Characterization of the Potential Probiotic Vibrio sp. V33 Antagonizing Vibrio Splendidus Based on Iron Competition.

BACKGROUND: Vibrio Splendidus Vs is an important aquaculture pathogen that can infect a broad host of marine organisms. In our previous study, an antagonistic bacterium Vibrio sp. V33 that possessed inhibitory effects on the growth and virulence of a pathogenic isolate V. splendidus Vs was identified.
OBJECTIVES: Here, we further explored the antagonistic substances and antagonistic effects from the viewpoint of iron competition.
MATERIALS AND METHODS: The main antagonistic substances in the supernatants from Vibrio sp. V33 were identified using the bioassay-guided method. The response of V. splendidus Vs under the challenge of cell-free supernatant from Vibrio sp. V33 was determined via sodium dodecyl sulfate-polyacrylamide gel electrophoresis and real-time reverse-transcription PCR.
RESULTS: The main antagonistic substances produced by Vibrio sp. V33 have low molecular weights, are water soluble, and are heat-stable substances. Meanwhile, the iron uptake rate of Vibrio sp. V33 was higher than that of V. splendidus Vs. In the presence of cell-free supernatant from Vibrio sp. V33, expressions of two functional genes, viuB and asbJ related to ferric uptake processes in V. splendidus Vs, were up-regulated, whereas furVs coding the ferric uptake repressor was suppressed below 0.5-fold. One gene coding phosphopyruvate hydratase does not change at mRNA level, but was up-regulated at protein level.
CONCLUSIONS: Our results suggested that antagonistic effect of Vibrio sp. V33 on the pathogenic isolate V. splendidus Vs was partially due to the stronger ability of Vibrio sp. V33 to seize iron. This cell-free supernatant from Vibrio sp. V33 created an iron-limited milieu for V. splendidus Vs, which led to the changed expression profiles of genes that were related to iron uptake in V. splendidus Vs. Copyright:

1. Background

Along with the rapid expansion of aquaculture, various bacterial infections have increasingly emerged and hindered healthy aquaculture development ( 1 ). Usually, antibiotics are used to treat bacterial disease, but more and more study showed that probiotics, such as Aeromonas sp., Agarivorans sp., Bacillus sp. and Vibrio sp. have increasingly played their important roles in protection of cultured animals from pathogenic bacterial infection ( 2 , 3 ).

Iron is essential for many key enzymes, particularly those involved in citric acid cycle, glycolysis, and oxidative phosphorylation ( 4 , 5 ). Therefore, it is required by nearly all known organisms for survival and growth. Siderophores are specific ferric ion chelators that were composed of low-molecular-weight molecules, and have high affinity to iron ( 6 ). It was ubiquitously secreted by bacteria, including Gram-negative Escherichia coli, Salmonella typhimurium, Pseudomonas sp. and Vibrio sp., and Gram-positive Staphylococcus sp. and Bacillus sp. ( 6 ), leading to competition for iron between different kinds of bacteria ( 7 ). Antagonistic effects of siderophore-producing bacteria when they were exposed to iron-depleted conditions were reported previously ( 8 , 9 ). Bacillus cereus inhibited the growth of a fish pathogen Aeromonas hydrophila partly by competing for iron through siderophore release ( 10 ). Similarly, seven strains of Streptomyces sp. isolated from the marine sediments of a shrimp farm were found to inhibit the growth of Vibrio sp. in vitro by producing siderophores ( 11 ).

Ferric uptake regulators are conserved iron-uptake-related proteins that regulate siderophore biosynthesis and corresponding receptors in most prokaryotic organisms, and they function as regulators when bound with Fe2+ ( 12 ). The expression of ferric-uptake-related genes in bacteria was changed with varied iron levels. The mRNA expression of the gene coding furVs is down-regulated in the presence of iron chelator 8-hydroxyquinoline ( 13 ). However, the expression of other iron-uptake-related genes in V. splendidus under iron-limited conditions remains largely unknown.

2. Objective

In our previous study, the antagonistic bacterium Vibrio sp. V33 was identified and found to inhibit the growth and virulence of a pathogenic bacterium V. splendidus ( 14 ). In the present study, the main antagonistic substances in the supernatant from Vibrio sp. V33 were characterized, and differential expressions of iron-uptake-related genes in V. splendidus Vs at both mRNA and protein levels in the presence of supernatant containing antagonistic substances were determined.

3. Materials and Methods

3.1. Bacteria, Culture Media

Vibrio sp. V33 was isolated from healthy cuttlefish, Sepia pharaonis in our previous study ( 14 ). V. splendidus Vs was isolated from Apostichopus japonicus suffering from skin ulceration syndrome in the indoor farms of the Jinzhou Hatchery, and its pathogenicity was determined in our previous studies ( 15 ). V. splendidus Vs and Vibrio sp. V33 were cultured at 28 °C in modified 2216E media consisting of 5 g.L-1 of tryptone and 1 g.L-1 of yeast extract in aged seawater. Unless otherwise stated, all chemicals used in this study were purchased from Sangon (Shanghai, China).

3.2. Growth Inhibition Assay

Vibrio sp. V33 was grown to OD600 of 0.7, 0.8, or 0.9, measured using a UV-Vis spectrophotometer (Beckman, USA). Then, supernatants were collected by centrifuged at 12000 × g and 4 °C for 5 min and they were filtered through 0.22-μm polycarbonate membrane filters (Millipore) to prepare the cell-free supernatant. The same volumes of different cell-free supernatants were separately added into cell pellet of V. splendidus Vs to make cell suspensions of the same concentrations. V. splendidus Vs suspensions in fresh media were used as a control. After incubated for 12 h, OD600 of each culture was measured using a UV-Vis spectrophotometer.

3.3. Characterization of Antagonistic Substances

Water solubility and organic solubility of the antagonistic substances were determined using the method developed by Jorquera et al. ( 16 ). Briefly, 500 mL cell-free supernatant of Vibrio sp. V33 was extracted with the same volumes of ethyl acetate and then evaporated at 42 °C to concentrate 100-fold. Then, 5 μL concentrated solution and the un-concentrated supernatant were separately added to the V. splendidus Vs suspensions. After incubated for 8 h, OD600 of each culture was measured using a UV-Vis spectrophotometer.

The molecular weight range of the antagonistic substances was estimated using the method of Chythanya et al. ( 17 ). Briefly, the cell-free supernatant from Vibrio sp. V33 was centrifuged using a 3-kDa molecular weight cut-off (MWCO) tubes at 6000 × g for 30 min at 4 °C. Two fractions (>3 kDa residue and <3 kDa filtrate) were collected. However, because the centrifuge tubes can only concentrate molecules with molecular mass >3 kDa, and the small molecular substances (<3 kDa) remained in the >3 kDa residue fraction. Moreover, the <3 kDa substances in both the residue filtrate fractions were the same. The same volume of the two fractions was added into V. splendidus Vs. Fresh media was also added into V. splendidus Vs and used as a control. After incubated for 8 h, OD600 of each culture was measured using a UV-Vis spectrophotometer. The inhibitory activity (I.A.) was measured as follows:

, where (a) is the treatment and (b) is the control.

3.4. Iron Uptake Rates Measurement

Ion uptake rates of Vibrio sp. V33 and V. splendidus Vs were measured using the method described by Lalloo et al. ( 18 ). Subsequently, 500 μL Vibrio sp. V33 or V. splendidus Vs were separately inoculated into 2216E media containing 0.01 g.L-1 FePO4 and each culture was grown at 28 °C. Samples were obtained every 12 h and centrifuged to collect supernatants for the remaining iron contents. Iron concentration was determined using a serum iron assay kit purchased from Nanjing Jiancheng Biochemistry (China).

3.5. Identification of differentially expressed proteins by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) combined with MALDI-TOF/TOF mass spectrometry

V. splendidus Vs was grown in 2216E media to an OD600 of approximately 0.2. Then, 1 mL cell-free supernatant of Vibrio sp. V33 was added to 3 mL of V. splendidus Vs culture. Fresh media was also added into V. splendidus Vs culture and was used as a control. After grown for another 2 h, the culture was centrifuged at 8000 × g for 10 min. The cell pellets were collected and suspended in cell lysate. The distinct bands between the sample treated with and without cell-free supernatant of Vibrio sp. V33 on the gel were collected and subjected to MALDI-TOF/TOF MS for protein identification. MS was performed on an ABI 5800 MALDI-TOF/TOF Plus mass spectrometer (Applied Biosystems, Foster City, USA). Data were acquired through a positive MS reflector by using a CalMix5 standard for the calibration of the instrument (ABI5800 calibration mixture).

3.6. Real-Time Reverse-Transcription Pcr (Real-Time Rt-Pcr)

V. splendidus Vs was grown in 2216E media to an OD600 of approximately 0.2. Then, 10 mL <3 kDa filtrate from Vibrio sp. V33 supernatant (described above) was added to 40 mL culture of V. splendidus Vs. The same volume of fresh media was added to another aliquot of V. splendidus Vs and was used as a control. After inoculated for 10, 20, and 30 min, 2 mL cell pellet was collected. RNA was extracted from cells collected at different time points using a Bacterial RNA Isolation kit (Omega) and then treated with RNase-free DNase. cDNA generated from 1 µg of DNase-treated RNA with PrimeScript reverse transcriptase (Takara) was used for real-time RT-PCR. Each assay was performed in triplicate with the 16S rRNA as internal control. The primers shown in Table 1 were designed on the basis of our genomic sequence of V. splendidus Vs. SYBR Premix Ex Taq (Takara) was used for the real-time RT-PCR reactions in an ABI 7500 real-time detection system (Applied Biosystems) as described previously ( 13 ). Dissociation analysis was performed on the amplification products at the end of each PCR to make sure that only one PCR product was amplified and detected. The comparative threshold cycle method (2−ΔΔCT method) was used for the analysis of the mRNA levels. The expression of fur gene (the ferric uptake regulator, Supporting Information Fig. S1), asbJ gene (siderophore ABC transporter, Supporting Information Fig. S2) and viuB gene (one siderophore-interacting protein, Supporting Information Fig. S3) from V. splendidus Vs was compared with that of expression in the cells before treatment, which was used as 100% expression.

Table 1

Primers used in this study.

Primer name	Nucleotide sequence (5ʹ→3ʹ)	Products
RTF3	ACAACAACCAGACTGCCAACA	223 bp
RTR3	GATAACTTCACCGCAGTCTAAACAT
ASBJF	GCGTTATTCATCGGGCAGGG	168 bp
ASBJR	GCCCAGGATGGATGGGGTTAG
ViuBF	CAGAATGTGCTGGCGGCTAC	216 bp
ViuBR	CCACTTCTGCGTTCATTGCC
ENOF	AAGAAGGTCTGCGTATCGGC	205 bp
ENOR	GCACAGTCCATAGCAAGAGTAAC
933F	GCACAAGCGGTGGAGCATGTGG	300 bp
16SRTR1	CGTGTGTAGCCCTGGTCGTA

3.7 Strain No. of Bacterial Strains

The isolates of Vibrio sp. V33 and V. splendidus Vs were deposited into the China General Microbiological Culture Collection (CGMCC, Beijing, China) http://www.cgmcc.net/english/ ( 19 ) with strain No. 12561 and 7.242, respectively.

3.8 Database Search and Statistical Analysis

Statistical analyses were performed by using the two-tailed Student t-test. Statistical significance was determined by one-way ANOVA ( 20 ). In all cases, the significance level was defined as * P < 0.05 and ** P < 0.01. The promoter prediction was conducted as the method described previously ( 21 ). The Fur binding DNA sequence used was 5′-GATAATGAT(A/T)ATCATTATC-3′ ( 22 ).

4. Results

4.1. Inhibitory Effect of Cell-Free Supernatant from Vibrio sp. V33

All the three supernatants collected from Vibrio sp. V33 at different densities showed inhibitory effects on the growth of V. splendidus Vs. OD600 was reduced to approximately 41.2%, 21.1% and 50.8% when the added supernatants were collected from Vibrio sp. V33 at the cell densities of 7.0×108, 8.0×108, and 9.0×108 CFU.mL-1, respectively (Fig. 1). This result suggests that the Vibrio sp. V33 secreted the antagonistic substances into its extracellular milieu, which showed a strong inhibitory effect on growth of V. splendidus Vs.

Fig 1

Antagonistic activity of the cell-free culture supernatants from Vibrio sp. V33 collected at different cell densities on the growth of V. splendidus Vs. The growth of V. splendidus Vs without inoculation of supernatant from Vibrio sp. V33 was used as control. Data are the means for at least three independent experiments and are presented as the means±SE.

4.2. Characterization of Antagonistic Substances

The relative I.A. of the antagonistic substances in the un-concentrated water phase and 100-fold concentrated organic phase were approximately 0.5- and 0.4-fold of that of the untreated cell-free supernatant (Fig. 2A). Thus, the substances in the water phase showed stronger antagonistic effect on growth of V. splendidus Vs than that in the 100-fold concentrated substance in the ethyl acetate extract, suggesting that the main antagonistic component was in the water-soluble phase. Similarly, when the cell-free supernatant collected from Vibrio sp. V33 was departed using the cut-off method at 3 kDa, the relative I.A. of the substances in the <3 kDa filtrate (un-concentrated) and >3 kDa (50-fold concentrated) residue were approximately 0.4 and 0.25-fold, respectively (Fig. 2B). The result suggested that the <3 kDa filtrate substances showed a much stronger antagonistic effect on V. splendidus Vs than the >3 kDa residue that contained the concentrated >3 kDa plus the unconcentrated <3 kDa substances. Thus, the main antagonistic substances in the supernatant from Vibrio sp. V33 were water soluble substances with molecular mass of less than 3 kDa.

Fig 2

(A) I.A. ratio of the substances in the water phase and 100-fold concentrated organic phase. (B) I.A. ratio of the substances in the <3 kDa filtrate and >3 kDa residue.

4.3. Determination of Iron Uptake Rates of Vibrio sp. V33 and V. splendidus Vs

When Vibrio sp. V33 and V. splendidus Vs were grown for 12 h, the remaining iron concentrations in both cultures were around 2.09 mg.L-1 and no significant difference was observed (Fig. 3). Both the growth of Vibrio sp. V33 and V. splendidus Vs needed iron, because the remaining iron concentrations in the Vibrio sp. V33 and V. splendidus Vs cultures after cultured for 24 h were 1.79 and 1.91 mg.L-1, respectively. The iron uptake rate of V. splendidus Vs was slower than that of Vibrio sp. V33 by approximately 0.01 mg.L-1·h-1. The significant difference in iron content appeared after cultured to 36 h, at this time point the iron uptake rates of Vibrio sp. V33 and V. splendidus Vs were 0.039 and 0.014 mg.L-1·h-1, respectively. Vibrio sp. V33 possessed a higher iron uptake rate of 0.025 mg.L-1·h-1 than that of V. splendidus Vs. Our results suggested that Vibrio sp. V33 was stronger in the ability to compete for iron and caused an iron deprivation the environment for V. splendidus Vs.

Figure 3

Iron concentrations remained in the culture after growth of V. splendidus Vs and Vibrio sp. V33 at different time points. Data are the means for at least three independent experiments and are presented as the means±SE.

4.4. Analysis of Differentially Expressed Proteins

The differentially expressed proteins in V. splendidus Vs cell grown in the presence or absence of the cell-free supernatant of Vibrio sp. V33 were determined using SDS-PAGE. One distinctive band with an approximate molecular weight of 50 kDa was detected (Fig. 4B). The expression of this protein was up-regulated after treated for 2 h. The protein was identified based on the fragments obtained from MS and our genomic sequence of V. splendidus Vs, as well as its blast in NCBI. It showed a 99.8% sequence identity to a phosphopyruvate hydratase from V. splendidus (gi|490873392|ref|WP_004735393.1|; E=0) (Fig. 4A), except that F422 was substituted by Y422 in phosphopyruvate hydratase from V. splendidus Vs. The nucleotide sequence of its coding gene corresponds to the eno gene (Supporting Information Fig. S4). However, no significant difference was noted at the mRNA level between the cells treated with cell-free supernatant of Vibrio sp. V33 and the control sample (Fig. 4C), implying that the cell-free supernatant of Vibrio sp. V33 influenced the expression of phosphopyruvate hydratase at the protein level rather than at the mRNA level under the tested conditions.

Figure 4

(A) Protein sequences of phosphopyruvate hydratase from Vibrio sp. (B) SDS-PAGE analysis of the differentially expressed protein in the cells treated with and without cell-free supernatant form Vibrio sp. V33. Lane 1, the protein marker; lane 2, the proteins from the V. splendidus Vs without supernatant treatment (control); lane 3, the proteins from the V. splendidus Vs treated with cell-free supernatant collected from Vibrio sp. V33. (C) Expression of the eno gene in the presence of the cell-free culture supernatants from Vibrio sp. V33. Data are the means for at least three independent experiments and are presented as the means±SE.

4.5. Expression of Iron-Uptake-Related Genes in V. splendidus Vs

Expression of furVs was down-regulated to 0.06-, 0.48-, and 0.30-fold at the mRNA level after treatment with the cell-free supernatant from Vibrio sp. V33 for 10, 20 and 30 min, respectively (Fig. 5A). In V. splendidus Vs, there also existed an asbJ gene coding a protein with a 97% sequence identity to a siderophore ABC transporter from V. splendidus 12B01 (gi|490869579|ref|WP_004731594.1; E=0), and a viuB gene coding a protein with a 93% sequence identity to a siderophore-interacting protein viuB from Vibrio cyclitrophicus (gi|498114642|ref|WP_010428798.1; E=1.2416E-180). The mRNA level of asbJ in the cells treated with the cell-free supernatant from Vibrio sp. V33 was down-regulated to 0.51- and 0.6-fold at 10 and 20 min, respectively, but up-regulated to 1.69-fold at 30 min (Fig. 5B). Similarly, the mRNA level of viuB in cells treated with the cell-free supernatant from Vibrio sp. V33 was down-regulated to 0.91- and 0.65-fold after 10 and 20 min, respectively, but up-regulated to 1.55-fold after 30 min (Fig. 5C). Apparently, the core regulator in the iron uptake process, Fur, was the most affected by iron level and was down-regulated earlier, but the expressions of functional genes had a time lag to be up-regulated. These results suggest that the expression profiles of the iron-uptake-related genes were affected by the cell-free supernatant from Vibrio sp. V33. This observation strengthened our speculation that the cell-free supernatant from Vibrio sp. V33 may inhibit the growth of V. splendidus Vs through creating an iron deficient environment.

Figure 5

Expression of iron-homeostasis-related genes in the presence of the cell-free culture supernatants of Vibrio sp. V33. Data are the means for at least three independent experiments and are presented as the means±SE.

4.6. Sequence Analysis of Up-Stream of the Functional Genes

The promoter regions and transcription factor binding sites of the functional genes, viuB, , and eno were analyzed. The 1 kb upstream regions from the start codon ATG of the three functional genes were used for analyses. BPROM prediction suggested that the upstream regions of the three functional genes contained the typical promoter regions that include the −35 and −10 domains (Fig. S4). The sequence of the Fur binding site was also searched adjacent to the −35 and −10 domains. FurVs binding site was presented in both the promoters of P and P, but not in the promoter of P. This notion further indicated that the regulator FurVs may control the expression of viuB gene and asbJ gene, however, the expression regulation of eno may not occur at mRNA level.

5. Discussion

Competition for iron via siderophore piracy is an important antagonistic mechanism that is utilized by potential probiotic bacteria to inhibit various pathogenic infections ( 7 , 10 , 23 ). In our previous study, we found that the antagonistic bacterium Vibrio sp. V33 produced more siderophores than that by V. splendidus Vs ( 14 ). In the present study, we further pointed out that the main antagonistic substances of Vibrio sp. V33 were thermostable and water soluble, and had low molecular weights. Combined with the higher iron uptake rate of Vibrio sp. V33, we speculated that the inhibitory effect of Vibrio sp. V33 was similar to that of Vibrio sp. E, which inhibited a non-siderophore producing strain of V. splendidus, Vibrio P, also partly due to the high ability to compete for iron ( 24 ).

In the present study, the protein level of one phosphopyruvate hydratase (enolase) was up-regulated in V. splendidus Vs in the presence of the cell-free supernatant from Vibrio sp. V33. The involvement of phosphopyruvate hydratase in the iron uptake process was consistent with the up-regulation expression profiles of α-enolase in Bacteroides fragilis under iron-limited conditions ( 25 ). This finding is also supported by other reports that correlate dietary iron deficiency with the regulation of the glycolytic pathway ( 26 , 27 ). The increased enolase expression in glycolysis suggests the necessity for sufficient ATP production under iron-limited conditions ( 28 ).

A previous study suggested that competition for iron via siderophore piracy can affect the gene expression patterns during bacterial interactions ( 29 ). Fur affects siderophore production in many bacterial species and controls the expression of most iron-uptake-related functional genes, such as those codes the siderophore ATP-binding cassette (ABC) transporter asbJ and the siderophore-receptor viuB, which were involved in ferric Vibriobactin uptake or utilization ( 30 - 33 ). In the present study, the expressions of iron-related genes in V. splendidus were determined in the presence of antagonistic substrates. FurVs expression was reduced in the presence of the cell-free supernatant of Vibrio sp. V33, which was the same to the phenomena that observed under iron-limited conditions created by 8-hydroxyquinoline ( 13 ). This result further strengthened our speculation that Vibrio sp. V33 inhibited the pathogenic isolate V. splendidus Vs through creation of the iron-limited circumstance. Along with the reduced mRNA level of furVs, the up-regulated mRNA levels of viuB and asbJ and the presence of furVs binding sites in the two promoters suggested that the antagonistic effect of Vibrio sp. V33 may perform through iron competition. Such effect was similar to the probiotic influence of Pseudomonas fluorescens toward Vibrio anguillarum ( 34 ) and Aeromonas salmonicida ( 35 , 36 ) and that of Pseudomonas sp. and Psychrobacter sp. toward Vibrio anguillarum and A. salmonicida subsp. Salmonicida ( 36 , 37 ). In our present study, under iron-limited circumstance created by Vibrio sp. V33, the up-regulation of phosphopyruvate hydratase, a multifunctional protein contributing to glycolysis/gluconeogenesis and other biological pathophysiological processes, was consistent with that observed in Cryptococcus gattii ( 38 ). However, our study further highlighted that the up-regulation occurred at the protein level and not directly regulated at mRNA level by the most important iron uptake regulator Fur.

6. Conclusion

Vibrio sp. V33 has previously been identified to be an antagonistic bacterium of a pathogenic isolate V. splendidus Vs, but none of its antagonistic substances has been characterized. It was determined that the iron uptake rate of Vibrio sp. V33 was higher than that of V. splendidus Vs, which was also supported by the following points: on one hand, the active tracking method showed that the main antagonistic substances produced by Vibrio sp. V33 were of low molecular weights, water soluble, and heat-stable, which belonged to the characteristics of siderophores. On the other hand, the expressions of two functional genes, viuB and asbJ related to iron uptake processes in V. splendidus Vs were upregulated, which meant that the iron uptake pathway was involved in the interaction between Vibrio sp. V33 and V. splendidus Vs. All of the data indicated that competition for iron may be the main antagonistic process of the Vibrio sp. V33