The purpose of this article is to bring attention to the potential importance of extracellular vesicles (EVs) of microbial origin in the production of fermented foods. We anticipate that understanding the role of the EVs in these processes will contribute to the development of new tools in food biotechnology. The reasons that lead us to make such claim are given as: (i) the production of EVs is a widespread feature in all domains of life; (ii) a growing number of functions are being identified for EVs, the most prominent one being biological communication; (iii) the study of EVs has become important for the understanding of biological processes and as a diagnostic and therapeutic tool in the biomedical sciences; (iv) many fermented foods require the activity of microbial consortia involving multiple species; (v) an increasing number of studies are showing that interactions between microorganisms may be relevant in the development of fermentation processes and would therefore be of biotechnological interest (Curiel et al., 2017; Tronchoni et al., 2017; Conacher et al., 2019).

Extracellular vesicles are structures that have been described in all domains of life. They are surrounded by a lipid bilayer and show a broad range of sizes, from 20 to 500 nm. EVs can have diverse biogenesis routes (Colombo et al., 2014) and, most importantly, they can be carriers of molecules with high information capacity, such as proteins, and diverse types of RNAs, including notably miRNAs (Rodrigues and Casadevall, 2018). In humans, EVs have been shown to participate in a wide variety of biological processes, from immunomodulation to cancer development. They are also known to be involved in interactions between pathogenic microorganisms and their host animals or plants. See Bielska et al. (2019) for an overview on the most recent advances in the field of EVs. EVs are specifically involved in communication (intra‐ and interspecific) between living cells in many different contexts (Raposo and Stahl, 2019).

On the other side, with some remarkable exceptions such as beer and white bread, most of the fermented foods consumed around the world are the result of the activity of consortia that incorporate multiple microbial species (and strains), with yeasts and lactic acid bacteria as the main players in most cases. These consortia can be spontaneous, or more or less domesticated, as in the case of kefir or sourdough (Lhomme et al., 2015; Walsh et al., 2016). Both main microbial players and associated microbiota (including potential spoilage microorganisms) might get involved in microbial interactions during food fermentation. These interactions are classified according to different criteria by different authors. But, for the purpose of this article, we will distinguish only between targeted and untargeted interactions. Untargeted interactions are the consequence of cellular metabolic activities that would take place in the same way in pure cultures. The simplest example is competition for nutrients that are found in limited amounts. Other examples are, the toxic effect on other species of some metabolic end products, such as ethanol or CO₂, or the constitutive production of unspecific toxic substances, unless its production is enhanced in response to other species. In contrast, targeted interactions involve communication between cells. For such interactions to exist there should be a recognition mechanism; that is, a cell perceives the message (presence) of cells from other species (or the same one, as in the case of quorum sensing). This should be followed by integration of the information mediated by, for example, transcriptional regulation cascades. And they will result in a physiological response from the recipient cell. Interspecies cross‐feeding might be targeted or untargeted, depending on whether specific signals for the recognition of the other partner are involved. The targeted mechanisms of interaction are intensely studied in parasitism, pathogenesis and some examples of symbiosis, but are much less known in the case of weaker interactions between microorganisms sharing the same environments.

There are often methodological difficulties in distinguishing between targeted and untargeted interactions. In the area of food fermentation, untargeted mechanisms of interaction between different microorganisms have often been described, but there are few examples of targeted interactions (Conacher et al., 2019). The clearest examples of targeted interactions are those involving physical contact between cells, such as co‐aggregation, biofilm building, or contact‐dependent killing, that some authors describe as ‘direct’ interactions (Rossouw et al., 2015; Pérez‐Torrado et al., 2017). However, the molecules or structures that mediate many targeted interactions are often poorly or not characterized. We consider that, in order to fully understand the interactions between the different microorganisms involved in food fermentation, it is necessary to focus on both the targeted and untargeted interactions. And as for the targeted mechanisms of interaction, it is necessary to fill the gaps in knowledge about molecules and cascades mediating intra‐ and interspecific recognition.

In this context, it seems reasonable to expect that EVs will play an important role in the mechanisms of interaction between different microbial species, as well as between strains of the same species. To confirm this hypothesis, it would be necessary to be able to isolate EVs produced by fermentation microorganisms, characterize their properties and composition and demonstrate an impact of the isolated vesicles on the behaviour of cells of a recipient strain (Fig. 1). In fact, there are already some reports of EVs produced by yeasts and lactic acid bacteria of food interest (Dean et al., 2019; Mencher et al., 2020). The case of lactic acid bacteria has been more studied, especially in the context of probiotic activities (also a form of biological interaction), which on several occasions were associated with the production of EVs (Liu et al., 2020). This includes those released by lactobacilli isolated from kefir (Seo et al., 2018), a clear example of a stable consortium of microorganisms involved in a food fermentation process. Yeast EVs have been studied mainly in species pathogenic to humans (Gil‐Bona et al., 2015; Brown et al., 2015). S. cerevisiae has provided very interesting results for understanding the mechanisms of EV biogenesis in yeasts (Oliveira et al., 2010; Winters et al., 2020), although it has been relatively under exploited as a model system in this context. More recently, EVs have been isolated under winemaking‐like conditions from six wine yeast species, including S. cerevisiae (Mencher et al., 2020). In addition, there are a number of studies on yeast‐yeast interactions that address the requirement for physical contact of the cells, using double‐compartment culture devices separated by semi‐permeable membranes. Among them, we can find either works that confirm contact is necessary or works that conclude the opposite (Renault et al., 2013; Taillandier et al., 2014; Wang et al., 2015). These differences can be due to many factors, such as the species and strains of yeast used, or the lack of standardization in experimental conditions and cultivation devices. However, none of these designs take into account the possibility that EVs may be involved in the interaction. It cannot be ruled out that, in some instances, a requirement for transfer of EVs might have been mistaken as a requirement for cell‐to‐cell contact. Note that even when relatively large pore sizes are used, the diffusion of EVs may be restricted by the membranes being used. Similar experimental designs, but taking into account EVs, and setting appropriate controls, could also shed light on their role in some yeast interactions taking place during fermentation processes.

There is an undeniable academic interest in knowing the role of EVs in the interactions between fermentative microorganisms. For example, the analysis of EV’s macromolecular composition (miRNAs, proteins, lipids and other metabolites) and of the physiological responses of yeast cells to EVs might unveil novel intracellular signalling systems, or at least new functions for those already known. This should improve our understanding of the ecology of food fermentations, whether they are spontaneous, use established communities (kefir, yoghurt, sourdough), or use de novo assemblies, with selected microbial strains. All this knowledge could contribute to the design of new microbial consortia, optimized for specific applications. Beyond food fermentations, it is likely that these studies will also have an impact on other industrial processes, driven by mixed starter cultures (combinations of microbial strains and species). Certainly, the study of EVs is already becoming relevant in the understanding of the probiotic activity of various microorganisms (Seo et al., 2018).

We can venture that the analysis of the EVs isolated from fermentative processes will probably provide relevant information on their status and dynamics. The analogy would be the use of EVs isolated from biological fluids as diagnostic markers (Lianidou and Pantel, 2019). This information would be complementary to others such as metagenomic analysis in its different forms. To be able to interpret and exploit this type of data we still need gaining much more knowledge about the EVs from pure microbial cultures and, in a second step, their binary combinations (Fig. 1). Therefore, the short‐term expectation is to increase this body of knowledge by getting a growing number of research groups involved on the study of microbial interactions in fermented food processing and paying attention to the potential impact of EVs on their experimental results.

本文的目的是引起人们对微生物来源的细胞外囊泡 (EV) 在发酵食品生产中的潜在重要性的关注。我们预计，了解电动汽车在这些过程中的作用将有助于食品生物技术新工具的开发。我们做出这样的说法的原因如下：（i）电动汽车的生产是生活各个领域的普遍特征； (ii) 越来越多的电动汽车功能被确定，其中最突出的功能是生物通讯； (iii) EVs的研究对于理解生物过程以及作为生物医学的诊断和治疗工具变得非常重要； (iv) 许多发酵食品需要涉及多个物种的微生物群落的活性； (v) 越来越多的研究表明，微生物之间的相互作用可能与发酵过程的发展有关，因此具有生物技术意义（Curiel等人， 2017 年；Tronchoni等人， 2017 年；Conacher等人， 2019 ）。

细胞外囊泡是生命各个领域中已被描述的结构。它们被脂质双层包围，尺寸范围广泛，从 20 到 500 nm。 EV 可以具有多种生物发生途径（Colombo等， 2014 ），最重要的是，它们可以是具有高信息容量的分子的载体，例如蛋白质和各种类型的 RNA，尤其包括 miRNA（Rodrigues 和 Casadevall， 2018 ） 。在人类中，EV 已被证明参与多种生物过程，从免疫调节到癌症发展。还已知它们参与病原微生物与其宿主动物或植物之间的相互作用。参见 Bielska等人。 （ 2019 ）了解电动汽车领域的最新进展。 EV 特别参与许多不同环境下活细胞之间的通信（种内和种间）（Raposo 和 Stahl， 2019 ）。

另一方面，除了啤酒和白面包等一些显着的例外之外，世界各地消费的大多数发酵食品都是由多种微生物物种（和菌株）组成的联合体活动的结果，其中包括酵母菌和乳酸菌。大多数情况下都是主要参与者。这些联合体可以是自发的，也可以或多或少是驯化的，就像开菲尔或酸面团的情况一样（Lhomme等人， 2015 年；Walsh等人， 2016 年）。主要微生物参与者和相关微生物群（包括潜在的腐败微生物）都可能参与食品发酵过程中的微生物相互作用。不同作者根据不同标准对这些相互作用进行分类。但是，出于本文的目的，我们将仅区分有针对性和无针对性的交互。非目标相互作用是细胞代谢活动的结果，在纯培养物中也会以相同的方式发生。最简单的例子是对有限数量的营养素的竞争。其他例子是，某些代谢终产物（例如乙醇或CO ₂ ）对其他物种的毒性作用，或非特定有毒物质的组成型产生，除非其产量因响应其他物种而增强。相反，有针对性的相互作用涉及细胞之间的通信。要使这种相互作用存在，就应该有一个识别机制；也就是说，细胞感知来自其他物种（或同一物种，如群体感应）的细胞的信息（存在）。随后应该整合由转录调控级联等介导的信息。 它们会导致受体细胞产生生理反应。种间交叉喂养可能是有针对性的，也可能是无针对性的，具体取决于是否涉及识别其他伙伴的特定信号。在寄生、发病机制和一些共生实例中，人们对相互作用的目标机制进行了深入研究，但对于共享相同环境的微生物之间的相互作用较弱的情况却知之甚少。

区分有针对性和无针对性的相互作用通常存在方法上的困难。在食品发酵领域，经常描述不同微生物之间的非靶向相互作用机制，但很少有靶向相互作用的例子（Conacher et al ., 2019 ）。靶向相互作用最明显的例子是那些涉及细胞之间物理接触的相互作用，例如共聚集、生物膜形成或接触依赖性杀伤，一些作者将其描述为“直接”相互作用（Rossouw等人， 2015 ；Pérez-Torrado等人）等， 2017 ）。然而，介导许多靶向相互作用的分子或结构通常表征不佳或没有表征。我们认为，为了充分了解食品发酵中不同微生物之间的相互作用，有必要同时关注靶向和非靶向相互作用。至于相互作用的目标机制，有必要填补有关介导种内和种间识别的分子和级联的知识空白。

在这种背景下，似乎可以合理地预期 EV 将在不同微生物物种之间以及同一物种菌株之间的相互作用机制中发挥重要作用。为了证实这一假设，有必要能够分离发酵微生物产生的 EV，表征其特性和组成，并证明分离的囊泡对受体菌株细胞行为的影响（图 1）。事实上，已经有一些关于食品用酵母和乳酸菌产生 EV 的报道（Dean等， 2019 ；Mencher等， 2020 ）。乳酸菌的情况得到了更多的研究，特别是在益生菌活性（也是生物相互作用的一种形式）的背景下，这在多种情况下与 EV 的生产有关（Liu等人， 2020 ）。这包括从开菲尔中分离出的乳酸杆菌释放的物质（Seo et al ., 2018 ），这是参与食品发酵过程的稳定微生物群落的明显例子。酵母EV主要在对人类致病的物种中进行研究（Gil-Bona等人， 2015 ；Brown等人， 2015 ）。酿酒酵母为理解酵母中 EV 生物发生机制提供了非常有趣的结果（Oliveira等人， 2010 ；Winters等人， 2020 ），尽管在这方面它作为模型系统的开发相对较少。最近，在类似酿酒的条件下，从六种葡萄酒酵母中分离出了 EV，其中包括S. 酿酒酵母（Mencher等人， 2020 ）。此外，还有许多关于酵母与酵母相互作用的研究，这些研究使用由半透膜分隔的双室培养装置来满足细胞物理接触的要求。其中，我们可以找到要么证实接触必要性的著作，要么得出相反结论的著作（Renault等， 2013 ；Taillandier等， 2014 ；Wang等， 2015 ）。这些差异可能是由于许多因素造成的，例如所用酵母的种类和菌株，或者实验条件和培养设备缺乏标准化。然而，这些设计都没有考虑到电动汽车参与交互的可能性。不能排除在某些情况下，电动汽车转移的要求可能被误认为是细胞间接触的要求。请注意，即使使用相对较大的孔径，EV 的扩散也可能受到所使用的膜的限制。类似的实验设计，但考虑到 EV 并设置适当的控制，也可以阐明它们在发酵过程中发生的一些酵母相互作用中的作用。

不可否认，学术界对了解 EV 在发酵微生物之间相互作用中的作用有着浓厚的兴趣。例如，对 EV 的大分子组成（miRNA、蛋白质、脂质和其他代谢物）以及酵母细胞对 EV 的生理反应的分析可能会揭示新的细胞内信号系统，或者至少是已知的新功能。这应该可以提高我们对食品发酵生态学的理解，无论它们是自发的、使用已建立的群落（开菲尔、酸奶、酸面团）还是使用选定的微生物菌株从头组装。所有这些知识都可以有助于设计新的微生物群落，并针对特定应用进行优化。除了食品发酵之外，这些研究也可能对混合发酵剂（微生物菌株和物种的组合）驱动的其他工业过程产生影响。当然，EV 的研究已经与了解各种微生物的益生菌活性相关（Seo等人， 2018 ）。

我们可以大胆地说，对从发酵过程中分离出来的电动汽车的分析可能会提供有关其状态和动态的相关信息。类比是使用从生物体液中分离出的 EV 作为诊断标记（Lianidou 和 Pantel， 2019 ）。这些信息将与其他信息相补充，例如不同形式的宏基因组分析。为了能够解释和利用此类数据，我们仍然需要从纯微生物培养物中获得更多关于 EV 的知识，第二步是它们的二元组合（图 1）。因此，短期期望是通过让越来越多的研究小组参与发酵食品加工中微生物相互作用的研究并关注 EV 对实验结果的潜在影响来增加知识体系。