A Mitochondrial Etiology of Neuropsychiatric Disorders.
A Mitochondrial Etiology of Neuropsychiatric Disorders.
复制标题
DOI：
10.1001/jamapsychiatry.2017.0397
复制
发表时间：
2017-09
期刊：
影响因子：
25.8
通讯作者：
D. Wallace
中科院分区：
文献类型：
--
作者：
D. Wallace
利用强大的基因组学技术，包括全基因组关联研究（GWAS）、全外显子组测序（WES）和全基因组测序，已经投入了大量资源来分析神经精神疾病，以寻找与这些疾病相关的核DNA （nDNA）基因变异。然而，没有一致的病理生理病因精神障碍已经出现。例如，在GWAS和WES对数千例自闭症病例进行分析后，发现了许多拷贝数变异和功能丧失突变，但没有一种变异占很大比例。此外，在自闭症患者中发现的携带功能丧失突变的基因与那些与先天性心脏病和代谢紊乱相关的基因重叠“脑部”疾病与先天性心脏病和代谢紊乱有什么关系？可能的问题不在于数据，而在于我们试图解释数据的视角。在西方医学哲学中，假设一种症状来自某个器官，那么问题的“原因”一定是该器官的缺陷。行为、学习和记忆方面的问题与大脑有关。因此，我们认为，要发现这些问题的原因，我们必须研究大脑。然而，还有另一种观点。有可能是大脑对某些系统性缺陷特别敏感，从而导致大脑特有的症状。但系统性缺陷怎么会优先影响大脑呢？大脑只占我们体重的2%到3%，但消耗了我们线粒体能量的20%，这是通过耗氧量来衡量的。因此，随着全身线粒体能量代谢的下降，大脑是第一个能量受到限制的器官，低于其生物能量阈值并导致神经系统症状。生物能量缺陷越轻微，症状就越具有大脑特异性，多动症或抑郁症就是很可能的例子。如果线粒体氧化磷酸化（OXPHOS）减少20%就足以引起抑郁症，那么使用当前技术在个体中不会明显证明生化缺陷。相比之下，对患者队列的分析显示，在多种神经系统疾病中，与对照组相比，大脑和外周组织中OXPHOS的平均差异具有统计学意义。如果线粒体功能障碍是精神疾病的主要原因，为什么还没有发现受影响的线粒体基因？他们有。线粒体基因组由1000到2000个DNA编码基因和数千个母系遗传的线粒体DNA （mtDNA）组成。线粒体基因组的nDNA成分编码了线粒体解剖成分、线粒体中间代谢和线粒体生物发生以及线粒体调控的所有基因。mtDNA组件为OXPHOS及其表达的最关键基因编码，mtDNA本质上是发电厂的接线图。mtDNA的突变率非常高。因此，新的突变有规律地出现，最初产生突变体和正常mtdna的混合物（异质性）。突变体的异质百分率的变化可以产生分级缺陷和可变表型。由于多个基因编码OXPHOS的每种酶复合物，多个不同的基因改变可导致相似的生物能量缺陷和相关表型。虽然大脑是对线粒体功能障碍最敏感的器官，但心脏、肾脏、肌肉和内分泌系统也可能是高能量需求的组织。像大脑一样，心脏处于慢性能量负荷之下，并且将是对慢性能量缺乏做出反应的第一个器官之一，从而导致心肌病。轻度线粒体功能障碍也与糖尿病和代谢综合征有关因此，nDNA和mtDNA基因突变导致部分线粒体功能障碍，可引起高能器官功能障碍，导致神经、心脏和内分泌症状。有相当多的证据支持线粒体功能障碍与神经精神疾病有关的假设。例如，在核苷酸4336A>G的mtDNA tRNAGln基因中发现了一个古老的mtDNA单核苷酸多态性（SNP），约3%的阿尔茨海默病患者，约5%的帕金森病患者，约7%的阿尔茨海默病和帕金森病患者，但仅占欧洲人口的0.4%左右具有功能变异（单倍群）的古老mtDNA谱系与广泛的神经、心脏和代谢疾病的易感性相关线粒体DNA突变也随着年龄在组织中积累，导致线粒体功能与年龄相关的下降。这些体细胞mtDNA突变导致的进行性线粒体缺陷可加剧遗传性nDNA或mtDNA线粒体基因缺陷，最终导致复合能量缺乏，低于神经元生物能量阈值，导致神经精神症状然而，据我所知，在GWAS或WES临床研究中很少分析mtDNA变异。对于GWAS来说，问题在于Illumina对mtDNA snp的选择。而不是选择mtDNA snp的观点
Enormous resources have been invested in the analysis of neuropsychiatric disorders using powerful genomics techniques, including genome-wide association studies (GWAS), whole-exome sequencing (WES), and whole-genome sequencing, to search for nuclear DNA (nDNA) gene variants associated with these disorders. Yet, no coherent pathophysiological etiology for psychiatric disorders has emerged. For example, after analysis of thousands of autism cases by GWAS and WES, numerous copy number variants and loss-offunction mutations have been identified, but no single variant accounts for a significant proportion of cases. Moreover, the genes that have been found to harbor loss-of-function mutations in patients with autism overlap with those associated with congenital heart disease and metabolic disorders.1 What do “brain” diseases have to do with congenital heart disease and metabolic disorders? The likely problem is not with the data but with the perspective from which we are attempting to interpret the data. In Western medical philosophy, it is assumed that if a symptom emanates from an organ, then the “cause” of the problem must be a defect in that organ. Problems with behavior, learning, and memory relate to the brain. Therefore, we assume that to discover the causes of these problems, we must look into the brain. However, there is an alternative perspective. It is possible that there might be systemic defects to which the brain is uniquely sensitive, thus causing brainspecific symptoms. But how could a systemic defect preferentially affect the brain? The brain is only 2% to 3% of our body’s weight but expends 20% of our mitochondrial energy, as measured by oxygen consumption. Therefore, as systemic mitochondrial energy metabolism declines, the brain is the first organ to become limited for energy, dropping below its bioenergetic threshold and resulting in neurological symptoms. The milder the bioenergetic defect, the more brain-specific the symptoms, with hyperactivity or depression being likely examples. If a 20% reduction in mitochondrial oxidative phosphorylation (OXPHOS) were sufficient to cause depression, the biochemical defect would not be significantly demonstrable in an individual using current technology. By contrast, analysis of cohorts of patients has revealed statistically significant mean differences in OXPHOS relative to control individuals in the brain as well as in peripheral tissues for a variety of neurological diseases. If mitochondrial dysfunction is a major cause of psychiatric disorders, why haven’t the affected mitochondrial genes been found? They have. The mitochondrial genome consists of 1000 to 2000 nDNA-coded genes plus thousands of copies of the maternally inherited mitochondrial DNA (mtDNA). The nDNA component of the mitochondrial genome codes for all of the genes for the anatomical components of the mitochondrion, for mitochondrial intermediate metabolism and mitochondrial biogenesis, and for the regulation of the mitochondrion. The mtDNA component codes for the most critical genes for OXPHOS and for their expression, the mtDNA being in essence the wiring diagram of the power plant. The mtDNA has a very high mutation rate. Hence, new mutations arise regularly, initially giving rise to mixtures of mutant and normal mtDNAs (heteroplasmy). Changes in the heteroplasmy percentage of a mutant can give graded defects and variable phenotypes. Because multiple genes code for each of the enzyme complexes of OXPHOS, multiple different gene alterations can result in similar bioenergetic defects and related phenotypes. While the brain is the organ most sensitive to mitochondrial dysfunction, the heart, kidney, muscle, and endocrine systems can also be high energy demand tissues. Like the brain, the heart is under chronic energy load and will be one of the first organs to respond to chronic energy deficiency, resulting in cardiomyopathy. Mild mitochondrial dysfunction is also associated with diabetes and metabolic syndrome.2 Hence, nDNA and mtDNA gene mutations that result in partial mitochondrial dysfunction can cause dysfunction in high energy organs leading to neurological, cardiac, and endocrine symptoms. There is considerable evidence supporting the hypothesis that mitochondrial dysfunction is associated with neuropsychiatric disorders. For example, an ancient mtDNA single-nucleotide polymorphism (SNP) in the mtDNA tRNAGln gene at nucleotide 4336A>G is found in about 3% of patients with Alzheimer disease, about 5% of those with Parkinson disease, and about 7% of those with Alzheimer and Parkinson diseases, but only about 0.4% of the European population.3 Ancient mtDNA lineages harboring function variants (haplogroups) have been associated with predisposition to a broad range of neurological, cardiac, and metabolic diseases.2 Mitochondrial DNA mutations also accumulate with age in tissues, resulting in the age-related decline in mitochondrial function. The progressive mitochondrial defects resulting from these somatic mtDNA mutations can exacerbate inherited nDNA or mtDNA mitochondrial gene defects, ultimately causing a composite energy deficiency that falls below neuronal bioenergetic thresholds resulting in neuropsychiatric symptoms.2 Yet, to my knowledge, mtDNA variation is rarely analyzed in GWAS or WES clinical studies. For GWAS, the problem has been the selection by Illumina of the mtDNA SNPs to be interrogated. Rather than select mtDNA SNPs VIEWPOINT