电喷雾萃取及电离质谱法分析大豆水提液快速鉴别种子活力

Rapid identification of seed vigor of soybean water extracts by electrospray ionization mass spectrometry

为快速评价大豆种子活力,该文以人工加速老化后的大豆种子为原料,采用电喷雾萃取电离质谱(extractive electrospray ionization-mass spectrometry,EESI-MS)直接检测不同活力大豆种子水提液,获得其正离子模式下指纹图谱,结合主成分分析(principal component analysis,PCA)、聚类分析(cluster analysis,CA)和判别分析(discriminant analysis,DA),建立鉴别大豆种子活力的方法。结果表明,在正离子模式下,电喷雾萃取电离质谱结合多变量分析方法可以有效鉴别不同活力大豆种子。主成分分析提取了前3个主成分因子,累计贡献率达到94.80%,聚类分析和判别分析的正确率都为100%,对外部验证样本进行判别分析,正确率为95%。电喷雾萃取电离质谱能快速鉴别不同活力大豆种子,为种子活力检测提供了一种快速、准确、高效的新方法。

In order to evaluate the vigor of soybean seeds quickly and accurately, soybean seeds after artificial accelerated aging were taken as material, and extractive electrospray ionization mass spectrometry（EESI-MS） was performed to obtain chemical fingerprints directly from the water extracts of soybean seeds with different vigor. According to the principle of electrical conductivity method, the water extracts from 8 different degrees of deterioration of soybean seed samples were tested and distinguished without complex sample pretreatments. The raw mass spectra data were analyzed using multivariate analysis, including principal component analysis（PCA）, cluster analysis（CA） and discriminant analysis（DA） to establish a method to identify different soybean seed vigor. The seeds were randomly picked from a local cultivated variety, Gandou 8. Every aqueous extract of soybean seed sample was determined 6 times in parallel. EESI-MS was performed on a commercial linear ion trap mass spectrometer installed with a homemade EESI-MS ion source under the positive ion detection mode. A gentle nitrogen（N2） sheath gas flow（1.2 MPa） was driven into liquid methanol to form N2/methanol gas flow as spray reagent with an ionization voltage of 3.5 k V. The mass scanning range was 50-800 and the ion transfer tube temperature was 180 ℃. The mass spectra were rapidly recorded by EESI-MS and the data were processed by multivariate analyses. PCA was performed with Matlab 7.0 software, and CA and DA were performed with statistical program for SPSS19.0 software. The results showed that in the positive ion mode, EESI-MS combined with multivariate analysis method could effectively identify the soybean seed with different vigor. For the first 80 components, 3 principal components were extracted using PCA method, with the contribution rates of 83.70%（PC1）, 10.10%（PC2） and 1%（PC3）, respectively, and the cumulative contribution rate of 94.80%, maintaining most of the original information of the samples. All the seeds had a regular pattern of aggregation and dispersion. The seeds with the same aging time were clustered together, and the seeds with different aging time could be completely separated. CA program was used to calculate the Euclidean distance between 40 soybean seed samples for stratified CA. The seeds of different vigor could also be completely separated with the correct cluster rate of 100%. Eighty soybean seed samples were analyzed by stepwise discriminant analysis with Wilk＇lambda method, and Fisher discriminant analysis was conducted. Soybean seed samples of different aging time could be completely separated, and the soybean seeds with same vitality were clustered together with the correct rate of 100%. Furthermore, the correct rate of 40 external validation samples was 95% for DA. Therefore, EESI-MS can quickly identify the soybean seeds with different vigor, providing a new method for the fast, accurate and efficient seed vigor detection. The experiment provides a theoretical basis for the vigor detection of soybean seed.