CASIA OpenIR  > 脑网络组研究
基于磁共振影像与多组学数据解析精神分裂症功能失连接生物学机制的研究
王萌
2022-05-14
Pages126
Subtype博士
Abstract

精神分裂症是一种常见的重性精神疾病,全球的终生患病率约为1%,具有高复发性、高异质性和高致残性等特点。精神分裂症患者的主要临床表现为妄想、幻觉和概念紊乱等阳性症状;以及被动/淡漠、快感缺失和社交回避等阴性症状;同时伴有思维、注意和记忆等方面的认知功能障碍。由于精神分裂症通常发病于青春期晚期或成年早期并持续终生,给个人、家庭及社会都造成了巨大的负担。然而,由于目前为止精神分裂症的病理机制尚不清楚,严重阻碍了患者的精准诊断与有效治疗。失连接假说是精神分裂症病理机制的核心假说之一,系列基于活体的多模态脑影像技术研究,为精神分裂症的宏观尺度失连接提供了重要证据。然而,这种宏观尺度失连接背后的生物学机制尚不清楚。随着近年来多尺度脑科学相关技术的发展和基因组、转录组等多组学数据的涌现,结合机器学习和统计分析模型的跨尺度数据综合分析研究,可能为深入理解精神分裂症的病理机制开辟了新的途径。本文基于多中心、大样本和跨种族的精神分裂症数据集,系统考察了精神分裂症的稳定宏观功能失连接模式,进而结合磁共振影像、人脑转录组、单细胞测序和基因组数据,创新性地通过结合多种统计分析方法发展了跨尺度研究策略。论文的主要工作和创新点归纳如下:

1. 基于磁共振影像的精神分裂症功能失连接模式研究

尽管已有研究表明精神分裂症中存在广泛的功能失连接,然而目前仍缺乏一致的明确证据。本文基于3个独立的、多中心和跨种族的精神分裂症磁共振影像数据集,采用功能梯度和全脑功能连接的分析方法,揭示了精神分裂症患者中存在稳定的皮层–皮层和纹状体–皮层功能失连接模式,并且两种功能失连接模式在3个独立数据集中高度复现。两种功能失连接模式还具有高度的稳健性,不受头动和不同预处理步骤等影响。该稳定一致的失连接模式为精神分裂症的失连接假说提供了重要的影像学证据支持,也为临床关联和生物学机制的理解奠定了基础。

2. 功能失连接与临床症状和认知功能的关联研究

在发现精神分裂症中存在稳定可靠的皮层–皮层和纹状体–皮层功能失连接基础上,本文接下来试图揭示两种功能失连接对精神分裂症的可能临床贡献。通过发展基于广义加性模型的影像症状关联分析方法,本文考察了两种不同的功能失连接模式与患者的临床症状的关系。在3个数据集中的研究结果一致表明:皮层–皮层显著的功能失连接能够预测患者的阴性和一般病理学症状,但不能预测阳性症状;而纹状体–皮层显著的功能失连接能够预测患者的阴性和阳性症状,但不能预测一般病理学症状。针对3个数据集的功能失连接模式,基于认知功能注释的统计关联分析,本文发现:皮层–皮层功能失连接与面部/情感处理相关,纹状体–皮层功能失连接与注意力聚焦、抑制错误和疼痛相关。以上研究建立了精神分裂症中两种宏观功能失连接与个体临床症状和认知功能的关系,对于精神分裂症患者临床症状的客观评估与脑失连接模式的理解提供了新的视角。

3. 基于转录组与单细胞测序数据的功能失连接生物学机制研究

上述基于神经影像学的研究结果发现了精神分裂症中稳定的宏观尺度功能失连接模式,然而其背后的细胞机制完全未知。在影像学研究的基础上,本文结合人脑转录组和两种单细胞测序数据,通过发展影像转录组和单细胞富集的联合分析方法,建立了精神分裂症的宏观功能失连接与微观的细胞类型环路间的联系。研究结果发现,皮层–皮层功能失连接与第2/3层的兴奋性神经元和少突胶质细胞构成的皮层–皮层细胞投射环路有关;而纹状体–皮层功能失连接与第4/5层的兴奋性神经元和少突胶质细胞构成的皮层–皮下细胞投射环路有关。两种宏观失连接–微观细胞环路间的联系在两个独立的单细胞数据集中具有高可复现性。此外,基于抑郁症数据集的对比分析表明我们所发现的跨尺度环路损伤在精神分裂症中可能具有一定的疾病特异性。本文跨尺度多组学的研究策略为系统研究精神疾病的病理机制开辟了新的研究思路,研究结果为理解精神分裂症宏观功能失连接的细胞环路机制提供了科学证据。

4. 基于基因组数据的功能失连接生物学机制验证研究

以上研究建立了精神分裂症细胞环路改变-宏观失连接-临床症状的关联,然而已有证据表明精神分裂症是多基因遗传疾病,遗传变异与细胞环路间的关系,进而与宏观失连接-临床症状的关联尚不清楚。因此,本文在细胞层次研究的基础上,结合多个全基因组关联分析(GWAS)研究和个体的基因组数据,通过多组学联合分析发现,GWAS研究发现的精神分裂症基因风险位点显著富集于第2/3层和第4/5层兴奋性神经元的特异性基因集中,富集结果在两个独立的单细胞数据集中具有高度的一致性。基于个体的基因组数据,本文计算了单细胞特异性的多基因风险分数并考察了在精神分裂症患者与健康对照中的分布差异,研究发现精神分裂症患者呈现显著增加的细胞类型特异性的多基因风险分数,主要包括第2/3层和第4/5层兴奋性神经元和少突胶质细胞特异性的多基因风险分数。基因组层次上的研究发现从遗传角度为细胞环路改变提供了证据支持。

综上,本文揭示了精神分裂症的宏观功能失连接与微观的细胞类型环路间稳定可靠的联系,同时建立了遗传变异-细胞环路-宏观功能失连接-临床症状四者之间的关联,为全面理解精神分裂症的病理机制提供了科学依据。

Other Abstract

Schizophrenia is a common severe mental illness with a lifetime prevalence of about 1% worldwide, characterized by high relapse, heterogeneity, and disability. The main clinical manifestations of patients with schizophrenia are positive symptoms such as delusions, hallucinations, and conceptual disorganization; as well as negative symptoms such as passivity/apathy, anhedonia, and social avoidance; and cognitive dysfunction in thinking, attention, and memory et al. Schizophrenia usually occurs in late adolescence or early adulthood and lasts for life, hence, it places a huge burden on individuals, families, and society. However, since the pathological mechanism of schizophrenia is still unclear, it seriously hinders the accurate diagnosis and effective treatment of patients. The dysconnectivity hypothesis is one of the core hypotheses of the pathological mechanism of schizophrenia, and a series of in vivo multimodal brain imaging studies have provided important evidence for the macroscale dysconnectivity of schizophrenia. However, the biological mechanism behind this macroscale dysconnectivity is unclear. With the development of multi-scale brain science-related technologies and the emergence of multi-omics data such as genome and transcriptome in recent years, the comprehensive analysis of cross-scale data combined with machine learning and statistical analysis models may open up a new way for the in-depth understanding of the pathological mechanism of schizophrenia. Based on multicenter, large sample, and cross-racial schizophrenia datasets, this paper systematically investigates the consistent macroscale functional dysconnectivity in schizophrenia, then combines magnetic resonance imaging, whole-brain transcriptomic data, single-cell sequencing data, and subject-specific genome-wide genotyping data, and innovatively develops cross-scale research strategies by combining multiple statistical analysis methods. Based on this strategy, this paper consistently and robustly reveals the connection between the macroscale functional dysconnectivity and the microscale cell-type-specific circuit in schizophrenia and establishes the association between genetic variation–cellular circuit–macroscale functional dysconnectivity–clinical symptoms, which provides a scientific basis for a comprehensive understanding of the pathological mechanism of schizophrenia. The main contributions and innovation points of the paper are summarized as follows:

1. Study on the functional dysconnectivity pattern in schizophrenia based on magnetic resonance imaging.

Although existing studies have shown brain-wide functional dysconnectivity in schizophrenia, there is a lack of consistent and definitive evidence. In this paper, based on three independent, multicenter, and cross-racial schizophrenia magnetic resonance image datasets, the functional gradient and whole-brain functional connectivity are investigated and we find that there exist consistent corticocortical and corticostriatal dysconnectivity patterns in schizophrenia, which are highly reproducible in three independent datasets. Besides, the two dysconnectivity patterns are also highly robust and are not affected by head motion and different preprocessing steps. The reproducible and robust dysconnectivity pattern provides important neuroimaging evidence for the dysconnectivity hypothesis of schizophrenia and also lays the foundation for the understanding of clinical associations and biological mechanisms.

2. Study on the association between functional dysconnectivity and clinical symptoms and cognitive functions.

Based on the discovery of consistent and robust corticocortical and corticostriatal dysconnectivity patterns in schizophrenia, this paper then attempts to reveal the possible clinical contributions of the two dysconnectivity patterns. By developing a neuroimaging-symptom association method based on generalized additive models, this paper investigates the relationships between the two dysconnectivity patterns and the clinical symptoms in schizophrenia. The findings in the three datasets consistently show that the significant corticocortical dysconnectivity can predict negative and general psychopathology symptoms but not positive symptoms, while the significant corticostriatal dysconnectivity can predict negative and positive symptoms but not general psychopathology symptoms. Based on cognitive function annotations using the Neurosynth database, in three independent datasets, this paper consistently finds that corticocortical dysconnectivity is associated with face/affective processing, and corticostriatal dysconnectivity is associated with cued attention, inhibition error, and pain. The above analysis establishes the relationships between the two macroscale dysconnectivity patterns and individual clinical symptoms and cognition in schizophrenia and may provide a new perspective on the objective assessment of clinical symptoms in patients with schizophrenia and the understanding of macroscale dysconnectivity.

3. Study on the biological mechanism of functional dysconnectivity based on transcriptome and single-cell sequencing data.

The above neuroimaging-based findings have found reproducible macroscale functional dysconnectivity patterns in schizophrenia, yet the cellular mechanisms behind them are completely unknown. Based on the neuroimaging study, this paper combines whole-brain transcriptome and two single-cell sequencing data to establish the connection between the macroscale functional dysconnectivity and microscale cell-type-specific circuit by developing a joint method of imaging transcriptome and single-cell enrichment. The results of the study show that corticocortical dysconnectivity is related to corticocortical cellular projection composed of excitatory neurons in layer 2/3 and oligodendrocytes, while corticostriatal dysconnectivity is related to corticosubcortical cellular projection composed of excitatory neurons in layer 4/5 and oligodendrocytes. The connections between the two macroscale functional dysconnectivity and microscale cell circuits are highly reproducible in two independent single-cell datasets. In addition, a control analysis based on the major depression disorder dataset suggests that the cross-scale disruption is specific in schizophrenia. The cross-scale multi-omics strategy developed in this paper opens up new research ideas for the systematic study of the pathological mechanism of mental illness, and the results provide scientific evidence for understanding the cellular circuit mechanism of macroscale functional dysconnectivity in schizophrenia.

4. Validation of the biological mechanism of functional dysconnectivity based on genomic data.

The above study has established the association between cellular circuit disruption–macroscale functional dysconnectivity–clinical symptoms. There is evidence that schizophrenia is a polygenic genetic disorder, however, the relationship between genetic variation and cellular circuit, and thus the association with macroscale functional dysconnectivity–clinical symptoms is unclear. Therefore, on the basis of cell-level research, combined with multiple genome-wide association analysis (GWAS) studies and individual genomic data, we perform a multi-omics joint analysis and found that the schizophrenia gene risk alleles found in GWAS research are significantly enriched in the specific gene sets of excitatory neurons at layers 2/3 and 4/5, and the enrichment results are highly consistent in two independent single-cell datasets. Based on the genomic data of individuals, the single-cell-specific polygenic risk scores are calculated and the differences between patients with schizophrenia and healthy controls are examined, and it is found that the cell-type-specific polygenic risk scores of schizophrenia patients showed a significant increase, mainly including the polygenic risk scores specific for excitatory neurons at layers 2/3 and 4/5 and oligodendrocytes. The findings at the genomic level provide genetic evidence for cell circuit disruption in schizophrenia.

Keyword精神分裂症 磁共振影像 多组学 跨尺度 失连接
Language中文
Document Type学位论文
Identifierhttp://ir.ia.ac.cn/handle/173211/48743
Collection脑网络组研究
毕业生_博士学位论文
Recommended Citation
GB/T 7714
王萌. 基于磁共振影像与多组学数据解析精神分裂症功能失连接生物学机制的研究[D]. 中国科学院自动化研究所. 中国科学院自动化研究所,2022.
Files in This Item:
File Name/Size DocType Version Access License
博士学位论文_王萌.pdf(7219KB)学位论文 开放获取CC BY-NC-SA
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[王萌]'s Articles
Baidu academic
Similar articles in Baidu academic
[王萌]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[王萌]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.