Effects of the 5-HT2A and DRD3 genotypes on cortical morphology and functional connectivity density in drug-naïve first episode schizophrenia
Yafei Kang a,1, Wei Zhang a,1, Yahui Lv a, Suping Cai a, Hanxiao Xu a, Jijun Wang b,⁎, Liyu Huang a,⁎⁎
a SchoolofLifeSciencesandTechnology,XidianUniversity,Xi'an,Shaanxi710071,PRChina b ShanghaiMentalHealthCenter,ShanghaiJiaotongUniversity,Shanghai200030,PRChina
article info
Article history:
Received 18 April 2019
Received in revised form 21 November 2019 Accepted 29 November 2019
Available online xxxx
Keywords:
First episode schizophrenia 5-Hydroxytryptamine 2A receptor Dopamine D3 receptor
Cortical morphology
Functional connectivity density
1. Introduction
Schizophrenia is a highly complex disease with a high heritability rate (Hall et al., 2006; Hilker et al., 2018). Considerable evidence from family, twin, and adoption research have demonstrated that many genes significantly contribute to the etiology and pathogenesis of schizophrenia (Zheng et al., 2017). Investigating genetic influence on the variation of brain attributes may contribute to improve clinical man- agement and open new avenues to understand the schizophrenia pathophysiology.
Serotonin and dopamine are two of the most significant neurotrans- mitters affecting mental health (Zheng et al., 2017). Convergent evi- dence reveals that both serotonin and dopamine play crucial roles in the pathophysiology of schizophrenia (Kapur and Remington, 1996). The 5-hydroxytryptamine 2A receptor (5-HT2A) and dopamine D3 re- ceptor (DRD3) genes belong to the family of biogenic amine serotonin and dopamine neurotransmitter receptors that have been frequently
⁎ Correspondingauthor.
⁎⁎ Correspondence to: L. Huang, School of Life Science and Technology, Xidian
University, Xi'an, Shaanxi 710071, PR China.
E-mail addresses: jijunwang27@163.com (J. Wang), huangly@mail.xidian.edu.cn
(L. Huang).
1 Yafei Kang and Wei Zhang contributed equally to this study.
https://doi.org/10.1016/j.schres.2019.11.058
0920-9964/© 2018 Published by Elsevier B.V.
abstract
The 5-hydroxytryptamine 2A receptor (5-HT2A) and dopamine D3 receptor (DRD3) have been extensively stud- ied as promising candidate genes for schizophrenia. Magnetic resonance imaging studies have demonstrated that schizophrenia is associated with widespread structural and functional abnormalities in the brain. Serotonin and dopamine receptors play crucial roles in the development of the human cerebral cortex and brain activity. How- ever, how the 5-HT2A and DRD3 genes impact brain structure and function in schizophrenia remains unknown. In the present study, we investigated the main effect of disease state and the interaction effect between disease state and genotype of these two genes on cortical volume, thickness, surface area and functional connectivity density (FCD) in fifty-five drug-naïve first episode schizophrenia patients and fifty-three healthy controls. We found that the differences in local FCD (lFCD) and global FCD (gFCD) between patients and healthy controls were predominantly located in brain hub regions. The significant interaction effects of disease state and 5- HT2A and DRD3 genes on brain structure and function were mainly located in the temporal cortex. Our findings may help to improve the understanding of the relationship between 5-HT2A and DRD3 genotypes and schizo- phrenia pathogenesis.
© 2018 Published by Elsevier B.V.
studied as classical candidate genes for susceptibility to schizophrenia (Domínguez et al., 2007). The 5-HT2A receptors are highly expressed in the frontal, parietal, temporal and occipital, anterogenual cortexes, and entorhinal area (Forutan et al., 2002; Hall et al., 2000; Roth, 2011). DRD3 are principally located in mesolimbic regions like the ven- tral striatum, islands of Calleja, nucleus accumbens, globus pallidus, and thalamus, but they are also found in the frontal and other cortical re- gions (such as temporal) as well as the cerebellum (Herroelen et al., 1994; Meador-Woodruff et al., 1996; Sokoloff et al., 2006). Recently, re- search into 5-HT2A and DRD3 genes has focused on the polymorphisms T102C (rs6313) and Ser9Gly (rs6280). These two single nucleotide polymorphisms (SNPs) have been extensively analyzed regarding their association with susceptibility to schizophrenia (Ni et al., 2013; Lochman et al., 2013; Özçetin et al., 2014). The CC genotype of rs6313 in 5-HT2A gene has been found to be linked to decreased post- synaptic serotonin receptor expression and more frequent suicidal ide- ation (Jakubczyk et al., 2012; Myers et al., 2007). As illustrated by a pre- vious study, polymorphism of rs6280 in the DRD3 gene has been reported to be implicated in altered dopamine binding affinity (Lundstrom and Turpin, 1996). Serotonin and dopamine receptors play crucial roles in the development of the human cerebral cortex and brain activity (Lewis, 1997; Pieramico et al., 2012; Tagliazucchi et al., 2016; Whitakerazmitia, 2001). However, the relationship be- tween polymorphisms of rs6313 and rs6280 and brain structural and
Please cite this article as: Y. Kang, W. Zhang, Y. Lv, et al., Effects of the 5-HT2A and DRD3 genotypes on cortical morphology and functional connectivity density ..., Schizophrenia Research, https://doi.org/10.1016/j.schres.2019.11.058
SCHRES-08605; No of Pages 9
2 Y. Kang et al. / Schizophrenia Research xxx (xxxx) xxx
functional abnormalities has not been thoroughly investigated in schizophrenia.
Compelling evidence from structural and functional magnetic reso- nance imaging (MRI) studies has suggested that schizophrenia is related to widespread abnormalities in brain structure and function (Friston and Frith, 1995; Pol et al., 2001; Thompson et al., 2001; Zhou et al., 2007). Meta-analyses of structural studies have suggested that de- creased cortical brain volume may serve as an anatomical biomarker of the pathophysiology of schizophrenia (Ellisonwright and Bullmore, 2010; Steen et al., 2006). Additionally, widespread cortical thinning and reduced surface area have been reported in frontal and temporal re- gions in schizophrenia (Crespofacorro et al., 2000; Goldman et al., 2009; Rimol et al., 2010). The 5-HT2A and DRD3 genes may affect cortical mor- phology through modulation of serotonin and dopamine levels (Daubert and Condron, 2010; Wang and Deutch, 2008b; Woodward et al., 2009b). However, to the best of our knowledge, no research has assessed the association between 5-HT2A and DRD3 gene polymor- phisms and cortical volume, thickness and surface area in schizophrenia patients.
A large body of studies has indicated that schizophrenia patients ex- hibit aberrant functional connectivity in their brains (Venkataraman et al., 2012; Wang et al., 2017). However, it remains unclear how genetic variants at 5-HT2A rs6313 and DRD3 rs6280 impact resting-state brain activity. We evaluated local/global functional connectivity tissue prop- erties from resting-state functional MRI (fMRI) datasets using functional connectivity density (FCD) mapping (Tomasi and Volkow, 2010). FCD mapping is a recently proposed, data-driven method that quantifies the number of resting state functional connectivity between a given voxel with other voxels (Tomasi and Volkow, 2010). This method al- lows the identification of functional hubs (brain regions densely con- nected to other regions) with high sensitivity and discrimination (Tomasi and Volkow, 2012a). Unlike the common hypothesis-driven seed-voxel correlation approach, FCD does not require specific regional hypotheses about seed locations and is a powerful graph theory tool for exploratory studies. The greater the number of voxels functionally con- nected to other brain voxels, the higher the FCDs they have. Such a high FCD value indicates that this voxel plays a more prominent role in the brain (Tomasi et al., 2015; Tomasi and Volkow, 2011b). In general, local FCD (lFCD) measures the number of voxels in a local functional connectivity cluster, and global FCD (gFCD) measures the number of whole-brain functional connections to each voxel (Tomasi et al., 2015). Recent studies have reported disruptions in FCD in various psy- chiatric disorders, such as major depressive disorder (Gong et al., 2017), autism spectrum disorder (Tomasi and Volkow, 2017) and schizophrenia (Liu et al., 2015; Tomasi and Volkow, 2014; Zhuo et al., 2014). Functional studies have revealed that serotonin and dopamine are implicated in functional brain activity (Pieramico et al., 2012; Tagliazucchi et al., 2016), whereas to the best of our knowledge, the re- lationship between the genetic effects of 5-HT2A and DRD3 genes and FCD has not been studied in schizophrenia.
In the current study, we attempt to analyze the main effect of disease and the interaction effect between disease state and genotype (rs6313 and rs6280) on cortical morphology (cortical volume, thickness and surface area) and FCD (lFCD and gFCD) in drug-naïve first episode schizophrenia (FES) patients and matched healthy controls (HCs). Such an unbiased study will inform our understanding of the genetic in- fluences of rs6313 and rs6280 genotypes on whole-brain structural and functional alterations in FES. More importantly, this study will provide a better understanding of the relationship between the 5-HT2A and DRD3 genotypes and schizophrenia pathogenesis.
2. Materials and methods
2.1. Ethical statement
We confirm that all procedures performed in our study involving human participants were in accordance with the ethical standards of the Shanghai Mental Health Center and with the 1964 Helsinki Declara- tion and its later amendments or comparable ethical standards. Each participant gave written informed consent after experimental proce- dures had been fully explained.
2.2. Participants
We recruited a total of 108 Han Chinese in the present study, includ- ing 55 drug-naïve FES patients and 53 gender-, age-, and education- matched HCs. All patients were diagnosed with the Structured Clinical Interview for DSM-IV, Patient Version (SCID-I/P), for schizophrenia. Specific inclusion criteria for FES patients included (1) age between 16 and 40 years, (2) first episode illness (patients who have only recently formally presented, been evaluated, and been treated for schizophrenia within the treatment services of mental health system) (Weiden et al., 2007), and (3) completion of nine-year compulsory education. The ex- clusion criteria for patients were (1) existence of a neurological disor- der, (2) pregnancy, (3) a history of alcohol or drug abuse assessed based on the Mini-International Neuropsychiatric Interview (MINI); a history of nicotine abuse assessed based on self-report and the Fagerstrom Test (Fagerstrom and Schneider, 1989) or (4) a history of suicide risk. The severity of symptoms in FES patients was assessed using the scale for the assessment of negative symptoms (SANS) rating scale and the brief psychiatric rating scale (BPRS) scores. Fifty-three HCs were recruited by advertisements in the same region as the patients. To verify the absence of psychiatric illnesses in controls and their first- degree relatives, all HCs were administered the SCID, Non-Patient Edi- tion (SCID-I/NP). Other inclusion criteria for HCs were age between 16 and 40 years and completion of 9-year compulsory education. The ex- clusion criteria were the same for controls as for patients.
2.3. Genotyping
All participants underwent blood collection for genetic analysis. The plasma samples were stored at −80 °C. For each participant, genomic DNA was extracted using the Flexi Gene DNA Kit. The extracted DNA was quantified using a Nano DropTM 2000 (Thermo Fisher ScientificTM, Waltham, MA, USA). The DNA was stored at −80 °C prior to SNP analy- sis. SNP analysis was performed using the Kompetitive Allele Specific PCR genotyping system. The primer sequences for the 5-HT2A rs6313 polymorphism were as follows: forward: 5′-AAATGATGACACCAGG
Table 1
Demographic information of FES and HC. FES
HC t p
t1:1 t1:2
t1:3 t1:4
t1:5 t1:6 t1:7 t1:8 t1:9 t1:10 t1:11 t1:12 t1:13 t1:14 t1:15
t1:16 t1:17 t1:18 t1:19 t1:20
N
Male/female
Age (SD) (years) Education (years) SANS
Affect flattening Alogia Avolition-apathy Anhedonia-asociality Attention
SANS total BPRS
52
20/32 26(6.88) 12.6(2.99)
5.14(5.26) 2.21(3.82) 3.87(3.35) 5.81(4.58) 1.81(2.11) 18.85(16.23) 47(10.32)
51
22/29 0.574 26(6.51) 0.066 13(2.81) 0.24
– – – – – – – – – – – – – –
0.629a 0.798 0.626
– – – – – – –
Data are means (standard deviation) for age, education, SANS and BPRS scores. Abbreviations: FES: first episode schizophrenia, HC: healthy controls, N: number, SD: stan- dard deviation, SANS: scale for the assessment of negative symptoms, BPRS: brief psychi- atric rating scale.
a Chi-square tests for categorical measures.
Please cite this article as: Y. Kang, W. Zhang, Y. Lv, et al., Effects of the 5-HT2A and DRD3 genotypes on cortical morphology and functional connectivity density ..., Schizophrenia Research, https://doi.org/10.1016/j.schres.2019.11.058

t2:1 t2:2
t2:3 t2:4
t2:5 t2:6 t2:7 t2:8 t2:9
t2:10 t2:11
Table 2
Y. Kang et al. / Schizophrenia Research xxx (xxxx) xxx 3 3000 ms, slice thickness = 3 mm, slice gap = 1.05 mm, FOV =
Genotype and allele frequencies of rs6313 and rs6280 in FES and HC.
216 × 216 mm2, data matrix size = 64 × 64, voxel size = 3 × 3 × 3 mm3 and 170 image volumes. All participants were instructed to keep their eyes closed, to relax and to avoid falling asleep, and not to think about anything in particular during the scan.
2.5. Structural MRI data processing
Structural MRI data were analyzed in an automated manner with the FreeSurfer software suite (http://surfer.nmr.mgh.harvard.edu, version 4.0.1) (Fischl, 2012; Fischl and Dale, 2000). The implemented process- ing stream included removal of non-brain tissue, transformation to Talairach-like space, and segmentation of white matter, grey matter and cerebrospinal fluid. The white and grey matter boundary was tes- sellated, and topological defects were automatically corrected. After normalization of intensity and transition of white and grey matter, the reconstructed pial boundary was indicated by determining the maxi- mum shift in intensity through surface deformation. Then, the entire ce- rebral cortex of each subject was visually inspected, and inaccuracies in segmentation were manually corrected. After the construction of the cortical representations, parcellation of the cerebral cortex was con- ducted to classify each brain vertex as either sulcal or gyral, and these vertexes were then subparcellated into 148 cortical regions (74 per hemisphere) via the aparc.a2009s template (Destrieux et al., 2010). We selected three common structural characteristics, namely, cortical volume, thickness and surface area, for further statistical analysis on the basis of this brain template.
Genotype p(χ2)
Allele
T 57(0.548) 57(0.558) T 81(0.778) 74(0.725)
p(χ2) HWE MAF
C
47(0.451) 0.876 0.597 0.447(C) 45(0.441)
C
23(0.221) 0.375 0.936 0.247(C) 21(0.274)
rs6313 TT FES 16 HC 13
rs6280 TT FES 32 HC 27
CT/CC
25/11 0.355 31/7
CT/CC
17/3 0.778 20/4
FES: first episode schizophrenia, HC: healthy controls, HWE: Hardy–Weinberg equilib- rium, MAF: minor allele frequency.
CTCTACAGT-3′, reverse: 5′-TGTCCAGTTAAATGCATCAGAAGTG-3′. The primer sequences for the DRD3 rs6280 polymorphism were as follows: forward: 5′-CCTCTGGGCTATGGCATCTC-3′, reverse: 5′-GCTGGCACCTG TGGAGTTCT-3′.
2.4. Imaging acquisition
The MRI images were obtained on a 3.0 Tesla Siemens Trio scanner using an echo planar imaging sequence at the Shanghai Mental Health Center in Shanghai, China. Structural T1-weighted images were col- lected with the following parameters: echo time (TE) = 2.69 ms, repe- tition time (TR) = 2.3 s, field of view (FOV) phase = 93.8%, voxel size = 1 × 1 × 1 mm3, slice thickness = 1 mm, acquisition matrix = 256 × 256 mm2, and 192 sagittal slices (without gap). The detailed pa- rameters of the functional data were as follows: TE = 30 ms, TR =
Fig. 1. Spatial distribution of the average resting-state local FCD and global FCD superimposed on the middle axial MRI plane for each group. These maps reflect the average number of functional connections per voxel. In resting conditions, the average local FCD and global FCD peak in occipital lobe for FES patients and controls. Abbreviations: FCD: functional connectivity density, FES: first episode schizophrenia, HC: healthy controls, MRI: magnetic resonance imaging.
Please cite this article as: Y. Kang, W. Zhang, Y. Lv, et al., Effects of the 5-HT2A and DRD3 genotypes on cortical morphology and functional connectivity density ..., Schizophrenia Research, https://doi.org/10.1016/j.schres.2019.11.058

4 Y. Kang et al. / Schizophrenia Research xxx (xxxx) xxx
2.6. fMRI data processing
Preprocessing of the resting-state fMRI data was carried out on MATLAB 2012a using the Data Processing Assistant for Resting-State fMRI (DPARSF) (Yan and Zang, 2010). The preprocessing steps for the resting-state fMRI data, were as follows: (1) discarding the first ten image volumes to allow for adaptation to scanning environment, (2) slice timing adjustment, (3) realigning each volume to the first vol- ume for movement correction, (4) spatial normalization to the Montreal Neurological Institute (MNI) template (resampling with a voxel size of 3 × 3 × 3 mm3), (5) detrending and filtering (0.01–0.1 Hz), and (6) ap- plication of a multiple regression model with white matter, cerebrospi- nal fluid signal, and head motion parameters as covariates based on the Friston 24 model (Friston et al., 2015). The data from 5 subjects (3 pa- tients and 2 HCs) were excluded for head motion exceeding 3.0 mm in any dimension (x, y, and z) or 3.0° of angular motion. Since resting- state functional connectivity can be influenced by micromovement from volume to volume (Power et al., 2012), the frame-wise displace- ment values for each participant were computed, and the mean frame-wise displacement was included as a covariate in the group comparisons.
2.7. FCD
We calculated the strength of lFCD and gFCD using an in-house script according to a method described by Tomasi and Volkow (2010, 2011b). The number of local functional connections was computed through Pearson correlations between time-varying signals at voxel x0 and those from its closest neighbor voxels using an arbitrary threshold r N 0.6 (Tomasi and Volkow, 2011a, 2011b). This threshold was selected in previous studies because r N 0.4 led to a higher false positive rate and increased CPU time and r N 0.7 yielded lFCD maps with decreased dy- namic range and lower sensitivity; hence, r = 0.6 was selected for all calculations (Tomasi and Volkow, 2010). The lFCD at a given voxel(x0) was calculated as the local k(x0) between voxel(x0) and its adjacent voxels. A voxel was added to the list of voxels functionally connected with x0 when it was adjacent to a voxel that was connected to x0 by a continuous pathway of functionally connected voxels with Pearson cor- relation r N 0.6. This calculation was performed for all voxels that belonged to the list of the functionally connected voxels until no new voxels could be added to the list. This calculation was initiated sepa- rately for each x0 and repeated for all x0 voxels in the brain. The gFCD at a given voxel x0 was calculated as the global k(x0) between the voxel x0 and all other voxels in the entire brain. We selected the corre- lation threshold to be consistent with that used to compute the lFCD maps because gFCD was stable under threshold variations in the range 0.5 b R b 0.7 (Tomasi and Volkow, 2012b). This computation for gFCD was performed for all x0 voxels in the whole brain. The use of a single scaling factor for lFCD and gFCD, 1/k0, reflecting the mean lFCD and gFCD across subjects and voxels in brain, k0, allowed us to normalize the distribution of lFCD and gFCD. Finally, average lFCD and gFCD maps were smoothed with an isotropic Gaussian kernel of 6 mm full width at half maximum.
2.8. Statistical analysis
The two-sample t-test was employed to compare demographic char- acteristics, including age and years of education, and the χ2 test was im- plemented to examine the differences in gender, allele frequencies, and genotypes. The minor allele homozygotes and the heterozygotes were merged into a group to further assess the interaction effect of disease state and genotype. We included the gender/age/years of education (and estimated total intracranial volume (eTIV) for volume and surface area measures) as covariates to control for differences in these variables in the following statistical analysis. Firstly, a two-sample t-test was used by Statistical Product and Service Solutions (SPSS 20.0) to explore the
effect of the genotype on SANS and BPRS scores in the FES group. Mor- phological data analyses were run with outliers excluded. As defined in the boxplot, outliers were calculated as the value that higher than third quartile + 1.5 times the interquartile range or lower than first quar- tile − 1.5 times the interquartile range. Normal distribution and homo- geneity of values of thickness, volume, and surface area were checked using Kolmogorov–Smirnov tests before statistical analysis. To assess disease × genotype interaction effects, a two-way analysis of covariance (ANCOVA) was implemented for data with equal variances and a nor- mal distribution; if these assumptions were not fulfilled, a Scheirer- Ray-Hare test (a nonparametric equivalent of two-way ANCOVA) was used (p b 0.05, false discovery rate (FDR) corrected). Another two- way ANCOVA implemented in DPARSF was employed to investigate dis- ease × genotype interaction effects on lFCD/gFCD maps (p b 0.05, FDR corrected). Subsequently, a post hoc t-test analysis was applied to ana- lyze the genotype-related differences in interactive effects.
3. Results
3.1. No significant differences in demographic and genotypic characteristics There were no significant differences between patients and HCs with
respect to gender, age and years of education (p N 0.05, Table 1).
Fig. 2. Main effects of disease on local FCD and global FCD (p b 0.05, FDR corrected). (a and b) There were significant disease main effects on local FCD in the precuneus and occipital cortex. (c and d) There were significant main effects of disease on global FCD in the orbitofrontal cortex and inferior temporal cortex. Abbreviations: FCD: functional connectivity density, FES: first episode schizophrenia, HC: healthy controls.
Please cite this article as: Y. Kang, W. Zhang, Y. Lv, et al., Effects of the 5-HT2A and DRD3 genotypes on cortical morphology and functional connectivity density ..., Schizophrenia Research, https://doi.org/10.1016/j.schres.2019.11.058

Detailed information on demographic characteristics is summarized in Table 1.
There were no significant differences for rs6313 and rs6280 in the patients and HCs with respect to genotype frequencies, allele frequen- cies, Hardy–Weinberg equilibrium (HWE) and minor allele frequency (MAF) (p N 0.05, Table 2). The p-values of genotype frequencies, allele frequencies, HWE, and MAF are shown in Table 2.
When merging the minor allele homozygotes and the heterozygotes into a group of minor carriers, there were no significant differences in demographic variables between the two genotypic subgroups with re- spect to gender, age and years of education (p N 0.05, Table S1). The scores of SANS and BPRS in FES were not influenced by the individual's genotypes of rs6313 and rs6280 (p N 0.05, Table S2).
3.2. The distribution of lFCD/gFCD
Fig. 1 shows the average distribution of lFCD/gFCD for patients and HCs. Hub regions of lFCD were bilaterally located in the occipital lobe (Brodmann areas [BAs] 17–19), inferior parietal lobe (BA 7) and precuneus (BA 23) for patients and HCs. The regions with high gFCD value in the two groups were mainly bilaterally located in the occipital lobe (BAs 17–19), frontal lobe (BAs 8–11, 46–48), temporal lobe (BAs
20–22, and 38), and inferior parietal lobe (BA 7). In particular, the bilat- eral occipital lobe (BAs 17–19) was the region with the highest lFCD and gFCD in patients and HCs.
3.3. The main effect of disease
When analyzing the main effect of disease, significantly decreased cortical thickness in FES patients was observed in the left middle frontal lobe (p b 0.05, FDR corrected), which has been reported in our previous study (Kang et al., 2018). No other disease-related differences were de- tected in cortical morphology. Regarding the main effect of disease on lFCD/gFCD, we found that patients had significantly decreased lFCD in the left precuneus (BA 23) as well as increased lFCD in the left occipital cortex (BA 17) and reduced gFCD in the bilateral orbitofrontal cortex (BAs 10 and 11) and right inferior temporal cortex (BA 20) (p b 0.05, FDR corrected, Fig. 2, Table S3).
3.4. Disease × genotype interaction effects and post hoc analysis
A significant interaction effect of disease state and 5-HT2A rs6313 genotypes was detected in left planum temporale (L.PT) cortical thick- ness (p b 0.001, F = 14.08, FDR corrected, Fig. 3(a)). Specifically, post
Y. Kang et al. / Schizophrenia Research xxx (xxxx) xxx 5
Fig. 3. The disease × genotype interaction effects of rs6313 and rs6280 on cortical volume, thickness and surface area (p b 0.05 FDR corrected). (a) There was a significant interaction effect of disease state and genotype of rs6313 in thickness of L.PT. (b) There were significant interaction effects of disease state and genotype of rs6280 on cortical volume and surface area of R. ITS. Abbreviations: FES: first episode schizophrenia, HC: healthy controls, L.PT: left planum temporale, R.ITS: right inferior temporal sulcus. *p b 0.05, **p b 0.01.
Please cite this article as: Y. Kang, W. Zhang, Y. Lv, et al., Effects of the 5-HT2A and DRD3 genotypes on cortical morphology and functional connectivity density ..., Schizophrenia Research, https://doi.org/10.1016/j.schres.2019.11.058

6 Y. Kang et al. / Schizophrenia Research xxx (xxxx) xxx
hoc tests showed that patients who were C carriers had reduced thick- ness in the L.PT compared with TT homozygous patients. Conversely, there were no significant differences between C carriers and TT homo- zygotes in the HCs. We also observed significant interaction effects of disease state and DRD3 rs6280 genotypes on cortical volume and sur- face area in the right inferior temporal sulcus (R.ITS) (BA 20) (p b 0.001, F = 13.78; p b 0.001, F = 13.90, FDR corrected, Fig. 3(b)). Specifically, compared with TT homozygous patients, decreased volume and surface area in the R.ITS were observed in patients who were C car- riers, while the significant differences disappeared in the HCs.
Regarding the interaction effects on lFCD/gFCD, significant interac- tions between disease state and 5-HT2A rs6313 were detected in the gFCD of the bilateral inferior temporal gyrus (ITG) (p b 0.05, FDR corrected, Fig. 4, Table S4). In particular, the interaction between disease status and 5-HT2A rs6313 on gFCD of R.ITG overlapped with the disease-related main effect. Post hoc analyses showed a significantly de- creased gFCD of the bilateral ITG in C carriers relative to TT homozygotes in the patient group, while in the HCs, C carriers exhibited increased gFCD compared with TT homozygotes.
4. Discussion
The present study investigated the potential main effect of disease state and interaction effects of disease state and two important genetic risk factors (rs6313 and rs6280) in serotonin and dopamine neurotrans- mitter systems on whole-brain cortical volume, thickness, surface area, lFCD and gFCD in drug-naïve FES patients and HCs. There were two major findings: (1) abnormalities of lFCD and gFCD in patients were lo- cated in functional hubs (regions with high FCD), which included the precuneus, occipital cortex, orbitofrontal cortex, and inferior temporal cortex; (2) interactions between disease state and 5-HT2A and DRD3 genes in cortical morphology (cortical volume, thickness and surface
area) and FCD (lFCD and gFCD) were mainly observed in the temporal cortex.
Previous research has suggested that energy-efficient regions in the brain (densely connected nodes) are thought to serve as hubs (Bassett et al., 2008; Buckner et al., 2009). In the present study, one of the prin- cipal observations is that the distributions of lFCD and gFCD in patients and HCs are highly localized in the occipital cortex, precuneus, parietal lobe, frontal lobe, and temporal lobe. In particular, the occipital cortex, the region with the highest lFCD and gFCD values in patients and con- trols, was the most prominent functional hub. We found hubs appear to be distributed similarly between FES and HCs. To the best of our knowledge, no consistent conclusion was reached in the distribution paradigm of hub regions between schizophrenia patients and HCs. Van den Heuvel et al. found reduced hubs organization in the schizo- phrenia patients compared with HCs, which was most pronounced in the cortical networks (van den Heuvel et al., 2013). In contrast, Liu et al. found similar distributions of hubs between chronic schizophrenia patients and HCs (Liu et al., 2017). The inconsistency may generate form multiple factors, such as medication, methodological differences, sam- ple size, and the heterogeneity of schizophrenia. The present findings from first-episode, drug-naive patients may be informative. Our result suggests that the distribution of the hub region for local and global FCD was not affected by schizophrenia in the early stage of this illness. The hub vulnerability hypothesis demonstrates that psychiatric diseases have been linked to abnormalities in brain hubs (Bassett et al., 2008; Buckner et al., 2009). This hypothesis suggests that hubs of the human brain have higher metabolic demands and longer-distance connections than other regions (Crossley et al., 2014; Langbaum and Chen, 2009; Raichle and Gusnard, 2002), and therefore may be more likely to exhibit abnormalities than non-hubs in many brain psychiatric disorders, in- cluding schizophrenia and Alzheimer's disease (Crossley et al., 2014). Similar to replications in previous studies, we found that abnormal lFCD in patients was located in the precuneus and occipital cortex, and
Fig. 4. The disease × genotype interaction effects of rs6313 on global FCD (p b 0.05, FDR corrected). (a) There was a significant interaction effect of disease state and genotype of rs6313 on global FCD of L.ITG and R.ITG. Abbreviations: FES: first episode schizophrenia, HC: healthy controls, ITG: inferior temporal gyrus, L: left, R: right. *p b 0.05, **p b 0.01.
Please cite this article as: Y. Kang, W. Zhang, Y. Lv, et al., Effects of the 5-HT2A and DRD3 genotypes on cortical morphology and functional connectivity density ..., Schizophrenia Research, https://doi.org/10.1016/j.schres.2019.11.058

abnormal gFCD was located in the frontal lobe and temporal lobe. These brain regions serve as functional hubs, a finding which is consistent with the pathological brain lesions observed in schizophrenia patients (Crossley et al., 2014; Rubinov, 2013). Specifically, the precuneus has been extensively studied over the past decade as a central hub of the de- fault mode network (Zhang and Li, 2012). Multiple structural and func- tional deficits in schizophrenia patients in this area have been observed in previous studies (Zhao et al., 2017). The occipital lobe has also been investigated as a brain hub in previous studies (Jiang et al., 2016; Wang et al., 2015); this area is an important region in the central ner- vous system and the site of a wide variety of lesions (Flores, 2002). A re- cent study showed that schizophrenia lesions were mainly located in frontal, temporal cortical, precuneus hubs (Crossley et al., 2014), which is in agreement with our results. Taken together, the disease- related differences in the brain regions mentioned above are explicable and logical. These important findings provide strong evidence that the high cost/high value hubs of the brain are more likely to be pathologi- cally lesioned than non-hub regions. In addition, recent studies have shown that global functional connectivity operates at a higher time and metabolic cost as compared to short-range functional connectivity (Wang et al., 2018). Thus, gFCD hubs could be more pathologically le- sioned than lFCD hubs because of increased time-related and metabolic costs, and greater importance should be attached to the abnormalities of gFCD observed in frontal and temporal cortical hubs.
One of the major observations in our study is that significant interac- tion effects of disease state and genotypes of rs6313 and rs6280 on cor- tical morphology and lFCD/gFCD were located in the temporal cortex, which indicated that these two SNPs were mainly associated with ab- normalities of temporal lobe regions in patients with schizophrenia. In the current study, a significant interaction between disease state and 5-HT2A rs6313 was observed on cortical thickness of the L.PT and gFCD of the bilateral ITG. In particular, the interaction effect on gFCD of the R.ITG overlapped with a disease-related decrease in gFCD in pa- tients. This result suggested that C carriers may contribute to further gFCD decreases in schizophrenia patients. Further post hoc analysis found that C carriers exhibited significantly reduced thickness in the L. PT and decreased gFCD in the R.ITG compared with TT homozygotes in the patients, the HCs showed the reverse pattern. This phenomenon revealed that depending on disease status, the genetic influence of the rs6313 genotype may affect cortical thickness and gFCD of the temporal cortex in a totally different pattern, which might be caused by the spe- cific regulation of 5-HT2A rs6313 expression on temporal lobe regions in patients. Previous studies focusing on molecular mechanisms have reported that the temporal lobe is a crucial region for the expression of rs6313 (Bray et al., 2004; Polesskaya and Sokolov, 2002). Specifically, researchers found that the expression of the “C” allele in the temporal cortex in healthy individuals and schizophrenia patients was lower than the expression of the “T” allele at both the mRNA and protein levels (Polesskaya and Sokolov, 2002). Furthermore, reduced receptor binding related to the “C” allele has been reported in postmortem brains (Turecki et al., 1999). We found thinner L.PT cortical thickness and de- creased gFCD in the bilateral ITG in C carriers in the FES group. It is not surprising that carriers of the risk allele had decreased functional con- nectivity and cortical thinning in FES. Consisting with our results, the study by Schaefer et al. (2014) supports a conceptual framework of serotonergically modulated long-range functional connectivity. Studies using electrophysiological data provide further support suggesting sero- tonin to drive long-range functional connectivity (Avesar and Gulledge, 2012; Dieudonné and Dumoulin, 2000). 5-HT2A receptors play a direct role in neuronal development, cell proliferation, and dendritic branching. An effect on synaptic plasticity might explain the influence of the C allele on temporal structures. Speculatively, the C allele of rs6313 in FES related to a lower binding of the serotonergic receptor, re- duced numbers of serotonergic tracts, and thus less cortical targets, which results in decreased cortical thickness. These results provide evi- dence for the risk of C allele, showing a relationship of the 5-HT2A
receptor with the physiological process closely related to synaptic plas- ticity in schizophrenia.
We noticed that an interaction effect related to SNP rs6280 on corti- cal morphology was observed in temporal lobe structure. The post hoc analysis showed that patients who were C carriers had significantly re- duced cortical volume and surface area in R.ITS compared with TT ho- mozygous patients, but these genotype-related differences disappeared in the HCs. This phenomenon indicated that the genetic ef- fect of the rs6280 genotype on temporal lobe regions was more notable when people were in a disease state. We found that patients who were C carriers had significantly reduced cortical volume and surface area in R. ITS compared with TT homozygous patients. The mechanism through which DRD3 could impact cortical volume and surface area is not fully clear. As suggested by the previous studies (Wang and Deutch, 2008a; Woodward et al., 2009a), abnormal dopamine activity may lead to changes at the neuronal (i.e. dendritic length) and macrostructural (i.e. cortical volume) levels of cerebral morphology. A previous study re- ported the genetic effect of gene-gene interactions between DRD3 rs6280 and brain-derived neurotrophic factor Val66Met on brain mor- phology in temporal lobe structures in schizophrenia (Takahashi et al., 2008). Our findings of an interaction effect provide support for a role for the genetic effect of DRD3 rs6280 on the temporal cortex in schizo- phrenia. Additionally, a recent study reported that the DRD3 rs6280 polymorphism is significantly linked to long-term auditory memory (Yeh et al., 2012). This memory has been widely accepted to be related to structures in the temporal lobe (Yeh et al., 2012), which indirectly supports an association between rs6280 and the temporal lobe. Geno- types of the rs6280 variant have been found to show a differential affin- ity for dopamine (Lundstrom and Turpin, 1996), rendering it a functional candidate polymorphism. These important findings improve our understanding of the possible genetic influence of the DRD3 rs6280 genotype on whole-brain function in FES patients. The implications of the current results are largely methodological until the causal relation- ship between cerebral morphology and the expression of DRD3 and 5- HT2A receptors is better understood in non-human primates and humans.
5. Limitations
There are several limitations in the present study. First, these find- ings need to be replicated with a larger sample. Second, we analyzed the independent interaction effect of disease state and genotype on whole-brain characteristics in each SNP, but the effects of multiple SNPs or genes on the structure and function of the brain are more likely to be epistatic or additive. Third, neuroimaging evidence in our study is based on resting-state fMRI, hence, further studies using different ex- perimental paradigms are required to validate our findings. Finally, pre- vious studies suggested nicotine use can affect the serotonin and dopamine receptors. Although subjects with a history of nicotine abuse were excluded, ultra-light smokers were included. This con- founding factor could influence our results.
6. Conclusions
The present study investigated the potential genetic effects of 5- HT2A and DRD3 genes on cortical volume, thickness, surface area and lFCD/gFCD in FES patients and matched HCs. We found that abnormal lFCD and gFCD in patients were located in hub regions, these results provide evidence for the hub vulnerability hypothesis. In addition, we found that the temporal lobe is significantly influenced by the geno- types of 5-HT2A rs6313 and DRD3 rs6280. This study improves our un- derstanding of the possible genetic influence of 5-HT2A and DRD3 on brain structure and function in patients with FES, as well as identifies underlying neural mechanisms linking 5-HT2A rs6313 and DRD3 rs6280 with the risk of developing schizophrenia via the intermediate phenotype of abnormalities in brain structure and function.



Here, compared to healthy controls, patients with SZ were less likely to be heterozygous for the DRD3 ser/gly allele and more likely to have the homozygous state for the A trait. As to functional brain connectivity, the risk allele was the gly allele, the N allele when homozygous. This may reflect that in a rather sanguine habitancy, those with schizophrenia with A NPA type show up more.

As to 5HT2A, a possible N trait gene, the SZ patients are less likely to be heterozygous. But in functional brain connectivity, the risk allele was the minor allele (the African allele).

So this study does not demonstrate that the A or N candidate genes are different in SZ patients but that an endophenotype of SZ is distinguishable in patients with SZ by genotypes of the A and N candidate genes.