Enhanced Acid Tolerance in Bifidobacterium longum by Adaptive Evolution: Comparison of the Genes between the Acid-Resistant Variant and Wild-Type Strain

Abstract

Acid stress can affect the viability of probiotics, especially Bifidobacterium. This study aimed to improve the acid tolerance of Bifidobacterium longum BBMN68 using adaptive evolution. The stress response, and genomic differences of the parental strain and the variant strain were compared by acid stress. The highest acid-resistant mutant strain (BBMN68m) was isolated from more than 100 asexual lines, which were adaptive to the acid stress for 10^th, 20^th, 30^th, 40^th, and 50^th repeats, respectively. The variant strain showed a significant increase in acid tolerance under conditions of pH 2.5 for 2 h (from 7.92 to 4.44 log CFU/ml) compared with the wild-type strain (WT, from 7.87 to 0 log CFU/ml). The surface of the variant strain was also smoother. Comparative whole-genome analysis showed that the galactosyl transferase D gene (cpsD, bbmn68_1012), a key gene involved in exopolysaccharide (EPS) synthesis, was altered by two nucleotides in the mutant, causing alteration in amino acids, pI (from 8.94 to 9.19), and predicted protein structure. Meanwhile, cpsD expression and EPS production were also reduced in the variant strain (p < 0.05) compared with WT, and the exogenous WT-EPS in the variant strain reduced its acid-resistant ability. These results suggested EPS was related to acid responses of BBMN68.

Introduction

Probiotics are commonly recommended for health promoting effects on humans and other wide applications [13,20,24,28], and should be active to fulfill their physiological roles in the gut [20]. However, probiotics suffer from reduced viability because of harsh environments, including food manufacture, storage, and the gastrointestinal tract [13,20,24,28]. These stresses include extremes in temperature, osmosis, acid, and oxidation [20]. Bifidobacterium, one kind of probiotic, is weak in stress resistance [13,20,24,28]. Acid stress results in severe damage to Bifidobacterium species, as these organisms cannot grow under conditions of pH < 4.0 [13,29]. Therefore, it is vital to strengthen the acid tolerance of Bifidobacterium to increase its survival rate in probiotic products and in the gastrointestinal tract.

Adaptation to selection pressures is a convenient and effective means to achieve applicable microorganisms [18]. One previous study selected 115 Escherichia coli mutations as targets for selecting high temperature tolerance cell lines by adaptive evolution [26]. Another evolution of an E. coli MG1655 derivative led to increased tolerance to octanoic acid, and produced carboxylic acids at a 5-fold higher titer than its parent strain [23]. Adaptive evolution is also applied to probiotic, Zhang et al. [33] exhibited an evolved mutant (lb-2) of Lactobacillus casei, which has advantages in biomass and concentrations of lactate and acetate. Bifidobacterium longum biotype longum 8809dpH, isolated by low pH pressure, showed enhanced acid resistance compared with the wild-type strain [24].

B. longum subsp. longum BBMN68 is a new strain isolated from the feces of a centenarian living in Bama, Guangxi, China [35]. Strain BBMN68 plays vital roles in maintaining digestive function, improving constipation, and increasing the immunologic function of the host, per studies carried out in mice [9,13,31]. This study aimed to improve the acid resistance of BBMN68 through adaptive evolution. The stable resistant mutant of BBMN68 was screened under acid stress. The differences in the phenotype and genome between the WT and variant strains were compared to determine the acid tolerance mechanisms.

 

Materials and Methods

Bacterial Strains, Medium, and Growth Conditions

Bifidobacterium longum subsp. longum BBMN68 [9] was maintained at –80℃ in skimmed milk containing 30% glycerol. All the bacterial strains were cultured in modified de Man-Rogosa-Sharpe medium supplemented with 0.5 g/l ʟ-cysteine (MRSC) at an initial pH of 6.5 at 37℃, under anaerobic conditions of the anaerobic glass tube in 100% N2.

Isolation of Acid-Resistant Mutants

To achieve adaptive evolution of BBMN68, cells from an overnight BBMN68 strain, previously cultured in standard conditions, were washed in phosphate-buffered saline and used to inoculate (1%, 0.1 ml in a total volume of 10 ml) in fresh modified MRSC adjusted to pH 2.5 with 1 N HCl. These cultures were incubated at 37℃ for 16 h. The cells were recovered in liquid medium at the initial pH of 6.5 for 24-48 h until the OD600 was 1.0. Then the above two steps was repeated 50 times, and colonies were isolated from the 10th, 20th, 30th, 40th, and 50th repeats.

Then aliquots of the isolated acid-resistant mutants were plated on MRSC agar at pH 6.5 and incubated in a 2.5 L anaerobic jar (Oxoid, UK) with one Anero Pack-Anaero (Mitsubishi, Japan) at 37℃ for 48 h to screen 20 single clones. The sachet of Anero Pack-Anaero absorbs O2 in the jar, generates about 16% of CO2 or more simultaneously, and produces an anaerobic atmosphere (the concentration of O2 will be zero within 2 h). More than 100 clones were tested for the stability of the acid resistance at low pH (3.0 and 2.5) for 2 h, and the highest acid resistance strain was isolated.

To test the stability of the acid resistance phenotype of potential mutants, the survival ability was determined at low pH (3.0 and 2.5) after daily cultivation in MRSC for 1,000 generations (6.25 generations per 12 h) [17]. Bacterial growth was monitored by measuring the optical density at 600 nm (OD600) using a UV-visible spectrophotometer (Agilent, USA), and the pH was measured in the culture supernatant using a standard pH electrode. For identification of the Bifidobacterium mutant strain, primers F-CGATGCAACGCGAAGAACC and R-AATCCGGCTGGCAAC ACG, specific for the 16S rRNA gene sequence, were used [29].

Acid Tolerance Test

To test the acid resistance of the Bifidobacterium strains, both strains were grown in MRSC liquid medium at an initial pH of 6.5 at 37℃, in anaerobic glass tubes in 100% N2 for 14 h. The inocula were then diluted with fresh liquid MRSC medium at pH values of 6.5, 3.0, or 2.5 until the OD600 was 0.1 (in a total volume of 10 ml per glass tube). Anaerobic growth of these inoculated cultures was monitored f or 2 h at 3 7℃. T he s urviving c ultures were counted by plate count on MRSC agar medium.

Microscopy

For scanning electron microscopy (SEM) [12], bacteria cultured for 14 h were centrifuged at 10,000 ×g for 1 min. The precipitate was fixed with 2.5% (v/v) glutaraldehyde (pH 7.2) for 2-3h after removing the supernatant, and precipitated at 4℃ for 12 h. After washing three times in phosphate buffer, the samples were dehydrated in a dilution series of ethanol solutions (10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, and 3× 100%).

For transmission electron microscopy (TEM) [6], cells were fixed in 2.5% glutaraldehyde and negatively stained with uranyl acetate. The samples were rapidly frozen for ultrastructural analysis with acetone containing 0.1% tannic acid and 0.5% glutaraldehyde, and sequentially with acetone containing 1% osmium tetroxide and 0.1% uranyl actetate, and the sections were then embedded in resin at room temperature. Images were taken on an H-7650B 80-kV transmission electron microscope.

Genome Sequencing and Comparison

Total DNA was extracted using a bacterial genomic DNA kit (Tiangen, China) according to the manufacturer’s instruction. Sequencing was performed at Yuanquanyike Co. Ltd., China, using an Illumina Hiseq2000, using the 35 bp paired-end approach according to standard Illumina protocols. Of the more than 230 million 35-bp-long reads generated for BBMN68m, 97.67% were successfully mapped to the BBMN68 (GenBank Accession No. NC_014656.1) complete genome sequence using Maq software (default parameters). This resulted in 100× genome coverage. Single nucleotide polymorphisms (SNPs) were then identified and filtered using the pileup command for the SNP filter [19].

Computational Three-Dimensional (3D) CpsD Protein Structure Prediction

The 3D structures of the cpsD proteins from the mutant and wild-type strains were predicted by protein folding recognition with Phyre ver. 2.0 (http://www.sbg.bio.ic.ac.uk/phyre2/html) [8,16]. The protein images (.pdb format) were visualized by RasMol ver. 2.2.

CpsD Gene Expression by Quantitative PCR

Total cellular RNA was isolated from the acid-stressed cells (pre-incubated at pH 6.5 for 12 h before challenge at pH 2.5 for 2 h), in pH 5.0 at 48 h, and cells in pH 6.5 at 12 h and 48 h, using Trizol. The cDNAs were synthesized using the Prime Script RT reagent Kit (TaKaRa, Japan). In brief, first, gDNA was erased with 2 μl of gDNA Eraser buffer, 1 μl of gDNA Eraser, 1 μg of RNA, and RNase free water up to 10 μl, for 2 min at 42℃. Second, 4 μl of buffer 2, 1 μl of reverse transcriptase Mix, and RNase free water up to 20 μl were added and incubated at 37℃ for 15 min and then 85℃ for 15 sec.

The PCR primers were designed using Primer-premier 5 software. The primer pairs were 16S rRNA: F –CGATGCAACGCGAAGAACC, R-AATCCGGCTGGCAACACG; and BBMN68_1012: F–CATCAA GTCGGCGTCTATCA, R-CCATCCAGATGCGACCAT. The amplification efficiency of each primer set was determined using the standard curve. Each qPCR mixture with a total volume of 20 μl contained 10 μl of SYBR Green Real-time PCR Master Mix (Tiangen, China), 0.5 μl of the forward and reverse primers, 2 μl of cDNA template, and 7 μl of nuclease-free water. The qPCR reaction was performed using a Light Cycler 96 cycler (Roche, Switzerland). The amplification program was optimized and consisted of 35 cycles of 94℃ for 30 sec, 60℃ for 20 sec, 72℃ for 20 sec, and final extension at 72℃ for 5 min. Determination of expression ratios was performed in duplicates for all the six biological replicates.

Exopolysaccharide Isolation and Content Determination

Fresh MRSC medium was inoculated (2%) with overnight culture in MRSC medium and incubated at 37℃, in initial pH of 6.5 for 6, 12, 24, 48, and 72 h, and acidic condition of initial pH 5.0 for 48 h. The bacteria and supernatant were then separated by centrifugation (6,000 ×g for 20min), and the supernatant was precipitated at 4℃ for 48 h by adding 3 volumes of chilled absolute ethanol. The sedimentation was collected by centrifugation (10,000 ×g, 10 min, 4℃), dissolved in distilled water, and resuspended in 3 volumes of chilled absolute ethanol (repeated 3 times). The precipitate was then dissolved in distilled water as the sample. The samples were dialyzed for 72 h through a 3.5 kDa Visking dialysis membrane at 4℃, with renewal of water per 12 h. The exopolysaccharide (EPS) concentration was determined using the phenol-sulfuric acid colorimetric method.

Statistical Analysis

Statistical analyses were performed using SPSS 16.0 (Chicago, USA). Comparisons were done using one-way analysis of variance, followed by the least significant differences comparison test (set at 5%).

 

Results

Isolating Acid-Resistant Variant Strain of BBMN68

More than 100 asexual BBMN68 lines were adapted to acid stress for 10, 20, 30, 40, and 50 times in MRSC under acid condition (pH 3.0 and 2.5). One colony showed a significantly higher survival rate than the WT at pH 2.5 after 2 h of incubation, which was designated as BBMN68m. No viable cells (< 1.3 log CFU/ml) were detectable in the WT strain. The mutant strain showed a significant increase in survival rate after incubation at pH 2.5 for 2 h (from 7.92 to 4.44 log CFU/ml) and decreased for 1 log CFU/ml when incubated at pH 3.0 (from 7.76 to 6.62 log CFU/ml; Fig. 1A). No significant variation was observed in the survival rate after 1,000 generations of BBMN68m (p > 0.05), indicating the stability of acid resistance of the BBMN68m strain (data not shown).

Fig. 1.The survival cells of bifidobacteria after acid stresses and the growth curve in pH 6.5. (A) The survival cells of the WT strain and low pH-resistant mutant strain. (B) Growth curve of WT and mutant strains. The live bacterial cell counts (WT, —; mutant, —) and pH value (WT, ---; mutant, ---).

There was no difference in the survival rate in liquid culture between the two strains (Fig. 1B). Both strains exhibited a high growth rate after incubation for 8 h and then entered the stationary phase of growth at 14 h, with cell counts of approximately 9.62 (mutant) and 9.41 (WT) log CFU/ml. The final pH values of the two strains also showed a difference of approximately 0.25 units (WT, 4.35, and mutant, 4.10).

The Variant Showed an Altered Surface Structure

After cultured for 14 h, the WT strain maintained a more even distribution throughout the suspension, whereas in the mutant strain there was a tendency for a proportion of the cells to settle (Fig. 2A). Viewed by SEM, the two strains seemed similar but the WT showed thread-like and more small lumps on the surface, whereas the mutant variant was smoother and slender (Fig. 2B). In addition, WT cells appeared to have a more furcation shape than those of the mutant strain.

Fig. 2.Different phenotypes of the WT and mutant strains. (A) The WT and mutant strains over 14 h. (B) Scanning electron microscopy images of WT and mutant cells. The WT cells showed thread-like structures (red circle) and several lumps (blue circle) on the surface, whereas the mutant cells were smoother and more slender.

Gene Differences Between the Mutant Strain and Wild-Type Strain

To determine the genetic changes present in the mutant strain, we sequenced the genome of BBMN68m. Genome comparison identified 19 SNPs in the mutant strain when compared with the WT strain (data not shown). Six genes showed changes in amino acid sequences (bbmn68_286, bbmn68_375, bbmn68_825, bbmn68_1012, bbmn68_1211, and bbmn68_1462), of which three showed positional changes of amino acid within the predicted protein functional domains (bbmn68_286, bbmn68_375, and bbmn68_1012; Table 1). A key gene involved in the synthesis of EPS in Bifidobacterium, bbmn68_1012 (galactosyl transferase, CpsD), changed two nucleotides in the mutant strain, which resulted in amino acid sequence changes (Q164 to R164 and E210 to K210), and a pI shift of predicting protein (from 8.94 to 9.19; Table 1). These SNPs were confirmed by conventional PCR using five different primer pairs (Table 2), and all the results were consistent with the complete genomic sequence.

Table 1.Differential SNPs in the BBMN68m genome with respect to strain BBMN68.

Table 2.Conventional PCR primers of genes containing SNPs.

Computational 3D CpsD Protein Structure Prediction

The overall architecture of the mutant and the wild-type cpsD protein fold contains two abutting Rossmann-like folds (β/α/β domains) (Fig. 3) [8,16]. The N-terminal part is a typical nucleotide-binding domain with Asp-S-Asp (referred to as a DxD motif) in the carboxylates [16]. In contrast, the two folds in the mutant and wild-type cpsD proteins were found to have considerably different structures. The first α-helix in the N-terminal of the wild-type cpsD protein extended from the catalytic domain, but the mutant one showed a backward turn (Fig. 3). The active sites of the metal-binding DxD motif in the wild-type protein formed an α-helix, whereas the mutant showed a coiled structure.

Fig. 3.The 3D predicted structure of CpsD proteins. The pink and yellow parts are the α-helix and β-sheet of the N-terminal, respectively, relative to the donor substrate. The green structure is the conserved side chain carboxylate (C-terminal), and the hydrogen bonds of the nucleophilic hydroxyl of the accepting sugar molecule [15]. D199xD201 may be the active sites of metal binding.

Effects of Acid on cpsD Expression

To address whether the acid impacts the cpsD expression, the change folds of mRNA expression in both strains to the acid stress were investigated induction for 2 h at low pH (2.5) after being incubated under the initial pH of 5.0 for 48 h (Fig. 4). There was no difference in expression in both strains, incubated under initial pH 6.5 for 14 h, and followed by treatment induced for 2 h at pH 2.5. This correlated with the RNA-Seq result, of which cpsD expression did not change after pH 3.5 treatment for 2 h [13]. Gene cpsD exhibited expression changes in both wild-type and mutant strains when under acid stress of pH 5.0. The mRNA expression of cpsD was induced in the wild-type strain (3.9 fold), and suppressed in the mutant strain (-2.7 fold).

Fig. 4.The change folds of cpsD gene expression in both WT and mutant strains.

Different EPS Contents in the Mutant Strain and Wild-Type Strain

The EPS production properties of both strains were investigated, as shown in Fig. 5. The EPS production gradually increased with bacterial growth. The amount of EPS produced in the culture peaked in the late stationary phase of growth (48 h), and the mutant strain produced less EPS than the wild-type strain (Fig. 5A). After 48 h culture, the mutant variant contained 191 mg/l EPS whereas the wild type contained 332 mg/l. However, the OD value in the mutant strain was higher. TEM showed that both strains produce an outer cell surface layer, presumed to be a capsule consisting of surface EPS [6]; the mutant cell had a smaller/slimmer capsule and the EPS content was less (Fig. 5B).

Fig. 5.Contents of the EPS WT and mutant strains. (A) EPS production by the WT and mutant strains. The EPS contents (WT, —; mutant, —) and OD600 (WT,---; mutant, ---) of both strains. Each value represents the average of triplicate measurements. (B) TEM images of WT and mutant cells. Scale bars, 500 nm.

Effects of Low pH on EPS Content

At initial pH of 5.0, compared with under pH 6.5, the EPS content showed a dramatic decline in both WT (105 mg/l) and mutant strains (90 mg/l; Fig. 6A), respectively. However, the growth of the WT was severely inhibited, as the OD600 was 0.85 in pH 5.0, lower than the 2.01 in pH 6.5 (Fig. 6B). There was no obvious variation in mutant strain (2.16 and 2.12). There was a difference between the pH values in the inocula of both strains at initial pH 5.0 (Fig. 6C). After 48 h, the pH values of both strains were 4.21 (WT) and 3.56 (mutant), respectively. It suggests the mutant strain with lower EPS content than WT had higher acid tolerance.

Fig. 6.The EPS production in the mutant and WT strains under acidic conditions. (A) EPS production by the WT and mutant strains under pH 5.0 and 6.5 at 48 h. (B) OD600 of both strains under pH 5.0 and 6.5 at 48 h. (C) The final pH values of the two strains under pH 5.0 and 6.5 at 48 h.

Exogenous Addition of WT-EPS on the Survival of Mutant Strain

To explore whether the accumulation of EPS under acid stress was related to acid tolerance, the affect of EPS from the wild-type strain (WT-EPS) addition on the survival of the mutant strain was tested. As shown in Fig. 7, addition of exogenous WT-EPS (80% purity) was found to reduce the number of viable mutant cells under acid stress. The survival rate of mutant strain decreased 1 unit of log CFU/ml in the presence of 300 and 500 mg/l (final concentration) WT-EPS at pH 2.5 for 2 h, respectively. However, addition of mutant-EPS could not change the density of WT cells in pH 2.5 for 2 h. Therefore, EPS production of both strains may have differences in acid stress response.

Fig. 7.Influence of exogenous EPS addition on viability of cells under acid stress. Mid-exponential phase cells were incubated with different concentrations of exogenous EPS in MRSC for 6 h. The cells were challenged at pH 2.5 for 2 h and CFU values were determined.

 

Discussion

The commercialized probiotics frequently suffer from low survival rate, for which one key limiting factor is the inherently low tolerance of stress during food manufacture, storage, and the gastrointestinal tract [20]. Adaptive evolution is the process of the microorganism’s adaptation to selection pressures in a controlled manner [7]. In this study, we isolation the highest acid-resistant mutant strain (BBMN68m) from more than 100 asexual B. longum subsp. longum BBMN68 lines, which were adapted to acid stress for the 10th, 20th, 30th, 40th, and 50th repeats, respectively.

Adaptive evolution has been proven to be effective in improving cellular tolerance against disadvantageous environments and the foundation is beneficial mutant genes [24]. Brennan et al. [2] reported that Saccharomyces cerevisiae had a 4-fold increased tolerance toward the biojet fuel blend AMJ-700t because of one gene mutant (tTcb3p). The mutant E72 from Bacillus subtlis168 exhibited higher growth rate and xylose consumption rate with three mutations (araR, sinR, and comP) [32]. In this study, comparative whole-genome analysis showed that BBMN68m contained 19 mutant genes, of which six genes represented changes in the amino acid sequences. One of the mutant genes participated in EPS synthesis and changed the content of the production, which may be the reason of affecting the acid tolerance in the strains.

EPS may be one critical element for the surroundings tolerance of Bifidobacterium. Previous studies indicated that EPS is produced as a response result to acid, osmotic, or oxidative stresses [1,22,27]. The EPS-1 from Bacillus amyloliquefaciens strain C-1 exhibited tolerance to oxidative stress, including strong reducing power, superoxide radicals, and hydroxyl free radicals scavenging activities [30]. Alp et al. [1] reported that Bifidobacterium strains showed a positive correlation between EPS production and resistance to bile salts and low pH. The bacterial EPS is often negatively charged, which appears to affect cell aggregation [14]. A Lactobacillus johnsonii FI9785 mutant, with a reduction in EPS content, exhibited a big increase of cell aggregation [12]. The cells with EPS formed an “aggregative” phenotype, which affected the stress resistance [14]. Vibrio cholerae O1 strain TSI-4 produced a kind of amorphous EPS that was more osmotic and showed higher oxidative stress resistance compared with the corresponding non-aggregation strains [27]. The aggregative Lactobacillus rhamnosus produced EPS, which indicated as a response to acid stress [22]. Pseudomonas aeruginosa with “aggregative” phenotype tolerated higher stress by making two kinds of EPSs (Pel and Psl) [4,5]. A wild-type Lactobacillus crispatus showed a cell aggregation phenotype and had improved gastrointestinal persistence compared with its spontaneous non-aggregating mutant [3]. In this study, the BBMN68m strain was more aggregative and was lower in EPS producing content compared with that of the wild-type strain (Fig. 2). Addition of exogenous WT-EPS reduced the number of viable mutant cells, but mutant-EPS could not affect the WT cell density under the acid stress (Fig. 7). Perhaps the high level of EPS originating from the WT strain itself caused its low aggregation, and the effect of exogenous mutant-EPS was limited. Moreover, there were differences in the main molecular weight and structure but not in monomer composition functional groups (data unpublished). Yang et al. [30] reported that two variant EPSs from B. amyloliquefaciens C-1 had different antioxidant activities and effect on oxidative stress. However, there were no reports that demonstrated the relationship between the structure, monomer composition, molecular weight, and acid tolerance. The results suggest the EPS may be a key influencing factor for the response to acid stress in BBMN68 and BBMN68m strains.

The cpsD gene mutation causes the EPS change in the BBMN68m strain, which may play a key role for acid stress tolerance. Fourteen of the 48 sequenced Bifidobacterium contained genes related to EPS biosynthesis and were found to be relevant to EPS function [10]. The EPS gene clusters harbor one putative priming GTF gene (bbmn68_1012, p-gtf) annotated as CpsD “galactosyl transferase” (cpsD), which coded a 502 amino acid protein. CpsD is an enzyme that catalyzes the initial step of the EPS-unit synthesis and should be present in all eps-clusters [10]. In Streptococcus pneumoniae, deletion of the gene cpsD will disable capsular polysaccharide biosynthesis [11,21]. BBMN68 CpsD and homologous proteins had <75% identity with B. longum genomes, and only B. longum subsp. longum CECT 7347 (WP_022527352.1) showed 100% identity with it. The 3D structure of the CpsD proteins were diverse. The metal-binding DxD motif of wild-type protein formed an α-helix, whereas the mutant one showed a coiled structure (Fig. 3). The function of the metal-binding DxD motif appeared to be for the coordination of the divalent cation, most commonly Mn2+, which is essential for the binding of the nucleotide sugar-donor substrate [8,16]. A slight shift in the DxD motif, the active-site positioning of the cation, could undergo attack by an incoming nucleotide sugar, leading to a product with retained stereochemistry [16]. Therefore, the two proteins may have different functions for binding the sugar substrate, as the content or structure of EPS was changed [6,11,12,16]. The cpsD expression was also regulated by acid. In E. coli, the transcription of the regulators diguanylate cyclase and phosphodiesterase produces and degrades the second messenger c-di-GMP, respectively, which were also translational control components in EPS synthesis [15,25]. In eps-clusters of Bifidobacterium, there was unawareness of the transcriptional regulators of EPS synthesis. However, the cpsD expression was suppressed in BBMN68m, with lower EPS content and higher acid tolerance at initial pH 5.0 (Fig. 4), which may be a response to acid stress. Thereby, the cpsD gene mutant was the important cause of the EPS change in the BBMN68m strain, and may be the key to acid stress tolerance.

In conclusion, an acid-resistant mutant strain (BBMN68m) of B. longum subsp. longum BBMN68 was obtained using the adaptive evolution method. The EPS change may be the critical factor for the acid tolerance of BBMN68m, which is caused by the cpsD gene. Taken together, the results presented here indicate that we not only obtained an acid-resistant strain for the food industry, but also contributed to revealing the possible mechanisms of acid tolerance of Bifidobacterium.