Microbial Dynamics and Phage Composition Reveal Key Transitions Driving Product Stability in Natural Vinegar Fermentation

Abstract

This study employed shotgun metagenomics to investigate microbial dynamics, phage-bacteria interactions, and functional genes throughout a three-month apple vinegar fermentation process. A total of 5621 microbial species were identified, revealing three distinct phases: (i) Enterobacteria and non-Saccharomyces species dominated the initial substrate; (ii) S. cerevisiae and Leuconostoc pseudomesenteroides prevailed in the intermediate phase; and (iii) acetic acid bacteria (Acetobacter ghanesis and Gluconobacter spp.), alongside non-Saccharomyces species (Pichia kudriavzevii and Malassezia restricta), dominated the final stages. Bacteriophage analysis revealed the presence of phages targeting spoilage bacteria, such as Pseudomonas and Erwinia, suggesting a role in regulating microbial stability and enhancing fermentation control. Functional metagenomic analysis highlighted key pathways associated with microbial growth and metabolite production, including carbohydrate and amino acid metabolism, energy production, and glycan biosynthesis. Enzymes involved in stress adaptation and secondary metabolism, including oxidative phosphorylation and phenolic compound synthesis, demonstrated microbial resilience and their potential role in shaping the product’s sensory and functional properties. Moreover, Enterobacteriaceae species were associated with pectin degradation during the early stages, aiding substrate breakdown. These findings are crucial for microbial and phage management in fermentation technology, offering valuable insights for innovation in the vinegar industry.

1. Introduction

Vinegar fermentation is a complex bioprocess driven by a dynamic core microbiome, primarily composed of yeasts, acetic acid bacteria (AAB), and lactic acid bacteria (LAB) [1]. However, the presence of less-characterized microorganisms and their enzymatic functions also plays a crucial role [2]. These microbial transitions not only influence the final acidity and flavor profile but also affect the nutritional content and health benefits of the vinegar [3]. Furthermore, the microbial diversity during fermentation is closely linked to the accumulation of secondary metabolites and aroma compounds, which enhance the overall quality of the vinegar [4]. Understanding how microbial succession impacts these processes can lead to the development of targeted strategies to optimize fermentation conditions and create vinegar with specific flavor profiles and health-promoting properties [5].

Advances in metagenomics have significantly expanded our understanding of the microbial diversity that drives fermentation processes [6]. Among these tools, shotgun metagenomics stands out as a high-throughput sequencing approach that enables the simultaneous identification of microbial species, functional gene annotation, and reconstruction of metabolic networks [7,8]. Unlike other metagenomic methods, shotgun sequencing achieves species-level resolution and reveals genes associated with a wide range of biological functions [7]. This approach has been instrumental in elucidating key genes involved in the metabolism of diverse bioactive compounds, including essential amino acids, various vitamins, and functional enzymes present in fermented foods [9]. As a result, it has facilitated the development of targeted strategies to simulate and optimize microbial consortia, contributing to the production of fermented products with improved quality, stability, and health-promoting properties [9]. The quality of traditional fermented foods is closely linked to their highly complex and dynamic microbiota. Shotgun metagenomics not only enables functional prediction but also provides the molecular basis for understanding microbial contributions to the sensory and nutritional attributes of these products [10].

In vinegar research, shotgun metagenomics was initially employed to investigate the microbiota of cereal vinegar, revealing the metabolic networks responsible for its distinct flavor [7,8]. This analysis highlighted the microbial diversity and functional contributions of key species, especially yeast and AAB, in shaping the sensory characteristics of the product. Additionally, it facilitated the reconstruction of metabolic pathways and provided insights into microbial interactions and the adaptation mechanisms of the vinegar microbiome under specific production conditions [8]. Despite these advancements, critical knowledge gaps remain, particularly regarding the complex interactions between bacteria, viruses, and other microorganisms involved in vinegar fermentation. This is particularly evident in vinegars produced from diverse substrates and origins, where differences in raw materials and environmental conditions further complicate the understanding of microbial dynamics and their impact on fermentation processes.

Understanding bacteria-virus interactions in fermentation is critical due to their dual role in microbial ecosystems [9]. Controversial hypotheses suggest that bacteriophages could serve as natural biocontrol agents, controlling spoilage microbes and enhancing microbial stability [11]. However, their presence has also been linked to reduced fermentation efficiency by infecting key bacterial species. Studies have shown that cider phages directly influence lactic acid bacteria (LAB) metabolism, whereas phages associated with cheese starter cultures enhance community stability and help prevent collapse during the fermentation process [12]. Moreover, the underlying mechanisms remain unclear, indicating a significant knowledge gap that warrants further investigation into phage–bacteria interactions. Additionally, the functional roles of underexplored microorganisms and their enzymatic contributions to traditional apple vinegar fermentation, such as pectin degradation and aromatic compound production, have yet to be fully understood. This issue can be addressed through Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis, which offers comprehensive insights into metabolic pathways, enzymatic functions, and microbial interactions [13].

This study aimed to provide a comprehensive metagenomic analysis of apple vinegar fermentation, focusing on microbial diversity, bacteria-phage interactions, and functional pathways. By exploring these elements, the work uncovers critical insights into the fermentation process, offering new perspectives on microbial ecosystem management and paving the way for the optimization of fermentation conditions. These findings contribute to enhancing both product stability and sensory quality, advancing the broader understanding of this ancient fermentation practice.

2. Materials and Methods

2.1. Vinegar Production and Sampling

Apple vinegar fermentation was carried out following the protocol described by Maske et al. [14], with all assays performed in triplicate according to artisanal methods traditionally practiced in Brazilian domestic settings. Fresh organic apples were sourced from a local market in Curitiba, Paraná, Brazil, and selected based on uniformity, ripeness, and the absence of visible defects. The apples were processed at a ratio of 0.5 kg per liter of must. The must was manually prepared by crushing the apples and mixing them with sterile water. Refined white sugar was then added to adjust the initial soluble solids concentration to 20 °Brix, as measured using a digital refractometer (Hanna Instruments, Woonsocket, RI, USA) thereby ensuring suitable conditions for spontaneous alcoholic fermentation.

Fermentations were conducted in sterile 3 glass containers and incubated for 90 days, comprising two distinct phases. The first phase consisted of spontaneous alcoholic fermentation, which lasted six weeks. Following this period, an acetic inoculum (known as “mother vinegar”) was introduced to initiate the second phase—induced acetic fermentation [14]. Samples of 50 mL were collected on days 0, 30, 60, and 90, immediately frozen to preserve their integrity, and stored at −20 °C until microbiological analysis.

2.2. Total DNA Extraction

Genomic DNA from the microbial community was extracted from 10 mL of each sample collected during fermentation. The samples were centrifuged at 4000× g, 4 °C, for 10 min to pellet the microbial cells, which were then resuspended in 200 µL of phosphate-buffered saline (PBS). To facilitate cell wall disruption, the resuspended material was treated with specific enzymes: lysozyme (10 mg/mL) to break down bacterial cell walls and lyticase (2 mg/mL) to degrade yeast cell walls. Following enzyme treatment, the DNA was purified using the DNeasy PowerSoil kit (Qiagen, Hilden, Germany), according to the manufacturer’s instructions.

The extracted DNA was quantified using a Nanodrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and assessed for integrity via 0.8% (w/v) agarose gel electrophoresis.

2.3. Viral DNA Extraction

To recover the viral particles, 10 mL of apple cider vinegar was centrifuged at 500× g for 5 min to remove the particulate fraction from the sample. The supernatant was then filtered using a 0.22 μm syringe filter. RNase A and DNase I were added to the filtrate at a final concentration of 1 μg/mL and left at 37 °C for 30 min to degrade the genomic DNA and RNA of the microbial cells. To precipitate the viral particles, a cold solution of PEG 8000 was added to the sample to achieve a final concentration of 10% (w/v) and incubated at 4 °C overnight. The mixture was then centrifuged at 12,000× g at 4 °C for 1 h, and the supernatant was discarded. The precipitated material was then resuspended in 2 mL of ice-cold SM buffer (200 mM NaCl, 10 mM MgSO₄, 50 mM Tris pH 7.5) and stored at 4 °C until use.

The nucleic acids of the purified viruses were extracted using the Viral DNA/RNA Extraction kit (Loccus, Cotia, Brazil). After treating part of the extract with DNAse, the RNA was converted into cDNA using the High-Capacity RNA-cDNA kit (ThermoFisher) and quantified using a Nanodrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA).

2.4. Library Construction and Shotgun Sequencing

For microbial DNA, 1 ng of genomic DNA was used to construct the library with the Nextera XT DNA Library Preparation Kit (Illumina, San Diego, CA, USA), following the manufacturer’s protocol. For viral DNA, both the extracted DNA and complementary DNA (cDNA) were combined in equimolar proportions to construct the viral library, using the same preparation procedure as described for microbial DNA. The quality and size distribution of the libraries were assessed using GelBot (Loccus). Sequencing was performed on the Illumina NextSeq platform, employing a paired-end protocol with a read length of 300 base pairs (2 × 150 bp).

2.5. Bioinformatics and Data Analysis

For each sample, quality control of the sequences generated by sequencing was carried out using Trimmomatic v.0.39 software (Usadel Lab, RWTH Aachen University, Aachen, Germany) [15]. This procedure aims to remove low-quality regions and adapter sequences that could affect the quality of the genome assembly. After this stage, the reads from each sample were used to assemble the metagenome using the SPAdes v.3.15.4 algorithm (Center for Algorithmic Biotechnology, Institute of Translational Biomedicine, St. Petersburg State University, St. Petersburg, Russia) with the -meta parameter [16]. SPAdes v.3.15.4 is a robust and widely used algorithm for assembling metagenomes, as it can handle sequences of different sizes and complexities. To ensure the accuracy and quality of the metagenome annotation, the Prokka v.1.14.6 tool (Torsten Seemann, University of Melbourne, Melbourne, Australia) [17] was used. This tool is capable of automatically annotating genes and other characteristics, providing information on the genetic composition of the sample. ences [18]. However, only sequences related to fungi, bacteria, and bacteriophages were considered, as they are key drivers of microbial interactions and functional dynamics in vinegar fermentation.

2.6. KEGG Pathway Annotation

The translated gene sequences were aligned against the eggNOG protein database using BLASTP (WU-BLAST 2.0) https://blast.ncbi.nlm.nih.gov/Blast.cgi (accessed on 11 March 2025). A single gene could be assigned to multiple functional categories within eggNOG, as described by Xie et al. [19]. Annotation based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) was performed using the KAAS tool with default parameters in Single Best Hit https://www.genome.jp/tools/kaas/ (accessed on 8 April 2025) [19]. Genes associated with KEGG orthology were subsequently submitted to iPath https://pathways.embl.de/ipath3.cgi (accessed on 9 April 2025) for metabolic pathway reconstruction in each sample. The 60 h time point was selected for functional analysis, as it represented the peak of microbial activity during fermentation.

3. Results and Discussion

3.1. Dynamic Succession of Bacteria and Fungi

3.1.1. Shotgun Reads

A total of 1,860,561 reads were obtained from traditional apple vinegar samples. Shotgun metagenomic sequencing identified a diverse microbial community, encompassing 44 phyla, 221 orders, and 5621 species across bacterial and fungal domains. The proportion of unclassified taxa significantly increased as taxonomic classification advanced to deeper levels (Figure 1). For bacteria, at the phylum level, only 4.1% of taxa remained unclassified by Day 90. However, at the order and species levels, the proportion of unclassified taxa reached 13.2% and 72.7%, respectively. Similarly, for fungi, unclassified taxa at the phylum level accounted for 37.3% by Day 90, increasing to 39.58% at the order level and 53.5% at the species level. The observation of increasing proportions of unclassified taxa at finer taxonomic levels in metagenomic studies is a well-documented phenomenon. For instance, a study evaluating metagenomic classifiers on soil microbiomes found a substantial proportion of unclassified reads [20]. Similarly, taxonomic profiling of ancient metagenomic samples reported that a significant proportion of reads remained unclassified across various metagenomic profilers [21]. This outcome is expected, as the proportion of unclassified taxa typically increases when moving from broader to more specific taxonomic levels. The high proportion of unclassified taxa emphasizes the need for expanded genomic databases and advanced sequencing technologies to improve taxonomic resolution. Despite this limitation, the functional potential of unclassified taxa remains significant, as they may harbor unique metabolic pathways crucial for food fermentation processes. This highlights the importance of integrating functional analyses with taxonomic studies to better understand and optimize microbial roles in enhancing food flavor, texture, and preservation during fermentation.

3.1.2. Microbial Diversity

Across all fermentation stages (Day 0, Day 30, Day 60, Day 90), the bacterial and fungal communities were dominated by distinct phyla (Figure 1). In bacteria, Proteobacteria and Firmicutes were the most abundant, with Proteobacteria being dominant initially (58.62%) and gradually decreasing by Day 90 (46.41%), while Firmicutes increased significantly over time (from 14.65% at Day 0 to 72.73% at Day 90). For fungi, Ascomycota was consistently dominant, particularly at Day 30 (89.02%), accounting for most fungal taxa across all time points. These shifts influence metabolic pathways and sensory outcomes in vinegar production, with the roles of key microorganisms explored further in the following sections.

3.1.3. Order-Level Trends

At the bacterial order level, the initial stages of fermentation (Day 0) were dominated by Enterobacterales (38.6%) and Pseudomonadales (33.1%), which are taxa typically associated with nutrient-rich environments and rapid growth [22]. However, this composition changed markedly by Day 30, with Lactobacillales emerging as the dominant phylum (27.9%), likely due to its ability to thrive in the acidic conditions created during fermentation [23]. By Day 60, a new shift occurred, with Rhodospirillales becoming dominant (31.5%), reflecting the adaptive capacity of taxa such as Acetobacteraceae, a family well-suited for survival and activity in highly acidic and oxygenated environments [24,25]. By Day 90, the microbial community displayed significant diversity, with no single phylum dominating. Diverse orders, including Hyphomicrobiales, Lactobacillales, and Rhodospirillales, were present. This diversification likely resulted from the depletion of easily fermentable substrates and the accumulation of secondary metabolites [3], which may have inhibited dominant taxa and facilitated the coexistence of less competitive yet specialized microbes.

For fungi, the initial stage (Day 0) was dominated by Saccharomycetales (68.4%), which includes many yeast species that are adapted to high-sugar environments [26]. However, by Day 30, there was near-complete dominance of Saccharomyces cerevisiae within this phylum, comprising 89.02% of the total fungal population. This aligns with the rapid proliferation of fermentative yeasts during the initial stages of fermentation, when sugar availability is high [27]. As fermentation progressed, the fungal community exhibited high diversity, with major contributions from Hypocreales (Day 60), Eurotiales (Day 90), and other minor phyla, suggesting a stabilization phase in which specialized fungi occupy diverse ecological niches.

3.1.4. Species-Level Interactions

The succession of microbial species during natural fermentation for vinegar production revealed dynamic shifts in bacterial and fungal communities (Figure 1), strongly influenced by the evolving fermentation environment. At the bacterial level, the early dominance of Erwinia billingiae (14.65%) and Pseudomonas rhodesiae (4.76%) reflects the initial presence of environmental and opportunistic taxa. These species are often associated with fresh substrates and aerobic conditions but tend to diminish as fermentation progresses [28].

The first significant microbial shift occurred after 30 days, marked by the consistent dominance of fermentative microorganisms such as Saccharomyces cerevisiae (89%) and Leuconostoc pseudomesenteroides (21.2%). The intricate nature of LAB–yeast interactions is underscored by two key mechanisms: (i) yeast autolysis releases essential nutrients, including amino acids, polysaccharides, and riboflavin, which support bacterial growth, and (ii) the acidification of the fermentation medium by LAB creates a favorable environment for yeast proliferation [29]. These synergistic interactions have been shown to enhance the sensory characteristics of products like wine, sourdough, and yogurt [30]. However, detailed insights into these mechanisms during vinegar fermentation are still limited.

The rise of Leuc. pseudomesenteroides during days 30 and 60, peaking at 21.5% and 13.11%, respectively, highlights its critical role in the early and intermediate stages of fermentation. This species is known for its lactic acid production, which contributes to pH reduction and creates a more selective environment for acid-tolerant species [3]. The dominance of S. cerevisiae on Day 30 highlights its key role in ethanol fermentation and its competitiveness in sugar-rich environments. The high production of ethanol and lactic acid during this phase creates strong selective pressures favoring organisms capable of maintaining activity under extreme environmental conditions [31].

By Day 60, the emergence of acid-tolerant AAB becomes crucial for the oxidation of ethanol to acetic acid, which is a key step in vinegar production [32]. The use of a submerged culture (“mother vinegar”) to induce acetic fermentation after six weeks in this study may have influenced the growth dynamics of AAB. The dominant species was Acetobacter ghanensis, accounting for 13.07% of the reads; however, a broad diversity of AAB was reported, including Gluconobacter oxydans, G. albidus, G. thailandicus, and G. sphaericus. The high microbial diversity observed can lead to the accumulation of a broad spectrum of metabolites beyond acetic acid, playing a crucial role in shaping the complex aroma profile of natural vinegar. During this phase, ethanol levels decline as AAB actively oxidizes it into acetic acid, which progressively accumulates throughout the acetous fermentation [3]. A previous study conducted with same vinegar fermentation method [14] reported a peak acetic acid concentration of approximately 23.30 g/L, highlighting the efficient conversion of ethanol and the successful completion of the fermentation process. Exploring individual AAB species could facilitate the development of optimized microbial consortia, enhancing vinegar production by improving sensory attributes, enriching aromatic complexity, and tailoring functional properties to meet diverse consumer and industrial demands.

Interestingly, Pichia kudriavzevii emerged with a transient dominance (31.5%) following the alcoholic phase dominated by S. cerevisiae. The reduction in ethanol, driven by AAB activity, may have allowed this yeast to thrive. Known for its resilience, P. kudriavzevii can tolerate high levels of acetic acid, low pH, and other fermentation stressors [33], making it a competitive organism in this phase. Its presence has been linked to the production of desirable aromatic compounds that enhance the sensory profiles of fermented products [34]. Further investigations could focus on selecting this species as an adjunct in vinegar fermentation to enhance aromatic complexity and improve sensory attributes.

Malassezia restricta emerged strongly at 90 days, coinciding with reduced ethanol levels and the availability of complex substrates like lipids [14]. Traditionally associated with the skin microbiota [35], M. restricta is a lipophilic yeast capable of utilizing lipid-rich substrates, likely derived from cell debris or fermentation byproducts. Its growth indicates that the fermentation environment has shifted to favor microbes adapted to such conditions. These microbial shifts have direct implications for the quality of the final vinegar, as they may influence the development of unique aromatic compounds, contribute to the complexity of the sensory profile, and affect the overall stability and functionality of the product.

3.2. Bacteriophage Diversity

The virome analysis revealed over 25 bacteriophage sequences classified within the order Caudovirales (Figure 2), with the viral families Myoviridae, Podoviridae, and Siphoviridae being the most abundant. The Myoviridae family, recognizable by their long, contractile tails, included notable bacteriophages such as Pseudomonas phage PMBT3, Pseudomonas phage UJF_PDIM6, and Erwinia phage vB_Ems49. These phages, capable of infecting a diverse range of bacterial hosts, play a pivotal role in shaping microbial communities that are vital for vinegar production [36]. On the other hand, the Podoviridae family, defined by their short, non-contractile tails, includes phages such as Pseudomonas phage phiAH14a, Lactococcus phage Tuc2009, and Gordonia phage Puppar, which are known for their efficiency in rapidly lysing bacteria and significantly influencing microbial succession dynamics during the fermentation process [37]. In contrast, the Siphoviridae family, distinguished by their long and flexible tails, includes phages such as Erwinia phage vB_Ems58, Mycobacterium phage Noxifer, and Mycobacterium phage PurpleHaze. These phages often establish stable, host-specific interactions that help sustain microbial diversity within the fermentation environment. Additionally, several bacteriophages could not be assigned to these families and were categorized as “unclassified Caudoviricetes”. This group included less abundant phages from the genera Pagevirus and Lessievirus, along with others exhibiting unique or unidentified characteristics.

Predictive Role of Phages in Product Stability

Phages displayed a wide spectrum of host specificities, with the majority targeting a single bacterial host. The predominant phages identified were Pseudomonas phage PMBT3 (12% of total phage reads), Erwinia phage vB_Ems49 (10%), Erwinia phage vB_Ems58 (8%), and Lactococcus phage Tuc2009 (6%) (Figure 2). To the best of the available knowledge, this is the first study to identify Pseudomonas phage PMBT3 within the virome diversity of vinegar samples. This phage has garnered attention for its potential as a biocontrol agent against P. aeruginosa and its biofilms, which are of particular concern in the food industry [38,39]. Moreover, its application could be further explored in vinegar production, where controlling microbial contaminants is critical to ensuring product quality and safety. By targeting specific bacterial pathogens, PMBT3 may offer a sustainable and efficient approach to enhance the overall microbiological stability of vinegar fermentation processes. A diverse array of other lytic bacteriophages targeting Pseudomonas species was identified in this study. Among these, the lytic bacteriophage Pseudomonas UFJF_PfDIW6 (5%) has gained significant attention for its reported antimicrobial activity against various Pseudomonas strains under different pH levels and temperatures [40,41]. Additionally, the presence of phages such as 201ϕ2-1-(targeting both Pseudomonas chlororaphis and P. aeruginosa), Noxifer (P. aeruginosa), and Nickie (Pseudomonas syringae pv. avii) underscores the genus’ phage diversity and its potential for biocontrol and quality improvement [42,43,44,45].

The second most prevalent group of phages targeted the Erwinia species, representing 13% of the total reads. Among these, two temperate phages, Erwinia phage Vb_EhrS_59 (10%) and Erwinia phage Vb_EhrS_49 (3%), were identified (Figure 2). These phages share substantial nucleotide sequence identity, particularly in genes associated with head assembly, DNA packaging, and lysis [46]. Notably, both phages infect Erwinia horticola, the pathogen responsible for beech black bacteriosis, highlighting their potential as biocontrol agents. Moreover, their lack of significant similarity to previously characterized viruses within the Enterobacteriaceae family underscores their uniqueness and opens avenues for novel biotechnological applications.

Interestingly, no viruses infecting yeast and key AAB were detected in this study. However, the analysis identified phages targeting LAB, such as Lactococcus phage Tuc2009 (6%) and another Lactococcus phage (1%), which could threaten LAB persistence. Although Lactococcus does not play a central role in the final stages of vinegar production, it may contribute to the vinegar’s flavor profile. The presence of LAB phages in fermentation systems poses risks by disrupting bacterial growth and delaying fermentation, potentially leading to lower-quality products [11].

The investigation of the minor viral population revealed phages targeting pathogenic bacteria in apple cider vinegar, including Pahexavirus (3%), which infects Enterobacteriaceae, as well as Serratia phage BF (1%) and Klebsiella phage N1M2 (1%). Pahexavirus, in particular, has demonstrated anti-biofilm activity against Cutibacterium acnes—a bacterium known to contaminate traditional fermented foods through human handling [47,48]. This finding underscores the broader impact of phage populations in fermentation environments, not only in shaping microbial communities but also in offering potential applications for targeted biocontrol strategies.

3.3. Predictive Functional Features

The functional annotation of the metagenomic sequences revealed that 57% of the genes were assigned to Clusters of Orthologous Groups (COG; a database categorizing genes based on orthology and conserved functions across species), while the remaining 43% of open reading frames were mapped to the Kyoto Encyclopedia of Genes and Genomes (KEGG; a database linking genes to metabolic pathways and higher-level biological systems). The most abundant COG category, “general function prediction only”, highlighted genes with versatile and broadly defined roles [49]. Amino acid and carbohydrate transport were also significant, underscoring their roles in microbial growth, flavor compound production, and energy utilization during fermentation [50]. Additional COG categories, such as transcription, ribosomal biogenesis, energy production, and cell envelope biogenesis, emphasize the community’s capacity to adapt and maintain structural resilience in the harsh fermentation environment [51].

The KEGG metabolic pathway analysis, as shown in Figure 3, categorized genes into three hierarchical levels: Level 1, representing broad high-level functions such as metabolism, environmental information processing, and genetic information processing (100%); Level 2, encompassing specific functional categories such as carbohydrate metabolism, amino acid metabolism, and energy metabolism (5–30%); and Level 3, detailing individual sub-pathways that contribute to metabolic and functional diversity (1–4.9%).

At Level 1, metabolism emerged as the most dominant high-level function, followed by environmental information processing, genetic information processing, and cellular processes (Figure 3). A smaller fraction of sequences was classified as poorly characterized, reflecting limited functional information for certain genes Within the high-level metabolism category, carbohydrate metabolism represented the most abundant function, accounting for 13.54% of annotated genes, followed by amino acid metabolism at 11.89% (Figure 3). Similarly, a study by Liu et al. [52] found that genes related to carbohydrate and amino acid metabolism were highly expressed during soybean paste fermentation, emphasizing their roles in energy generation, microbial growth, and flavor compound production [52]. Energy metabolism accounted for 6.80%, while lipid metabolism and xenobiotics biodegradation and metabolism contributed 2.70% and 2.59%, respectively. Glycan biosynthesis and metabolism (3.55%) and the metabolism of cofactors and vitamins (6.52%) also made significant contributions. The significant roles of energy metabolism, lipid metabolism, and xenobiotics biodegradation further highlight the microbial community’s ability to adapt to the acidic and oxidative conditions characteristic of vinegar fermentation. Additionally, contributions from glycan biosynthesis and the metabolism of cofactors and vitamins reflect the metabolic versatility required to maintain functionality throughout the fermentation process.

At Level 2, carbohydrate metabolism pathways were the most abundant, including fructose and mannose metabolism (5.08%), pyruvate metabolism (4.78%), and glycolysis/gluconeogenesis (4.07%) (Figure 3). These pathways play a central role in converting sugars into intermediates for acetic acid production. Among amino acid pathways, alanine, aspartate, and glutamate metabolism (3.24%) and glycine, serine, and threonine metabolism (5.32%) were prevalent. These pathways contributed significantly to microbial growth and the synthesis of flavor and aroma compounds [53]. Other relevant pathways within the Level 2 classification (e.g., oxidative phosphorylation, sulfur metabolism, unsaturated fatty acid biosynthesis, and benzoate degradation) demonstrate significant roles in supporting energy efficiency, microbial adaptation, and the processing of complex organic substrates, thereby ensuring a clean and efficient fermentation matrix.

At Level 3, sub-pathways provided further detail (Figure 3). Within the energy metabolism sub-pathways, oxidative phosphorylation, methane metabolism, and sulfur metabolism were prominently enriched. Several amino acid metabolism pathways (e.g., alanine, aspartate, glutamate, glycine, serine, and threonine) were reported, while pathways with a relative abundance of less than 1%, including certain lipid biosynthesis and secondary metabolite pathways, contributed to the system’s metabolic versatility. These sub-pathway genes emphasize the importance of enriched amino acid and energy metabolism in supporting microbial growth, energy production, and adaptation, while less abundant genes highlight the metabolic versatility essential for natural vinegar fermentation [54].

3.4. Correlation Between Predominant Species and Predictive Functions

The activities of A. ghanensis, Leuc. pseudomesenteroides, and S. cerevisiae were analyzed across critical metabolic pathways related to vinegar fermentation quality and the final product (

Table 1. Predicted metabolic pathway contributions of Acetobacter ghanensis, Leuconostoc pseudomesenteroides, and Saccharomyces cerevisiae during natural apple vinegar fermentation.

Pathway/Metabolite *	Acetobacter ghanensis	Leuconostoc pseudomesenteroides	Saccharomyces cerevisiae
Amino Acids
Alanine	High **	Moderate	High
Valine	High	Moderate	None
Leucine	High	Moderate	None
Isoleucine	Moderate	Moderate	None
Glutamate	Moderate	None	High
Proline	None	None	High
Lysine	None	None	Moderate
Phenylalanine	None	High	None
Tyrosine	None	High	None
Tryptophan	None	High	None
Methionine	Moderate	None	Moderate
Cysteine	Moderate	None	Moderate
Pyruvate Metabolism
Acetyl-CoA Production	High	Moderate	High
Lactate Production	None	High	Moderate
Ethanol Production	None	None	High
Acetate Production	High	None	None
Oxaloacetate Formation	Moderate	None	Moderate
Formate Production	Moderate	None	None
Succinate Production	Moderate	None	None
Leucine/Isoleucine Synthesis	High	Moderate	None
Fructose and Mannose
Fructose Utilization	High	Moderate	High
Mannose Utilization	High	High	Moderate
D-Mannitol Production	High	High	None
Sorbitol Pathway	Moderate	High	None
Glycolysis Linkage	High	Moderate	High
L-Fucose Utilization	None	Moderate	None
L-Rhamnose Utilization	None	None	Moderate
Fructose-6P Conversion	High	High	High
Starch and Sucrose
Sucrose Utilization	High	High	High
Starch Degradation	Moderate	None	High
Maltose Utilization	None	None	High
Trehalose Metabolism	High	Moderate	High
Levan Biosynthesis	High	High	None
Inulin Utilization	Moderate	None	None
Cellobiose Utilization	High	None	None
Glucose-6P Conversion	High	High	High
Secondary Metabolism
Phenolic Compound Synthesis	High	Moderate	High
Terpenoid Biosynthesis	None	None	Moderate
Polyketide Biosynthesis	None	None	High
Alkaloid Synthesis	None	None	None
Non-ribosomal Peptide Biosynthesis	Moderate	None	High
Flavonoid Synthesis	High	Moderate	None
Isoprenoid Biosynthesis	Moderate	None	High
Steroid Biosynthesis	None	None	High
Pentose Phosphate Pathway
Glucose-6P Dehydrogenase	High	Moderate	High
Ribulose-5P Epimerase	Moderate	High	High
Transketolase Activity	High	High	High
D-Ribose Synthesis	High	Moderate	High
Sedoheptulose-7P Synthesis	High	High	High
Nucleotide Sugar Biosynthesis	Moderate	Moderate	High

* The functional annotation is based on KEGG orthology, and pathway predictions correspond to the 60 h time point, which represented the peak of microbial activity during fermentation. ** The classification of High (≥80%), Moderate (20–79%), and None (<20%) is based on the percentage of pathway enzymes identified for each microorganism.

). These pathways included amino acid biosynthesis, pyruvate metabolism, fructose and mannose metabolism, starch and sucrose metabolism, secondary metabolism, and the pentose phosphate pathway. The functional potential was classified based on the percentage of enzymes present within each metabolic pathway for the three major microbial groups, where “High” (≥80%) indicates a strong contribution, “Moderate” (10–79%) indicates a partial contribution, and “None” (<10%) indicates minimal or no involvement.

In amino acid biosynthesis, A. ghanensis demonstrated high activity in the biosynthesis of alanine, valine, and leucine, which are precursors to branched-chain esters known for imparting fruity and sweet notes to vinegar [53,54,55]. Leuc. pseudomesenteroides exhibited high activity in aromatic amino acid pathways, including phenylalanine, tyrosine, and tryptophan. These aromatic amino acids are vital precursors for volatile aromatic compounds that significantly influence vinegar’s sensory profile. For example, phenylalanine is converted into phenethyl alcohol, imparting floral notes, while tyrosine and tryptophan contribute to the formation of phenolic volatiles and indoles, which add complexity and depth to the aroma [56]. Although S. cerevisiae exhibited limited activity in aromatic amino acid pathways, it played a pivotal role in ethanol production. During the fermentation process, ethanol acts as a precursor for the formation of esters and other volatile compounds that enhance aroma production [56].

In starch and sucrose metabolism, all three microorganisms showed high activity in sucrose utilization, emphasizing their role in breaking down disaccharides into fermentable sugars. A. ghanensis demonstrated significant activity in trehalose metabolism, potentially enhancing microbial tolerance to stress during fermentation [57] Leuc. pseudomesenteroides, which exhibits high activity in the sorbitol pathway, may enhance the sweetness profile and act as a precursor for the synthesis of minor aromatic compounds [58].

S. cerevisiae complements this by facilitating glycolysis and mannose utilization (Table 1), maintaining metabolic balance and supporting overall fermentation efficiency. The integration of these pathways ensures a consistent conversion of sugars into flavor precursors. In secondary metabolism, A. ghanensis displays high activity in phenolic compound synthesis and flavonoid biosynthesis. These processes contribute to the antioxidant properties of vinegar and enhance its health benefits [59]. Leuc. pseudomesenteroides played a moderate role in the synthesis of phenolic compounds, with a potential function in stabilizing the final product.

The pentose phosphate pathway showed high activity across all three microorganisms. This activity supports the generation of precursors for secondary metabolites and aromatic compounds. For instance, ribulose-5P and sedoheptulose-7P are critical for synthesizing nucleotide sugars, which play a role in stabilizing the flavor and consistency of vinegar [60]. S. cerevisiae exhibited high activity in nucleotide sugar biosynthesis, further highlighting its role in enhancing the textural properties and flavor of the final product.

These findings highlight the complementary roles of AAB, LAB, and Saccharomyces yeast in vinegar fermentation. Their coordinated metabolic activities not only ensure efficient substrate conversion and the production of essential compounds like acetic acid but also contribute to the formation of a diverse array of aroma and flavor compounds. This synergy results in high-quality vinegar with a complex sensory profile that balances acidity, sweetness, and aromatic richness, while enhancing its nutritional and health-promoting properties.

3.5. Pyruvate Metabolism and Pectin Degradation

In this study, pyruvate metabolism and pectin degradation were selected as focal pathways due to their critical roles in fermentation and their impact on product quality (Figure 4 and Figure 5). Pyruvate metabolism links carbohydrate breakdown to key metabolites, while pectin degradation in apples releases sugars that drive microbial growth and efficiency [61].

In S. cerevisiae, the enzymes active in the pyruvate metabolism pathway underscore its pivotal role in ethanol production, primarily through pyruvate decarboxylase (K01568) activity (Figure 4). Ethanol serves as a critical substrate for acetic acid biosynthesis by acetic acid bacteria (AAB) and plays a significant role in shaping the aromatic profile of vinegar. Additionally, S. cerevisiae generates acetyl-CoA (K01895), a central metabolite involved in lipid metabolism and ester biosynthesis. Other enzymes, such as malate dehydrogenase (K00024) and phosphoenolpyruvate carboxykinase (K01610), contribute to oxaloacetate formation, supporting amino acid biosynthesis and enhancing flavor complexity in the final product.

A. ghanensis demonstrated high activity in acetate and acetyl-CoA production, both of which are critical for acetic acid synthesis, the defining component of vinegar’s sourness. Meanwhile, Leuc. pseudomesenteroides contributed significantly to lactate production, a key precursor for aroma compounds. For example, lactate can be converted into ethyl lactate, which enhances the vinegar’s fruity and buttery notes [62].

A. ghanensis showed active enzymes associated with acetic acid production through robust acetate and acetyl-CoA synthesis, primarily mediated by acetaldehyde dehydrogenase (ALDH, K00128) and acetyl-CoA synthetase (ACS, K01895) (Figure 4). These enzymes facilitate the conversion of acetaldehyde into acetyl-CoA, which is subsequently used in acetic acid synthesis, ensuring a high conversion rate of substrates and a clean fermentation process. Additionally, enzymes like cytochrome c oxidase (K02274) and NADH dehydrogenase (K00330) are critical in the electron transport chain. They drive oxidative phosphorylation, generating the energy required for these processes [63]. The high enzymatic activity associated with oxidative metabolism in this pathway allows A. ghanensis to efficiently utilize energy while recycling metabolic byproducts, ensuring stability in the vinegar’s acidity during the oxidative stages of fermentation. This activity directly contributes to its characteristic sourness and overall quality.

Leuc. pseudomesenteroides active enzymes contribute significantly to lactate production via lactate dehydrogenase (LDH, K00016) (Figure 4). Lactate serves as a precursor for compounds like ethyl lactate, which imparts fruity and buttery notes to the vinegar’s aroma. Additionally, Leuc. pseudomesenteroides is involved in aromatic amino acid metabolism, including pathways for phenylalanine (K00832) and tyrosine (K00830), utilizing enzymes like phenylalanine ammonia-lyase (PAL, K01595) and tyrosine aminotransferase (K00810). These pathways produce volatile compounds, such as phenethyl alcohol and 4-hydroxyphenylacetaldehyde, which enhance the sensory profile of the product [62]. This microorganism plays an essential role in balancing acidity with aromatic precursors.

Pectin, a complex polysaccharide found in apple cell walls, serves as an essential carbon source during fermentation. Composed of galacturonic acid and various sugar residues, its breakdown requires specific enzymes, with the efficiency of this process varying across microorganisms. Among the key players, S. cerevisiae demonstrated moderate activity, exhibiting polygalacturonase (EC 3.2.1.15) activity (Figure 5) but playing only a partial role in pectin degradation. In addition, Leuc. pseudomesenteroides and Acetobacter ghanensis showed no detectable pectinolytic activity, suggesting a limited role in pectin breakdown. Consequently, their ability to fully depolymerize pectin is limited, requiring the presence of additional microorganisms or enzymes produced by the fruit itself to achieve complete hydrolysis.

Given the high abundance of Enterobacteriaceae at the start of fermentation (Figure 1) and their known enzymatic activity [64,65,66], pentose and glucuronate interconversion were explored in Enterobacter sp. (Figure 5). This microorganism showed strong potential for pectin degradation due to the presence of genes encoding key enzymes, including pectin lyase (EC 4.2.2.10), which breaks pectin into unsaturated digalacturonides; polygalacturonase (EC 3.2.1.15), which hydrolyzes polygalacturonic acid; and pectinesterase (EC 3.1.1.11), which removes methoxy groups, thereby facilitating further hydrolysis [67]. In apple fermentation for vinegar production, this enzymatic activity promotes the breakdown of pectin into galacturonic acid and other fermentable monosaccharides, supporting glucuronate interconversion pathways and enhancing fermentation efficiency. However, the use of Enterobacteriaceae species requires caution due to risks such as undesirable metabolite production, competition with beneficial microorganisms, and safety concerns associated with certain strains [68]. Balancing enzymatic efficiency with proper control measures is crucial for maximizing benefits while ensuring fermentation quality and product safety.

4. Conclusions

This comprehensive study sheds light on the microbial dynamics, bacteriophage interactions, and functional pathways that drive apple vinegar fermentation. We observed a rich microbial diversity throughout the fermentation process, including key vinegar-related species such as A. ghanensis, Leuc. pseudomesenteroides, and S. cerevisiae. The early stages were dominated by Enterobacteriaceae, followed by consecutive phases enriched in yeasts, LAB, and AAB. This succession is key to fermentation success, shaping the vinegar’s acidity and flavor.

The virome analysis revealed the presence of bacteriophages, particularly those targeting spoilage bacteria like Pseudomonas and Erwinia, indicating a regulatory role in maintaining microbial stability. These findings underscore the potential of using bacteriophages to enhance fermentation control and product consistency.

From a functional perspective, the metagenomic analysis highlighted the critical roles of carbohydrate and amino acid metabolism, energy production, and glycan biosynthesis in supporting microbial growth and the production of key metabolites, including acetic acid, ethanol, and aromatic compounds. The significant contributions of A. ghanensis, Leuc. pseudomesenteroides, and S. cerevisiae to these metabolic pathways were evident in their specific enzymatic activities, which directly influence the vinegar’s sensory profile and health benefits.

In addition to providing new insights into microbial and viral interactions in vinegar fermentation, this study highlights the importance of optimizing microbial consortia and phage dynamics. These findings offer practical applications for improving the quality, stability, and nutritional value of vinegar, paving the way for future innovations in fermentation biotechnology.