Systems Medicine of Metabolism and Signaling, Laboratory of Pediatrics, University Medical Center Groningen, University of Groningen, 9713, AV, Groningen, the Netherlands; Systems Biology Centre for Energy Metabolism and Ageing, University of Groningen, 9713, AV, Groningen, the Netherlands.
Systems and Synthetic Biology, Wageningen University & Research, 6708, WE, Wageningen, the Netherlands.
Biochim Biophys Acta Mol Basis Dis. 2019 Feb 1;1865(2):360-370. doi: 10.1016/j.bbadis.2018.10.021. Epub 2018 Oct 29.
Flavin adenine dinucleotide (FAD) and its precursor flavin mononucleotide (FMN) are redox cofactors that are required for the activity of more than hundred human enzymes. Mutations in the genes encoding these proteins cause severe phenotypes, including a lack of energy supply and accumulation of toxic intermediates. Ideally, patients should be diagnosed before they show symptoms so that treatment and/or preventive care can start immediately. This can be achieved by standardized newborn screening tests. However, many of the flavin-related diseases lack appropriate biomarker profiles. Genome-scale metabolic models can aid in biomarker research by predicting altered profiles of potential biomarkers. Unfortunately, current models, including the most recent human metabolic reconstructions Recon and HMR, typically treat enzyme-bound flavins incorrectly as free metabolites. This in turn leads to artificial degrees of freedom in pathways that are strictly coupled. Here, we present a reconstruction of human metabolism with a curated and extended flavoproteome. To illustrate the functional consequences, we show that simulations with the curated model - unlike simulations with earlier Recon versions - correctly predict the metabolic impact of multiple-acyl-CoA-dehydrogenase deficiency as well as of systemic flavin-depletion. Moreover, simulations with the new model allowed us to identify a larger number of biomarkers in flavoproteome-related diseases, without loss of accuracy. We conclude that adequate inclusion of cofactors in constraint-based modelling contributes to higher precision in computational predictions.
黄素腺嘌呤二核苷酸（FAD）及其前体黄素单核苷酸（FMN）是氧化还原辅因子，它们是一百多种人类酶活性所必需的。编码这些蛋白质的基因突变会导致严重的表型，包括能量供应不足和有毒中间产物的积累。理想情况下，应该在患者出现症状之前进行诊断，以便立即开始治疗和/或预防护理。这可以通过标准化的新生儿筛查测试来实现。然而，许多与黄素有关的疾病缺乏适当的生物标志物特征。基于基因组规模的代谢模型可以通过预测潜在生物标志物的改变特征来辅助生物标志物研究。不幸的是，当前的模型，包括最近的人类代谢重建 Recon 和 HMR，通常不正确地将结合酶的黄素视为游离代谢物。这反过来又导致了严格耦合途径中人为的自由度。在这里，我们提出了一个经过精心整理和扩展的黄素蛋白组的人类代谢重建。为了说明功能后果，我们表明，与早期的 Recon 版本相比，经过整理的模型的模拟可以正确预测多种酰基辅酶 A 脱氢酶缺乏症以及系统性黄素耗竭的代谢影响。此外，新模型的模拟使我们能够在不降低准确性的情况下，在与黄素蛋白组相关的疾病中识别出更多数量的生物标志物。我们得出结论，在基于约束的建模中充分包含辅助因子有助于提高计算预测的精度。
