Xinjiang Laboratory of Native Medicinal and Edible Plant Resources Chemistry, College of Chemistry and Environmental Science, Kashi University, Kashi 844000, China.
Xinjiang Laboratory of Native Medicinal and Edible Plant Resources Chemistry, College of Chemistry and Environmental Science, Kashi University, Kashi 844000, China.
Spectrochim Acta A Mol Biomol Spectrosc. 2023 Dec 5;302:123072. doi: 10.1016/j.saa.2023.123072. Epub 2023 Jun 24.
Candida rugosa lipase (CRL, EC3.1.1.3) is one of the main enzymes synthesizing esters, and ZIF-8 was chosen as an immobilization carrier for lipase. Enzyme activity testing often requires expensive reagents as substrates, and the experiment processes are time-consuming and inconvenient. As a result, a novel approach based on near-infrared spectroscopy (NIRs) was developed for predicting CRL/ZIF-8 enzyme activity. The absorbance of the immobilized enzyme catalytic system was evaluated using UV-Vis spectroscopy to investigate the amount of CRL/ZIF-8 enzyme activity. The powdered samples' near-infrared spectra were obtained. The sample's enzyme activity data were linked with each sample's original NIR spectra to establish the NIR model. A partial least squares (PLS) model of immobilized enzyme activity was developed by coupling spectral preprocessing with a variable screening technique. The experiments were completed within 48 h to eliminate inaccuracies between the reduction in enzyme activity with increasing laying-aside time throughout the test and the NIRs modeling. The root-mean-square error of cross-validation (RMSECV), the correlation coefficient of validation set (R) value, and the ratio of prediction to deviation (RPD) value were employed as assessment model indicators. The near-infrared spectrum model was developed by merging the best 2nd derivative spectral preprocessing with the Competitive Adaptive Reweighted Sampling (CARS) variable screening method. This model's root-mean-square error of cross-validation (RMSECV) was 0.368 U/g, the correlation coefficient of calibration set (R_cv) value was 0.943, the root-mean-square error of prediction (RMSEP) set was 0.414 U/g, the correlation coefficient of validation set (R) value was 0.952, and the ratio of prediction to deviation (RPD) was 3.0. The model demonstrates that the fitting relationship between the predicted and the reference enzyme activity value of the NIRs is satisfactory. The findings revealed a strong relationship between NIRs and CRL/ZIF-8 enzyme activity. As a result, the established model could be implemented to quantify the enzyme activity of CRL/ZIF-8 quickly by including more variations of natural samples. The prediction method is simple, rapid, and adaptable to be the theoretical and practical basis for further studying other interdisciplinary research work in enzymology and spectroscopy.
皱褶假丝酵母脂肪酶（CRL，EC3.1.1.3）是合成酯的主要酶之一，ZIF-8 被选择作为脂肪酶的固定化载体。酶活性测试通常需要昂贵的试剂作为底物，并且实验过程耗时且不方便。因此，开发了一种基于近红外光谱（NIRs）的新方法来预测 CRL/ZIF-8 酶活性。使用紫外-可见光谱法评估固定化酶催化体系的吸光度，以研究 CRL/ZIF-8 酶活性的量。获得粉末样品的近红外光谱。将样品的酶活性数据与每个样品的原始 NIR 光谱相关联，以建立 NIR 模型。通过结合光谱预处理和变量筛选技术，建立了固定化酶活性的偏最小二乘（PLS）模型。实验在 48 小时内完成，以消除测试过程中随着搁置时间的增加而导致的酶活性降低与 NIRs 建模之间的不准确性。交叉验证均方根误差（RMSECV）、验证集相关系数（R）值和预测偏差比（RPD）值被用作评估模型指标。通过合并最佳二阶导数光谱预处理和竞争自适应重加权采样（CARS）变量筛选方法，建立了近红外光谱模型。该模型的交叉验证均方根误差（RMSECV）为 0.368 U/g，校准集相关系数（R_cv）值为 0.943，预测集均方根误差（RMSEP）值为 0.414 U/g，验证集相关系数（R）值为 0.952，预测偏差比（RPD）值为 3.0。该模型表明，NIRs 与 CRL/ZIF-8 酶活性之间的预测值与参考酶活性值之间存在良好的拟合关系。研究结果表明，NIRs 与 CRL/ZIF-8 酶活性之间存在很强的关系。因此，通过包含更多自然样本的变化，可以建立该模型来快速定量 CRL/ZIF-8 的酶活性。预测方法简单、快速，适用于进一步研究酶学和光谱学等其他跨学科研究工作的理论和实践基础。
