The intricate symbiotic relationship between lactic acid bacterial starters in the milk fermentation ecosystem

Abstract

Fermentation is one of the most effective methods of food preservation. Since ancient times, food has been fermented using lactic acid bacteria (LAB). Fermented milk is a very intricate fermentation ecosystem, and the microbial metabolism of fermented milk largely determines its metabolic properties. The two most frequently used dairy starter strains are Streptococcus thermophilus (S. thermophilus) and Lactobacillus delbrueckii subsp. bulgaricus (L. bulgaricus). To enhance both the culture growth rate and the flavor and quality of the fermented milk, it has long been customary to combine S. thermophilus and L. bulgaricus in milk fermentation due to their mutually beneficial and symbiotic relationship. On the one hand, the symbiotic relationship is reflected by the nutrient co-dependence of the two microbes at the metabolic level. On the other hand, more complex interaction mechanisms, such as quorum sensing between cells, are involved. This review summarizes the application of LAB in fermented dairy products and discusses the symbiotic mechanisms and interactions of milk LAB starter strains from the perspective of nutrient supply and intra- and interspecific quorum sensing. This review provides updated information and knowledge on microbial interactions in a fermented milk ecosystem.

HIGHLIGHTS

The symbiotic relationship between Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus is reviewed.

Their nutrient co-dependence is discussed.

The role of quorum sensing in their interaction is discussed for the first time.

This review is of interest to colleagues interested in exploiting LAB starter cultures.

Keywords:

• Fermented milk
• lactic acid bacteria
• symbiosis
• nutrient co-dependence
• quorum sensing
• Streptococcus thermophilus
• Lactobacillus delbrueckii subsp. bulgaricus

1. Introduction

Fermented foods are an important part of long-standing traditional food cultures around the world. (Wang et al. 2023). Fermented foods are generally defined as foods produced through controlled microbial growth and enzymatic conversion of major and minor food components (Marco et al. 2017). They provide and preserve nutrients and sensory qualities (flavors, aromas, and textures), enriching the human diet. Fermentation improves the sensory properties, shelf life, and nutritional value of foods while reducing toxic and anti-nutritional food factors. Humans, microbial fermentation, and traditional fermented foods have always had a close relationship (Kabak and Dobson 2011; Tamang et al. 2020; Xu et al. 2021; Yang, Fan, and Xu 2020). More importantly, fermented foods promote human health primarily through the fermentative activities of the associated food microbes (Marco et al. 2021). Consumers are becoming more health conscious and aware of the benefits of fermented dairy products, which is driving demand for fermented dairy products (Sakandar and Zhang 2021). Major known health-promoting mechanisms of fermented foods, including improving the nutritional value of raw food materials, synthesizing active compounds, modulating the gut microbiota, and boosting host immunity, are summarized in Figure 1.

Figure 1. Beneficial effects of fermented foods on human health promotion. The beneficial effects of fermented foods on human health are observed when food ingredients, fermentation products, and live microorganisms are consumed and enter the gastrointestinal tract. These components, along with the resident gut microflora, are transformed into bioactive substances, such as peptides, bacteriocins, amino acids, conjugated linoleic acids, short-chain fatty acids (SCFAs), and other organic acids. Furthermore, microorganisms and their byproducts related to fermentation have the potential to interact with the natural gut microflora, epithelial cells in the intestines, and the immune system of the host.

Fermented milk products are widely produced and consumed across the globe. For thousands of years, milk fermentation has been one of the oldest and most practical techniques utilized to extend the shelf-life of milk (Sakandar and Zhang 2022). Back-slopping is a traditional production method of naturally fermented foods. In this method, some material from a previous fermentation is inoculated into the fresh substrate, preserving the natural microbiota that imparts the properties to the new batch of the fermented product (Zannini et al. 2016). Some traditional starter cultures have started to be exploited and used in the large-scale industrial production of fermented foods in the last century (Macori and Cotter 2018).

A starter culture comprises living microorganisms intentionally used to initiate fermentation to modify the chemical composition and sensory qualities of the substrate, leading to more standardized products. Such cultures include one or more species of lactic acid bacteria (LAB) fermenting together as a co-culture or mixed culture (Hill et al. 2017; Medina-Pradas et al. 2017). Inoculating microbial starter cultures during fermentation initiates biochemical reactions in the raw materials, thereby optimizing the traditional fermentation and production process. Lactic acid bacteria are phylogenetically located in the Clostridia branch of Gram-positive bacteria. They are non-sporing cocci, coccobacilli or rods, and aero-tolerant anaerobes, with a molar DNA base composition of less than 50% G ± C (George et al. 2018; Pasolli et al. 2020). They are among the most widely studied microorganisms worldwide. Given the important role that LABs in different biotechnological processes, it is not surprising that they have received much attention from the scientific community for decades (De Filippis, Pasolli, and Ercolini 2020).

Lactic acid bacteria are the main starter cultures used for producing fermented dairy products and have great application market potential and value. Fermented milk is an intricate fermentation ecosystem that regulates microbial fermentative metabolism and nutrient interactions (Gesudu et al. 2016; Yu et al. 2018). In this ecosystem, microorganisms interact directly through physical contact, quorum sensing (QS), symbiosis, parasitism, predation, and inhibition, while extracellular metabolites mediate indirect interactions through commensalism, neutralism, amensalism, mutualism, and competition (Wang et al. 2023). Streptococcus thermophilus (S. thermophilus) and Lactobacillus delbrueckii subsp. bulgaricus (L. bulgaricus) are the most commonly used starter cultures to manufacture fermented dairy products. The interaction between L. bulgaricus and S. thermophilus directly affects the quality of the fermented milk. Thousands of years ago, humans began to use fermentation to sour milk (McAuliffe 2018). For thousands of years, S. thermophilus and L. bulgaricus have been used in combination to ferment yogurt; and these two mutually beneficial symbiotic bacteria could support each other to improve the growth rate, taste, and quality of fermented milk (Iyer et al. 2010). This review discusses their mutualistic interaction and mechanism during milk fermentation from multiple perspectives.

2. Lactic acid bacteria and fermented milk

2.1. Definition and biochemical metabolic characteristics of LAB

Although humans have used LAB for producing naturally fermented dairy products and other fermented foods for thousands of years, there is still a limited understanding of LAB (Sakandar and Zhang 2022). The typical LAB are Gram-positive, nonspore-forming, nonmotile, catalase-negative, devoid of cytochromes, anaerobic to aerotolerant cocci or rods, which are acid-resistant and produce lactic acid as the main end product of carbohydrate fermentation (Teneva-Angelova et al. 2018). Lactic acid was first discovered in 1780 by C.W. Scheele in sour milk, which was later used in its industrial production (Castillo Martinez et al. 2013). Through the progress of lactic acid fermentation, microorganisms break down numerous nutrients, including proteins, lipids, and carbohydrates, into simpler and absorbable forms, such as peptides, free amino acids, free fatty acids, and lactic acid (Figure 2A; Sharma et al. 2023; Fernandez, Picard-Deland, and Anne 2019). Therefore, the available nutrients in fermented dairy products are higher than in non-fermented food due to the greater protein solubility, release of micronutrients, and restrictive amino acids (Sharma et al. 2020).

Figure 2. Milk fermentation by LAB through homolactic and heterolactic fermentation pathways. A. Degradation of main milk nutrient components by lactic acid bacteria during fermentation is a critical process in the production of various dairy products. These pathways result in the degradation of protein, fat, and carbohydrates in milk, leading to the development of unique flavors, textures, and other desired characteristics. B. Lactic acid bacteria utilize homolactic or heterolactic fermentation pathways (designated I and II, respectively) to break down lactose into lactic acid with and without other by-products.

In 2020, a taxonomic revision of the LAB categorized more than 300 species in seven genera and two families into a single family, the Lactobacillaceae, with 31 genera, comprising 23 new genera to encompass microorganisms previously grouped as Lactobacillus species (Qiao et al. 2022). A merger of Lactobacillaceae and Leuconostocaceae into Lactobacillaceae as one family has been recommended. Furthermore, the genus Lactobacillus has been reclassified into 25 genera (Zheng et al. 2020). Currently, 12 LAB genera are utilized in the production of fermented foods (Table 1). The primary LAB found in starter cultures for producing fermented dairy products primarily comprise Streptococcus, Lactobacillus, Bifidobacterium, Lactococcus lactis, and Leuconostoc species (Chen et al. 2017; Karaman and Ozcan 2021; Macori and Cotter 2018; Makarova et al. 2006).

Table 1. Lactic acid bacteria used in fermented food production.

Genus	Species or subspecies	Type of fermentative pathway
Lactobacillus	Lactobacillus delbrueckii subsp. bulgaricus	Homofermentative
	Lactobacillus delbrueckii subsp. lactis	Homofermentative
	Lactobacillus acidophilus	Homofermentative
	Lactobacillus helveticus	Homofermentative
	Lactobacillus johnsonii	Homofermentative
	Lactobacillus crispatus	Homofermentative
	Lactobacillus gasseri	Homofermentative
	Lactobacillus kefiranofaciens subsp. kefiranofaciens	Homofermentative
Lacticaseibacillus	Lacticaseibacillus casei	Facultatively heterofermentative
	Lacticaseibacillus paracasei	Homofermentative
	Lacticaseibacillus rhamnosus	Homofermentative
Lactiplantibacillus	Lactiplantibacillus plantarum	Facultatively heterofermentative
Limosilactobacillus	Limosilactobacillus fermentum	Obligately heterofermentative
	Limosilactobacillus reuteri	Obligately heterofermentative
Ligilactobacillus	Ligilactobacillus salivarius	Homofermentative
Latilactobacillus	Latilactobacillus curvatus	Homofermentative
	Latilactobacillus sakei	Homofermentative
Streptococcus	Streptococcus thermophilus	Homofermentative
Bifidobacterium	Bifidobacterium longum subsp. infantis	Bifidus pathway
	Bifidobacterium longum subsp. longum	Bifidus pathway
	Bifidobacterium breve	Bifidus pathway
	Bifidobacterium animalis subsp. animalis	Bifidus pathway
	Bifidobacterium animalis subsp. lactis	Bifidus pathway
	Bifidobacterium adolescentis	Bifidus pathway
	Bifidobacterium bifidum	Bifidus pathway
Leuconostoc	Leuconostoc mesenteroides subsp. Mesenteroides	Obligately heterofermentative
	Lactococcus lactis subsp. lactis	Homofermentative
Lactococcus	Lactococcus lactis subsp. cremoris	Homofermentative
	Lactococcus lactis subsp. lactis biovar diacetylactis	Homofermentative
Pediococcus	Pediococcus acidilactici	Facultatively heterofermentative
	Pediococcus pentosaceus	Facultatively heterofermentative
Weizmannia	Weizmannia coagulans	Homofermentative

Lactic acid bacteria are divided into homotypic and heterotypic fermentation according to the types of homofermentative end products (Salim-ur-Rehman, Paterson, and Piggott 2006). During LAB fermentation, sugars are converted to lactic acid (Bintsis 2018). Homolactic fermentation refers to the fermentation pathway that only produces two molecules of lactic acid by fermenting one molecule of glucose, using the glycolytic pathway (also known as the Embden-Meyerhof-Parnans pathway, EMP pathway; Figure 2B) (Gänzle 2015). Lactobacillus delbrueckii, S. thermophilus, and Lactococcus lactis are the most common edible LAB fermented by homolactic acid (Macori and Cotter 2018; Mokoena 2017). Heterolactic fermentation refers to the fermentation pathway in which glucose produces not only lactic acid but also ethanol, acetic acid, formic acid, and carbon dioxide after fermentation (Figure 2B) (Gänzle 2015). Leuconostoc mesenteroides and L. bulgaricus are representative LAB fermented by the heterolactic acid fermentation pathway (Macori and Cotter 2018). Other LAB are considered to be “facultatively” heterofermentative, meaning that they produce carbon dioxide and other products under certain conditions or from specific substrates (Cousin et al. 2012; Hutkins 2006).

2.2. Fermented milk as a carrier of LAB starter

Dairy and milk consumption are frequently included as important elements in a healthy and balanced diet (Pereira 2014). There is a long history of human production and consumption of fermented dairy products worldwide (Akdeniz and Akalın 2019). The historical development of fermented milk starters is closely related to that of fermented milk, passing through two stages: empirical starter and selective starter (Shrivastava and Ananthanarayan 2015; Tamime 2002). The practice of introducing old fermented yogurt, or “back-slopping,” into fresh milk for fermentation has been utilized by humans for centuries, unbeknownst to them, in fermenting various dairy foods (McAuliffe 2018; Steensels et al. 2019). Technological advancements have allowed for the isolation and cultivation of pure LAB to produce desired fermentation traits, such as acid, fragrance, and aroma from various raw materials.

The choice of starter cultures is one of the most important factors determining the success and outcome of fermentation. In the process, fermented milk is produced as a curd-like product due to the fermentative activities of LAB. The milk matrix acts as a carrier, safeguarding the vitality of LAB during gastrointestinal transit and enabling them to contribute to host health upon ingestion (De Filippis, Pasolli, and Ercolini 2020). Under appropriate conditions, dairy starter cultures can produce lactic acid, peptides, vitamins, extracellular polysaccharides, and other substances through metabolism. Lactic acid bacteria convert lactose in milk into lactic acid, and protein is degraded into polypeptides or amino acids. Such a process produces organic acids, alcohols, esters, ketones, and other flavor components, endowing fermented milk with specific flavors and rich nutrients. The fermentation process of naturally fermented milk is a dynamic ecological succession of the complex community of milk-associated microbiota that transforms the raw materials. This is mainly driven by the LAB community under specific fermentation conditions (Figure 3). Factors that control the fermentation process include the inoculum size, starter culture interaction with the raw milk and the endogenous microbiota, time of fermentation, oxygen, pH, temperature, and so on. By adjusting these parameters, the microecological succession in the milk fermentation process can be finely tuned to yield a variety of fermented milk products including wurimo, kefir, cheese, cultured butter, matzoon, cultured buttermilk, yogurt, milk curd, and others (Wang et al. 2023).

Figure 3. Driving factors of the dynamic succession of lactic acid bacteria in fermented dairy products and typical fermented dairy products. The inner ring shows various driving factors that control the dynamic succession of lactic acid bacteria in fermented dairy products. These driving factors include the initial microbial inoculum, its interaction with the endogenous microbiota, and environmental factors like raw materials, temperature, oxygen content, pH, and moisture. By tuning these variables in the fermentation process, a variety of fermented dairy products can be produced. Some of the representative fermented dairy products are shown in the outer ring.

2.3. Interaction and symbiosis of LAB starter in fermented milk

The interaction of commensal microorganisms in nature is indispensable to global carbon and nitrogen cycling and has been used for food fermentation and preservation for thousands of years (Sieuwerts et al. 2008). Interactions among microorganisms during fermentation can result in beneficial, neutral, or detrimental effects on the overall community. These interactions include mutualism (in which the microorganisms interact to their mutual benefit), favoritism symbiosis (in which one microorganism has negative effects on the other without inhibiting its growth), competition (in which the microorganisms compete for energy and nutrients), and parasitism (in which one microbe benefits at the expense of the other, and the suffering microbe provides nutrients to the benefiting microbe while its growth is inhibited). These interactions are important drivers for the assembly and evolution of microbial communities in the fermented food ecosystem. Moreover, they contribute to determining the safety and quality of the final product to some extent (Ivey, Massel, and Phister 2013; Sieuwerts et al. 2008; Smid and Lacroix 2013; Winters et al. 2019). However, classical theories alone are insufficient in explaining the intricate relationships and interdependence among microorganisms within complex networks (Frey-Klett et al. 2011; Ponomarova et al. 2017; Wang, Xu, et al. 2016; Wolfe and Dutton 2015). Most known natural and industrial food fermentation processes are driven by either simple or complex microbial communities. Microbial interaction is an important factor in maintaining the stable coexistence of microbial communities (Wang et al. 2023). Positive interactions between associated microbes are crucial in achieving improved fermentation processes and substrate conversion in fermented food production. The mechanisms behind positive interactions have garnered considerable research attention in recent decades (Canon et al. 2020). A thorough comprehension of the interaction mechanisms is necessary to enhance the utilization of multi-starters and natural fermentation in the food production process (Albergaria and Arneborg 2016).

Microbial communities are not simple aggregates of individual microorganisms, but rather complex network ecosystems formed by the interactions among microbial species, other organisms, and their environments through a variety of means, including resource competition, growth inhibition, cross-feeding, QS, and horizontal gene transfer (Settachaimongkon et al. 2014; Sieuwerts 2016). The interaction of microflora is driven by more than just exchanging electron donors and substance metabolism pathways. In fact, over 98% of the microorganisms are auxotrophic and require specific exogenous nutrients for growth, including amino acids, vitamins, and other nutrients (Zengler and Zaramela 2018). Differences in auxotrophs have a significant impact on the interaction between microorganisms in the microbial community, especially in terms of their ability to acquire and assimilate nutrients despite having similar energy requirements (Mu et al. 2022).

3. Effects of metabolites on the interaction between L. bulgaricus and S. thermophilus

3.1. Interdependence of LAB at the metabolic level

Food fermentation often depends on mutualistic interaction. The most notable example of this interaction is the symbiotic relationship between fermented milk and its leavening agents, specifically S. thermophilus and L. bulgaricus (Ge et al. 2023). The mutualism of the fermented milk system is mainly reflected in the respective metabolism of the two bacteria, utilizing metabolites released by each other to meet their own metabolic needs. Such interaction is also known as cross-feeding, which refers to the utilization and catabolism of metabolites (including carbon source, nitrogen source, amino acid, vitamin, nucleotide, electron donor, electron receptor, and other growth factors) secreted by one bacterium by another bacterium. Such metabolite transfer and consumption are widespread in microbial communities to improve the survival of microorganisms in a resource-poor environment, shaping the composition, structure, and evolution of microbial communities (Mu et al. 2022). The metabolic interaction between S. thermophilus and L. bulgaricus in milk fermentation is intricate and extensively researched (Sieuwerts 2016). The two bacteria can grow together through mutual exchange and utilization of metabolites and growth-promoting factors generated during their metabolism. These interactions depend on the release of chemical molecules into the environment, or direct contact between bacterial cells. Despite intense competition for resources, microbial communities exhibit metabolic interdependence among populations within the community, even across different habitats (Zelezniak et al. 2015). Since each strain has unique metabolic potential, the presence and intensity of interactions between L. bulgaricus and S. thermophilus depend on the specific combination of strains. In milk, these strains are symbiotic and mutually beneficial (Figure 4A). Typically, yogurt fermentation involves two exponential growth phases separated by a transition phase of slower growth (Figure 4B) (Sieuwerts 2016).

Figure 4. The symbiosis between Streptococcus (S.) thermophilus and Lactobacillus (L.) delbrueckii subsp. bulgaricus in milk fermentation. A. Metabolite-level interactions between and nutrient codependency of S. thermophilus (red) and L. bulgaricus (green) during the milk fermentation process

production or enzymatic activity;

positive effect of the component;

negative effect;

neutral or yet-to-be-confirmed effect. B. Schematic diagram showing the different bacterial growth phases during milk fermentation. The growth of S. thermophilus and L. bulgaricus are shown by the red and green curves, respectively. AA: amino acid; EPS: exopolysaccharide; LCFA: long-chain fatty acid.

3.2. Interaction between L. bulgaricus and S. thermophilus in the first growth phase

During the initial phase of co-culture growth, S. thermophilus grows more rapidly than L. bulgaricus due to its higher tolerance to neutral pH and higher efficiency in absorbing amino acids and trace elements compared with L. bulgaricus (Sieuwerts 2016). During this stage, S. thermophilus is capable of producing formic and folic acids while growing. In contrast, the growth of L. bulgaricus is limited in a milk environment due to its inability to independently produce formic and folic acids. However, when the two bacteria are mixed during fermentation, S. thermophilus provides formic acid to L. bulgaricus through the use of pyruvic acid in the EMP pathway under anoxic conditions (Lecomte et al. 2016). At the same time, S. thermophilus can provide L. bulgaricus with the necessary folic acid for growth and compensate for its lack of folic acid due to the incomplete folate biosynthesis pathway in L. bulgaricus (Sybesma et al. 2003; Van de Guchte et al. 2006; Wegkamp et al. 2004). Furthermore, formic acid and folic acid can serve as a precursor and a cofactor, respectively, enabling purine biosynthesis in L. bulgaricus and thereby providing purine substrates for S. thermophilus (Sybesma et al. 2003; Vinderola, Mocchiutti, and Reinheimer 2002). In addition, S. thermophilus hydrolyzes lactose in milk with β-galactosidase (LacZ) to produce glucose and galactose (Cui et al. 2016; Tarrah et al. 2018; Wu, Cheung, and Shah 2015). The glucose produced then undergoes glycolysis to generate lactic acid (Iskandar et al. 2019; Vaillancourt et al. 2008). Moreover, S. thermophilus consumes oxygen and produces carbon dioxide, facilitating the anaerobic growth of L. bulgaricus (Sasaki et al. 2014).

During the transition stage of milk fermentation, the growth of S. thermophilus is restricted due to the limited content of free amino acids, especially sulfur and branched-chain amino acids (Sieuwerts 2016). The nitrogen sources required for the growth of S. thermophilus and L. bulgaricus are mainly small molecular peptides and amino acids. To fulfill the requirements of cell growth, casein is first degraded into peptides and amino acids, serving as the primary nitrogen source (Liu et al. 2016). At this stage, the growth of L. bulgaricus and the expression of its extracellular protease gene prtB are initiated, and the extracellular protease prtB of L. bulgaricus can degrade casein to produce amino acids (such as methionine, histidine, and proline) and peptides (Garault et al. 2000; Zhao, Qu, and Cui 2015). These amino acids and peptides facilitate the second stage of exponential growth of S. thermophilus, which is also beneficial to the exponential growth of L. bulgaricus (Sieuwerts et al. 2010). However, the protease of S. thermophilus, prtS, has a weak ability to degrade casein and cannot directly obtain enough amino acids needed for growth from the milk environment. The hydrolysis of proteins by L. bulgaricus supplies an ample quantity of amino acids by up-regulating the amiABCDE operon, which is responsible for oligopeptide transport to support the growth of S. thermophilus. In addition, S. thermophilus can also synthesize three branched-chain amino acids that L. bulgaricus cannot synthesize (Ng et al. 2003). Consequently, the amino acid requirements of both bacteria are met under symbiotic conditions, facilitating fermentation.

3.3. Interaction between L. bulgaricus and S. thermophilus in the second growth phase

In the second exponential stage, genes participating in the production of long-chain fatty acids are usually up-regulated in S. thermophilus. Since L. bulgaricus lacks some biological pathways needed for the de novo synthesis of long-chain fatty acids (Partanen, Marttinen, and Alatossava 2001), S. thermophilus can provide long-chain fatty acids to L. bulgaricus (Sieuwerts et al. 2008). Glutathione (GSH) is an antioxidant. Streptococcus thermophilus can absorb exogenous GSH and synthesize GSH through its glutathione synthetase, gshF, which can protect the bacteria from stressors like oxygen and acids (Cui et al. 2021). At the same time, S. thermophilus exports GSH from the cell and provides it to the non-GSH synthesizing L. bulgaricus, enhancing its growth and ability to resist acid stress (Wang et al. 2015; Wang, Xu, et al. 2016). In addition, the biosynthesis of extracellular polysaccharides is upregulated due to the increased availability of casein-derived nitrogen (Zhang et al. 2014).

Urease is a multi-subunit urea amide hydrolase, accounting for 0.9% of the estimated core genome of S. thermophilus (Rasmussen et al. 2008). Among all LAB involved in dairy fermentation, only S. thermophilus exhibits urease activity (Mora et al. 2005; Yamauchi et al. 2019). The urease activity of S. thermophilus has a metabolic relationship with the biosynthesis pathways involved in the metabolism of aspartic acid, glutamine, arginine, and carbon dioxide (Arioli et al. 2007). During the process, the release of ammonia raises the pH levels both inside and outside the cells, facilitating bacterial acid production and growth. Urease is the key growth limiting factor whether S. thermophilus is cultured alone or with L. bulgaricus. Urease promotes lactose utilization by S. thermophilus and the production of lactic acid and acetaldehyde (Mora, Arioli, and Compagno 2013; Scala et al. 2019; Yu et al. 2020). At the same time, the urease activity of S. thermophilus facilitates the homolactic acid fermentation of L. bulgaricus (Arioli et al. 2017). The urease activity of S. thermophilus would affect the acidification of L. bulgaricus by regulating the carbon dioxide content, thereby promoting acidification, yogurt fermentation, and reducing production costs (Yamauchi et al. 2019).

The interaction between Lactobacillus bulgaricus and S. thermophilus in milk co-fermentation is well-studied at the metabolic level. Recent investigations into the role of probiotics in milk fermentation ecosystems have broadened our understanding of microbial interactions. Vinderola, Mocchiutti, and Reinheimer (2002) studied the growth kinetics of co-culturing 48 LAB strains, including S. thermophilus, L. bulgaricus, Lactobacillus acidophilus, Lacticaseibacillus paracasei, and Bifidobacterium. They identified four modes of growth interactions between the basic starters and probiotics: growth stimulation, growth inhibition, growth delay, and no impact (Vinderola, Mocchiutti, and Reinheimer 2002). Although the primary metabolic characteristics of fermented milk are formed by the fermentation of L. bulgaricus and S. thermophilus (Sieuwerts et al. 2010), the metabolic and protein hydrolysis activities of probiotics also affect the quality of fermented milk, including its sensory characteristics, nutritional value, and probiotic function (Li et al. 2019; Yu et al. 2018). A study compared the volatile and nonvolatile metabolomic characteristics of fermented milk produced singly by Bifidobacterium animalis subsp. lactis Probio-M8 or co-fermented together with Lacticaseibacillus paracasei Zhang. Analyses by LC-MS and GC-MS revealed that co-fermentation with Lacticaseibacillus paracasei Zhang had a greater impact on the volatile and nonvolatile milk metabolomes compared with single fermentation with Bifidobacterium animalis subsp. lactis Probio-M8, particularly enhancing the volatile metabolites of aroma. Thus, the combined use of these two strains in the fermentation exerted an additive effect on the metabolomics changes (Wang et al. 2021). Peng et al. (2022) studied the effects of the combined application of Lacticaseibacillus paracasei Zhang and Bifidobacterium animalis subsp. lactis V9 in milk fermentation on the bacterial growth and milk metabolomics changes during the fermentation and storage processes. The findings indicated that the co-cultivation of these two strains promoted the growth of Bifidobacterium animalis subsp. lactis V9, leading to an increase in both volatile and nonvolatile metabolites and short-chain fatty acids during storage (Peng et al. 2022).

In future research, it would be worthwhile to investigate the metabolomics effects of using different probiotic strains in the fermentation process, the interactions between traditional starter cultures and probiotics, and the contribution of probiotic implementation in improving the techno-functional properties of the product.

4. The role of QS in the interaction among LAB

4.1. Overview of quorum sensing and signal molecules

Microbes are ubiquitous on Earth, living in mixed or single-species communities (Weiland-Bräuer 2021). The quality of fermented food is highly dependent on the involved microorganisms their metabolic activities and interactions (Johansen and Jespersen 2017). In addition to metabolic-level interactions, researchers are investigating more intricate interaction mechanisms, such as cell-to-cell communication, in such ecosystems. Quorum sensing is a molecular signaling mechanism, through which microorganisms use cell communication to control population density and communicate within or between species through secreted signal molecules to coordinate gene expression and group behavior (Linciano et al. 2020; Mukherjee and Bassler 2019; Papenfort and Bassler 2016). It is a precise cell density-dependent microbial communication mechanism that acts via a QS molecule-mediated adaptive switch for gene transcription (Bai and Rai 2011; Sieuwerts et al. 2008; Smid and Lacroix 2013). In recent years, QS has also been found to involve in the regulation of many important physiological functions and features of LAB, such as bacterial biofilm formation (Liu et al. 2018), resistance to environmental pressures (Zhang 2022), autolysis (Sun 2010), and0 acid tolerance (Moslehi-Jenabian, Gori, and Jespersen 2009), biosyntheses of exopolysaccharides, antimicrobial peptides, and extracellular enzymes (Pezzulo et al. 2012; Wang 2022; Wasfi et al. 2018), adhesion to intestinal epidermal cells (Buck, Azcarate-Peril, and Klaenhammer 2009), and intestinal colonization (Medellin-Peña and Griffiths 2009).

During the QS process, bacteria secrete specific signal molecules (i.e., the language for microbial communication) into the environment while growing. As the microbial cell density increases, the concentration of these signaling molecules also rises. When the concentration of signal molecules reaches a threshold, these signal molecules combine with transcription factors to form a complex that triggers the expression of downstream genes, leading to the desired physiological effects and regulation (Waters and Bassler 2005). Quorum sensing is an important mechanism for interaction between bacterial cells and plays a critical regulatory role in bacterial population control and specific physiological functions (De Kievit and Iglewski 2000).

A recurring theme for QS has emerged after years of study. Briefly, LuxI homologs catalyze acylated homoserine lactone synthesis, while the luxR counterparts exhibit specificity for their cognate acylated homoserine lactone (Figure 5A (a)). When it reaches the threshold concentration, it specifically interacts with the luxR transcriptional regulator, leading to the expression of the luxICDABE operon and bioluminescence. In a complex ecosystem, bacterial communication can be viewed from a broad perspective of inter- or intramicrobial population communication, which includes three modes of QS: self-talk, where a species communicates within its own population; crosstalk, where different species communicate using common AI; and eavesdropping, where species that are unable to produce AI “listen” (Figure 5A (b)) (Coquant et al. 2021; Whiteley, Diggle, and Greenberg 2017).

Figure 5. Schematic diagrams showing the mechanisms of quorum sensing (QS) and autolysis of Lactobacillus (L.) bulgaricus. A. Illustration of QS bioluminescence mechanism. (a) a general QS mechanism – LuxI produces N-acyl-homoserine lactones (AHL, represented by yellow spheres). (b) QS signaling relies on the release of automatic inducers (AIs) into the environment for three modes of communication between bacteria, including self-talk, crosstalk, and eavesdropping. B. Common QS systems regulate three types of gene expression. (a) Quorum sensing in Gram-negative bacteria consists of LuxI and LuxR proteins. (b) The most common signal molecules in the QS system of Gram-positive bacteria are autoinducing peptides (AIPs). (c) LuxS/AI-2 is a quorum sensing system that can be used for interspecific information exchange. It is mediated by the signal molecule, furanosyl borate dister (AI-2), facilitating communications between Gram-negative and Gram-positive bacteria. C. Mechanism of autolysis in L. bulgaricus.

There are two categories of information exchange in LAB: intra- and interspecific. Notably, the AI-2 signaling molecule is different from all other QS signals in that it can achieve interspecific communication among bacteria, making it the “common language” for bacterial communication (Pereira, Thompson, and Xavier 2013). This common QS system regulates the expression of three types of genes, as shown in Figure 5B. Currently, QS system is mainly divided into three categories according to the bacterial signal molecules, including N-acyl-homoserine lactones in Gram-negative bacteria (Coquant, Grill, and Seksik 2020), autoinducing peptide in Gram-positive bacteria (Prazdnova et al. 2022), and furanosyl borate dister, AI-2, in both Gram-positive and Gram-negative bacteria (Pereira, Thompson, and Xavier 2013). Among them, the action mechanism of signal molecule N-acyl homoserine lactones existing in gram-negative bacteria is shown in Figure 5B (a). LuxI protein is an enzyme that regulates the synthesis of signal molecules, including N-acyl-homoserine lactones (AHLs). LuxR protein is a cytoplasmic signal molecule receptor and a DNA binding transcription activation element responsible for binding AHLs. Acyl-homoserine lactones are released into the environment through the cell membrane following synthesis. The level of AHLs increases with the number of bacteria. As AHL concentrations reach a certain threshold, these molecules will form AHL-LuxR complexes with the N-terminus of the receptor protein, LuxR, in the bacterial membrane, thus activating the expression specific downstream genes. There are also some other QS signaling molecules, such as quinolone in Pseudomonas (Dubern and Diggle 2008), the diffusion signaling factor (Majdura et al. 2023; Deng et al. 2011), γ-butyrolactone (Shi et al. 2022; Takano 2006), 2-amino acetophenone (Kesarwani et al. 2011), and bradyoxetin (Loh et al. 2002).

AI-2 is a furanone compound, which is a symmetrical double-pentacyclic-structure furanose boratediester, a by-product of the methionine cycle-methyl cycle (Zhang et al. 2020). AI-2 is a signal molecule that exists in both Gram-positive bacteria and Gram-negative bacteria. The specific mechanism of action is shown in Figure 5B (c). The synthesis of AI-2 is catalyzed by the luxS gene, which is essential for producing AI-2 for intercellular signal transduction (Tannock et al. 2005). The AI-2/luxS QS system plays a crucial role in the information exchange of many bacteria (Chen et al. 2002; Keller and Surette 2006). At present, the genomes of over 70 known bacterial species contain homologous sequences of luxS and the biosynthetic pathways of AI-2 (Schauder et al. 2001).

AI-1, a small autoinducing signaling peptide, is a crucial signaling molecule for intraspecific information exchange. It is derived from the two-component phosphokinase protein and is secreted after post-translational processing. However, unlike AHLs, it cannot pass through the cell membrane by diffusion, but relying on certain transport systems for membrane translocation, such as the ATP-binding-cassette. The growth of LAB constantly synthesizes and excretes AI-1, which is recognized by its specific receptors when accumulating to a certain level. Autoinducing peptide is recognized by the histidine kinase protein of the two-component system (TCS), and it binds to and interacts with the corresponding protein to induce the transcription of QS regulatory genes and other downstream genes, forming a dynamic communication system (Sturme et al. 2002; Waters and Bassler 2005). The molecular synthesis pathway of typical autoinducing peptides is shown in Figure 5B (b). Bacterial TCS are widely found signal transduction pathways that facilitate environmental adaptation; they are mainly composed of a detection membrane-reactive histidine protein kinase and a response regulator protein (Groisman 2016). These two proteins transmit signals through phosphorylation and dephosphorylation (Stock, Robinson, and Goudreau 2000).

4.2. The role of QS in the autolysis of LAB

Although numerous studies have been conducted on the QS of LAB, there is insufficient reporting on the effects of QS on the interactions between LAB during the milk fermentation process. Notably, autolysis is an interesting feature of LAB that can influence the outcome of the milk fermentation process and the microbial interactions in the milk fermentation ecosystem. Further research has found that QS systems play an active role in regulating bacterial interactions within co-cultures, as demonstrated by significant changes in the protein-level expression of luxS during the co-cultivation of different LAB strains (Di Cagno et al. 2009).

4.2.1. Autolysis of LAB affects the milk fermentation process

Autolysis of LAB is a frequent phenomenon, in which spontaneous cellular lysis occurs to maintain a favorable environment (Blaya, Barzideh, and LaPointe 2018; Lazzi et al. 2016). Autolytic enzymes causing bacterial autolysis hydrolyze the peptidoglycan network structure of cell walls under certain conditions. The expression of these enzymes leads to the dissolution and disappearance of bacterial murein, releasing intracellular metabolites into the surrounding environment (Gasson 1996). Autolysis is a vital and natural physiological process in cellular growth (Buist, Venema, and Kok 1998). During milk fermentation, autolysis is a mechanism for LAB to prevent excessive cell density in the fermentation medium. As cells lyse, intracellular enzymes, such as proteases and lipases, will be released into the fermentation medium, which will decompose the inherent components of protein and fat in milk, shaping the texture, flavor, physical and chemical properties, and shelf life of fermented dairy products (Cibik and Chapot-Chartier 2000; Ortakci et al. 2015). Accelerating the mechanical autolysis of LAB can improve the quality of fermented dairy products (Pang et al. 2017). Rapid autolysis of mixed strain starters endows dairy products with unique sensory properties, reducing the need to add flavors and spices. At the same time, rapid autolysis can effectively control the adverse effects of acidification on the quality of yogurt after shelf life, improving its acceptance by consumers (Sun 2010). Accelerating autolysis also reduces the detrimental effects of acidification on the fermented milk quality during shelf life, improving consumer acceptance rates. In cheese fermentation, rapid autolysis of LAB shortens the ripening period and reduces the production cost, and the post-autolysis release of proteases and lipases enhances the flavor of the cheese products (Pang et al. 2014). However, excessively rapid autolysis of starter bacteria can result in a decline in viable bacteria, insufficient acid production, poor curd, and excessive lactose residue (Pillidge et al. 2002). Despite this, autolysis of starter strains is generally considered a desirable property in terms of improving the quality of fermented dairy products. Therefore, the mechanism and regulation of autolysis in LAB have attracted much attention.

4.2.2. Quorum sensing regulates the autolysis of LAB

Bacteria typically sense and respond to environmental changes via a TCS (Cui et al. 2012). Genome sequencing and gene function analysis found that LAB genomes contain multiple TCS that are responsible for various functions, such as bacteriocin synthesis, acid resistance, bile tolerance, and bacterial adhesion (Borland et al. 2015; Marx, Meiers, and Brückner 2014; Guo and Sun 2017). The TCS in QS of LAB can regulate the autolysis in L. bulgaricus BAA-365 by secreting signal molecules to the surrounding environment during growth (Pang et al. 2017, 2020). When the concentration of the signal molecule in the surrounding environment reaches or exceeds the threshold, the cell wall-located histidine protein kinase, vicK, will initiate protein phosphorylation of vicR. Phosphorylated vicR binds to RNA polymerases and promotes the gene expression of lytM, ssaA, and atlA, accelerating cell wall rupture (Figure 5C). It was shown that the QS activity of L. bulgaricus regulates its autolysis (Pang et al. 2017). Another study analyzed the autolysis of L. bulgaricus LJJ at different growth stages under different treatment conditions and found that the bacteria were highly susceptible to autolysis during its logarithmic growth phase, and the susceptibility was further increased by elevating ambient temperature (4 ∼ 50 °C) and prolonging the incubation time. Additionally, the cellular lysozyme activity and thus the autolysis are influenced by environmental pH, temperature, incubation time, the concentrations of sodium dodecyl sulfate and Triton X-100, and the bacterial growth stage (Cui et al. 2013). Another studied bacterium is S. thermophilus strain DN-001065, which shows extensive autolysis in the later stage of growth in milk (Husson-Kao et al. 1999). The autolysis of S. thermophilus is triggered by unfavorable environmental conditions, such as lactose deprivation and the addition of salt or organic solvents (Husson-Kao et al. 2000)

4.2.3. Autolysis of LAB affects the microbial interaction in the milk fermentation ecosystem

Autolysis of Lactobacillus bulgaricus is related to the drop of environmental acidity (Chen et al. 2021). The study conducted by Xu (2017) investigated the role of two key genes of the acid adaptation-related TCS, hpk1 and rr1, in the bacterial symbiosis between Lactobacillus bulgaricus CH3 and s Streptococcus thermophilus strain by inactivating these two genes. It was found that the growth ability, acetaldehyde production ability, and viscosity production ability of the double gene-inactivated mutant were lower than that of the wild-type CH3 strain (Xu 2017). The findings highlight the active involvement of TCS in the interaction between L. bulgaricus and S. thermophilus. Moreover, it was found that L. bulgaricus and S. thermophilus strains with a higher rate of autolysis have shorter coagulation times and a better acid production capacity in fermentation experiments compared with those with a lower autolysis rate (Sun 2010). However, excessive autolysis of LAB would substantially decrease the extent of post-acidification of fermented milk, as high autolytic strains are prone to lyse in the later stages of fermentation. The damaged cell wall and membrane of autolytic cells reduce environmental acidity, curbing bacterial growth during fermentation. Additionally, the lower number of viable cells restricts lactose metabolism and acid production, resulting in a considerable reduction of post-acidification of fermented milk. Thus, a high autolytic rate of LAB is desirable for fermentation.

The autolysis of two most commonly used dairy starters, L. bulgaricus and S. thermophilus, will certainly affect the microbial interaction in the fermented milk ecosystem, and the autolysis characteristics of LAB starters are of practical importance in improving the quality of fermented dairy products. Lactic acid bacteria regulate their autolysis through the TCS in QS. Moderate autolysis will positively affect the fermentation characteristics of LAB starter strains, and the release of intracellular constituents during autolysis will benefit the microbial interaction in milk fermentation, although the complex mechanism requires further exploration. The autolysis of LAB is strain-specific and is influenced by different fermentation conditions. Therefore, selecting autolytic strains and optimizing fermentation conditions would enhance the properties of fermented dairy products.

4.3. The role of QS in the acid tolerance of LAB

4.3.1. Acid tolerance of LAB

Lactic acid bacteria need to survive different environmental conditions, such as different temperatures, acid levels, and salt concentrations (Rao et al. 2004). Acid stress is the most important factor affecting the growth and metabolism of LAB. The acid stress of LAB in the process of growth and metabolism mainly comes from two aspects, i.e., its acid production and the external environment from food processing and storage and the gastrointestinal tract after being ingested (Gu 2017). The acidic metabolites, such as lactic acid and acetic acid, produced by LAB in the process of milk fermentations have dual effects. On the one hand, they may inhibit the growth and reproduction of other microorganisms, resulting in better storage properties of fermented milk. Moreover, they may react with aldehydes and ketones to produce aromatic compounds, enhancing the flavor of fermented milk (Cogan et al. 2007). On the other hand, excessive acid accumulation during fermentation inhibits bacterial growth due to the elevated intracellular proton levels and the decreased intracellular pH, affecting the transmembrane δ pH, consuming the proton driving force in the bacteria, and deteriorating their proliferation ability and biological activity (Lorca and de Valdez 1999). Therefore, the acid adaptability and tolerance capacities of LAB, especially those used as fermentation starter strains, are crucial (Wang, Cui, and Qu 2018).

4.3.2. Quorum sensing regulates the acid tolerance of LAB

The QS system can regulate the acid tolerance of LAB and influence the interaction of starter strains in milk. The ability of LAB to survive in an acidic environment is very important for the stability and quality of fermented milk. Lactic acid bacteria use luxS/AI-2 to respond to environmental stimuli and regulate their growth and metabolism (Park et al. 2016). Under the conditions of acid shock, high temperature, and hunger stress, the AI-2 activity of LAB increases, while a decline in AI-2 activity is observed under conditions of low temperature and high osmotic pressure stress (Gu et al. 2018). At the same time, acid shock, low temperature, high osmotic pressure, and nutritional stress induce the over-expression of the luxS gene (Yeo et al. 2015). Under acid stress, bacteria can detect environmental changes and release AI-2 to regulate gene expression. This mechanism aids bacterial adaptation to the fluctuating environment (Gu 2017).

A previous study used quantitative polymerase chain reaction to show that the gene expression of a TCS (JN675228/JN675229) of the L. bulgaricus CH3 strain was up-regulated with the acid adaptation ability (Cui et al. 2012). To confirm the causal relationship, gene knockout mutants of JN675228 and JN675229 were constructed. The growth ability of the two mutants was substantially lower than the wild type, and the mutant lacking the JN675229 gene had reduced acid adaptability. These outcomes indicate that JN675228 and JN675229, particularly the latter, play a critical role in the acid adaptability of this strain. To investigate the signal transduction mechanism of JN675228/JN675229 in the acid adaptation in L. bulgaricus, differential protein expression among the two mutants and wild-type strain was screened using two-dimensional gel electrophoresis, transcriptional analysis, and time-of-flight mass spectrometry under natural fermentation and acid adaptation conditions (Wang, Cui, and Qu 2018). The results suggested that the TCS (comprising JN675228 and JN675229) regulates the acid adaptability of L. bulgaricus in many ways, including regulating proton pump-related proteins, classical stress shock proteins, carbohydrate metabolism, nucleotide biosynthesis, DNA repair transcription and translation, peptide transport and degradation, and cell wall biosynthesis.

Thermophilic bacteria can use lsrB and rbsB genes for AI-2 signal transduction, accomplishing biofilm formation and environmental stress adaptation (Kaur, Capalash, and Sharma 2018). The number of TCS in S. thermophilus varies between strains, usually ranging from six to eight (Hols et al. 2005). While S. thermophilus Sfi39 exhibits excellent fermentation proficiency, it possesses poor stress resilience (Lemoine et al. 1997). This bacterial strain is sensitive to acid and oxygen in the logarithmic growth period (Zotta et al. 2008), and its rr01 gene is involved in heat and acid tolerance (Zotta et al. 2009). A study isolated, identified, and analyzed the TCS of 22 strains of S. thermophilus from traditional yogurt samples in Inner Mongolia. The results showed that all 22 strains of S. thermophilus had eight pairs of greatly variable TCS, and their stress response corresponded to the unique TCS gene sequences. Therefore, it was speculated that the strain characteristics are linked to the differences in TCS (Hu et al. 2018). The LMD-9 genome contains six complete TCS, each with a sensor kinase and response regulator (RR; TCS 2, 4, 5, 6, 7, and 9), and two orphan RRs (RR01 and 08) with truncated sensor kinase. The RRs in the TCS were systematically inactivated, and subsequent experimentation using a fermented milk model revealed that among all the RRs tested, only RR05 was essential for the growth of S. thermophilus LMD-9 (Thevenard et al. 2011).

The AI-2 activity of Lactobacillus acidophilus NCFM (L. acidophilus NCFM) and Lacticaseibacillus rhamnosus GG (L. rhamnosus GG) significantly increased with pH reduction under acid stress (Moslehi-Jenabian, Gori, and Jespersen 2009). In addition, the transcriptional changes in the luxS gene of L. acidophilus NCFM and L. rhamnosus GG were consistent with the AI-2 activity, suggesting that the acid tolerance of this strain could be associated with the expression of luxS and other related genes. Moreover, the TCS mutant of L. acidophilus NCFM was more sensitive to acid than the wild-type, with an upregulation of expression of more than 80 genes, including luxS, under acid stress, indicating TCS and the related genes like luxS play a role in adapting to an acidic environment (Azcarate-Peril et al. 2005). The Limosilactobacillus fermentum 2-1 utilizes a similar acid stress coping mechanism (Gu 2017). In Lacticaseibacillus casei 12 A, it was found that the nicotinamide adenine dinucleotide dehydrogenase, the adenosine triphosphate binding protein (pstB), and the histidine protein kinase in the TCS can work together to increase the acid tolerance (Overbeck et al. 2017). This could have a significant impact on how bacteria survive in the gastrointestinal tract and their interactions with the local gut microbial community. During the logarithmic phase, the acid-producing ability of Limosilactobacillus fermentum 332 was found to be directly proportional to the activity of signal molecule AI-2 (Qiao et al. 2022). Furthermore, adding exogenous AI-2 significantly enhanced the activity of LAB and exopolysaccharide production (Wang 2022).

4.3.3. Acid tolerance of LAB affects the microbial interaction in the milk fermentation ecosystem

Quorum sensing systems, particularly the luxS/AI-2 system, play an important role in the interactions among LAB by aiding in environmental stress tolerance. AI-2 has been demonstrated to improve the acid resistance of LAB and promote the metabolism of S. thermophilus, thus enhancing its interaction with L. bulgaricus (Meng, Zhao, and Lu 2022).

Glutathione is an important signaling molecule in QS. It is a tripeptide compound capable of scavenging free radicals and enhancing cell stress resistance (Wang et al. 2019). Bacterial GSH levels are regulated by QS in response to changes in cell density and the cellular redox state (Zhou et al. 2019). It is interesting to note that GSH synthesized by S. thermophilus not only improves its acid resistance but also the vitality of L. bulgaricus, enhancing the growth of both starter cultures and the performance of milk fermentation (Wang, Xu, et al. 2016; Xue et al. 2023). Thus, the QS activity based on GSH is beneficial to the mutual interaction of the starter cultures during milk fermentation.

The acid tolerance of LAB is an adaptive response to environmental changes. These research findings suggest that LAB can increase their acid tolerance under acid stress by using a luxS/AI-2 signaling system or a TCS, enabling the bacteria to survive in an acidic environment. The QS of L. bulgaricus, S. thermophilus, or other LAB promotes their growth, metabolism, and interactions in fermented milk. The acid tolerance of LAB is undoubtedly regulated by QS, but its precise mechanism needs to be investigated in the future. Currently, the study of this aspect remains largely unexplored.

5. Concluding remarks and future perspectives

One of the earliest efficient food preservation techniques is fermentation, which has a long history and is ingrained in global food culture. People have used LAB to produce naturally fermented dairy products and other fermented foods for thousands of years. These bacteria are widely distributed in nature and have rich biodiversity, making them an essential biological resource with practical applications in human everyday life. Dairy fermentation ecosystems comprise complex microbial communities, affecting microbial proliferation, metabolic activity, and interactions to determine the quality of fermented dairy products. The most frequently employed starter species for producing fermented dairy products, such as yogurt, are S. thermophilus and L. bulgaricus. These microbes interact with each other in the fermentation process through symbiosis at the nutrient and metabolomics levels. In the process of microbial interaction, the role of QS through TCS, especially the luxS/AI-2 signal system, is crucial in governing microbial population and dynamics and regulating some growth characteristics of LAB, such as autolysis and acid stress tolerance. The regulation of autolysis and acid tolerance by QS in fermented food starter cultures is beneficial to their growth, metabolism, and interactions during fermentation, ultimately enhancing the functional and sensory quality of the final products. This article reviews the effects and mechanisms of nutrient assimilation and intra- and interspecific QS on the interaction of LAB during milk fermentation.

The food fermentation process constitutes a micro-ecosystem co-fermented by multiple microorganisms, encompassing intricate and dynamic microbial interactions. The stability, safety, flavor, and nutritional value of the final product are reliant on the microbial fermentation process and interactions. High-throughput DNA sequencing, RNA-seq, functional genomics, and metabolomics technologies have been widely used in exploring the microbiology behind a variety of traditional fermentation processes in recent years. These technologies can be used to track the changes in the taxonomic and functional microbiome. Furthermore, the fermented food industry has gradually evolved from a small, family-based workshop production mode to a highly integrated industrial factory production scale, in light of the rapid development and innovation of modern biotechnology. The ability to characterize and trace the microbiome asso­ciated with traditional fermented foods is made possible by implementing the omics tools. Bioinformatics, machine learning, and other cutting-edge techniques can predict microbial community evolution and metabolic outcomes, shedding light on the fermentation process. By bringing together biotechnology advancements, the nutritional value and safety of fermented foods can also be improved. In the future, a deeper understanding of the intricate microbial interactive mechanism can be explored with the aid of more sophisticated technological tools.

Author contributions

Shujuan Yang Writing – Original Draft. Mei Bai Conceptualization. Lai-Yu Kwok: Writing – Reviewing and Editing. Zhi Zhong Writing – Reviewing and Editing. Zhihong Sun Conceptualization, Project administration.

Competing interests

The authors have no conflicts of interest to declare. All authors have read and agreed to the published version of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

We are profoundly thankful to the Research Fund for the National Key R&D Program of China (2022YFD2100700), the National Natural Science·Foundation of China (32325040 and 32101918), the Inner Mongolia Science & Technology Planning Project (2022YFSJ0017), and the Earmarked Fund for China Agricultural Research System (CARS-36).