中国科学院海洋研究所机构知识库

牡蛎是世界性大宗贝类，其风味和品质受糖原、氨基酸和脂肪酸等营养成分的影响较大。在牡蛎相关基础研究中，营养品质性状解析是分子育种的一个重要方向。随着测序等技术的快速发展，基因等序列资源积累迅速；但传统的营养物质含量检测方法成本普遍很高，加上大规模基因分型成本也较高，成为制约相关性状遗传解析等研究的障碍。本研究首先利用傅里叶变换近红外光谱仪建立了较为可靠的长牡蛎中糖原等营养成分的近红外模型，然后通过近红外光谱法获取了采集自不同环境的1278个野生长牡蛎个体的表型数据，选取极端表型个体进行重测序和转录组联合分析寻找营养品质性状的关键基因和位点。

本研究中，开发并优化了基于近红外反射光谱快速高效地测定长牡蛎的营养成分含量的方法。将来自 19 个沿海地点的长牡蛎样品冷冻干燥、研磨并使用傅里叶变换近红外光谱仪扫描以收集光谱数据，使用偏最小二乘法、乘法散射校正、一阶导数和 Norris 平滑建立了糖原、氨基酸和脂肪酸等营养成分的近红外模型。糖原、脂肪酸、氨基酸和牛磺酸近红外模型的R_C 值分别为0.9678、0.9312、0.9132和0.8928，对应的RPD_V 值分别为3.33、2.04、3.10和1.62，表明预测值和实测值之间具有高度相关性，模型准确稳定可进行实践应用。

利用上述模型评估了1278个牡蛎样品的营养成分，结果表明糖原含量随季节而变化，并且在一月份含量更高，糖原含量的变化与脂肪酸趋势一致，与氨基酸趋势相反。在潮下带和潮间带培养的长牡蛎的糖原、氨基酸和牛磺酸水平也存在显著差异。与潮下带相比，生活在潮间带的双壳贝类面临更大的环境压力，摄食时长更短。为了维持体内平衡，潮间带长牡蛎的糖原等营养物质被消耗较快，这表明潮位可能是影响长牡蛎营养成分含量的重要因素。

选取不同月份及不同潮位样本中的极端表型个体，分为高糖原含量组与低糖原含量组，进行组学分析。发现一月份低糖原与高糖原样本间的差异表达基因共4206个，三月份低糖原与高糖原样本间的差异表达基因共2243个，共有差异表达基因601个。根据重测序SNP等位基因频率的差异进行筛查，三月份个体中差异SNP所在的基因有2535个，一月份个体中差异SNP所在的基因有2971个，其中相同的基因有891个。一月份基因表达和SNP频率均存在差异的基因361个，三月份基因表达和SNP频率均存在差异的基因230个。以上结果表明在不同月份糖原含量调控基因及其表达模式可能存在较大差异。荧光定量实验结果进一步验证发现，与低糖原含量组相比，糖类反应元件结合蛋白(ChREBP)和溶质载体家族13A3（SLC13A3）基因在高糖原含量组的表达均为显著下调（p-value<0.01）。SLC13A3基因可能参与糖原相关的跨膜转运活动，转运柠檬酸循环的代谢中间体。ChREBP基因可以协同调节葡萄糖利用和脂肪酸合成，所以该基因的高表达造成了相应糖原含量的降低，即目的基因表达量对糖原含量起抑制作用。此外，重测序和转录组联合分析还富集到了与ATP的水解以及作用于修饰蛋白的催化活性相关的通路，这些通路中的基因可能通过糖酵解和三羧酸循环等作用于涉及磷酸基团的化学反应和途径，参与糖原代谢以及氨基酸合成等许多生物学过程。

Oyster is a major shellfish in the world, and its flavor and quality are greatly affected by the nutritional components such as glycogen, amino acid and fatty acid. The analysis of nutritional quality traits is an important direction of molecular breeding in the basic research of oyster. With the rapid development of sequencing technology, gene sequence resources accumulate rapidly. However, the high cost of traditional nutrient content detection methods and the high cost of large-scale genotyping have become obstacles to the genetic analysis of related traits. In this study, more reliable near-infrared models of glycogen and other nutrients in Crassostrea gigas was established by Fourier transform NIR spectrometer. Then, phenotypic data of 1278 wild Crassostrea gigas individuals collected from different environments were obtained by NIR spectroscopy. The individuals with extreme phenotypes were selected for resequencing and transcriptome analysis to search for key genes and loci of nutritional quality traits.

In this study, a fast and efficient method was developed for the determination of nutrient content in Crassostrea gigas based on near-infrared reflectance spectroscopy. Samples of Crassostrea gigas from 19 coastal sites were freezed-dried, ground, and scanned by Fourier transform near-infrared spectrometer to collect spectral data. NIR models of glycogen, amino acids, and fatty acids were established using partial least squares, multiplication scattering correction, first derivatives, and Norris smoothing. The R_C values of glycogen, fatty acid, amino acid and taurine NIR models were 0.9678, 0.9312, 0.9132 and 0.8928, respectively. And the corresponding RPD_V values were 3.33, 2.04, 3.10 and 1.62, respectively, indicating that the predicted values and measured values were highly correlated, and the model was accurate and stable for practical application.

The nutrient composition of 1278 oyster samples was evaluated using the above model, and the results showed that glycogen content varied seasonally and was higher in January. The change in glycogen content was consistent with the fatty acid trend and opposite to the amino acid trend. There were also significant differences in glycogen, amino acid and taurine levels between Crassostrea gigas cultured in subtidal and intertidal zones. Compared with the subtidal zone, bivalves living in the intertidal zone face greater environmental stress and feed for a shorter period of time. In order to maintain homeostasis, glycogen and other nutrients are consumed in intertidal oysters, suggesting that tide level may be an important factor affecting the nutrient content of Crassostrea gigas.

Individuals with extreme phenotype in different month and different tidal level samples were divided into high and low glycogen content groups for omics analysis. There were 4206 differentially expressed genes between low and high glycogen samples in January, 2243 differentially expressed genes between low and high glycogen samples in March, and 601 common differentially expressed genes. Screening based on differences in allele frequencies of resequenced SNPS showed that there were 2,535 differentially encoded genes in March and 2,971 differentially encoded genes in January, of which 891 were identical. There were 361 genes with different gene expression and SNP frequency in January, and 230 genes with different gene expression and SNP frequency in March. These results indicate that there may be significant differences in glycogen content regulation genes and their expression patterns in different months. Fluorescence quantitative PCR further verified that carbohydrate response element binding protein (ChREBP) and solute carrier family 13A3 (SLC13A3) genes were significantly down-regulated in the high glycogen content group compared with the low glycogen content group (p-value<0.01). The SLC13A3 gene may be involved in glycogen-related transmembrane transport activities and transport the metabolic intermediates of the citric acid cycle. ChREBP gene can synergistically regulate glucose utilization and fatty acid synthesis, so the high expression of this gene leads to the decrease of corresponding glycogen content, that is, the expression level of the target gene inhibits the glycogen content. In addition, the resequencing and transcriptome analysis also enriched with the ATP hydrolysis and the catalytic activity of modified proteins related pathways. The genes in these pathways may act on chemical reactions and pathways involving phosphate groups through glycolysis and the tricarboxylic acid cycle, and participate in many biological processes such as glycogen metabolism and amino acid synthesis.