Cournoyer Aurore, Bazinet Mathieu, Clément Jean-Pierre, Plante Pier-Luc, Fliss Ismail, Bazinet Laurent
Department of Food Science, Université Laval, Québec G1V 0A6, Canada; Laboratoire de Transformation Alimentaire et Procédés ÉlectroMembranaires (LTAPEM, Laboratory of Food Processing and ElectroMembrane Processes), Université Laval, Québec G1V 0A6, Canada; Institute of Nutrition and Functional Foods (INAF), Université Laval, Québec G1V 0A6, Canada.
Department of Computer Science and Software Engineering, Université Laval, Québec G1V 0A6, Canada.
Food Res Int. 2025 Jan;200:115417. doi: 10.1016/j.foodres.2024.115417. Epub 2024 Nov 28.
Industrial wastewaters are significant global concerns due to their environmental impact. Yet, protein-rich wastewaters can be valorized by enzymatic hydrolysis to release bioactive peptides. However, achieving selective molecular differentiation and eventually enhancing peptide bioactivities require costly cascades of membranes. In this study, a complex porcine cruor hydrolysate, containing 150 well-characterized peptides and demonstrating only an antifungal activity, was used as a model solution to evaluate the impact of current modes (continuous electrical current (CC), pulsed electric field (PEF) and polarity reversal (PR)) and the combination of pulse/pause-reversal pulse duration (10 s/1 s and 1 s/1 s) during peptides separation by an electromembrane process. The data analysis was assisted by a machine learning (ML)-based peptidomic approach to identify which of the 45 physicochemical characteristics of the peptides explain migration, or lack thereof, during electrodialysis with filtration membrane, a generic electromembrane process. The results demonstrated, for the first time, that electric current conditions modulate the population of recovered peptides and their associated fraction bioactivities. ML models identified the main features correlated to peptide migration, allowing tentative explanations of the underlying peptide selective migration phenomena. For CC-PEF 10 s/1 s-PR 10 s/1 s, isoelectric point (pI) (importance of 63.1%) and molecular weight (MW) (17.7%) were most important. For PEF 1 s/1 s, pI (53.9%), MW (23%) and GRAVY score (6.2%) played major roles. Finally, for PR 1 s/1 s, MW (82.5%), GRAVY score (5.5%) and tyrosine content (1.1%) were the key features. In addition, CC, PEF 10 s/1 s and PR 10 s/1 s allowed the production of two reusable fractions, an antibacterial recovery fraction and a feed fraction retaining antifungal activity, which aligns with the concept of circular economy.
工业废水因其对环境的影响而成为全球关注的重大问题。然而，富含蛋白质的废水可以通过酶水解进行增值利用，以释放生物活性肽。然而，要实现选择性分子分化并最终提高肽的生物活性，需要成本高昂的多级膜分离过程。在本研究中，一种复杂的猪凝血酶水解产物被用作模型溶液，该水解产物含有150种特征明确的肽，且仅表现出抗真菌活性，用于评估当前模式（连续电流（CC）、脉冲电场（PEF）和极性反转（PR））以及在通过电膜过程分离肽期间脉冲/暂停-反转脉冲持续时间（10秒/1秒和1秒/1秒）的组合的影响。数据分析借助基于机器学习（ML）的肽组学方法，以确定肽的45种物理化学特征中的哪些特征可以解释在带有过滤膜的电渗析（一种通用的电膜过程）期间肽的迁移或不迁移情况。结果首次表明，电流条件会调节回收肽的种类及其相关馏分的生物活性。ML模型确定了与肽迁移相关的主要特征，从而对潜在的肽选择性迁移现象给出了初步解释。对于CC-PEF 10秒/1秒-PR 10秒/1秒，等电点（pI）（重要性为63.1%）和分子量（MW）（17.7%）最为重要。对于PEF 1秒/1秒，pI（53.9%）、MW（23%）和亲水性氨基酸平均性质（GRAVY）得分（6.2%）起主要作用。最后，对于PR 1秒/1秒，MW（82.5%）、GRAVY得分（5.5%）和酪氨酸含量（1.1%）是关键特征。此外，CC、PEF 10秒/1秒和PR 10秒/1秒能够产生两种可重复使用的馏分，一种抗菌回收馏分和一种保留抗真菌活性的饲料馏分，这与循环经济的理念相符。
