentation of stimuli, matrix C). Non-diagonal
elements of matrices A and B indicate the rate of change in response of one region that is caused by activity in another region without
iScience 28, 111909, February 21, 2025 e3
iScienc
e
Article
ll
OPEN ACCESS

external perturbations or modulated by contextual factors, respectively (in Hz). Positive connectivity strength describes excitatory
influences, while negative value represents inhibitory influences. The diagonal elements refer to self-inhibition processes.
To infer the effective connectivity parameters at the first and second level analyses, we used the Parametric Empirical Bayes (PEB)
framework as implemented in SPM12.
47
This hierarchical approach allows estimations of the fully connected DCM models using
network node time-series and evaluations of the nested (reduced) network models using Bayesian model reduction (BMR).
At the first level, we estimated individual fully connected bilinear DCM models of six key nodes. Based on previous
research
12,21,22,31,40
and current findings, six ROIs centered around the individual functional activity peaks referenced to the corre-
spondent anatomical structures and group ROIs were considered as key nodes of emotion regulation and social behavior network.
These nodes included arvlPFC, bilateral posterior ventrolateral prefrontal cortex (pvlPFC), bilateral amygdala (AMY), dmPFC along
with adjacent bilateral superior frontal gyrus (SFG), sgACC, and vmPFC. For each individual ROI average time-series, we regressed
out head motion parameters, linear trend, and constant term using GLM, and high-pass filtered with 0.008Hz cutoff. We averaged
bilateral SFG, pvlPFC and AMY time-series across left and right hemispheres.
21,101
For fully connected DCM models, the external
stimuli influenced all six model nodes (i.e., picture presentation conditions), as well as contextual modulations (i.e., emotion regula-
tion condition) influenced bidirectional connections and not self-inhibitory.
At the second level, we identified the PEB network models and their parameters that describe group average endogenous
connections and their contextual modulations for emotion regulation processes, as well as network architectures associated with
individual sociality scores applied as covariates. In the PEB framework, this was accomplished by comparing the evidence of the
reduced models with certain combinations of parameters switched off, which could be derived analytically from the full model using
Bayesian Model Reduction (BMR).
47
Specifically, network models varied separately for endogenous connections between nodes
(i.e., comparing evidence for reduced endogenous connectivity models, matrix A) and contextual modulations of functional coupling
strengths between nodes (i.e., comparing evidence for reduced contextual connectivity models, matrix B). We used an automatic
model search over all possible reduced network models to identify parameters that did not contribute to the model evidence.
Bayesian Model Averaging (BMA) over 256 best models from the last iteration of the model search procedure was used to reveal
the most plausible endogenous connections and contextual modulations, respectively. The posterior probability (Pp) for each
PEB parameter was calculated by comparing the evidence of all models which had the corresponding connectivity parameter
switched on, versus all models which had that parameter switched off. The Pp > 0.95 corresponds to the strong evidence.
Analyses of the behavioral and psychometric data
For the second experiment, we performed a cross-correlation analysis of behavioral ratings and psychometric scores using Pearson
correlation. We also investigated associations between individual brain activity, connectivity, psychometric scores, and behavioral
ratings with focus on sociality scores of neutral and positive social pictures. For ROI activity estimates, we extracted whole-brain
GLM contrast betas (‘‘upregulate positive social - view neutral social pictures’’, ‘‘upregulate positive social - fixation’’). For ROI con-
nectivity estimates, we performed the follow-up correlations for illustration purposes. The statistical significance was corrected for
multiple comparisons using false discovery rate (FDR, p< 0.05). For average values, we reported the mean and standard devia-
tion (SD).
e4 iScience 28, 111909, February 21, 2025
iScienc
e
Article
ll
OPEN ACCESS