A genetics-first, rare variant approach to understanding the neurobiological substrates of schizophrenia and associated disorders Public
Sefik, Esra (Spring 2022)
Published
Abstract
It has long been known that schizophrenia has a substantial genetic component, but large-scale genomic investigations have only recently begun to illuminate the exact nature of this genetic architecture. There are now at least eight copy number variants (CNVs) with genome-wide significance for association with schizophrenia, which provide new substrates to investigate the etiology and pathogenesis of this complex and heterogeneous disorder. Among these, the 3q29 deletion (3q29Del) confers the highest known risk for schizophrenia (>40-fold increase); but the neurobiological mechanisms contributing to the abnormal neurodevelopmental phenotypes are not yet understood. The objective of this dissertation was to address this knowledge gap by employing a multidisciplinary approach that integrates tools from high-throughput RNA-sequencing, network analysis, neuroimaging, behavioral and clinical phenotyping, and statistical modeling. First, we used a systems-biology approach to interrogate the network-level behavior of 3q29 interval genes within the global protein-coding transcriptome of the healthy human prefrontal cortex. The modular properties and connectivity patterns yielded key predictions about novel biological roles, functional interactions and putative disease associations for individual 3q29 genes. Next, we performed the first known in vivo quantitative neuroimaging study in individuals with 3q29Del, and assessed the relationship between neuroanatomical findings and standardized measures of cognitive and sensorimotor abilities. Our findings showed that abnormal development of posterior fossa structures, particularly the cerebellum, may be a neuroimaging-based biomarker in 3q29Del. We additionally found that cerebellar volumetric changes are associated with cognitive disability, and diminished visual-motor integration skills, suggesting that the cerebellum is a possible mechanistic intermediary between this genetic lesion and motor and non-motor syndromic phenotypes. Furthermore, we conducted the first in-depth evaluation of psychotic symptoms in subjects with 3q29Del, compared this profile to 22q11.2Del, and investigated the relationship between psychotic symptoms and findings from structural brain imaging. Results from this work established the unique and shared profiles of psychotic symptoms across two high-impact CNVs and revealed cerebellar involvement in elevated psychosis-risk in 3q29Del. Altogether, the presented findings substantially advance our understanding of the role that 3q29Del plays in vulnerability for severe neurodevelopmental and psychiatric disorders and provide novel insights into neurogenetic mechanisms shaping human behavior and development.
Table of Contents
CHAPTER 1. Introduction 1
A clinical and epidemiological introduction to schizophrenia 1
Schizophrenia neuroanatomy 3
The neurodevelopmental hypothesis of schizophrenia 6
Genetic architecture of schizophrenia 7
Copy number variation 11
Research aims 14
Figures 16
Figure 1. The canonical 3q29 deletion locus, modified from the National Center for Biotechnology Information’s Genome Data Viewer. 16
References 17
CHAPTER 2. Convergent and distributed effects of the 3q29 deletion on the human neural transcriptome 35
Abstract 36
Introduction 37
Methods and Materials 38
Results 43
Discussion 50
Data availability 55
Figures 56
Figure 1. Unbiased weighted gene co-expression network analysis (WGCNA) of the human transcriptome in the healthy adult prefrontal cortex (PFC). 56
Figure 2. Network-based inference of the functional impact of 3q29Del on the adult human prefrontal cortex (PFC). 59
Figure 3. 3q29 interval genes form transcriptomic subnetworks enriched for known schizophrenia, autism and intellectual / developmental disability-risk genes. 61
Figure 4. Network of prioritized drivers predicted to contribute to the neuropsychiatric sequelae of 3q29Del. 63
Figure 5. Weighted gene co-expression network analysis (WGCNA) predicts differentially expressed genes in the mouse mode of 3q29Del. 65
References 67
Supplemental Materials 78
Extended Methods 78
Extended Results 97
Supplemental Figures 114
Figure S1. Tissue sample attributes and donor phenotypes of the GTEx dataset used for reference network construction. 114
Figure S2. Pre-processing of the reference dataset: outlier removal. 116
Figure S3. Determination of the soft-thresholding power (β) in weighted gene co-expression network analysis (WGCNA). 117
Figure S4. Determination of network reproducibility and module preservation in an independent test dataset. 119
Figure S5. Individual module preservation and quality statistics underlying composite Zsummary scores. 122
Figure S6. Publication numbers for 3q29 genes and historic schizophrenia spectrum disorder candidate genes. 123
Figure S7. Overlap between known protein-protein interactions (PPI) and gene co-expression patterns of 3q29 interval genes. 124
Figure S8. STRING protein-protein interaction (PPI) networks of genes co-expressed in 3q29 modules. 126
Figure S9. Amalgamated illustration of top biological processes and pathways enriched in 3q29 modules. 133
Supplemental References 134
CHAPTER 3. Structural deviations of the posterior fossa and the cerebellum and their cognitive links in a neurodevelopmental deletion syndrome 142
Abstract 143
Introduction 144
Methods and Materials 145
Results 151
Discussion 154
Acknowledgements 162
Data availability 162
Tables 163
Table 1. Demographic characteristics of the study sample in volumetric analyses, stratified by diagnostic group. 163
Table 2. Summary of multiple linear regression results testing for differences in volumetric measures of interest between 3q29Del and control subjects. 164
Table 3. Post hoc analysis of the sex-specific effects of diagnostic group on eICV. 166
Table 4. Exploratory analysis of the relationship between volumetric measures of interest and posterior fossa arachnoid cyst and mega cisterna magna findings among 3q29Del subjects. 167
Table 5. Summary of multiple linear regression results showing the relationships between cerebellar white matter volume and standardized test scores for sensorimotor and cognitive abilities among 3q29Del subjects. 169
Figures 171
Figure 1. Histogram showing the age distribution of study participants in volumetric analyses, stratified by sex and diagnostic group. 171
Figure 2. Scatter plots showing the distribution of A) total cerebellum volume, B) cerebellar cortex volume, C) cerebellar white matter volume, and D) eICV as a function of age among male and female subjects in each diagnostic group. 172
Figure 3. Predictor effect plots showing the effect of diagnostic group on volumetric measures of interest. 173
Figure 4. Predictor effect plot showing the moderating effect of sex on the relationship between diagnostic group and eICV. 175
Figure 5. Prevalence of posterior fossa arachnoid cyst and mega cisterna magna findings in structural MRI scans of 3q29Del and control subjects. 176
Figure 6. Predictor effect plots showing the relationships between cerebellar white matter volume and visual-motor integration skills, composite IQ, verbal IQ and non-verbal IQ among 3q29Del subjects. 178
References 180
Supplemental Materials 198
Extended Methods 198
Supplemental Tables 209
Table S1. Demographic characteristics of the study sample in volumetric analyses, stratified by diagnostic group and sex. 209
Table S2. Comparison of structural magnetic resonance imaging (MRI) protocols. 210
Table S3. Extended linear regression results testing the effect of diagnostic group on volumetric measures of interest and polynomial modeling of age. 212
Table S4. Summary of supplemental results from penalized cubic spline models testing the effect of diagnostic group on volumetric measures of interest. 221
Table S5. Exploratory modeling of diagnostic group by sex interaction effects on volumetric measures. 223
Table S6. Post hoc analysis of the suggestive sex by diagnostic group interaction effect on eICV-adjusted cerebellar white matter volumes. 226
Table S7. Summary of multiple linear regression findings and descriptive statistics for volumetric measures of interest in 3q29Del and control groups. 229
Table S8. Demographic and relevant clinical characteristics of 3q29Del subjects with versus without posterior fossa arachnoid cyst or mega cisterna magna findings. 230
Table S9. Descriptive statistics for standardized test scores for sensorimotor and cognitive abilities among 3q29Del subjects. 232
Table S10. Extended multiple linear regression results testing the relationships between tissue-specific cerebellar volumes and sensorimotor and cognitive abilities among 3q29Del subjects. 234
Table S11. Standardized test scores for sensorimotor and cognitive abilities in 3q29Del subjects with versus without posterior fossa arachnoid cyst or mega cisterna magna findings. 235
Supplemental Figures 236
Figure S1. Example cerebellar segmentation masks for representative age- and sex-matched 3q29Del and healthy control pairs. 237
Figure S2. Relationships between estimated total intracranial volume, total brain volume and head circumference among 3q29Del subjects. 239
Figure S3. Relationships between eICV and A) total cerebellum, B) cerebellar cortex, and C) cerebellar white matter volumes among 3q29Del subjects versus controls. 241
Figure S4. Correction of cerebellar volumes for head size variation in volumetric case-control analyses. 243
Figure S5. Scatter plots showing the distribution of eICV-adjusted A) total cerebellum volume, B) cerebellar cortex volume, and C) cerebellar white matter volume as a function of age among male and female subjects in each diagnostic group. 245
Figure S6. Regression diagnostics: testing the assumptions of ordinary least squares regression for best-fitting models from Table S3. 247
Figure S7. Developmental trajectories for volumetric measures of interest estimated by the penalized cubic spline approach. 248
Figure S8. Predictor effect plot showing a suggestive interaction effect between diagnostic group and sex on eICV-adjusted cerebellar white matter volumes. 249
Figure S9. Estimated normative percentile curves for cerebellar volumes and eICV, stratified by sex. 252
Figure S10. Scatter plots showing the distribution of A) total cerebellum volume, B) cerebellar cortex volume, C) cerebellar white matter volume, and D) eICV as a function of age among 3q29Del subjects with versus without posterior fossa arachnoid cyst or mega cisterna magna findings. 253
Figure S11. Violin plots with box plots visualizing the distribution of standardized test scores for sensorimotor and cognitive abilities among 3q29Del subjects. 254
Figure S12. Heatmap visualization of pairwise Pearson’s correlations between standardized test scores for sensorimotor and cognitive abilities among 3q29Del subjects. 255
Figure S13. Cerebellar protein expression profiles of 3q29 interval genes annotated by the Human Protein Atlas. 257
Supplemental References 259
CHAPTER 4. Psychosis spectrum symptoms among individuals with schizophrenia-associated copy number variants and evidence of cerebellar correlates of symptom severity 262
Abstract 263
Introduction 264
Methods and Materials 265
Results 269
Discussion 273
Acknowledgements 278
Data availability 278
Tables 278
Table 1. Demographic and relevant clinical information for individuals with 3q29Del. 279
Table 2. Rates of clinically significant psychotic symptoms (i.e., at least one SIPS item rated ≥ 3) among subjects with 3q29Del and 22q11.2Del. 280
Table 3. The overall results of the ANCOVA between the SIPS ratings of each diagnostic group and pairwise comparisons. 281
Figures 283
Figure 1. The unadjusted means of individual SIPS ratings for each diagnostic group. 283
Figure 2. The relationships between cerebellar structure and psychosis-risk symptoms in 3q29Del. 284
References 286
Supplemental Materials 300
Extended Methods 300
Extended Results 304
Supplemental Tables 306
Table S1. Demographic and relevant clinical information for the HC and 22q11.2Del samples and comparison with 3q29Del. 306
Table S2. Correlations between SIPS symptom ratings and age at visit, stratified by symptom domain and diagnostic group. 308
Table S3. Sex-adjusted means and standard errors of SIPS ratings after log-transformation. 309
Table S4. Demographic, clinical, and MRI-derived volumetric information for the 3q29Del subsample with available neuroimaging and SIPS data. 310
Table S5. Extended linear regression results: The effect of cerebellar volumetric measures on domain-specific symptom ratings in 3q29Del and polynomial modeling of age. 311
Table S6. Extended logistic regression results: The effect of cerebellar volumetric measures on the probability of a psychotic disorder or APSS diagnosis in 3q29Del and polynomial modeling of age. 320
Table S7. Nested comparison of diagnostic and dimensional phenotypes in 3q29Del subjects with versus without posterior fossa arachnoid cyst and mega cisterna magna findings. 323
Supplemental Figures 325
Figure S1. Age-stratified prevalence rates of florid or attenuated psychotic symptoms in 3q29Del. 325
Figure S2. Scatter plots of the relationship between SIPS symptom ratings and age at visit, stratified by symptom domain and diagnostic group. 327
Figure S3. Sex-specific SIPS ratings among deletion groups and HCs. 328
Figure S4. Comparison of hemisphere-specific ROI volumes in the 3q29Del subsample with available neuroimaging and SIPS data. 329
Figure S5. Sensitivity analysis: The relationship between cerebellar cortex volume and positive symptom severity in 3q29Del after removal of an extreme data point. 330
Supplemental References 331
CHAPTER 5. Conclusions 332
References 340
About this Dissertation
| School | |
|---|---|
| Department | |
| Subfield / Discipline | |
| Degree | |
| Submission | |
| Language |
|
| Research Field | |
| Mot-clé | |
| Committee Chair / Thesis Advisor | |
| Committee Members |
Primary PDF
| Thumbnail | Title | Date Uploaded | Actions |
|---|---|---|---|
|
|
A genetics-first, rare variant approach to understanding the neurobiological substrates of schizophrenia and associated disorders () | 2022-04-07 19:42:27 -0400 |
|
Supplemental Files
| Thumbnail | Title | Date Uploaded | Actions |
|---|