Development of a standard food authenticity testing workflow for honey using non-targeted LC/MS analysis

食品欺诈(包括替代、掺假或贴错标签--例如原产地)对我们的食品系统构成威胁，给正宗食品的消费者和生产者造成经济损失，并可能造成食品安全后果。作为最古老的甜味剂之一，蜂蜜一直受到欺诈和掺假的影响。目前需要更新工具来测试蜂蜜的真实性，特别是在欺诈性做法迅速演变的情况下。仅针对有限类型掺假的传统方法的替代方法是使用基于高分辨率质谱学的非靶标分析(NTA)对食品进行更全面的多参数筛选。这种方法的新颖性依赖于对每个样本的大量化学指纹的收集和分析。NTA现在被认为是下一阶段的食品监测工具。非靶向指纹分析在食品真实性方面很有希望，但结果并不总是可复制的，需要标准化。在这个项目中，来自麦吉尔大学、拉瓦尔大学、安捷伦技术公司、加拿大卫生部、加拿大食品检验局和沃贡实验室服务公司的研究人员，结合他们的专业知识，协调用于食品真实性的非目标指纹工作流程，并以蜂蜜为初始案例，开发可有效用于市场上实际销售的食品的可靠的非目标分析标准协议。更具体地说，该项目旨在：(I)了解热处理和过滤对蜂蜜化学指纹的影响；(Ii)为450个蜂蜜样本建立指纹图谱；(Iii)系统地比较和排序各种数据处理、数据分析工具，用于蜂蜜样本的化学计量学分类；(Iv)制定快速、可靠的半常规端到端工作流程，以跟踪蜂蜜的质量属性(主要是花卉来源)，并通过中试实验室间测试验证标准协议的性能。

Food fraud (including substitution, adulteration or mislabeling - e.g. origin) is a threat to our food system, resulting into economic losses for consumers and producers of authentic food, with potential food safety consequences. As one of the oldest sweeteners, honey has always been subject to fraud and adulteration. Updated tools are currently needed to test the authenticity of honey, especially since fraudulent practices evolve quickly. An alternative to conventional methods targeting only limited types of adulteration is a more comprehensive multiparameter screening of food products, using high resolution mass spectrometry based non-targeted analysis (NTA). The novelty of the approach relies on the collection and the analysis of a large chemical fingerprint for each sample. NTA now regarded as the next stage of surveillance tools for food. Non-targeted fingerprinting is promising for food authenticity but results are not always replicable and requires standardization. In this project, researchers from McGill University, Laval University, Agilent Technologies, Health Canada, the Canadian Food Inspection Agency and Vogon Laboratory Services, bring together their expertises to harmonize non-targeted fingerprinting workflows used for food authenticity, and develop robust standard protocols for non-targeted analysis which can be used effectively on actual food products sold in the market, using honey as an initial case study. More specifically, this project intends to (i) understand the impact of thermal processing and filtration on the chemical fingerprint of honey; (ii) Establish the fingerprints for >450 honey samples; (iii) Systematically compare & rank a wide range of data processing, data analysis tools for the chemometric based classification of honey samples, (iv) Develop a rapid, robust semi-routine end-to-end workflow to track quality attributes of honey (primarily floral origin) and validate the performance of the standard protocol through a pilot interlaboratory test.