Hostname: page-component-78c5997874-8bhkd Total loading time: 0 Render date: 2024-11-10T17:25:04.618Z Has data issue: false hasContentIssue false

The interaction between early life complications and a polygenic risk score for schizophrenia is associated with brain activity during emotion processing in healthy participants

Published online by Cambridge University Press:  02 February 2024

Veronica Debora Toro
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy Department of Humanities, University of Foggia, Foggia, Italy
Linda A. Antonucci
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy
Tiziana Quarto
Affiliation:
Department of Humanities, University of Foggia, Foggia, Italy
Roberta Passiatore
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy
Leonardo Fazio
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy Department of Medicine and Surgery, Libera Università Mediterranea “Giuseppe Degennaro”, Bari, Italy
Gianluca Ursini
Affiliation:
Lieber Institute for Brain Development, Johns Hopkins Medical Campus, Baltimore, MD, USA Department of Psychiatry and Behavioral Sciences, Johns Hopkins University, Baltimore, MD, USA
Qiang Chen
Affiliation:
Lieber Institute for Brain Development, Johns Hopkins Medical Campus, Baltimore, MD, USA
Rita Masellis
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy U.O.C. Psichiatria Universitaria, Azìenda Ospedaliero-Universitaria Consorziale Policlinico, Bari, Italy
Silvia Torretta
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy
Leonardo Sportelli
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy
Gianluca Christos Kikidis
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy
Francesco Massari
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy
Enrico D'Ambrosio
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy Department of Psychosis Studies, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, SE5 8AF, UK
Antonio Rampino
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy U.O.C. Psichiatria Universitaria, Azìenda Ospedaliero-Universitaria Consorziale Policlinico, Bari, Italy
Giulio Pergola
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy Lieber Institute for Brain Development, Johns Hopkins Medical Campus, Baltimore, MD, USA
Daniel R. Weinberger
Affiliation:
Lieber Institute for Brain Development, Johns Hopkins Medical Campus, Baltimore, MD, USA
Alessandro Bertolino
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy U.O.C. Psichiatria Universitaria, Azìenda Ospedaliero-Universitaria Consorziale Policlinico, Bari, Italy
Giuseppe Blasi*
Affiliation:
Psychiatric Neuroscience Group, Department of Translational Biomedicine and Neuroscience, University of Bari “Aldo Moro”, Bari, Italy U.O.C. Psichiatria Universitaria, Azìenda Ospedaliero-Universitaria Consorziale Policlinico, Bari, Italy
*
Corresponding author: Giuseppe Blasi; Email: giuseppe.blasi@uniba.it
Rights & Permissions [Opens in a new window]

Abstract

Background

Previous evidence suggests that early life complications (ELCs) interact with polygenic risk for schizophrenia (SCZ) in increasing risk for the disease. However, no studies have investigated this interaction on neurobiological phenotypes. Among those, anomalous emotion-related brain activity has been reported in SCZ, even if evidence of its link with SCZ-related genetic risk is not solid. Indeed, it is possible this relationship is influenced by non-genetic risk factors. Thus, this study investigated the interaction between SCZ-related polygenic risk and ELCs on emotion-related brain activity.

Methods

169 healthy participants (HP) in a discovery and 113 HP in a replication sample underwent functional magnetic resonance imaging (fMRI) during emotion processing, were categorized for history of ELCs and genome-wide genotyped. Polygenic risk scores (PRSs) were computed using SCZ-associated variants considering the most recent genome-wide association study. Furthermore, 75 patients with SCZ also underwent fMRI during emotion processing to verify consistency of their brain activity patterns with those associated with risk factors for SCZ in HP.

Results

Results in the discovery and replication samples indicated no effect of PRSs, but an interaction between PRS and ELCs in left ventrolateral prefrontal cortex (VLPFC), where the greater the activity, the greater PRS only in presence of ELCs. Moreover, SCZ had greater VLPFC response than HP.

Conclusions

These results suggest that emotion-related VLPFC response lies in the path from genetic and non-genetic risk factors to the clinical presentation of SCZ, and may implicate an updated concept of intermediate phenotype considering early non-genetic factors of risk for SCZ.

Type
Original Article
Copyright
Copyright © The Author(s), 2024. Published by Cambridge University Press

Introduction

Schizophrenia (SCZ) is a brain disorder associated with an high heritability reaching up to 80% (Owen, Sawa, & Mortensen, Reference Owen, Sawa and Mortensen2016; Sullivan, Reference Sullivan2005) and complex genetic architecture. In particular, the largest genome-wide association study (GWAS) to date has tackled the genetic complexity of this disorder, identifying 287 independent genetic loci associated with its diagnosis (Trubetskoy et al., Reference Trubetskoy, Pardiñas, Qi, Panagiotaropoulou, Awasthi, Bigdeli and O'Donovan2022). A commonly used strategy to shape a profile of such a complex genetic risk is the computation of a SCZ-related polygenic risk score (PRS) (Consortium, Reference Consortium2014; Purcell et al., Reference Purcell, Wray, Stone, Visscher, O'Donovan, Sullivan and Sklar2009; Trubetskoy et al., Reference Trubetskoy, Pardiñas, Qi, Panagiotaropoulou, Awasthi, Bigdeli and O'Donovan2022; Wray, Goddard, & Visscher, Reference Wray, Goddard and Visscher2007), which sums the weighted effect size of GWAS-significant alleles on diagnosis and indexes the cumulative polygenic risk for SCZ for each individual. This method has often been used to investigate PRS association with disease-related phenotypes at biological and clinical levels (Cosgrove et al., Reference Cosgrove, Harold, Mothersill, Anney, Hill, Bray and Donohoe2017; Fusar-Poli et al., Reference Fusar-Poli, Pries, van Os, Erzin, Delespaul, Kenis and Guloksuz2022; Lancaster et al., Reference Lancaster, Ihssen, Brindley, Tansey, Mantripragada, O'Donovan and Linden2016; Walton et al., Reference Walton, Geisler, Lee, Hass, Turner, Liu and Ehrlich2014).

Other evidence suggests that non-genetic factors such as complications during pregnancy, labor, delivery, and neonatal periods, also referred to as early life complications (ELCs) (Ursini et al., Reference Ursini, Punzi, Chen, Marenco, Robinson, Porcelli and Weinberger2018) increase the risk for SCZ. (Cannon, Jones, & Murray, Reference Cannon, Jones and Murray2002; McNeil, Cantor-Graae, & Sjöström, Reference McNeil, Cantor-Graae and Sjöström1994). More importantly, a recent report has indicated that exposure to ELCs increases the liability for SCZ explained by cumulative genetic risk for the disease, as indexed by PRS (Ursini et al., Reference Ursini, Punzi, Chen, Marenco, Robinson, Porcelli and Weinberger2018). More recently, another study using a different method to evaluate ELCs failed to demonstrate their statistically significant interaction with PRS (Vassos et al., Reference Vassos, Kou, Tosato, Maxwell, Dennison, Legge and Murray2022).

An important topic in this field is the investigation of the relationship between risk factors of SCZ and biological phenotypes of relevance for the disorder. In this regard, previous literature has highlighted the concept of ‘intermediate phenotypes’, which are quantitative correlates of the illness linked with its genetic risk (Bertolino & Blasi, Reference Bertolino and Blasi2009), expressed in healthy relatives and individuals at clinical and genetic risk, and thus etiologically independent from state-related variables including pharmacological treatment, symptoms, and chronicity. Indeed, the relationship between putative intermediate phenotypes and genetic risk for SCZ should be best provided in healthy subjects to prevent the effects of confounding factors related to the state of the disease that might affect the biology of the brain, as done previously (Chen et al., Reference Chen, Ursini, Romer, Knodt, Mezeivtch, Xiao and Weinberger2018; Dimitriadis et al., Reference Dimitriadis, Lancaster, Perry, Tansey, Jones, Singh and Linden2021; Koch, Nyberg, Lundquist, & Kauppi, Reference Koch, Nyberg, Lundquist and Kauppi2022; Miller et al., Reference Miller, Scult, Conley, Chen, Weinberger and Hariri2018).

Crucial phenotypes of SCZ are those related to anomalies in emotion processing (Kohler & Martin, Reference Kohler and Martin2006). This domain is modulated by genetic variation (Drabant et al., Reference Drabant, Hariri, Meyer-Lindenberg, Munoz, Mattay, Kolachana and Weinberger2006; Hariri et al., Reference Hariri, Drabant, Munoz, Kolachana, Mattay, Egan and Weinberger2005, Reference Hariri, Mattay, Tessitore, Kolachana, Fera, Goldman and Weinberger2002; Lo Bianco et al., Reference Lo Bianco, Blasi, Taurisano, Di Giorgio, Ferrante, Ursini and Bertolino2013; Smolka et al., Reference Smolka, Schumann, Wrase, Grüsser, Flor, Mann and Heinz2005) possibly relevant to SCZ (Shifman et al., Reference Shifman, Bronstein, Sternfeld, Pisanté, Weizman, Reznik and Darvasi2004; Williams, Owen, & O'Donovan, Reference Williams, Owen and O'Donovan2007) and includes both simple and complex subprocesses, by which emotions are perceived, recognized, as well as regulated and managed (Green et al., Reference Green, Penn, Bentall, Carpenter, Gaebel, Gur and Heinssen2008). Several functional imaging studies have convergently indicated in SCZ abnormal activation of different brain regions during emotion processing, including the ventrolateral prefrontal cortex (VLPFC), as well as the dorsolateral prefrontal cortex, the amygdala and the hippocampus (Dyck, Loughead, Gur, Schneider, & Mathiak, Reference Dyck, Loughead, Gur, Schneider and Mathiak2014; Dzafic, Burianová, Periyasamy, & Mowry, Reference Dzafic, Burianová, Periyasamy and Mowry2018; Etkin, Büchel, & Gross, Reference Etkin, Büchel and Gross2015; Frank et al., Reference Frank, Dewitt, Hudgens-Haney, Schaeffer, Ball, Schwarz and Sabatinelli2014; Mier et al., Reference Mier, Lis, Zygrodnik, Sauer, Ulferts, Gallhofer and Kirsch2014; Spilka & Goghari, Reference Spilka and Goghari2017). However, the directionality of such abnormalities is not always consistent in the literature. In fact, either hyper-(Dyck et al., Reference Dyck, Loughead, Gur, Schneider and Mathiak2014; Dzafic et al., Reference Dzafic, Burianová, Periyasamy and Mowry2018; Taylor, Liberzon, Decker, & Koeppe, Reference Taylor, Liberzon, Decker and Koeppe2002) or hypo-activation (Anticevic et al., Reference Anticevic, Van Snellenberg, Cohen, Repovs, Dowd and Barch2012; Paradiso et al., Reference Paradiso, Andreasen, Crespo-Facorro, O'Leary, Watkins, Boles Ponto and Hichwa2003; Spilka & Goghari, Reference Spilka and Goghari2017; Taylor et al., Reference Taylor, Kang, Brege, Tso, Hosanagar and Johnson2012) of the emotion-related brain network have been reported, which may partly rely on differences in task and study design, as well as on state-related characteristics of patients. Furthermore, it is unclear if abnormal brain processing of emotions in SCZ is linked with genetic risk for the disorder and may be considered an intermediate phenotype for the disease. For example, results in at-risk individuals are not consistent and indicate either lack of anomalies (Martin et al., Reference Martin, Croft, Pitt, Strelchuk, Sullivan and Zammit2020; Rasetti et al., Reference Rasetti, Mattay, Wiedholz, Kolachana, Hariri, Callicott and Weinberger2009; van Donkersgoed, Wunderink, Nieboer, Aleman, & Pijnenborg, Reference van Donkersgoed, Wunderink, Nieboer, Aleman and Pijnenborg2015) or abnormal activity in brain regions included in the emotion-related brain network (Park et al., Reference Park, Yun, Shin, Kim, Jung, Shin and Kwon2016; van Buuren, Vink, Rapcencu, & Kahn, Reference van Buuren, Vink, Rapcencu and Kahn2011; Wang et al., Reference Wang, Li, Liu, Wei, Jiang, Lui and Chan2018; Wolf et al., Reference Wolf, Satterthwaite, Calkins, Ruparel, Elliott, Hopson and Gur2015). Furthermore, findings from a recent meta-analysis indicate a lack of differential brain response between at-risk individuals and healthy participants (HP) (Fiorito et al., Reference Fiorito, Aleman, Blasi, Bourque, Cao, Chan and Sescousse2022). Accordingly, results on the association between brain emotion processing and polygenic risk for SCZ are sparse. To our knowledge, such a link has been reported in only one study using a limited sample size, in which a positive correlation of prefrontal, amygdala and insula activity with a SCZ PRS was found during anger perception, which shifted to a negative correlation during perception of happiness (Dzafic et al., Reference Dzafic, Burianová, Periyasamy and Mowry2018). On the other hand, another study using a larger sample size did not find any correlation between PRS for this disorder and brain processing of emotional stimuli, which was instead associated with a single nucleotide polymorphism (SNP) (Erk et al., Reference Erk, Mohnke, Ripke, Lett, Veer, Wackerhagen and Walter2017). Taken together, all these findings suggest that it is doubtful that brain activity during emotion processing falls within the standard definition of intermediate phenotype for SCZ because of the lack of solid evidence on its association with at-risk conditions and disease-related genetics. On the other hand, another interpretation may be based on the hypothesis that phenotypes crucially linked with SCZ are expressed in presence of the interaction between genetic and non-genetic factors of risk (Ursini et al., Reference Ursini, Punzi, Chen, Marenco, Robinson, Porcelli and Weinberger2018). As a consequence, with the currently available resolution and statistical power, it is also possible that these phenotypes are not detectable in individuals at genetic risk when they are not exposed to such non-genetic factors.

The aim of the present study was to investigate the interaction between ELCs and polygenic risk factors for SCZ on physiological brain response during emotion processing. With this aim, we investigated in a discovery analysis whether a PRS, based on genetic risk variants derived from the largest SCZ GWAS to date (Trubetskoy et al., Reference Trubetskoy, Pardiñas, Qi, Panagiotaropoulou, Awasthi, Bigdeli and O'Donovan2022), interacts with ELCs in modulating brain response in HP during tasks eliciting either emotional perceptual processing (implicit processing) or explicit emotional evaluation (explicit processing) (Blasi et al., Reference Blasi, Hariri, Alce, Taurisano, Sambataro, Das and Mattay2009a; Blasi et al., Reference Blasi, Lo Bianco, Taurisano, Gelao, Romano, Fazio and Bertolino2009b; Quarto et al., Reference Quarto, Paparella, De Tullio, Viscanti, Fazio, Taurisano and Blasi2018). To strengthen the findings obtained in the discovery sample, we then performed two further parallel investigations: first, we used an independent sample of HP to replicate the results found in the discovery sample. Second, we performed a further analysis to verify whether the pattern of brain response found in the discovery and replication analyses in HP was coherent with those associated with the effect of the diagnosis of SCZ.

Materials and methods

Participants

Discovery sample of HP

One hundred sixty-nine HP (Table 1) were included in the discovery sample. All participants were white Caucasians from the region of Puglia, Italy (see online Supplementary material).

Table 1. Demographics and clinical variables of the samples of the study

HP, healthy participants; SCZ patients with schizophrenia; SES, socioeconomic status; premorbid IQ premorbid intelligence quotient; PANSS, Positive and Negative Syndrome Scale.

All HP were genotyped for variations associated with SCZ (Trubetskoy et al., Reference Trubetskoy, Pardiñas, Qi, Panagiotaropoulou, Awasthi, Bigdeli and O'Donovan2022) and the McNeil–Sjöström scale (McNeil et al., Reference McNeil, Cantor-Graae and Sjöström1994) was used to assess the presence of serious ELCs, as described in previous work (Ursini et al., Reference Ursini, Punzi, Chen, Marenco, Robinson, Porcelli and Weinberger2018). Furthermore, they underwent fMRI scanning during a task eliciting emotion processing (Blasi et al., Reference Blasi, Hariri, Alce, Taurisano, Sambataro, Das and Mattay2009a; Quarto et al., Reference Quarto, Blasi, Maddalena, Viscanti, Lanciano, Soleti and Curci2016; Taurisano et al., Reference Taurisano, Blasi, Romano, Sambataro, Fazio, Gelao and Bertolino2013).

Replication sample of HP

113 HP (76 with ELCs) (Table 1) were used as a replication sample in this study (see online Supplementary material). Genotyping and assessment of the presence of serious ELCs were performed as for the discovery sample. All these individuals performed a task eliciting emotion processing during fMRI (Hariri, Bookheimer, & Mazziotta, Reference Hariri, Bookheimer and Mazziotta2000).

Sample of SCZ

Seventy-five white Caucasian Apulian SCZ (Table 1), as diagnosed with the SCID, were used to investigate the effect of diagnosis on brain activity during emotion processing (see online Supplementary material). They also underwent fMRI scanning during the same emotion task used for the discovery sample (Blasi et al., Reference Blasi, Hariri, Alce, Taurisano, Sambataro, Das and Mattay2009a; Quarto et al., Reference Quarto, Blasi, Maddalena, Viscanti, Lanciano, Soleti and Curci2016; Taurisano et al., Reference Taurisano, Blasi, Romano, Sambataro, Fazio, Gelao and Bertolino2013).

All participants provided written informed consent to the study fully explained the procedure of them. The study was approved by the Italian local Institutional Review Board, as well as by the Institutional Review Board of the Intramural Program of the National Institute of Mental Health.

Computation of PRS

After genotyping (see online Supplementary material), PRS for each HP of the discovery and replication samples was calculated using the sum of an individual's statistically independent risk alleles, weighted by their effect size (odds ratio) derived from the Psychiatric Genomics Consortium Wave 3 (PGC3) SCZ GWAS (Trubetskoy et al., Reference Trubetskoy, Pardiñas, Qi, Panagiotaropoulou, Awasthi, Bigdeli and O'Donovan2022). To prevent any bias related to sex in the PRS calculation, only autosomal SNPs were included in the analysis, according to the standard procedure for PRS calculation (Consortium, Reference Consortium2014; Purcell et al., Reference Purcell, Wray, Stone, Visscher, O'Donovan, Sullivan and Sklar2009; Trubetskoy et al., Reference Trubetskoy, Pardiñas, Qi, Panagiotaropoulou, Awasthi, Bigdeli and O'Donovan2022). We performed a linkage disequilibrium (LD) clumping of the SNPs, according to parameters used by PGC (Consortium, Reference Consortium2014; Trubetskoy et al., Reference Trubetskoy, Pardiñas, Qi, Panagiotaropoulou, Awasthi, Bigdeli and O'Donovan2022), defining LD-independent SNPs as those with low LD (r2 < 0.1) within a 500 kb window. A PRS (PRS_GWA) was computed as done previously (Pergola et al., Reference Pergola, Papalino, Gelao, Sportelli, Vollerbergh, Grattagliano and Bertolino2019) using genome-wide significant (p < 5e-08) SNPs, according to PGC3 (Trubetskoy et al., Reference Trubetskoy, Pardiñas, Qi, Panagiotaropoulou, Awasthi, Bigdeli and O'Donovan2022), present in our dataset (n = 207). A further score was calculated in the discovery sample, using genetic variants whose statistics for the association with diagnosis of SCZ was at p < 0.05 (PRS_05,), in order to include in such PRS genetic variants that are likely not key for the pathophysiology of this brain disorder.

Since the discovery dataset was included in the PGC3 study, we used the leave-site-out summary statistics for this sample. All PRSs were adjusted for genotyping batch and for the first 10 genomic principal components to account for population stratification within the sample (Consortium, Reference Consortium2014; Trubetskoy et al., Reference Trubetskoy, Pardiñas, Qi, Panagiotaropoulou, Awasthi, Bigdeli and O'Donovan2022). Marginalized PRSs were transformed in units of standard deviations from the mean.

Assessment of ELCs: the McNeil- Sjöström scale

ELCs (Ursini et al., Reference Ursini, Punzi, Chen, Marenco, Robinson, Porcelli and Weinberger2018) were evaluated in both discovery and replication samples using a well-standardized and validated questionnaire (Nicodemus et al., Reference Nicodemus, Marenco, Batten, Vakkalanka, Egan, Straub and Weinberger2008), based on all the items scored with the McNeil–Sjöström scale for obstetric complications (McNeil et al., Reference McNeil, Cantor-Graae and Sjöström1994), administered to the mothers of HP. ELCs are classified according to time of onset and severity. With regard to the latter, a severity level between 1 and 6 was assigned to each ELCs, based on the degree of inferred potential harm to the offspring's central nervous system (Ursini et al., Reference Ursini, Punzi, Chen, Marenco, Robinson, Porcelli and Weinberger2018). ELCs with severity level ⩾4 are considered potentially harmful or relevant factors for fetal stress (Ursini et al., Reference Ursini, Punzi, Chen, Marenco, Robinson, Porcelli and Weinberger2018). Using a previously published method (Ursini et al., Reference Ursini, Punzi, Chen, Marenco, Robinson, Porcelli and Weinberger2018), we defined a positive history for serious ELCs based on the presence of at least one serious ELC (severity level ⩾4). On this basis, 111 out of 169 HP had a positive history of serious ELCs in discovery sample while in the replication sample those with a positive history of serious ELCs were 76 out of 113 (see online Supplementary table 1 for the description of available information on ELCs).

The frequency of ELCs in both samples is likely higher compared with the general population, as reported in previous work (Ursini et al., Reference Ursini, Punzi, Chen, Marenco, Robinson, Porcelli and Weinberger2018). This is likely related to the fact that we included in our samples only participants with certain information about the presence or absence of ELCs, since – when using a questionnaire with many items – it is often more challenging to exclude at least one serious ELCs than confirm its presence (for example, if a mother was not able to clarify whether her newborn was just ‘pale’ or ‘anoxic’, we had to exclude the subject from the sample if there are no other indicators of ELCs, while we were able to include him/her in the sample when other clear ELCs, like preeclampsia, IUGR, etc. are reported).

fMRI experimental paradigm and data acquisition for the discovery and SCZ samples

The event-related fMRI paradigm (Blasi et al., Reference Blasi, Lo Bianco, Taurisano, Gelao, Romano, Fazio and Bertolino2009b; Quarto et al., Reference Quarto, Blasi, Maddalena, Viscanti, Lanciano, Soleti and Curci2016; Taurisano et al., Reference Taurisano, Blasi, Romano, Sambataro, Fazio, Gelao and Bertolino2013) consisted of two runs: each run presented angry, fearful, happy, and neutral facial expressions from a validated set of facial pictures (NimStim, http://www.macbrain.org/resources.htm) (Tottenham et al., Reference Tottenham, Tanaka, Leon, McCarry, Nurse, Hare and Nelson2009). During one run (emotional perceptual processing: implicit processing), participants identified the sex of each face. In the other run (explicit emotional evaluation: explicit processing), they had to decide whether they would like to ‘approach’ or ‘avoid’ the face (see online Supplementary materials).

Blood Oxygen Level Dependent (BOLD) fMRI was performed on a GE Signa 3 T scanner while participants performed the task (see online Supplementary material).

fMRI experimental paradigm and data acquisition for the replication analysis

The fMRI paradigm for the replication sample consisted in a face matching task, which has been used in several previous studies (Hariri et al., Reference Hariri, Bookheimer and Mazziotta2000, Reference Hariri, Mattay, Tessitore, Kolachana, Fera, Goldman and Weinberger2002, Reference Hariri, Drabant, Munoz, Kolachana, Mattay, Egan and Weinberger2005). Participants are presented with blocks of trials that either ask them to decide which of two faces presented on the bottom of the screen match the face at the top of the screen, or which of two shapes presented at the bottom of the screen match the shape at the top of the screen. The faces have either an angry or fearful expression. A complete description of the task can be found in previous published papers (Hariri et al., Reference Hariri, Bookheimer and Mazziotta2000).

Also for the replication sample, BOLD fMRI was performed on a GE Signa 3 T scanner during task performance (see online Supplementary material).

Data analysis

Demographics and behavioral data

ANOVAs and χ2 were used to assess distribution of demographic variables between groups in the discovery, replication and SCZ samples. A repeated-measures ANCOVA, with ELCs and PRS as independent factors, and the two tasks as well as the four facial expressions as repeated-measures factors, was used to investigate the interaction between ELCs and PRS on reaction time data. Furthermore, a repeated-measures ANCOVA, with ELCs and PRS as independent factors, and the four facial expressions as repeated-measures factors, was performed to investigate the interaction between ELCs and PRS on the number of avoided faces. Age, sex, premorbid IQ, and SES were used as covariates of no interest.

fMRI data analysis

Discovery analysis

Analysis of the fMRI data was completed using Statistical Parametric Mapping 12 (SPM12, Wellcome Department of Cognitive Neurology, London, UK). After preprocessing of images (see online Supplementary material), in a first-level, single-subject analysis, linear contrasts were computed producing t statistical maps at each voxel for the four task conditions (corresponding to happy, angry, fearful, and neutral facial expressions of stimuli used in the event-related task) compared with the baseline condition (crosshair).

All individual contrast images were then entered in second-level random-effects analyses at the group level. First, a full factorial general linear model, covarying for age, sex, Premorbid IQ, and SES, was performed in HP to investigate the main effect of ELCs, of PRSs, their interaction, as well as their potential interactions with task and facial expressions. Such investigation was first performed using PRS_GWA, and then separately repeated with PRS_05.

All analyses were constrained by a mask obtained by combining group activation maps associated with the processing of each facial expression during each task and a gray matter mask implemented in the WFU PickAltas (http://fmri.wfubmc.edu/software/PickAtlas). The statistical threshold was set at p < 0.05 minimum cluster size (k) = 10, corrected for multiple comparisons using Threshold-Free Cluster Enhancement (TFCE-FWE<0.05) (Smith & Nichols, Reference Smith and Nichols2009). Then, percent signal change from the cluster associated with a PRS_GWA by ELCs interaction was extracted with MarsBar (https://www.nitrc.org/projects/marsbar/) and entered in Pearson's correlation tests, performed with IBM SPSS 25.0, to investigate its statistical relationship with PRS in individuals with and without severe ELCs separately.

Replication analysis

Also in the replication sample, analysis of the fMRI data was completed using SPM12. After preprocessing (see online Supplementary material), in the first-level analysis, matching vs resting were modeled using a box car convolved with the hemodynamic response function at each voxel. Head movement parameters obtained from the realignment procedure were included in the model as covariates, taking into account the effects of subject motion. A contrast map was generated for each individual using the beta value of face matching stimulus v. fixation cross-hair at each voxel.

Thus, we focused the group analysis on a region-of-interest (ROI) defined with MarsBar (https://www.nitrc.org/projects/marsbar/), using a sphere with a radius of 20 mm and centered on the activation peak of the cluster associated with a ELCs-PRS_GWA interaction in the discovery sample. In more detail, we extracted BOLD signal change from this ROI, which was used as a dependent variable, with ELCs (yes/no) and PRS as factors of interest and age, sex and premorbid IQ as covariates of no interest. Furthermore, Pearson's correlation test was used to investigate the relationship between BOLD signal change and PRS in individuals with and without severe ELCs separately.

Effect of diagnosis of schizophrenia

In a further analysis, HP of the discovery sample were compared with SCZ with ANCOVA to verify coherence of the brain imaging phenotype associated with the PRS_GWA by ELCs interaction in HP with those of full-blown disease. Again, we defined a 20 mm radius spherical ROI using MarsBaR, centered on the activation peak of the cluster associated with a ELCs-PRS_GWA interaction in the discovery sample. Given the aim of this analysis and the results of the ELCs-PRS_GWA interaction in HP, the interaction between diagnosis and task and/or emotion was not explored, even if these latter factors were inserted in the statistical design. Age, sex, premorbid IQ, and SES were also added as factors of no interest. Given that here the investigation was focused on the effect of diagnosis instead of the interaction between PRS and ELCs, a TFCE-FWE correction at p < 0.05 was used.

Results

Demographics and behavioral data

T-tests or χ2 in the discovery and replication sample revealed homogeneous distribution of age, sex, premorbid IQ and SES between HP with and without an history of serious ELCs (all p > 0.05). There were no effects of PRSs, ELCs, of their interaction, as well as of their interaction with task and/or facial expression on reaction time (all p > 0.05). Furthermore, no significant main effects of PRSs, ELCs, of their interaction, as well as of their interaction with facial expressions were found on the number of avoided faces (all p > 0.05). With regard to the analysis in the SCZ sample, age, sex, premorbid IQ and SES were significantly different between HP and SCZ (all p < 0.001).

Imaging data

Discovery analysis

A general linear model performed in HP did not reveal significant main effects of PRS_GWA or ELCs on brain activity. Instead, their interaction was significant in Left VLPFC (x, y, z: −52, 16, 16; k: 24; Z: 3.89) (BA44) (Fig. 1) (see also online Supplementary Figure 1 for whole brain results). No further interactions with tasks or/and facial expression were found. Post-hoc investigation on signal change extracted from the significant VLPFC cluster indicated a positive correlation between PRS_GWA and VLPFC activity in participants with a history of serious ELCs (r = 0.269; p = 0.008), while the correlation was negative in participants without history of ELCs (r = −0. 433; p < 0.001). Further separate ANCOVA using PRS_05 did not indicate significant PRS by ELCs interaction on brain activity during emotion processing.

Figure 1. Discovery sample: (A) Axial section of the brain showing the cluster in VLPFC where an interaction between PRS_GWA and ELCs was present; (B) scatterplot of % signal change extracted from the cluster represented in A. as a function of PRS_GWA and ELCs. See results for statistics. PRS_GWA: polygenic risk score genome-wide associated with diagnosis of schizophrenia.

Replication analysis

ANCOVA on signal change, extracted from the ROI centered on the peak of activity of the cluster associated with a ELCs-PRS_GWA interaction in the discovery analysis, revealed an interaction between PRS_GWA and ELCs also in replication sample (F = 7.42; p = 0.008). Moreover, Pearsons’ test indicated a positive correlation between PRS_GWA and VLPFC activity in participants with a history of serious ELCs (r = 0.277; p = 0.018), while such a relationship was not significant in participants without a history of ELCs (r = −0.198; p = 0.258) (Fig. 2) (see also online Supplementary Table 2 for whole brain results in this sample).

Figure 2. Replication sample: scatterplot of % signal change extracted from the 20 mm ROI centered on the peak voxel associated with a PRS_GWA by ELCs interaction in the discovery analysis. See results for statistics. PRS_GWA: polygenic risk score genome-wide associated with diagnosis of schizophrenia

Effect of diagnosis of schizophrenia

There was an effect of diagnosis on the ROI centered on the activation peak of the cluster associated with a ELCs-PRS_GWA interaction (PTFCE−FWE:0.016, x,y,z: −59, 8, 1) (BA 44) (Fig. 3a), with greater activity in SCZ than HP (Fig. 3b and online Supplementary figure 2).

Figure 3. (A) Axial section of the brain showing the cluster in left VIPFC associated with a main effect of diagnosis during emotion processing; (B) Mean BOLD signal change in SCZ and HP extracted from this cluster shown in A. See results for statistics. PRS_GWA: polygenic risk score genome-wide associated with diagnosis of schizophrenia. HP: healthy participants; SCZ: patients with schizophrenia

Discussion

The results of this study suggest that genetic and non-genetic factors interacting on diagnosis of SCZ (Ursini et al., Reference Ursini, Punzi, Chen, Marenco, Robinson, Porcelli and Weinberger2018) have also a combined effect on physiological brain response during emotion processing. Paramount in this context is VLPFC activity, which was associated in HP with the interaction between polygenic risk for SCZ and ELCs in two independent samples. Furthermore, VLPFC response during emotion processing was also anomalous in SCZ. Taken together, these results suggest a brain phenotype on the biological pathway connecting genetic and non-genetic risk with the clinical presentation of this brain disorder.

We found that the association between left VLPFC response during emotion processing and PRS is modulated in HP by ELCs in the discovery analysis. This interaction was present when investigating PRS_GWA, but not when looking at PRS_05. Importantly, we replicated such an interaction in an independent American sample of HP. PRS_GWA is a genetic score computed using genome-wide significant SCZ-associated alleles (Trubetskoy et al., Reference Trubetskoy, Pardiñas, Qi, Panagiotaropoulou, Awasthi, Bigdeli and O'Donovan2022), which raises the possibility that only cumulative effect of SNPs linked the most with pathophysiological mechanisms of the disorder are relevant in modulating brain activity during emotion processing in interaction with ELCs, while adding more ‘genetic noise’ to the genetic features most associated with SCZ implicates a loss of such an association. This is consistent with previous findings indicating that only the PRS computed with the variants with the most significant association with SCZ interacts with ELCs in affecting SCZ case–control status (Ursini et al., Reference Ursini, Punzi, Chen, Marenco, Robinson, Porcelli and Weinberger2018) and early neurodevelopmental outcomes (Ursini et al., Reference Ursini, Punzi, Langworthy, Chen, Xia, Cornea and Weinberger2021).

Further analysis of left VLPFC BOLD response indicated that, in the discovery and replication samples, it was positively correlated with PRS_GWA in presence of ELCs, which suggests that VLPFC activity increases with the increase of cumulative genetic risk for SCZ, but only in the context of perinatal, non-genetic risk factors for the disease (Ursini et al., Reference Ursini, Punzi, Langworthy, Chen, Xia, Cornea and Weinberger2021). On the other hand, we found in the discovery sample that the correlation between VLPFC response and PRS_GWA was negative in absence of ELCs. Differential relationships between PRS and brain activity in presence and absence of ELCs may be interpreted in the context of the differential susceptibility framework, (Belsky, Reference Belsky2007; Belsky & Pluess, Reference Belsky and Pluess2009), which is closely related to the concept of biological sensitivity to context (Ellis, Essex, & Boyce, Reference Ellis, Essex and Boyce2005). According to this framework, individual attributes (such as genetic features associated with brain disorders) implicate disadvantageous outcomes in high-stress environments (e.g. ELC), while they are associated with even advantageous outcomes in favorable environmental conditions (Belsky, Reference Belsky2007; Belsky & Pluess, Reference Belsky and Pluess2009). However, the negative correlation between left VLPFC activity and PRS_GWA in absence of ELCs was not statistically significant in the replication analysis, suggesting that further investigation is needed to confirm this relationship.

VLPFC is part of the emotion brain network (Etkin et al., Reference Etkin, Büchel and Gross2015; Phillips, Drevets, Rauch, & Lane, Reference Phillips, Drevets, Rauch and Lane2003) and previous literature posits that it plays a role in attributing emotional significance to environmental stimuli and in affective state production (Phillips et al., Reference Phillips, Drevets, Rauch and Lane2003). Indeed, abnormal activity in this brain region has been found in patients with SCZ in terms of either hyper- (Dyck et al., Reference Dyck, Loughead, Gur, Schneider and Mathiak2014) or hypo-activity (van der Meer et al., Reference van der Meer, Swart, van der Velde, Pijnenborg, Wiersma, Bruggeman and Aleman2014). Interestingly, BA 44 has also been specifically associated with severe ELCs in patients with SCZ, with bipolar disorder and in HP (Wortinger et al., Reference Wortinger, Engen, Barth, Andreassen, Nordbø Jørgensen and Agartz2022).Thus, our findings in HP suggest that the interaction between genetic and non-genetic risk factors for SCZ modulates activity in a brain region involved in basic processing of emotions and abnormally activated in patients with this brain disorder. To our knowledge, there are no studies to date investigating the interaction between SCZ-related polygenic risk and non-genetic factors of liability on neurobiological phenotypes relevant to this brain disease. On the other hand, results on the association between polygenic risk for SCZ and brain activity during emotion processing are sparse (Dzafic et al., Reference Dzafic, Burianová, Periyasamy and Mowry2018; Erk et al., Reference Erk, Mohnke, Ripke, Lett, Veer, Wackerhagen and Walter2017). In particular, only one study using a limited sample size of healthy individuals reported emotion-specific correlation between a PRS for SCZ and brain activity (Dzafic et al., Reference Dzafic, Burianová, Periyasamy and Mowry2018). Indeed, another study using a large sample of individuals did not find an association between emotion processing-related brain response and cumulative genetic risk for the illness (Erk et al., Reference Erk, Mohnke, Ripke, Lett, Veer, Wackerhagen and Walter2017). Results from a recent meta-analysis (Fiorito et al., Reference Fiorito, Aleman, Blasi, Bourque, Cao, Chan and Sescousse2022) appear to be consistent with the findings of this latter report. More to the detail, this study did not find any whole-brain difference in activity during emotion processing between HP and individuals at-risk for psychosis considering seventeen previous reports (Fiorito et al., Reference Fiorito, Aleman, Blasi, Bourque, Cao, Chan and Sescousse2022). Thus, most of the previous literature on this topic is in line with our findings indicating a lack of association between polygenic risk for SCZ and emotion-related brain activity. However, our results also indicate that this association can be detected when non-genetic factors of liability for this brain disorder are taken into account. This finding may be interpreted in the context of the concept of incomplete penetrance (Gottesman & Shields, Reference Gottesman and Shields1967). According to this explanatory framework, individuals with a genetic environment of risk may or may not express a related phenotype, which may be instead elicited when other factors of risk are present. Thus, it is possible that ELCs increase penetrance of polygenic risk on brain correlates associated with emotion processing in SCZ. Indeed, this aspect may implicate an updated concept of intermediate phenotype, which should also consider the interaction between genetic and non-genetic factors to discover crucial biological correlates of the disorder.

The comparison that we performed between HP and SCZ is consistent with, and adds further evidence to, the relevance of the PRS-ELCs interaction that we found in HP for full-blown SCZ. Indeed, it is well established that brain activity during emotion processing is anomalous in this brain disorder (Kohler & Martin, Reference Kohler and Martin2006). However, inconsistent directionality of the findings has been reported across studies, showing either hyper- or hypo-activation in SCZ compared with HP, which may be explained by different factors including task and patient characteristics. Here, we found that activity in left VLPFC is greater in SCZ than HP, which appears to be consistent with our finding that the interaction between genetic and non-genetic risk factors for SCZ tends to increase response in this brain region in HP. Taken together, these results suggest that VLPFC response during emotion processing is crucial for pathophysiological aspects of SCZ and linked with risk factors for the disorder. Furthermore, they are also consistent with the neurodevelopmental hypothesis of SCZ (Weinberger, Reference Weinberger1987), which posits that adverse events occurring during early life may lead to dysfunction of specific neural networks that would account for SCZ-related phenotypes in later life. This hypothesis calls for further studies aimed to elucidate the biological mechanisms underlying such interaction, which may include epigenetic signals as those related to methylation of DNA (Jaffe et al., Reference Jaffe, Gao, Deep-Soboslay, Tao, Hyde, Weinberger and Kleinman2016) and transcriptomic changes in placenta (Ursini et al., Reference Ursini, Di Carlo, Mukherjee, Chen, Han, Kim and Weinberger2023)

Some limitations should be considered in this study. First, we did not investigate the interaction between PRS and ELCs in SCZ. However, our choice was to investigate this interaction in HP to avoid state-related confounding effects, including psychopharmacological treatment, symptoms, and chronicity. A second limitation is that the frequency of ELCs in our sample is likely higher compared with those in the general population. This may be related to the fact that the inclusion of only participants with certain information about the presence or absence of ELCs may result in a sample with a higher frequency of ELCs compared with the general population, since it is often more challenging to exclude at least one serious ELC than confirm its presence, particularly when using maternal interviews. Indeed, because the objective of our study is to test a gene-environment interaction, we have decided to include in our sample only participants with certain information about the presence or absence of ELCs. Another limitation is that ELCs have been assessed using maternal interviews rather than medical records, which implicates a lack of detailed information about all the ELCs occurring in each individual. This aspect makes it not possible to disentangle the role of a specific ELC in the interaction with PRS, as well as to precisely weight the different level of ELCs exposure of each individual. However, consistent literature indicates that ELCs usually do not occur in isolation (Davies, Bell, & Bhattacharya, Reference Davies, Bell and Bhattacharya2016; Feghali, Caritis, Catov, & Scifres, Reference Feghali, Caritis, Catov and Scifres2016; Murata et al., Reference Murata, Yasuda, Imaizumi, Isogami, Fukuda, Kyozuka and Fujimori2022; Rasmussen, Ebbing, & Irgens, Reference Rasmussen, Ebbing and Irgens2017; Yang & Wu, Reference Yang and Wu2022). Thus, it is possible that the pooled effect of different complications acting during the peri-natal period is crucial to increase risk for pathophysiological mechanisms relevant to SCZ in interaction with the cumulative effect of genetic features linked with this brain disorder.

In conclusion, our results suggest that ELCs converge with genomic risk factors in affecting anomalies of emotional processing relevant to SCZ. In this sense, further studies should consider the interaction between genetic and non-genetic factors to identify intermediate phenotypes of SCZ, which are key to add new knowledge on pathophysiological mechanisms for this brain disorder.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/S0033291724000011.

Authors’ contributions

GB conceived and designed the study paradigm. LAA, TQ, and LF collected the fMRI data. GU collected the genetic and ELCs data. GU, QC, and DRW collected, analyzed and provide fMRI and genetic data of the replication sample. RM and ST contributed to genotypization procedures. FM and ED contributed to the enrollment. VDT, LF, RP, TQ, and GP analyzed the fMRI data. LS and GCK analyzed the genetic data. RP provided contributions to the statistical analysis. VDT drafted the initial manuscript, and GB provided critical revisions. AR, GP, and AB revised the article. All of the authors discussed the results, commented, and approved the final manuscript for submission.

Funding statement

This study was funded from the Apulian regional government for the project: ‘Early Identification of Psychosis Risk’ to AB; from the funding initiative Horizon Europe Seeds 2021 (Next Generation 8 EU - MUR D.M. 737/2021) for the project S68 CUP H99J21017550006 to LAA, GP, AR, GB, AB; from European Union funding within the MUR PNRR Extended Partnership initiative on Neuroscience and Neuropharmacology (Project no. PE00000006 CUP H93C22000660006 ‘MNESYS, A multiscale integrated approach to the study of the nervous system in health and disease’) to GP, AR, GB, AB; from the Research Projects of National Relevance (PRIN) 2017 Prot. 2017K2NEF4 awarded to GP that funded the position of RP; by The Italian Ministry of Economic Development with the grant #F/200044/01-03/X45 awarded to AB that founded GCK's scholarship. The collection of the MRI and genetic data for the Replication sample was supported by direct funding from the Intramural Research Program of the NIMH to the Clinical Brain Disorders Branch (PI: DRW, protocol 95-M-0150).

Competing interests

GB has received lecture fees by Lundbeck. AB has received speaker's fees from Otsuka, Lundbeck, Angelini, Janssen. The authors declare no competing interests.

References

Anticevic, A., Van Snellenberg, J. X., Cohen, R. E., Repovs, G., Dowd, E. C., & Barch, D. M. (2012). Amygdala recruitment in schizophrenia in response to aversive emotional material: A meta-analysis of neuroimaging studies. Schizophrenia Bulletin, 38(3), 608621. doi:10.1093/schbul/sbq131.CrossRefGoogle ScholarPubMed
Belsky, J. (2007). For better and for worse: Differential susceptibility to environmental influences. Current Directions in Psychological Science, 16(6), 300304.CrossRefGoogle Scholar
Belsky, J., & Pluess, M. (2009). Beyond diathesis stress: Differential susceptibility to environmental influences. Psychological Bulletin, 135(6), 885908. doi:10.1037/a0017376.CrossRefGoogle ScholarPubMed
Bertolino, A., & Blasi, G. (2009). The genetics of schizophrenia. Neuroscience, 164(1), 288299. doi:10.1016/j.neuroscience.2009.04.038.CrossRefGoogle ScholarPubMed
Blasi, G., Hariri, A. R., Alce, G., Taurisano, P., Sambataro, F., Das, S., … Mattay, V. S. (2009a). Preferential amygdala reactivity to the negative assessment of neutral faces. Biological Psychiatry, 66(9), 847853. doi:10.1016/j.biopsych.2009.06.017.CrossRefGoogle Scholar
Blasi, G., Lo Bianco, L., Taurisano, P., Gelao, B., Romano, R., Fazio, L., … Bertolino, A. (2009b). Functional variation of the dopamine D2 receptor gene is associated with emotional control as well as brain activity and connectivity during emotion processing in humans. The Journal of Neuroscience: the Official Journal of the Society for Neuroscience, 29(47), 1481214819. doi:10.1523/jneurosci.3609-09.2009.CrossRefGoogle ScholarPubMed
Cannon, M., Jones, P. B., & Murray, R. M. (2002). Obstetric complications and schizophrenia: Historical and meta-analytic review. The American Journal of Psychiatry, 159(7), 10801092. doi:10.1176/appi.ajp.159.7.1080.CrossRefGoogle ScholarPubMed
Chen, Q., Ursini, G., Romer, A. L., Knodt, A. R., Mezeivtch, K., Xiao, E., … Weinberger, D. R. (2018). Schizophrenia polygenic risk score predicts mnemonic hippocampal activity. Brain, 141(4), 12181228. doi:10.1093/brain/awy004.CrossRefGoogle ScholarPubMed
Consortium, S. W. G. o. t. P. G. (2014). Biological insights from 108 schizophrenia-associated genetic loci. Nature, 511(7510), 421427. doi:10.1038/nature13595.Google Scholar
Cosgrove, D., Harold, D., Mothersill, O., Anney, R., Hill, M. J., Bray, N. J., … Donohoe, G. (2017). MiR-137-derived polygenic risk: Effects on cognitive performance in patients with schizophrenia and controls. Translational Psychiatry, 7(1), e1012. doi:10.1038/tp.2016.286.CrossRefGoogle ScholarPubMed
Davies, E. L., Bell, J. S., & Bhattacharya, S. (2016). Preeclampsia and preterm delivery: A population-based case-control study. Hypertension in Pregnancy, 35(4), 510519. doi:10.1080/10641955.2016.1190846.CrossRefGoogle ScholarPubMed
Dimitriadis, S. I., Lancaster, T. M., Perry, G., Tansey, K. E., Jones, D. K., Singh, K. D., … Linden, D. E. (2021). Global brain flexibility during working memory is reduced in a high-genetic-risk group for schizophrenia. Biological psychiatry. Cognitive Neuroscience and Neuroimaging, 6(12), 11761184. doi:10.1016/j.bpsc.2021.01.007.CrossRefGoogle Scholar
Drabant, E. M., Hariri, A. R., Meyer-Lindenberg, A., Munoz, K. E., Mattay, V. S., Kolachana, B. S., … Weinberger, D. R. (2006). Catechol O-methyltransferase val158met genotype and neural mechanisms related to affective arousal and regulation. Archives of General Psychiatry, 63(12), 13961406. doi:10.1001/archpsyc.63.12.1396.CrossRefGoogle ScholarPubMed
Dyck, M., Loughead, J., Gur, R. C., Schneider, F., & Mathiak, K. (2014). Hyperactivation balances sensory processing deficits during mood induction in schizophrenia. Social Cognitive and Affective Neuroscience, 9(2), 167175. doi:10.1093/scan/nss120.CrossRefGoogle ScholarPubMed
Dzafic, I., Burianová, H., Periyasamy, S., & Mowry, B. (2018). Association between schizophrenia polygenic risk and neural correlates of emotion perception. Psychiatry Research. Neuroimaging, 276, 3340. doi:10.1016/j.pscychresns.2018.04.005.CrossRefGoogle ScholarPubMed
Ellis, B. J., Essex, M. J., & Boyce, W. T. (2005). Biological sensitivity to context: II. Empirical explorations of an evolutionary-developmental theory. Development and Psychopathology, 17(2), 303328. doi:10.1017/s0954579405050157.CrossRefGoogle ScholarPubMed
Erk, S., Mohnke, S., Ripke, S., Lett, T. A., Veer, I. M., Wackerhagen, C., … Walter, H. (2017). Functional neuroimaging effects of recently discovered genetic risk loci for schizophrenia and polygenic risk profile in five RDoC subdomains. Translational Psychiatry, 7(1), e997. doi:10.1038/tp.2016.272.CrossRefGoogle ScholarPubMed
Etkin, A., Büchel, C., & Gross, J. J. (2015). The neural bases of emotion regulation. Nature Reviews. Neuroscience, 16(11), 693700. doi:10.1038/nrn4044.CrossRefGoogle ScholarPubMed
Feghali, M. N., Caritis, S. N., Catov, J. M., & Scifres, C. M. (2016). Timing of delivery and pregnancy outcomes in women with gestational diabetes. American Journal of Obstetrics and Gynecology, 215(2), 243.e241–247. doi:10.1016/j.ajog.2016.03.006.CrossRefGoogle ScholarPubMed
Fiorito, A. M., Aleman, A., Blasi, G., Bourque, J., Cao, H., Chan, R. C. K., … Sescousse, G. (2022). Are brain responses to emotion a reliable endophenotype of schizophrenia? An image-based fMRI meta-analysis. Biological Psychiatry, 93(2), 167177. doi:10.1016/j.biopsych.2022.06.013.CrossRefGoogle Scholar
Frank, D. W., Dewitt, M., Hudgens-Haney, M., Schaeffer, D. J., Ball, B. H., Schwarz, N. F., … Sabatinelli, D. (2014). Emotion regulation: Quantitative meta-analysis of functional activation and deactivation. Neuroscience and Biobehavioral Reviews, 45, 202211. doi:10.1016/j.neubiorev.2014.06.010.CrossRefGoogle ScholarPubMed
Fusar-Poli, L., Pries, L. K., van Os, J., Erzin, G., Delespaul, P., Kenis, G., … Guloksuz, S. (2022). Examining facial emotion recognition as an intermediate phenotype for psychosis: Findings from the EUGEI study. Progress in Neuro-Psychopharmacology & Biological Psychiatry, 113, 110440. doi:10.1016/j.pnpbp.2021.110440.CrossRefGoogle ScholarPubMed
Gottesman, I. I., & Shields, J. (1967). A polygenic theory of schizophrenia. Proceedings of the National Academy of Sciences of the United States of America, 58(1), 199205. doi:10.1073/pnas.58.1.199.CrossRefGoogle ScholarPubMed
Green, M. F., Penn, D. L., Bentall, R., Carpenter, W. T., Gaebel, W., Gur, R. C., … Heinssen, R. (2008). Social cognition in schizophrenia: An NIMH workshop on definitions, assessment, and research opportunities. Schizophrenia Bulletin, 34(6), 12111220. doi:10.1093/schbul/sbm145.CrossRefGoogle ScholarPubMed
Hariri, A. R., Bookheimer, S. Y., & Mazziotta, J. C. (2000). Modulating emotional responses: Effects of a neocortical network on the limbic system. Neuroreport, 11(1), 4348. doi:10.1097/00001756-200001170-00009.CrossRefGoogle ScholarPubMed
Hariri, A. R., Drabant, E. M., Munoz, K. E., Kolachana, B. S., Mattay, V. S., Egan, M. F., & Weinberger, D. R. (2005). A susceptibility gene for affective disorders and the response of the human amygdala. Archives of General Psychiatry, 62(2), 146152. doi:10.1001/archpsyc.62.2.146.CrossRefGoogle ScholarPubMed
Hariri, A. R., Mattay, V. S., Tessitore, A., Kolachana, B., Fera, F., Goldman, D., … Weinberger, D. R. (2002). Serotonin transporter genetic variation and the response of the human amygdala. Science (New York, N.Y.), 297(5580), 400403. doi:10.1126/science.1071829.CrossRefGoogle ScholarPubMed
Jaffe, A. E., Gao, Y., Deep-Soboslay, A., Tao, R., Hyde, T. M., Weinberger, D. R., & Kleinman, J. E. (2016). Mapping DNA methylation across development, genotype and schizophrenia in the human frontal cortex. Nature Neuroscience, 19(1), 4047. doi:10.1038/nn.4181.CrossRefGoogle ScholarPubMed
Koch, E., Nyberg, L., Lundquist, A., & Kauppi, K. (2022). Polygenic risk for schizophrenia has sex-specific effects on brain activity during memory processing in healthy individuals. Genes, 13(3), 412. doi:10.3390/genes13030412.CrossRefGoogle ScholarPubMed
Kohler, C. G., & Martin, E. A. (2006). Emotional processing in schizophrenia. Cognitive Neuropsychiatry, 11(3), 250271. doi:10.1080/13546800500188575.CrossRefGoogle ScholarPubMed
Lancaster, T. M., Ihssen, N., Brindley, L. M., Tansey, K. E., Mantripragada, K., O'Donovan, M. C., … Linden, D. E. (2016). Associations between polygenic risk for schizophrenia and brain function during probabilistic learning in healthy individuals. Human Brain Mapping, 37(2), 491500. doi:10.1002/hbm.23044.CrossRefGoogle ScholarPubMed
Lo Bianco, L., Blasi, G., Taurisano, P., Di Giorgio, A., Ferrante, F., Ursini, G., … Bertolino, A. (2013). Interaction between catechol-O-methyltransferase (COMT) Val158Met genotype and genetic vulnerability to schizophrenia during explicit processing of aversive facial stimuli. Psychological Medicine, 43(2), 279292. doi:10.1017/s0033291712001134.CrossRefGoogle ScholarPubMed
Martin, D., Croft, J., Pitt, A., Strelchuk, D., Sullivan, S., & Zammit, S. (2020). Systematic review and meta-analysis of the relationship between genetic risk for schizophrenia and facial emotion recognition. Schizophrenia Research, 218, 713. doi:10.1016/j.schres.2019.12.031.CrossRefGoogle ScholarPubMed
McNeil, T. F., Cantor-Graae, E., & Sjöström, K. (1994). Obstetric complications as antecedents of schizophrenia: Empirical effects of using different obstetric complication scales. Journal of Psychiatric Research, 28(6), 519530. doi:10.1016/0022-3956(94)90042-6.CrossRefGoogle ScholarPubMed
Mier, D., Lis, S., Zygrodnik, K., Sauer, C., Ulferts, J., Gallhofer, B., & Kirsch, P. (2014). Evidence for altered amygdala activation in schizophrenia in an adaptive emotion recognition task. Psychiatry Research, 221(3), 195203. doi:10.1016/j.pscychresns.2013.12.001.CrossRefGoogle Scholar
Miller, J. A., Scult, M. A., Conley, E. D., Chen, Q., Weinberger, D. R., & Hariri, A. R. (2018). Effects of schizophrenia polygenic risk scores on brain activity and performance during working memory subprocesses in healthy young adults. Schizophrenia Bulletin, 44(4), 844853. doi:10.1093/schbul/sbx140.CrossRefGoogle ScholarPubMed
Murata, T., Yasuda, S., Imaizumi, K., Isogami, H., Fukuda, T., Kyozuka, H., … Fujimori, K. (2022). Association of labour duration in spontaneous deliveries with low neonatal apgar scores and foetal acidosis: The Japan Environment and Children's Study. Scientific Reports, 12(1), 21519. doi:10.1038/s41598-022-24359-3.CrossRefGoogle ScholarPubMed
Nicodemus, K. K., Marenco, S., Batten, A. J., Vakkalanka, R., Egan, M. F., Straub, R. E., & Weinberger, D. R. (2008). Serious obstetric complications interact with hypoxia-regulated/vascular-expression genes to influence schizophrenia risk. Molecular Psychiatry, 13(9), 873877. doi:10.1038/sj.mp.4002153.CrossRefGoogle ScholarPubMed
Owen, M. J., Sawa, A., & Mortensen, P. B. (2016). Schizophrenia. Lancet (London, England), 388(10039), 8697. doi:10.1016/s0140-6736(15)01121-6.CrossRefGoogle ScholarPubMed
Paradiso, S., Andreasen, N. C., Crespo-Facorro, B., O'Leary, D. S., Watkins, G. L., Boles Ponto, L. L., & Hichwa, R. D. (2003). Emotions in unmedicated patients with schizophrenia during evaluation with positron emission tomography. The American Journal of Psychiatry, 160(10), 17751783. doi:10.1176/appi.ajp.160.10.1775.CrossRefGoogle ScholarPubMed
Park, H. Y., Yun, J. Y., Shin, N. Y., Kim, S. Y., Jung, W. H., Shin, Y. S., … Kwon, J. S. (2016). Decreased neural response for facial emotion processing in subjects with high genetic load for schizophrenia. Progress in Neuro-Psychopharmacology & Biological Psychiatry, 71, 9096. doi:10.1016/j.pnpbp.2016.06.014.CrossRefGoogle ScholarPubMed
Pergola, G., Papalino, M., Gelao, B., Sportelli, L., Vollerbergh, W., Grattagliano, I., & Bertolino, A. (2019). Evocative gene-environment correlation between genetic risk for schizophrenia and bullying victimization. World psychiatry: Official Journal of the World Psychiatric Association (WPA), 18(3), 366367. doi:10.1002/wps.20685.CrossRefGoogle ScholarPubMed
Phillips, M. L., Drevets, W. C., Rauch, S. L., & Lane, R. (2003). Neurobiology of emotion perception I: The neural basis of normal emotion perception. Biological Psychiatry, 54(5), 504514. doi:10.1016/s0006-3223(03)00168-9.CrossRefGoogle ScholarPubMed
Purcell, S. M., Wray, N. R., Stone, J. L., Visscher, P. M., O'Donovan, M. C., Sullivan, P. F., & Sklar, P. (2009). Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature, 460(7256), 748752. doi:10.1038/nature08185.Google ScholarPubMed
Quarto, T., Blasi, G., Maddalena, C., Viscanti, G., Lanciano, T., Soleti, E., … Curci, A. (2016). Association between ability emotional intelligence and left insula during social judgment of facial emotions. PLoS One, 11(2), e0148621. doi:10.1371/journal.pone.0148621.CrossRefGoogle ScholarPubMed
Quarto, T., Paparella, I., De Tullio, D., Viscanti, G., Fazio, L., Taurisano, P., … Blasi, G. (2018). Familial risk and a genome-wide supported DRD2 variant for schizophrenia predict lateral prefrontal-amygdala effective connectivity during emotion processing. Schizophrenia Bulletin, 44(4), 834843. doi:10.1093/schbul/sbx128.CrossRefGoogle Scholar
Rasetti, R., Mattay, V. S., Wiedholz, L. M., Kolachana, B. S., Hariri, A. R., Callicott, J. H., … Weinberger, D. R. (2009). Evidence that altered amygdala activity in schizophrenia is related to clinical state and not genetic risk. The American Journal of Psychiatry, 166(2), 216225. doi:10.1176/appi.ajp.2008.08020261.CrossRefGoogle Scholar
Rasmussen, S., Ebbing, C., & Irgens, L. M. (2017). Predicting preeclampsia from a history of preterm birth. PLoS One, 12(7), e0181016. doi:10.1371/journal.pone.0181016.CrossRefGoogle ScholarPubMed
Shifman, S., Bronstein, M., Sternfeld, M., Pisanté, A., Weizman, A., Reznik, I., … Darvasi, A. (2004). COMT: A common susceptibility gene in bipolar disorder and schizophrenia. American Journal of Medical Genetics. Part B, Neuropsychiatric Genetics: the Official Publication of the International Society of Psychiatric Genetics, 128b(1), 6164. doi:10.1002/ajmg.b.30032.CrossRefGoogle Scholar
Smith, S. M., & Nichols, T. E. (2009). Threshold-free cluster enhancement: Addressing problems of smoothing, threshold dependence and localisation in cluster inference. Neuroimage, 44(1), 8398. doi:10.1016/j.neuroimage.2008.03.061.CrossRefGoogle ScholarPubMed
Smolka, M. N., Schumann, G., Wrase, J., Grüsser, S. M., Flor, H., Mann, K., … Heinz, A. (2005). Catechol-O-methyltransferase val158met genotype affects processing of emotional stimuli in the amygdala and prefrontal cortex. The Journal of Neuroscience: the Official Journal of the Society for Neuroscience, 25(4), 836842. doi:10.1523/jneurosci.1792-04.2005.CrossRefGoogle ScholarPubMed
Spilka, M. J., & Goghari, V. M. (2017). Similar patterns of brain activation abnormalities during emotional and non-emotional judgments of faces in a schizophrenia family study. Neuropsychologia, 96, 164174. doi:10.1016/j.neuropsychologia.2017.01.014.CrossRefGoogle Scholar
Sullivan, P. F. (2005). The genetics of schizophrenia. PLoS Medicine, 2(7), e212. doi:10.1371/journal.pmed.0020212.CrossRefGoogle ScholarPubMed
Taurisano, P., Blasi, G., Romano, R., Sambataro, F., Fazio, L., Gelao, B., … Bertolino, A. (2013). DAT by perceived MC interaction on human prefrontal activity and connectivity during emotion processing. Social Cognitive and Affective Neuroscience, 8(8), 855862. doi:10.1093/scan/nss084.CrossRefGoogle ScholarPubMed
Taylor, S. F., Kang, J., Brege, I. S., Tso, I. F., Hosanagar, A., & Johnson, T. D. (2012). Meta-analysis of functional neuroimaging studies of emotion perception and experience in schizophrenia. Biological Psychiatry, 71(2), 136145. doi:10.1016/j.biopsych.2011.09.007.CrossRefGoogle ScholarPubMed
Taylor, S. F., Liberzon, I., Decker, L. R., & Koeppe, R. A. (2002). A functional anatomic study of emotion in schizophrenia. Schizophrenia Research, 58(2–3), 159172. doi:10.1016/s0920-9964(01)00403-0.CrossRefGoogle ScholarPubMed
Tottenham, N., Tanaka, J. W., Leon, A. C., McCarry, T., Nurse, M., Hare, T. A., … Nelson, C. (2009). The NimStim set of facial expressions: Judgments from untrained research participants. Psychiatry Research, 168(3), 242249. doi:10.1016/j.psychres.2008.05.006.CrossRefGoogle ScholarPubMed
Trubetskoy, V., Pardiñas, A. F., Qi, T., Panagiotaropoulou, G., Awasthi, S., Bigdeli, T. B., … O'Donovan, M. C. (2022). Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nature, 604(7906), 502508. doi:10.1038/s41586-022-04434-5.CrossRefGoogle ScholarPubMed
Ursini, G., Di Carlo, P., Mukherjee, S., Chen, Q., Han, S., Kim, J., … Weinberger, D. R. (2023). Prioritization of potential causative genes for schizophrenia in placenta. Nature Communication, 14(1), 2613. doi:10.1038/s41467-023-38140-1.CrossRefGoogle ScholarPubMed
Ursini, G., Punzi, G., Chen, Q., Marenco, S., Robinson, J. F., Porcelli, A., … Weinberger, D. R. (2018). Convergence of placenta biology and genetic risk for schizophrenia. Nature Medicine, 24(6), 792801. doi:10.1038/s41591-018-0021-y.CrossRefGoogle ScholarPubMed
Ursini, G., Punzi, G., Langworthy, B. W., Chen, Q., Xia, K., Cornea, E. A., … Weinberger, D. R. (2021). Placental genomic risk scores and early neurodevelopmental outcomes. Proceedings of the National Academy of Sciences of the United States of America, 118(7), 792801. doi:10.1073/pnas.2019789118.Google ScholarPubMed
van Buuren, M., Vink, M., Rapcencu, A. E., & Kahn, R. S. (2011). Exaggerated brain activation during emotion processing in unaffected siblings of patients with schizophrenia. Biological Psychiatry, 70(1), 8187. doi:10.1016/j.biopsych.2011.03.011.CrossRefGoogle ScholarPubMed
van der Meer, L., Swart, M., van der Velde, J., Pijnenborg, G., Wiersma, D., Bruggeman, R., & Aleman, A. (2014). Neural correlates of emotion regulation in patients with schizophrenia and non-affected siblings. PLoS One, 9(6), e99667. doi:10.1371/journal.pone.0099667.CrossRefGoogle ScholarPubMed
van Donkersgoed, R. J., Wunderink, L., Nieboer, R., Aleman, A., & Pijnenborg, G. H. (2015). Social cognition in individuals at ultra-high risk for psychosis: A meta-analysis. PLoS One, 10(10), e0141075. doi:10.1371/journal.pone.0141075.CrossRefGoogle ScholarPubMed
Vassos, E., Kou, J., Tosato, S., Maxwell, J., Dennison, C. A., Legge, S. E., … Murray, R. M. (2022). Lack of support for the genes by early environment interaction hypothesis in the pathogenesis of schizophrenia. Schizophrenia Bulletin, 48(1), 2026. doi:10.1093/schbul/sbab052.CrossRefGoogle ScholarPubMed
Walton, E., Geisler, D., Lee, P. H., Hass, J., Turner, J. A., Liu, J., … Ehrlich, S. (2014). Prefrontal inefficiency is associated with polygenic risk for schizophrenia. Schizophrenia Bulletin, 40(6), 12631271. doi:10.1093/schbul/sbt174.CrossRefGoogle ScholarPubMed
Wang, Y., Li, Z., Liu, W. H., Wei, X. H., Jiang, X. Q., Lui, S. S. Y., … Chan, R. C. K. (2018). Negative schizotypy and altered functional connectivity during facial emotion processing. Schizophrenia Bulletin, 44(suppl_2), S491s500. doi:10.1093/schbul/sby036.CrossRefGoogle ScholarPubMed
Weinberger, D. R. (1987). Implications of normal brain development for the pathogenesis of schizophrenia. Archives of General Psychiatry, 44(7), 660669. doi:10.1001/archpsyc.1987.01800190080012.CrossRefGoogle ScholarPubMed
Williams, H. J., Owen, M. J., & O'Donovan, M. C. (2007). Is COMT a susceptibility gene for schizophrenia? Schizophrenia Bulletin, 33(3), 635641. doi:10.1093/schbul/sbm019.CrossRefGoogle ScholarPubMed
Wolf, D. H., Satterthwaite, T. D., Calkins, M. E., Ruparel, K., Elliott, M. A., Hopson, R. D., … Gur, R. E. (2015). Functional neuroimaging abnormalities in youth with psychosis spectrum symptoms. JAMA Psychiatry, 72(5), 456465. doi:10.1001/jamapsychiatry.2014.3169.CrossRefGoogle ScholarPubMed
Wortinger, L. A., Engen, K., Barth, C., Andreassen, O. A., Nordbø Jørgensen, K., & Agartz, I. (2022). Asphyxia at birth affects brain structure in patients on the schizophrenia-bipolar disorder spectrum and healthy participants. Psychological Medicine, 52(6), 10501059. doi:10.1017/s0033291720002779.CrossRefGoogle ScholarPubMed
Wray, N. R., Goddard, M. E., & Visscher, P. M. (2007). Prediction of individual genetic risk to disease from genome-wide association studies. Genome Research, 17(10), 15201528. doi:10.1101/gr.6665407.CrossRefGoogle ScholarPubMed
Yang, Y., & Wu, N. (2022). Gestational diabetes mellitus and preeclampsia: Correlation and influencing factors. Frontiers in Cardiovascular Medicine, 9, 831297. doi:10.3389/fcvm.2022.831297.CrossRefGoogle ScholarPubMed
Figure 0

Table 1. Demographics and clinical variables of the samples of the study

Figure 1

Figure 1. Discovery sample: (A) Axial section of the brain showing the cluster in VLPFC where an interaction between PRS_GWA and ELCs was present; (B) scatterplot of % signal change extracted from the cluster represented in A. as a function of PRS_GWA and ELCs. See results for statistics. PRS_GWA: polygenic risk score genome-wide associated with diagnosis of schizophrenia.

Figure 2

Figure 2. Replication sample: scatterplot of % signal change extracted from the 20 mm ROI centered on the peak voxel associated with a PRS_GWA by ELCs interaction in the discovery analysis. See results for statistics. PRS_GWA: polygenic risk score genome-wide associated with diagnosis of schizophrenia

Figure 3

Figure 3. (A) Axial section of the brain showing the cluster in left VIPFC associated with a main effect of diagnosis during emotion processing; (B) Mean BOLD signal change in SCZ and HP extracted from this cluster shown in A. See results for statistics. PRS_GWA: polygenic risk score genome-wide associated with diagnosis of schizophrenia. HP: healthy participants; SCZ: patients with schizophrenia

Supplementary material: File

Toro et al. supplementary material

Toro et al. supplementary material
Download Toro et al. supplementary material(File)
File 373.8 KB