Skip to main content
NCBI home page
Food Chemistry: X logoLink to Food Chemistry: X
. 2026 Jan 23;34:103577. doi: 10.1016/j.fochx.2026.103577

Microbial and metabolic basis of p-cresol formation in fermented bamboo shoots (suansun): The central role of Levilactobacillus zymae KUST4216

Zi Ye a,b,c,d,e, Xinyu Wang a,c,d,e, Tao Wang a,c,d,e, Chuanqi Chu a,c,d,e, Zhijia Liu a,c,d,e, Leilei Yu e,f, Ming Du g, Junjie Yi a,c,d,e,⁎
PMCID: PMC12876629  PMID: 41658355

Abstract

p-Cresol plays a pivotal role in shaping the flavor quality of suansun. This study examined the effect of inoculating Levilactobacillus zymae KUST4216, a high p-cresol-producing strain, on the fermentation of suansun. Comparative analyses between uninoculated and inoculated fermentation revealed that both followed similar acidification patterns, but inoculated fermented suansun accumulated significantly more p-cresol (6.62 ± 4.15 mg/kg vs. 0.20 ± 0.14 mg/kg). Inoculation with L. zymae KUST4216 rapidly established microbial dominance, reshaped community succession, and promoted early sugar utilization. Genomic analysis showed enrichment in carbohydrate and amino acid metabolism and abundant CAZymes, supporting efficient precursor supply for aromatic compound biosynthesis. Transcriptomic profiling confirmed coordinated regulation that directed metabolic flux toward p-cresol synthesis. Overall, L. zymae KUST4216 drives p-cresol formation via the tyrosine pathway and stabilizes fermentation ecology, providing mechanistic insight into flavor generation and offering a strategy for enhancing the sensory quality of fermented bamboo shoots through targeted microbial inoculation.

Keywords: Suansun, Levilactobacillus zymae, Inoculated fermentation, p-Cresol, Microbial community

Graphical abstract

Unlabelled Image

Open in a new tab

Highlights

  • •

    Levilactobacillus zymae KUST4216 was identified as a dominant p-cresol producer.

  • •

    KUST4216 inoculation reshaped microbial succession and enhanced fermentation.

  • •

    L. zymae KUST4216 efficiently produced p-cresol via the tyrosine metabolic pathway.

  • •

    Genomic and transcriptomic revealed coordinated regulation of p-cresol biosynthesis.

  • •

    Findings reveal a LAB-driven mechanism underlying flavor formation in suansun.

1. Introduction

Suansun, a traditional fermented bamboo shoot product, is a distinctive fermented food of China (Guan et al., 2020; Guan et al., 2022). It is commonly used in dishes like river snail rice noodles (Zhang et al., 2023) and is appreciated for its unique flavor and crisp texture. The characteristic flavor of suansun arises from a variety of volatile organic compounds such as p-cresol, acetic acid, hexanal, 2-methoxy-4-methylphenol, nonanal, and 1-octene-3-ol. (Li et al., 2022; Qian et al., 2024; Qin, Xiong, Xiong, et al., 2024). Among these, p-cresol stands out as a key aromatic metabolite, exerting a decisive influence on the flavor profile of suansun. Its relatively strong odor characteristics and high odor activity value contribute to the distinctive leather-like aroma, often described as the characteristic “stinky” odor of suansun (Guan et al., 2022; Qin, Xiong, Kang, et al., 2024; Qin, Xiong, Xiong, et al., 2024). This compound is synthesized via the microbial decarboxylation and methylation of tyrosine or phenylalanine (Li et al., 2025; Tang et al., 2022), with its production tightly coupled to the fermentation process and the metabolic activities of specific microorganisms. Previous studies have demonstrated a positive correlation between p-cresol concentration and the presence of genera such as Lactococcus, Lactobacillus, Enterococcus, Clostridium, Ruminococcaceae, Weissella (Chen et al., 2022; Guan et al., 2022; Hu et al., 2023). Collectively, these findings highlight the critical role of microbial fermentation in modulating the flavor characteristics of suansun.

Natural fermentation is traditionally employed in the production of fermented vegetables, including suansun, and remains the predominant method in commercial production (Li et al., 2023; Wang et al., 2023). However, this method faces several challenges, such as inconsistent flavor, limited process control, and potential food safety risks (Lu et al., 2022). To tackle these issues, recent investigations have concentrated on isolating functional strains from naturally fermented suansun—particularly lactic acid bacteria endowed with potent acidification capacity, versatile flavor metabolic capabilities, and antibacterial activities, which are further applied in inoculated fermentation (Li et al., 2026; Liang et al., 2025). For instance, Xue et al. (2025) found that inoculating Lactobacillus plantarum DACN768 during suansun fermentation increased the levels of key flavor compounds, demonstrating the benefits of controlled inoculation. While inoculated fermentation has been widely adopted in the industrial production of fermented foods such as yogurt, beer, and bread (Clark et al., 2024; Dilmetz et al., 2024; Santos et al., 2025), its application in suansun production remains limited. Therefore, investigating the potential of inoculated fermentation in suansun manufacturing is of considerable scientific and industrial significance. Our previous study has systematically identified the core microbial taxa associated with p-cresol biosynthesis, deciphered the key metabolic characteristics involved in its production process, and preliminarily clarified the link between microbial community structure and p-cresol accumulation (Wang et al., 2026). This work highlighted the crucial role of microbial fermentation in regulating the flavor characteristics of suansun. However, the specific metabolic mechanisms of individual functional strains remain to be further elucidated.

In this study, a Levilactobacillus zymae strain KUST4216, with a strong capacity for p-cresol production, was obtained from naturally suansun in Yunnan Province. Despite its promising potential, the metabolic stability of L. zymae KUST4216 in complex fermentation systems, as well as its specific impact on flavor regulation, remain poorly understood. To address these knowledge gaps, the present research compared the physicochemical properties, volatile compound profiles, and microbial communities of uninoculated suansun and suansun fermented with L. zymae KUST4216 inoculation. The aim was to investigate the influence of L. zymae KUST4216 on the overall quality of suansun and its effect on the microbial ecosystem during fermentation. Additionally, through comprehensive whole-genome and transcriptome analyses, this study sought to identify the specific genes and metabolic pathways involved in the biosynthesis of p-cresol by L. zymae KUST4216. This work provides a theoretical foundation for future efforts to regulate p-cresol production at both the genetic and molecular levels, thereby contributing to the optimization of flavor profiles in fermented foods.

2. Materials and methods

2.1. Materials and chemicals

The fresh bamboo shoots utilized in this study were obtained from a local farmer's market in Yunnan Province, China, and subsequently stored at 4 °C prior to analysis. MS-grade p-cresol and hexyl 2-methylbutyrate were purchased from Aladdin Reagents Co., Ltd. (Shanghai, China). The n-alkane (C5-C25) were acquired from ANPEL Laboratory Technologies (Shanghai, China). HPLC-grade organic solvents (acetonitrile and methanol) were obtained from Sigma-Aldrich (St. Louis, MO, USA). The E.Z.N.A.® Soil DNA Kit for DNA extraction was procured from Omega Bio-tek (Norcross, GA, USA). L. zymae strain with high p-cresol production capacity, designated L. zymae KUST4216, was preserved in the Guangdong Microbial Culture Collection Center (GDMCC No: 650660).

2.2. Screening of high-yield p-cresol strains

Fresh bamboo shoots were homogenized in a juicer (Joyoung, JYZ-E3C, Hangzhou, Zhejiang, China) to prepare bamboo shoot juice medium. The juice was filtered through cheesecloth and centrifuged at 12,000 ×g for 10 min at room temperature (22–25 °C), followed by sterilization using a 0.22 μm polyester membrane. The sterilized juice was stored at 4 °C until use, with approximately 250 mL of juice obtained per 1 kg of fresh bamboo shoots. Strains (n = 24) isolated from naturally fermented suansun were first activated three times on MRS medium. Subsequently, they were fully resuspended in sterile physiological saline. The resulting bacterial suspension was inoculated into sterilized bamboo shoots at a concentration of 1.0 × 108 CFU/mL, with fermentation thereafter conducted at 37 °C. After fermentation, the content of p-cresol was determined.

2.3. Fermentation of bamboo by inoculation with L. zymae KUST4216

The glycerol stock of L. zymae KUST4216 was inoculated into MRS broth, pre-cultured at 37 °C for 24 h, and subcultured twice under identical conditions. Cells were harvested by centrifugation (3800 ×g, 4 min), washed twice with sterile physiological saline, and resuspended to a final concentration of 108 CFU/mL (OD600nm = 1.0). Peeled and washed bamboo shoots were julienned, blanched in boiling water for 10 min, cooled, and evenly distributed into three 5 L fermentation vessels. Mineral water was added at a 1:1 (w/v) ratio. Inoculated suansun was prepared with a 2% (v/v) inoculum, while uninoculated suansun served as the control (Xue et al., 2025). Sealed vessels were fermented at 30 °C for 12 days. Samples collected on days 0, 2, 4, 8, and 12 of fermentation were analyzed for physicochemical properties and microbial diversity.

2.4. Analysis of physicochemical properties

The pH value was determined by a pH meter (Mettler Toledo, Shanghai, China). Titratable acidity (TA) was measured by titration to pH 8.2 ± 0.2 with 0.1 mol/L NaOH and the result was expressed as the percentage of lactic acid (w/w) (Ye et al., 2022).

2.5. Determination of organic acid content

Organic acid profiles in suansun samples across different fermentation stages were quantitatively analyzed using high-performance liquid chromatography (HPLC). The chromatographic analysis was based on the method of Liang et al. (2025) with slight modifications. Analytical conditions were as follows: an organic acid column (250 mm × 4.6 mm, 5 μm particle size; Alltech Grace, Deerfield, FL, USA) maintained at 30 °C; mobile phase comprising 25 mmol/L potassium phosphate buffer (pH 2.5) delivered at 0.8 mL/min; UV detection at 210 nm; injection volume of 30 μL. Each sample was analyzed in triplicate.

2.6. Volatile compounds analysis

Volatile compounds were analyzed using headspace solid-phase microextraction coupled with gas chromatography–mass spectrometry (HS-SPME-GC–MS), following the method Qin, Xiong, Xiong, et al. (2024) with appropriate modifications. Volatile compounds were enriched via HS-SPME by combining 3 mL of homogenized sample with an equal volume of saturated NaCl solution in a 20 mL headspace vial, supplemented with 5 μL of hexyl 2-methylbutyrate solution as an internal standard. A 50/30 μm divinylbenzene/carboxen/polydimethylsiloxane (DVB/CAR/PDMS) SPME fiber (Supelco, Bellefonte, USA) was employed for extraction under 50 °C water bath conditions for 15 min. Following extraction, the fiber was thermally desorbed in the GC injection port at 250 °C for 5 min to release adsorbed volatiles. Separation and detection were performed using an 8890/7000E GC–MS system (Agilent, Santa Clara, USA) equipped with an HP-INNOWAX capillary column (30 m × 0.25 mm × 0.25 μm; Agilent, Santa Clara, USA).

GC conditions were as follows: The initial temperature was set at 40 °C and held for 4 min, followed by a temperature ramp of 5 °C/min to 140 °C (held for 5 min), then increased at 3 °C/min to 155 °C (held for 1 min), further raised at 5 °C/min to 200 °C (held for 2 min), subsequently elevated at 1 °C/min to 205 °C, and finally increased at 7 °C/min to 240 °C. Helium was used as the carrier gas at a flow rate of 1 mL/min. The injector temperature was maintained at 250 °C. MS conditions were as follows: Ion source temperature of 230 °C, ionization energy of 70 eV, and full scan mode for data acquisition within a mass-to-charge ratio (m/z) range of 35–500. To ensure accurate compound identification, a homologous series of n-alkanes (C5-C20) was analyzed under identical experimental conditions using GC–MS. Retention indices (RIs) were calculated based on the n-alkane (C5-C20) retention times. Volatile components were compared with reference standards in the NIST database (NIST 14.0). The acquired data were cross-validated through mass spectral similarity matching and retention index comparisons.

2.7. Microbial communities

Genomic DNA from brine samples was extracted using the CTAB method as described by Kuerman et al. (2023). The purity and concentration of the extracted DNA were assessed by agarose gel electrophoresis and the purified genomic DNA was stored at −20 °C for subsequent analysis. Bacterial amplification was performed using primers designed based on the full-length 16S rRNA gene sequence: 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-GGTTACCTTGTTACGACTT-3′) (Ye et al., 2022). PCR products were purified using gel extraction, quantified with a Qubit 4.0 Fluorometer (Thermo Fisher Scientific, USA), and sequenced on an Illumina NextSeq 2000 platform (Illumina, San Diego, CA, USA). Raw paired-end reads were quality-controlled using fastp (v0.19.6) and assembled with FLASH (v1.2.11). Sequences were clustered into operational taxonomic units (OTUs) at a 97% similarity threshold via UPARSE (v7.1), followed by chimera removal. Taxonomic annotation was performed using the RDP Classifier (v2.11) against the SILVA database (v138) with a 70% confidence cutoff.

2.8. L. zymae KUST4216 genome sequencing, assembly, and annotation

Following revival and pre-treatment of the cryopreserved strain, the washed cell pellet was transferred to a 1.5 mL microcentrifuge tube and immediately flash-frozen in liquid nitrogen. The samples were subsequently submitted to Shanghai Majorbio Bio-pharm Technology Co., Ltd. (Shanghai, China) for DNA extraction and sequencing services. Raw sequencing reads underwent quality trimming to remove low-quality regions, yielding high-quality processed reads. De novo genome assembly was performed on the cleaned data to generate an optimized genomic contig. The Illumina platform (Illumina Inc., San Diego, CA, USA) is utilized to conduct PE150 (pair-end) sequencing. Functional annotation of the assembled genes was conducted using specialized databases, including the Kyoto Encyclopedia of Genes and Genomes (KEGG) and the carbohydrate-active enzymes (CAZy) database.

2.9. RNA extraction, Illumina sequencing and functional annotation

Samples were subjected to low-temperature centrifugation (10,000g, 5 min, 4 °C). The supernatant was stored at −80 °C for metabolite analysis, while the pellet was washed three times with physiological saline, flash-frozen in liquid nitrogen, and processed for transcriptomic studies (Xiao et al., 2022). RNA extraction, library construction, and high-throughput sequencing were performed by Shanghai Majorbio Bio-pharm Technology Co., Ltd. Quality-filtered raw data were aligned to the L. zymae reference genome. Differentially expressed genes (DEGs) were functionally annotated and pathway-enriched using the KEGG database.

2.10. Statistical analysis

Statistical analyses were performed using IBM SPSS Statistics (v26.0, IBM Corp., USA). One-way analysis of variance (ANOVA) followed by Tukey's multiple comparison was applied to evaluate significant differences, with p < 0.05 considered statistically significant. Experimental results are expressed as mean ± standard deviation (Mean ± SD). Data visualization and statistical plotting were conducted using Origin software (v2021, OriginLab Corporation, USA). Whole-genome, transcriptomic, and microbial community data analyses were processed through the Majorbio Cloud Platform (https://cloud.majorbio.com).

3. Results and discussion

3.1. Production and analysis of p-cresol by L. zymae KUST4216

3.1.1. Screening and identification of high p-cresol producing strains

p-Cresol is a core characteristic substance that shapes the unique flavor of suansun. Differences in its content represent one of the key factors contributing to variations in the flavor intensity and distinctness of suansun, and it can serve as an important reference index for evaluating the flavor quality of suansun (Wang et al., 2025; Wang et al., 2026). Consequently, screening for functional microbial strains capable of enhancing p-cresol formation during fermentation is of great significance for the targeted modulation of flavor development and the improvement of product quality. In this study, a total of 24 lactic acid bacterial strains were isolated and identified from naturally suansun samples collected from different regions of Yunnan Province (Fig. 1A). These isolates belonged to four genera—Levilactobacillus, Lactobacillus, Pediococcus, and Limosilactobacillus—indicating a high diversity of lactic acid bacteria. Among them, Lactiplantibacillus plantarum was the predominant species, accounting for 13 isolates (54.2%), consistent with patterns commonly observed in vegetable fermentation systems (Guan et al., 2020). Notably, Levilactobacillus zymae ranked as the second most abundant species (5 isolates, 20.8%), with a higher prevalence than that typically reported in other fermented vegetables, which may reflect the unique fermentation environment characteristic of Yunnan. After conducting single-strain fermentation of the aforementioned strains in sterile bamboo shoots, it was found that 5 strains could independently produce p-cresol in the fermentation system, and all these 5 strains were identified as L. zymae (Fig. 1B). As a lactic acid bacterium with important fermentation functions, L. zymae has been mainly isolated and identified from sourdough (Işık Doğan & Yılmaz, 2023). The quantitative results of p-cresol further revealed that L. zymae KUST4216 could significantly promote the production of p-cresol during the fermentation of bamboo shoots (p < 0.05) (Fig. 1C-E), thus being selected as the target functional strain for subsequent inoculation fermentation experiments.

Fig. 1.

Fig. 1

Open in a new tab

Isolation and screening of high-yield p-cresol-producing strain. (A) Experimental design. (B) Detection of p-cresol production among 24 isolated strains. (C) Quantification of p-cresol produced by the five positive strains. (D) Levilactobacillus zymae KUST4216 colony morphology. (E) Microscopic morphology of L.zymae KUST4216.

3.2. p-Cresol accumulation dynamics in inoculated suansun

Given the promoting effect of L. zymae KUST4216 on p-cresol production, we inoculated L. zymae KUST4216 into naturally fermented bamboo shoots, with uninoculated sample serving as control, to investigate the impact of L. zymae KUST4216 on the flavor characteristics of suansun. Specifically, quantitative analysis of p-cresol was conducted to verify whether the in vitro-screened high-yield p-cresol strain retains its high-yielding capacity in the complex fermentation system of suansun. Volatile compounds analysis via HS-SPME-GC–MS revealed a significant accumulation of p-cresol in the KUST4216 group compared to the uninoculated group. In the KUST4216 group, the p-cresol content remained at a low level in the early fermentation stage (days 0–4) and began to increase significantly from day 8, peaking at day 12 with a concentration of 6.62 ± 4.15 mg/kg (Fig. 2A). In contrast, the uninoculated group exhibited consistently low p-cresol levels throughout the entire fermentation process, with the maximum value only reaching 0.20 ± 0.14 mg/kg at day 12. This suggests that L. zymae KUST4216 inoculation promotes p-cresol synthesis, with its accumulation intensifying over time. L. zymae KUST4216 likely facilitates efficient phenolic compound production via phenylalanine or tyrosine metabolic pathways (Tang et al., 2024). The low content of p-cresol in uninoculated suansun suggests that the wild microbial community lacks the corresponding metabolic ability. This result confirmed that inoculation with L. zymae KUST4216 effectively promotes the synthesis and accumulation of p-cresol during suansun fermentation, and the accumulation intensity is positively correlated with fermentation time in the middle and late stages.

Fig. 2.

Fig. 2

Open in a new tab

Changes of volatile organic compounds in suansun. (A) p-Cresol. (B) Acetic acid. (C) 1-Hexanol. (D) Nonanal.

3.3. Effects of L. zymae KUST4216 inoculation on suansun quality

3.3.1. Profiling of key volatile compounds (excluding p-cresol)

Flavor quality is the core characteristic of suansun, determined by the synergistic effect of volatile compounds, organic acids, and other metabolites. The following sections focus on the changes in flavor-related substances (excluding p-cresol) in suansun after L. zymae KUST4216 inoculation, to clarify the regulatory role of the strain in suansun flavor formation. A total of 18 volatile compounds were identified from uninoculated and inoculated suansun. Among these, acetic acid, 1-hexanol, p-cresol and nonanal were the top four most abundant (Fig. 2). Acetic acid exhibited the highest abundance among all volatile compounds in both the uninoculated and inoculated fermented suansun and is also one of the characteristic flavor compounds of suansun. Acetic acid in KUST426 group accumulated rapidly during days 4–8 but declined sharply by day 12, indicating efficient carbon source utilization and acidification by L. zymae KUST4216 (Fig. 2B). In contrast, acetic acid accumulation in uninoculated suansun was slower, peaking only at day 12, reflecting metabolic lag in natural microbiota. 1-Hexanol in inoculated suansun showed an early peak followed by a decline, potentially linked to L. zymae KUST4216's lipid oxidation and reductive capacity (Fig. 2C). Conversely, uninoculated suansun exhibited erratic fluctuations (2–11 mg/kg), indicative of metabolic instability in wild microbial communities. Nonanal peaked on day 8 in both the inoculated and uninoculated suansun groups, with the uninoculated group exhibiting higher nonanal content than the inoculated suansun group at this time point (Fig. 2D). However, the inoculated suansun group demonstrated a more pronounced accumulation of nonanal in the early stages. Both groups showed a sharp decline in nonanal levels by day 12, possibly due to factors such as oxidation or degradation (Wang et al., 2023).

The production of volatile compounds in uninoculated suansun followed a slow, fluctuating pattern. Certain compounds (e.g., 1-hexanol) reached higher peaks but lacked temporal stability, likely due to competitive interactions among wild microbiota disrupting coordinated metabolic pathways (Guan et al., 2022). In contrast, inoculated suansun demonstrated rapid initiation of volatile compounds synthesis, with early and efficient accumulation of acetic acid and p-cresol, albeit with a simpler trend. Inoculated fermentation regulated the temporal dynamics and concentrations of key compounds (e.g., p-cresol, acetic acid, 1-hexanol), accelerating early-stage accumulation of specific flavor substances, thereby highlighting its role in enhancing flavor development in suansun (Xue et al., 2025).

3.3.2. Sensory characteristic analysis

3.3.2.1. Effect of L. zymae KUST4216 inoculation on the physicochemical properties of suansun

Beyond the key flavor compounds defining suansun's flavor quality, the fermentation system's fundamental physicochemical properties are equally critical. pH value and titratable acidity (TA) act as dual-functional indicators, directly contributing to the final flavor profile and serving as core metrics for monitoring fermentation dynamics (Ye et al., 2022). Their changing trends reflect microbial acidification, while constructing a stable acidic microenvironment essential for acid-tolerant microbial proliferation, substrate conversion and biosynthesis of characteristic flavorants like p-cresol, playing an indispensable role in suansun quality control (Qin, Xiong, Kang, et al., 2024). Both uninoculated and inoculated suansun exhibited an initial pH of 7.16 ± 0.01 at day 0 (Fig. 3A). During fermentation, the pH decreased rapidly in the early stage (0–4 days), stabilized between days 4–8, and reached the minimum values at day 12 (uninoculated suansun: 3.51 ± 0.01; inoculated suansun: 3.47 ± 0.01). This triphasic pH dynamic (rapid decline-stabilization-terminal equilibrium) is a typical characteristic of lactic acid fermentation: the sharp pH reduction in the early stage can effectively suppress the growth of spoilage microorganisms, while the stable acidic environment in the middle and late stages creates favorable conditions for the synthesis of flavor compounds (Wei et al., 2025).

Fig. 3.

Fig. 3

Open in a new tab

The physicochemical properties of suansun. (A) pH value. (B) Titratable acid. (C) Changes of organic acids in uninoculated and inoculated suansun.

The TA content showed a continuous upward trend during fermentation: it increased from 0.002% to 0.56% in uninoculated suansun and from 0.002% to 0.54% in inoculated suansun (Fig. 3B). There was no significant difference in TA between the two fermentation methods at the end of fermentation (p > 0.05), and both groups displayed similar acidification kinetics in the early stage. This finding suggests that L. zymae KUST4216 possesses robust metabolic capacity, enabling it to compete effectively in the complex fermentation environment, ensure the smooth progression of the fermentation process, and ultimately achieve an acidity level comparable to that of traditionally naturally fermented suansun.

3.3.2.2. Effect of L. zymae KUST4216 inoculation on organic acids content

Organic acids serve as the core flavor determinants of fermented vegetables (Xiong et al., 2024). They not only impart differentiated sour tastes but also neutralize undesirable off-flavors including bitterness, astringency, and fishiness, ultimately shaping a harmonious and rich flavor profile (Ye et al., 2022). Nine organic acids (oxalic, tartaric, quinic, malic, lactic, acetic, citric, succinic, and fumaric acids) were detected during suansun fermentation, consistent with previous studies (Lu et al., 2022; Qin, Xiong, Kang, et al., 2024). The total organic acid content in both uninoculated inoculated suansun exhibited a progressive increase throughout fermentation (Fig. S1), while there was no significant difference in the total organic acid content between uninoculated and inoculated suansun at the end of fermentation (Fig. 3A). Among these organic acids, lactic and acetic acids are the primary contributors to the sour taste of suansun (Wei et al., 2025). Both uninoculated group and L. zymae KUST4216 group produced lactic and acetic acids as the dominant organic acids during fermentation, a characteristic consistent with prior research (Tang et al., 2024). However, L. zymae KUST4216 significantly altered the contents of specific organic acids. The acetic acid content in inoculated suansun was significantly higher, while the contents of quinic acid and citric acid were significantly lower than those in uninoculated suansun (p < 0.05). This indicates that inoculation with L. zymae KUST4216 may enhance the refreshing sour taste by increasing acetic acid and reduce bitterness by decreasing quinic acid and citric acid, thereby regulating the flavor hierarchy of suansun.

3.3.2.3. Sensory characteristic variations

Electronic nose (e-nose) and electronic tongue (e-tongue) were used to comprehensively evaluate the sensory characteristics of suansun, so as to reflect the overall changes in flavor and taste after L. zymae KUST4216 inoculation. The detection of characteristic flavor compounds via each sensor of the e-nose enables the identification of subtle flavor variations between samples (Liu et al., 2025). For e-nose analysis, the response spectra of uninoculated and uninoculated suansun were similar across all sensors, but the response values differed slightly and varied dynamically with fermentation time (Fig. 4A). Specifically, L. zymae KUST4216 group showed higher response values at sensors S1 (sensitive to short-chain alkanes), S2 (sensitive to carbon-containing substances), S9 (sensitive to alkane alcohols/ketones), and S15 (sensitive to carbon-containing alcohols/aldehydes). This phenomenon is attributed to the rapid colonization of L. zymae KUST4216 in the inoculated fermentation system, which can synchronize metabolic activities and efficiently synthesize target flavor substances through glycolysis and fatty acid metabolism pathways (Qin, Xiong, Xiong, et al., 2024). In contrast, natural fermentation requires a longer adaptation period for microorganisms, and the inconsistent metabolism of the complex community leads to lower accumulation of these flavor substances. Principal component analysis (PCA) of the 18 e-nose sensor data showed that the cumulative contribution rate of PC1 (64.6%) and PC2 (15.1%) reached 79.7%, enabling clear separation of the two fermentation groups in the PCA plot, which confirms that there are significant differences in the volatile flavor profiles between inoculated and uninoculated suansun (Fig. 4B). For e-tongue analysis, the response spectra of inoculated and uninoculated suansun exhibited similar characteristics across all sensors, but there were slight differences in specific response values, which changed dynamically with the progression of fermentation (Fig. 4C). The response values of both groups at sensors S1 (sour taste), S3 (umami taste), and S7 (bitterness) were higher than those of unfermented bamboo shoots (d0), indicating that fermentation significantly enhances the taste intensity of suansun. PCA results (Fig. 4D) based on e-tongue data also showed clear separation between the uninoculated group and L. zymae KUST4216 group, confirming that L. zymae KUST4216 inoculation can modify the taste profile of suansun, which is consistent with the changes in organic acid composition.

Fig. 4.

Fig. 4

Open in a new tab

Electronic nose and electronic tongue analyses during suansun fermentation. (A) Radar plot of e-nose odor profiles. (B) Principal component analysis of e-nose data. (C) Radar plot of e-tongue taste profiles. (D) Principal component analysis of e-tongue data.

3.4. Effect of L.zymae KUST4216 inoculation on the bacterial community of suansun

The distinctive flavor of suansun is closely associated with the dynamic changes of microbial communities during fermentation. Throughout the process, diverse lactic acid bacteria, acetic acid bacteria, and other fermentative microorganisms produce a variety of metabolites, including organic acids, esters, aldehydes, ketones, and aromatic compounds, which collectively contribute to the sourness, freshness, and complex flavor profile of the product (Qin, Xiong, Xiong, et al., 2024; Shang et al., 2022). Moreover, interactions among microbial species and the succession of community structure directly influence the formation rate and composition of flavor compounds, ultimately determining the sensory quality of the final product.

3.4.1. α-diversity of microbial communities

Microbial community α-diversity, which reflects species richness and evenness, is closely linked to the stability of fermentation ecosystems and the consistency of metabolite synthesis (Wei et al., 2024). The α-diversity indices during suansun fermentation are shown in Fig. 5A. Among these, the Chao1 index— a core indicator for assessing species richness— increased during the early fermentation stage but began to decline from day 2 onwards. Changes in the Simpson index reflected competition and succession among dominant bacteria. The Simpson index of uninoculated suansun was relatively low at the early fermentation stage, peaked on day 4, and then gradually decreased. In contrast, the Simpson index of L. zymae KUST4216 group was generally higher than that of uninoculated suansun during the early stage and exhibited fluctuations throughout fermentation. The Shannon index, which reflects both species richness (quantity) and evenness (distribution equilibrium) (Wei et al., 2024), was higher in uninoculated suansun during the first two days compared to other groups. It decreased on day 4 and gradually increased thereafter. For L. zymae KUST4216 group, the Shannon index rose on day 2 and then slowly declined, aligning with the trend of the Chao1 index. The inoculated L. zymae KUST4216 initially occupied an extremely high proportion, resulting in low evenness. During mid-fermentation, transient equilibrium formed due to competition between dominant native strains and L. zymae KUST4216. Continuous dynamic competition among microbial populations caused unstable evenness. Overall, during the initial fermentation stage, mild environmental conditions and abundant resources allowed simultaneous proliferation of native communities and L. zymae KUST4216, leading to a transient increase in diversity. In mid-fermentation, lactic acid accumulation lowered the pH, creating strong selective pressure that favored acid-tolerant bacteria and eliminated pH-sensitive strains (Guan et al., 2022), causing a sharp decline in diversity. During late fermentation, the accumulation of metabolic byproducts and substrate conversion provided ecological niches for secondary bacteria, partially restoring community balance.

Fig. 5.

Fig. 5

Open in a new tab

Microbial community dynamics during suansun fermentation. (A) Changes in microbial α-diversity indices during fermentation. (B–C) PCoA showing β-diversity variation in uninoculated group (B) and L. zymae KUST4216 group (C) over time. (D) Comparison of microbial community structures between uninoculated and inoculated suansun at day 0 and day 12. (E) Circos plot of bacterial genera. (F–G) Changes in microbial species during suansun fermentation in uninoculated group (F) and L. zymae KUST4216 group (G). (H) Relative abundance of differential genera.

3.4.2. β-diversity of microbial communities

β-diversity analysis was used to compare the differences in microbial community structure between different fermentation groups and stages (Fig. 5B-D). As shown in Fig. 5B, samples from the initial stage of uninoculated fermentation (d0) clustered tightly, indicating high stability and minimal intra-group variation in microbial community structure. During mid- to late fermentation, sample dispersion increased, reflecting intensified dynamic changes and intra-group heterogeneity (Guan et al., 2022). Samples from different stages were distributed discretely along the PC1 axis, particularly with a significant separation between day 0 and subsequent stages, highlighting distinct microbial community structures between initial and later fermentation phases. For inoculated suansun (Fig. 5C), samples from the initial stage (d0) exhibited high clustering, suggesting uniform colonization and stability of the inoculated microbial community (Wu et al., 2025). Intra-group heterogeneity increased as fermentation progressed, likely due to the proliferation of native microorganisms. L. zymae KUST4216-d0 and uninoculated suansun-d0 were distinctly separated on the PC1 axis, indicating fundamental differences in initial microbial community structures caused by the introduction of exogenous dominant strains through inoculation (Qin, Xiong, Xiong, et al., 2024). At the completion stage (day 12), uninoculated and inoculated suansun samples were clearly separated on both PC1 and PC2 axes, demonstrating significant structural differences in their terminal microbial communities. This suggests that inoculated fermentation, driven by directional metabolism of dominant strains, follows a distinct succession pathway compared to natural fermentation, ultimately forming unique microbial community structures (Qin, Xiong, Xiong, et al., 2024).

3.4.3. Composition and succession of microbial communities

The composition and succession of bacterial communities are direct manifestations of complex interactions among microorganisms, and such interactions profoundly regulate the synthesis of fermentation metabolites and the formation of final product quality. To clarify the intrinsic link between microbial dynamics and suansun flavor development, the bacterial community composition and succession in natural fermented suansun and inoculated fermented suansun during fermentation were systematically investigated (Fig. 5E). In uninoculated suansun, 101 bacterial genera were identified at the genus level, with dominant genera being Lactiplantibacillus (54.47%), Bacillus (7.67%), and Pelomonas (6.53%). At the initial fermentation stage (day 0), the dominant genera in natural fermented suansun were Tardiphaga (21.83%), Pelomonas (20.64%), Bradyrhizobium (13.13%), and Sphingomonas (12.97%), accounting for 68.57% of the total bacterial population. As fermentation progressed, Tardiphaga and Pelomonas gradually decreased, comprising less than 1% after the mid-fermentation stage (day 4). Lactiplantibacillus, initially present at only 0.01%, gradually increased in abundance and became the dominant genus in later stages (day 8 and day 12), reaching 94.04% and 88.20%, respectively. At the species level, 132 bacterial species were identified in uninoculated suansun, with dominant species including L. plantarum (55.47%), Bacillus subtilis (7.35%), Pelomonas saccharophila (6.53%), Tardiphaga robiniae (5.96%), and L. zymae (3.16%). L. plantarum emerged as the predominant species in uninoculated suansun. This species is commonly found in various fermented vegetables, such as suancai, pickled chili, and kimchi (Li et al., 2023). Due to its high lactic acid production efficiency, which lowers environmental pH and inhibits other microorganisms, L. plantarum establishes dominance in fermentation systems (Lu et al., 2022). Notably, L. zymae was detected in uninoculated suansun, increasing from 1.52% to 6.58% during fermentation.

In inoculated suansun, 102 bacterial genera were identified at the genus level, with dominant genera being Levilactobacillus (62.52%), Lactiplantibacillus (29.15%), and Bacillus (2.25%). At the initial stage (day 0), Levilactobacillus (86.44%), Lactiplantibacillus (5.94%), and Pelomonas (1.45%) dominated, constituting 93.83% of the total bacterial population. Due to inoculation, Levilactobacillus exhibited high initial abundance. During fermentation, Levilactobacillus decreased from 86.44% (day 0) to 42.97% (day 4) before increasing again to 66.62% by day 12. Lactiplantibacillus, starting at 5.94%, gradually increased to 52.16% by day 4 but declined to 30.31% by the final stage (day 12). At the species level, 128 bacterial species were identified in inoculated suansun, with dominant species being L. zymae (62.52%), L. plantarum (29.15%), and Bacillus pumilus (2.01%). The inoculated strain L. zymae KUST4216 initially dominated due to artificial inoculation, rapidly consuming easily accessible carbon sources to establish ecological dominance. However, L. plantarum proliferated during mid-fermentation (days 4–8), temporarily surpassing L. zymae. After day 4, L. zymae regained dominance, reaching 67.15% by the final stage (day 12).

The succession patterns of uninoculated and inoculated suansun differed fundamentally (Fig. 5F-H). In uninoculated suansun, environmental genera (e.g., Tardiphaga, Pelomonas) dominated initially but were gradually replaced by L. plantarum (>88% in later stages), reflecting a unidirectional succession pattern. In inoculated suansun, the artificial introduction of L. zymae KUST4216 created initial dominance (86.44%), followed by dynamic competition with L. plantarum (L. plantarum surpassed L. zymae at day 4, but L. zymae KUST4216 regained dominance by day 12), demonstrating a bidirectional regulatory characteristic of “artificial intervention-microbial competition” functional traits of dominant species drove metabolic differences. In uninoculated fermentation, L. plantarum established dominance through rapid acidification, but its singular metabolic mode may limit flavor complexity (Qin, Xiong, Kang, et al., 2024). In inoculated suansun, the early dominance of L. zymae KUST4216 correlated with rapid acetate accumulation (86.47 mg/L by day 4), while its mid-stage metabolic adjustments likely triggered L. plantarum's utilization of secondary carbon sources, jointly influencing temporal flavor compound distribution (Li et al., 2025). L. zymae KUST4216's unique capability in p-cresol synthesis may compensate for metabolic deficiencies in natural fermentation. The temporary dominance reversal at day 4 may stem from differences in carbon source competition strategies: L. zymae KUST4216 preferentially consumes glucose before entering dormancy, whereas L. plantarum utilizes complex carbon sources. The recovery of L. zymae KUST4216 dominance by day 12 (66.62%) suggests adaptive mechanisms for oligosaccharide utilization or acid tolerance (Xue et al., 2025). Whether the naturally occurring L. zymae in natural fermented suansun contributes to phenol synthesis requires validation via genomic annotation.

The differential accumulation of p-cresol between the inoculated suansun dominated by L. zymae KUST4216 and the uninoculated suansun dominated by L. plantarum directly reflects their distinct types of fermentation and metabolism (Qin, Xiong, Kang, et al., 2024). This elevated p-cresol yield in the inoculated suansun is primarily driven by the specialized tyrosine metabolism of L. zymae KUST4216, which facilitates efficient production to p-cresol. In contrast, the uninoculated suansun depends on the more conserved metabolic pathways characteristic of L. plantarum and associated microbiota, leading to significantly lower p-cresol accumulation, ultimately leading to significantly greater p-cresol accumulation in the inoculated suansun. Future studies could develop mixed-starter cultures of L. zymae KUST4216 and L. plantarum to balance acidification speed and flavor complexity through ratio optimization.

3.5. Genomic and transcriptomic profiling of L.zymae KUST4216

3.5.1. Genomic characteristics of L. zymae KUST4216

Given that L. zymae KUST4216 exhibits a remarkable ability to produce p-cresol during fermentation, this characteristic is likely associated with its unique metabolic pathways or key functional genes (Xiao et al., 2023). Genomic information can not only elucidate the genetic basis underlying the biosynthesis of aromatic compounds but also provide insights into the strain's ecological adaptation and metabolic advantages within complex fermentation systems (Zhang et al., 2023). Therefore, whole-genome sequencing and functional annotation of L. zymae KUST4216 are of great significance.

3.5.1.1. Phylogenetic analysis

In the phylogenetic tree constructed based on genomic sequences, L. zymae KUST4216 and the conspecific type strain L. zymae GCF 001434115.1 clustered within a clade with high bootstrap support (Fig. 6A). The sequence identity reached 96.8%, and the coverage of the aligned sequences was 82.11%. With a score of 29,004.0 and a short branch length (genetic distance ≈ 0.02), these results indicate a close phylogenetic relationship between the two strains.

Fig. 6.

Fig. 6

Open in a new tab

Whole-genome analysis of L.zymae KUST4216. (A) Phylogenetic tree based on whole-genome sequences. (B) Functional annotation of the L.zymae KUST4216 genome according to KEGG pathway classification. (C) Distribution of carbohydrate-active enzyme (CAZyme) families in the L.zymae KUST4216 genome.

3.5.1.2. KEGG functional annotation

Whole-genome analysis holds the potential to offer insights into the genes associated with metabolic flavor substances in KUST4216. KEGG annotation of L. zymae KUST4216 (Fig. 6B) revealed genes associated with six major functional categories: Cellular Processes, Environmental Information Processing, Genetic Information Processing, Metabolism, Organismal Systems, and Human Diseases. A total of 1367 genes were annotated to metabolic pathways, with carbohydrate metabolism (201 genes, 14.7%) and amino acid metabolism (123 genes, 9.0%) being the most enriched categories. These findings suggest L. zymae KUST4216 preferentially utilizes carbon and nitrogen sources to support growth, consistent with the metabolic traits of lactic acid bacteria in traditional fermentation ecosystems (Hong et al., 2017). The enrichment of amino acid metabolism genes may correlate with flavor precursor biosynthesis, potentially influencing the aroma profile of suansun (Tang et al., 2022).

3.5.1.3. CAZyme encoding gene analysis

Carbohydrate-active enzymes (CAZymes) play a key role in the degradation and utilization of complex carbohydrates in fermentation substrates. CAZy annotation identified 72 genes in L. zymae KUST4216 (Fig. 6C), classified into four families: Auxiliary Activities (AA), Carbohydrate Esterases (CE), Glycoside Hydrolases (GH), and Glycosyl Transferases (GT). GHs exhibited the highest abundance (29 genes), followed by GTs (21 genes). GH enzymes catalyze glycosidic bond hydrolysis via water molecule insertion (Li et al., 2016), while GTs are essential for synthesizing complex glycoconjugates (Ryu et al., 2014). The predominance of GH and GT genes implies L. zymae KUST4216 may efficiently degrade bamboo shoot polysaccharides to generate glucose and other carbon substrates for fermentation (Bourne & Henrissat, 2001). These carbon sources provide the necessary energy for the metabolic pathways involving aromatic amino acids, such as tyrosine. This metabolic activity ultimately affects the production of p-cresol (Tang et al., 2024).

3.5.2. Transcriptomic dynamics and functional analysis of L. zymae KUST4216 during fermentation

3.5.2.1. Identification of differentially expressed genes (DEGs)

With the aim of further elucidating the molecular mechanisms underlying high p-cresol production in L. zymae KUST4216, transcriptomic analysis was conducted to characterize the expression profiles of key genes involved in aromatic compound metabolism (Xue et al., 2025). The difference in gene expression levels between the early fermentation stage (d2) and the mid-fermentation stage (d4) was analyzed using Principal Component Analysis (PCA) (Fig. 7A). The results showed that samples from the early fermentation stage were highly clustered in the plot, while samples from the mid-fermentation stage were relatively scattered. This indicated a greater inter-individual variation in gene expression on day 4, which may be attributed to the increased complexity of bacterial metabolic activities and changes in environmental stress during the late fermentation stage, leading to enhanced diversity in gene expression regulation (Zhang et al., 2024). There was a significant difference in the overall gene expression profiles between the early and mid-fermentation stages. This result demonstrated that as fermentation progressed, the bacterial gene expression pattern underwent systematic changes. These changes may involve the regulation of genes related to pathways such as energy metabolism, secondary metabolite synthesis, and stress response, reflecting the dynamic evolution of bacterial physiological states and functions at different fermentation stages (Qin, Xiong, Xiong, et al., 2024). In the comparative analysis between the two fermentation time points, a total of 826 differentially expressed genes (DEGs) with significant expression differences were identified, including 486 up-regulated genes and 340 down-regulated genes (Fig. 7B).

Fig. 7.

Fig. 7

Open in a new tab

Transcriptomic analysis of L.zymae KUST4216. (A) PCA analysis. (B) Differential gene volcano plot. (C) GO enrichment analysis of transcriptomic analysis. (D) KEGG pathway analysis of DEGs in transcriptomic analysis.

3.5.2.2. GO functional enrichment analysis

To elucidate the transcriptional regulatory features of L. zymae KUST4216 across fermentation stages, Gene Ontology (GO) enrichment analysis was performed. Results (Fig. 7C) revealed significant enrichment in three primary functional categories: Molecular Function (MF), Cellular Component (CC), and Biological Process (BP). GO enrichment analysis revealed the key metabolic regulatory network of L. zymae KUST4216 during fermentation. The DEGs across different fermentation stages were found to be primarily enriched in molecular functions. Upregulated DEGs were enriched in transport-related functions, with enrichment factors mostly ranging from 0.5 to 1.5, indicating substantial functional enrichment. The marked enrichment of transmembrane transport functions in day 4 upregulated genes suggest enhanced transmembrane substance transport during mid-fermentation (Harle et al., 2024). This adaptation likely reflects bacterial responses to nutrient depletion, requiring reinforced transport systems (e.g., ion/small-molecule transporters) to sustain survival and facilitate precursor uptake for secondary metabolite (e.g., flavor compound) biosynthesis. Concurrently, active transport mechanisms to excrete inhibitory metabolic byproducts may maintain intracellular homeostasis under accumulating environmental stress (Xiao et al., 2024). Downregulated DEGs were primarily linked to oxidoreductase activity, carboxylic acid catabolism, and organic acid degradation. The suppression of organic acid catabolic pathways during mid-fermentation may promote organic acid accumulation, a critical determinant of fermented product acidity. These findings collectively demonstrate L. zymae KUST4216's transcriptional strategy: upregulating transport genes to optimize nutrient and environmental adaptation, while downregulating catabolic genes to redirect metabolic fluxes toward biosynthetic pathways. Such regulatory dynamics ultimately shape the flavor profile and quality of fermented products.

3.5.2.3. KEGG pathway enrichment analysis

Given that the biological functions of the identified differentially expressed genes (DEGs) remain to be comprehensively clarified to decipher their regulatory roles in p-cresol synthesis, pathway enrichment analysis was conducted using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. This analysis aimed to annotate the DEGs to specific metabolic pathways and functional categories, thereby shedding light on how these genes coordinate to drive the high-yield trait of L. zymae KUST4216 (Fig. 7D). Statistical screening (p < 0.05) identified 13 significantly enriched metabolic pathways, visualized in a bubble plot to illustrate enrichment significance and functional distribution. ABC transporters and sulfur metabolism were significantly enriched in the up-regulated genes. This indicates that the ABC transporter pathway in bacteria was significantly activated in the later stage of fermentation, possibly for enhancing transmembrane substance transport and adapting to environmental changes in the later stage of fermentation (Harle et al., 2024). Pyruvate metabolism, lipoic acid metabolism, citrate cycle, and oxidative phosphorylation were significantly enriched in the down-regulated gene expressions. This suggests that the gene expressions of the core metabolic pathways in bacteria were down-regulated in the later stage of fermentation, possibly due to the consumption of nutrients in the later stage of fermentation, the bacteria reducing energy metabolism activities, or the metabolic pattern shifting from energy production to product synthesis (Yang et al., 2025). The pentose phosphate pathway, starch and sucrose metabolism, and other carbohydrate metabolism pathways were also significantly enriched in the down-regulated genes, indicating that the decomposition and utilization of sugars by bacteria weakened in the later stage of fermentation, further confirming the transformation of the metabolic pattern. These findings collectively highlight L. zymae KUST4216's transcriptional reprogramming during fermentation: upregulating transport systems to maintain homeostasis, while downregulating catabolic pathways to redirect resources toward product synthesis.

3.6. Underlying mechanism for high p-cresol production by L.zymae KUST4216 during suansun fermentation

To further investigate the key genes involved in the metabolic synthesis of p-cresol by L. zymae KUST4216, the results from both the whole-genome and transcriptome analyses were comprehensively analyzed. The network diagram of the formation pathway of key flavor substances (p-cresol, acetic acid, lactic acid) in suansun was constructed based on the whole-genome and transcriptome analyses (Fig. 8). Based on KEGG pathway annotation, carbohydrate metabolism-related genes were significantly downregulated at mid-fermentation stage. Genes involved in starch and sugar metabolism exhibited marked suppression. Downregulation of glycolytic pathway genes and the sugar transporter gene indicated reduced glycolytic activity. During early fermentation, the strain preferentially utilized readily degradable sugars (e.g., glucose and fructose), leading to mid-phase sugar depletion and a metabolic shift toward amino acid-driven secondary metabolism to sustain energy production (Wu et al., 2025). Thus, carbon metabolism in L. zymae KUST4216 transitions from attenuated carbohydrate utilization to enhanced amino acid metabolism, driving the accumulation of secondary metabolites such as p-cresol.

Fig. 8.

Fig. 8

Open in a new tab

The underlying mechanism for high p-cresol production by L.zymae KUST4216 during suansun fermentation.

4. Conclusions

This study pinpoints Levilactobacillus zymae KUST4216 as a pivotal functional strain that markedly enhances p-cresol production during suansun fermentation. Inoculation with this strain swiftly establishes ecological dominance in the microbial community, modulates community succession dynamics, and facilitates the preferential utilization of readily accessible sugars in the early fermentation stage. Upon depletion of these carbohydrates, L. zymae KUST4216 shifts to amino acid-driven secondary metabolism—with tyrosine metabolism as the core—thereby fueling the efficient synthesis and accumulation of p-cresol. Genomic analysis uncovers enriched carbohydrate and amino acid metabolic pathways, coupled with a repertoire of CAZymes, which collectively support substrate degradation and the provision of precursors for aromatic compound biosynthesis. Transcriptomic profiling further validates coordinated transcriptional reprogramming, encompassing the upregulation of nutrient uptake transporters and the downregulation of catabolic pathways to reroute metabolic flux toward p-cresol biosynthesis. Notably, further exploration is still needed regarding the applicability of L. zymae KUST4216 under large-scale industrial fermentation conditions and the specific functions of key candidate genes involved in p-cresol synthesis, which require additional verification through genetic manipulation. In conclusion, these findings provide mechanistic insights into the high p-cresol production of L. zymae KUST4216, elucidate the genetic and ecological basis of flavor formation, and support the potential application of inoculated fermentation for standardized industrial production of high-quality suansun. Addressing these limitations in future research will further promote the strain's practical industrial application.

The following are the supplementary data related to this article.

Fig. S1

Changes in the content of organic acids during the fermentation process of the uninoculated and inoculated suansun.

mmc1.pdf (126.7KB, pdf)

CRediT authorship contribution statement

Zi Ye: Writing – original draft, Methodology, Investigation, Data curation. Xinyu Wang: Methodology, Data curation. Tao Wang: Writing – review & editing, Conceptualization. Chuanqi Chu: Writing – review & editing, Conceptualization. Zhijia Liu: Writing – review & editing, Funding acquisition. Leilei Yu: Writing – review & editing, Investigation. Ming Du: Writing – review & editing, Supervision. Junjie Yi: Writing – review & editing, Supervision, Resources, Funding acquisition, Conceptualization.

Funding

The authors acknowledge the financial support of the National Natural Science Foundation of China (No. U23A20261 and 32572736), Excellent Youth Funding of Yunnan Province (YNQR-QNRC-2018-109), Yunnan Fundamental Research Projects (202301AW070015 and 202301AT070170) and Yunnan Academician Expert Workstation (202305AF150039).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.

References

  1. Bourne Y., Henrissat B. Glycoside hydrolases and glycosyltransferases: Families and functional modules. Current Opinion in Structural Biology. 2001;11(5):593–600. doi: 10.1016/S0959-440X(00)00253-0. [DOI] [PubMed] [Google Scholar]
  2. Chen C., Cheng G., Liu Y., Yi Y., Chen D., Zhang L., Wang X., Cao J. Correlation between microorganisms and flavor of Chinese fermented sour bamboo shoot: Roles of Lactococcus and Lactobacillus in flavor formation. Food Bioscience. 2022;50 doi: 10.1016/j.fbio.2022.101994. [DOI] [Google Scholar]
  3. Clark C.S., Ohstrom A., Rolon M.L., Smith M., Wolfe B.E., Wee J., Van Buiten C.B. Sourdough starter culture microbiomes influence physical and chemical properties of wheat bread. Journal of Food Science. 2024;89(3):1414–1427. doi: 10.1111/1750-3841.16957. [DOI] [PubMed] [Google Scholar]
  4. Dilmetz B.A., Desire C.T., Meneses J., Klingler-Hoffmann M., Young C., Hoffmann P. Impact of propagation time on yeast physiology during bottle conditioning of beer on an industrial scale. Food Chemistry. 2024;435 doi: 10.1016/j.foodchem.2023.137655. [DOI] [PubMed] [Google Scholar]
  5. Guan Q., Huang T., Peng F., Huang J., Liu Z., Peng Z., Xie M., Xiong T. The microbial succession and their correlation with the dynamics of flavor compounds involved in the natural fermentation of suansun, a traditional Chinese fermented bamboo shoots. Food Research International. 2022;157 doi: 10.1016/j.foodres.2022.111216. [DOI] [PubMed] [Google Scholar]
  6. Guan Q., Zheng W., Mo J., Huang T., Xiao Y., Liu Z., Peng Z., Xie M., Xiong T. Evaluation and comparison of the microbial communities and volatile profiles in homemade suansun from Guangdong and Yunnan provinces in China. Journal of the Science of Food and Agriculture. 2020;100(14):5197–5206. doi: 10.1002/jsfa.10569. [DOI] [PubMed] [Google Scholar]
  7. Harle O., Niay J., Parayre S., Nicolas A., Henry G., Maillard M.-B., Valence F., Thierry A., Guedon E., Falentin H., Deutsch S.-M. Deciphering the metabolism of Lactobacillus delbrueckii subsp. delbrueckii during soy juice fermentation using phenotypic and transcriptional analysis. Applied and Environmental Microbiology. 2024;90(3) doi: 10.1128/aem.01936-23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Hong C., Chen Y., Li L., Chen S., Wei X. Identification of a key gene involved in branched-chain short fatty acids formation in natto by transcriptional analysis and enzymatic characterization in Bacillus subtilis. Journal of Agricultural and Food Chemistry. 2017;65(8):1592–1597. doi: 10.1021/acs.jafc.6b05518. [DOI] [PubMed] [Google Scholar]
  9. Hu T., Zhu J., Chai Y., Qiao P., Yi K., Li S., Liu R., Zhong W. Characterization of microbial diversity and important off-odor compounds in fermented Ma bamboo shoots (Dendrocalamus latiflorus Munro) LWT- Food Science and Technology. 2023;188 doi: 10.1016/j.lwt.2023.115378. [DOI] [Google Scholar]
  10. Işık Doğan Ö., Yılmaz R. Tarhana Microbiota–Metabolome Relationships: An Integrative Analysis of Multi-Omics Data. Food Biotechnology. 2023;37(2):191–217. doi: 10.1080/08905436.2023.2204141. [DOI] [Google Scholar]
  11. Kuerman M., Wang R., Zhou Y., Tian X., Cui Q., Yi H., Zhang L. Metagenomic insights into bacterial communities and functional genes associated with texture characteristics of Kazakh artisanal fermented milk Ayran in Xinjiang. China. Food Research International. 2023;164 doi: 10.1016/j.foodres.2022.112414. [DOI] [PubMed] [Google Scholar]
  12. Li J., Liu Y., Xiao H., Huang H., Deng G., Chen M., Jiang L. Bacterial communities and volatile organic compounds in traditional fermented salt-free bamboo shoots. Food Bioscience. 2022;50 doi: 10.1016/j.fbio.2022.102006. [DOI] [Google Scholar]
  13. Li S., Sun M., Tian Y., Jian C., Lv B., Bai Y., Liu X., Guo Y. The potential correlation between microbial communities and flavors in fermented bamboo shoots. Food Bioscience. 2023;56 doi: 10.1016/j.fbio.2023.103066. [DOI] [Google Scholar]
  14. Li W., Fan H., He C., Zhang X., Wang X., Yuan J.…Xiao Y. PspAG97A: A halophilic α-glucoside hydrolase with wide substrate specificity from glycoside hydrolase family 97. Journal of Microbiology and Biotechnology. 2016;26(11):1933–1942. doi: 10.4014/jmb.1606.06047. [DOI] [PubMed] [Google Scholar]
  15. Li X., Wu Y., Zhang F., Zhou W.W., Zheng J. From bamboo shoot to suansun: A comprehensive review of fermentation mechanisms, metabolic pathways, and potential health implications. Food Chemistry. 2026 doi: 10.1016/j.foodchem.2025.147840. [DOI] [PubMed] [Google Scholar]
  16. Li X., You Y., Xue B., Chen J., Du M., Ibrahim A., Suo H., Zhang F., Zheng J. Decoding microbiota and metabolite transformation in inoculated fermented suansun using metagenomics, GC–MS, non-targeted metabolomics, and metatranscriptomics:Impacts of different Lactobacillus plantarum strains. Food Research International. 2025;203 doi: 10.1016/j.foodres.2025.115847. [DOI] [PubMed] [Google Scholar]
  17. Liang J., Wang Y., Wang T., Chu C., Yi J., Liu Z. Enhancing fermented vegetable flavor with Lactobacillus plantarum and Rhodotorula mucilaginosa. Food Research International. 2025;200 doi: 10.1016/j.foodres.2024.115500. [DOI] [PubMed] [Google Scholar]
  18. Liu W., Yu A., Xie Y., Zhang X., Guo B., Xu L.…Lei Z. Electronic nose, flavoromics, and lipidomics reveal flavor changes in longissimus thoracis of fattening Saanen goats by dietary Allium mongolicum regel flavonoids. Food Chemistry: X. 2025;29 doi: 10.1016/j.fochx.2025.102752. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lu H., Huang C., Yu K., Liu Z. Effects of mixed inoculation of Leuconostoc citreum and Lactobacillus plantarum on suansun (sour bamboo shoot) fermentation. Food Bioscience. 2022;47 doi: 10.1016/j.fbio.2022.101688. [DOI] [Google Scholar]
  20. Qian, Q., Chen, Z.-P., Li, K., Xiong, J.-w., Lu, Y.-H., Xu, H.-Y., & Xu, Z.-H. (2024). Guangxi sour bamboo shoots: A study on microbial diversity and flavor characteristics across regions. Journal of Food Composition and Analysis, 135. doi: 10.1016/j.jfca.2024.106641. [DOI]
  21. Qin X., Xiong T., Kang X., Liu Z., Xie M., Guan Q. Comparison of the quality of suansun product between starter inoculation and natural fermentation. Food Bioscience. 2024;58 doi: 10.1016/j.fbio.2024.103782. [DOI] [Google Scholar]
  22. Qin X., Xiong T., Xiong S., Liu Z., Xie M., Guan Q. Metatranscriptomics unravel the formation mechanism of key flavors during the natural fermentation of suansun, a Chinese traditional fermented bamboo shoot. Food Bioscience. 2024;57 doi: 10.1016/j.fbio.2023.103436. [DOI] [Google Scholar]
  23. Ryu J., Eom M.S., Ko W., Han M.S., Lee H.S. A fluorescence-based glycosyltransferase assay for high-throughput screening. Bioorganic & Medicinal Chemistry. 2014;22(8):2571–2575. doi: 10.1016/j.bmc.2014.02.027. [DOI] [PubMed] [Google Scholar]
  24. Santos C., Raymundo A., Moreira J.B., Prista C. Exploring the potential of lactic acid Bacteria fermentation as a clean label alternative for use in yogurt production. Applied Sciences. 2025;15(5) doi: 10.3390/app15052686. [DOI] [Google Scholar]
  25. Shang Z., Ye Z., Li M., Ren H., Cai S., Hu X., Yi J. Dynamics of microbial communities, flavor, and physicochemical properties of pickled chayote during an industrial-scale natural fermentation: Correlation between microorganisms and metabolites. Food Chemistry. 2022;377 doi: 10.1016/j.foodchem.2021.132004. [DOI] [PubMed] [Google Scholar]
  26. Tang H., Li P., Chen L., Ma J.-K., Guo H.-H., Huang X.-C., Zhong R.-M., Jing S.-Q., Jiang L.-W. The formation mechanisms of key flavor substances in stinky tofu brine based on metabolism of aromatic amino acids. Food Chemistry. 2022;392 doi: 10.1016/j.foodchem.2022.133253. [DOI] [PubMed] [Google Scholar]
  27. Tang H., Ma J.-K., Chen L., Jiang L.-W., Kang L.-Z., Guo Y.-Y., Men G.-Y., Nie D.-X., Zhong R.-M. Characterization of key flavor substances and their microbial sources in traditional sour bamboo shoots. Food Chemistry. 2024;437 doi: 10.1016/j.foodchem.2023.137858. [DOI] [PubMed] [Google Scholar]
  28. Wang D., Hu M., Tang X., Zhang Q., Zhao J., Mao B., Zhang H., Cui S. Characterization of physicochemical properties and flavor profiles of fermented Chinese bamboo shoots (suansun) from Liuzhou and Guilin. Food Bioscience. 2023;56 doi: 10.1016/j.fbio.2023.103125. [DOI] [Google Scholar]
  29. Wang X., Liu X., Liu Z., Chu C., Du M., Yi J. Unlocking the secrets of stinky fermented foods: A comprehensive overview of processes, flavor characteristics, fermentation microbiomes, and metabolic pathways. Food Bioscience. 2025 doi: 10.1016/j.fbio.2025.106408. [DOI] [Google Scholar]
  30. Wang X., Yu P., Wang T., Chu C., Liu Z., Du M., Yi J. Novel insights into p-cresol production in fermented bamboo shoots: Core microbiota and metabolic signatures. LWT. 2026 doi: 10.1016/j.lwt.2025.118940. [DOI] [Google Scholar]
  31. Wei J., Du J., Li K., Wang Y., Yang L., Li Z. The taste of pickled bamboo shoots: Evolution of metabolites based on electronic tongue, sensory evaluation, and non-targeted metabolomics. LWT- Food Science and Technology. 2025;223 doi: 10.1016/j.lwt.2025.117708. [DOI] [Google Scholar]
  32. Wei J., Li Z., Kuang J., Yan Z., Wang L., Lin Y., Du J., Li K., Wang Y., Yang L. Microbial community succession and changes of volatile compounds in the fermentation process of bamboo shoots. Food Microbiology. 2024;124 doi: 10.1016/j.fm.2024.104618. [DOI] [PubMed] [Google Scholar]
  33. Wu Y., You Y., Wu L., Du M., Ibrahim A., Suo H., Zhang F., Zheng J. Integrated metagenomics and metatranscriptomics analyses reveal the impacts of different Lactiplantibacillus plantarum strains on microbial communities and metabolic profiles in pickled bamboo shoots. Food Chemistry. 2025;464 doi: 10.1016/j.foodchem.2024.141772. [DOI] [PubMed] [Google Scholar]
  34. Xiao M., Peng Z., Hardie W.J., Huang T., Liu Z., Zhang Y.…Xiong T. Exploring the typical flavours formation by combined with metatranscriptomics and metabolomics during Chinese Sichuan paocai fermentation. LWT- Food Science and Technology. 2022;153 doi: 10.1016/j.lwt.2021.112474. [DOI] [Google Scholar]
  35. Xiao Y., Wang X., Song Z., Zhao X., Liu Z., Chu C., Wang Y., Hu X., Yi J. Enhancing phenylethyl alcohol production in Pichia fermentans via adaptive laboratory evolution under NaCl stress. Food Bioscience. 2024;62 doi: 10.1016/j.fbio.2024.105094. [DOI] [Google Scholar]
  36. Xiao Y., Zhang S., Liu Z., Wang T., Cai S., Chu C., Hu X., Yi J. Effect of inoculating Pichia spp. starters on flavor formation of fermented chili pepper: Metabolomics and genomics approaches. Food Research International. 2023;173 doi: 10.1016/j.foodres.2023.113397. [DOI] [PubMed] [Google Scholar]
  37. Xiong S., Xu X., Du T., Liu Q., Huang T., Ren H., Xiong T., Xie M. Organic acids drove the microbiota succession and consequently altered the flavor quality of Laotan Suancai across fermentation rounds: Insights from the microbiome and metabolome. Food Chemistry. 2024;450 doi: 10.1016/j.foodchem.2024.139335. [DOI] [PubMed] [Google Scholar]
  38. Xue B., You Y., Du M., Ibrahim A., Suo H., Zhang F., Zheng J. Metagenomic analysis of Lactobacillus plantarum DACN768 inoculation effects on volatile flavor compounds, microbial succession, and flavor metabolic network in suansun. Food Research International. 2025;199 doi: 10.1016/j.foodres.2024.115382. [DOI] [PubMed] [Google Scholar]
  39. Yang Y., Song X., Wang H., Liu C., Yang Y., Wang G., Xiong Z., Zhai Q., Ai L. Transcriptomics of Streptococcus thermophilus S-3: Patterns of nutrient requirements and metabolic responses under high-density fermentation conditions. Food Bioscience. 2025;69 doi: 10.1016/j.fbio.2025.107001. [DOI] [Google Scholar]
  40. Ye Z., Shang Z., Zhang S., Li M., Zhang X., Ren H., Hu X., Yi J. Dynamic analysis of flavor properties and microbial communities in Chinese pickled chili pepper (Capsicum frutescens L.): A typical industrial-scale natural fermentation process. Food Research International. 2022;153 doi: 10.1016/j.foodres.2022.110952. [DOI] [PubMed] [Google Scholar]
  41. Zhang S., Bao X., Xiao Y., Wang T., Chu C., Hu X., Yi J. Metagenomics reveals microbial communities and functional differences during chili pepper (Capsicum frutescens L.) natural fermentation: Effects of brines and containers. Food Frontiers. 2024;5(2):753–770. doi: 10.1002/fft2.355. [DOI] [Google Scholar]
  42. Zhang S., Shang Z., Liu Z., Hu X., Yi J. Flavor production in fermented chayote inoculated with lactic acid bacteria strains: Genomics and metabolomics based analysis. Food Research International. 2023;163 doi: 10.1016/j.foodres.2022.112224. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Fig. S1

Changes in the content of organic acids during the fermentation process of the uninoculated and inoculated suansun.

mmc1.pdf (126.7KB, pdf)

Data Availability Statement

Data will be made available on request.

Articles from Food Chemistry: X are provided here courtesy of Elsevier

RESOURCES