Orally Administered Thermostable N-Acyl Homoserine Lactonase from Bacillus sp. Strain AI96 Attenuates Aeromonas hydrophila Infection in Zebrafish

Abstract

N-Acylated homoserine lactone (AHL) lactonases are capable of degrading signal molecules involved in bacterial quorum sensing and therefore represent a new approach to control bacterial infection. Here a gene responsible for the AHL lactonase activity of Bacillus sp. strain AI96, 753 bp in length, was cloned and then expressed in Escherichia coli. The deduced amino acid sequence of Bacillus sp. AI96 AiiA (AiiA_AI96) is most similar to those of other Bacillus sp. AHL lactonases (∼80% sequence identity) and was consequently categorized as a member of the metallo-β-lactamase superfamily. AiiA_AI96 maintains ∼100% of its activity at 10°C to 40°C at pH 8.0, and it is very stable at 70°C at pH 8.0 for at least 1 h; no other Bacillus AHL lactonase has been found to be stable under these conditions. AiiA_AI96 resists digestion by proteases and carp intestinal juice, and it has broad-spectrum substrate specificity. The supplementation of AiiA_AI96 into fish feed by oral administration significantly attenuated Aeromonas hydrophila infection in zebrafish. This is the first report of the oral administration of an AHL lactonase for the efficient control of A. hydrophila.

INTRODUCTION

Bacteria communicate with each other by quorum sensing, a mechanism that is dependent on their population density and which was first reported to be the trigger for the bioluminescence of the aquatic bacterium Vibrio fischeri (39). Quorum-sensing bacteria can release, detect, and respond to the accumulated small signal molecules and then regulate the expressions of target genes (21, 33). Quorum-sensing systems have been found in pathogenic bacteria of plants, animals, and humans (13, 53). Gram-negative bacteria use N-acylated homoserine lactones (AHLs) as signal molecules (13). AHLs vary in their acyl chain lengths (4 to 14 carbons) and/or contain an oxo or a hydroxyl substituent in the acyl chain (21, 33). Many of the infection-related phenotypes associated with quorum sensing are controlled by AHL/LuxR/LuxI systems: LuxR is the AHL receptor, and LuxI is the AHL synthase. The quorum-sensing systems can induce antibiotic production; plasmid conjugation; nodulation; biocorrosion; biofilm formation; and the expression of virulence factors, e.g., proteases, lytic enzymes, toxins, siderophores, and adhesion molecules (1, 8, 13, 33, 43, 53).

Aeromonas hydrophila is a primary, secondary, and opportunistic Gram-negative bacterial pathogen (2). It is found mainly in water and water-related environments and causes a wide variety of symptoms in fish, including tissue swelling, necrosis, ulceration, and hemorrhagic septicemia (2, 9). A. hydrophila has the typical AHL/LuxR/LuxI quorum-sensing system and the signal molecules N-butyryl-l-homoserine lactone (C₄-HSL) (the major part) and N-hexanoyl-l-homoserine lactone (C₆-HSL) (the minor part). Many pathogenesis-related factors, including the expression of virulence factors (such as hemolysin, protease, S-layer proteins, DNase, and amylase), biofilm maturation, and the type II, III, and VI secretion systems, have been reported to be under the regulation of quorum sensing in Aeromonas spp. (4, 25, 26, 48, 49, 51). A. hydrophila is strongly resistant to multiple antibiotics and consequently results in significant economic losses to freshwater and warm-water fish farming worldwide (11). Thus, quorum quenching may provide an alternative efficient strategy to control A. hydrophila infection (10–12, 22, 40). AhyI/AhyR interference and AHL-degrading enzymes might be used to inhibit quorum-sensing systems, of which AHL-degrading enzymes are the most readily available tools (10, 22).

The first AHL-degrading enzyme was identified in Bacillus sp. strain 240B1 (16). The expression of its gene in Erwinia carotovora SCG1, a plant soft-rot pathogen, substantially reduced the level of AHL and consequently decreased proteolysis regulated by AHL and attenuated E. carotovora pathogenicity in plants (16). To date, more than 20 AHL-degrading enzymes have been identified in bacteria, fungi, and mammals (10, 14, 15). According to their AHL cleavage sites, these enzymes are classified as AHL lactonases (EC 3.1.1.81) or AHL acylases (synonym, AHL amidases) (EC 3.5.1.97) (http://www.chem.qmul.ac.uk/iubmb/enzyme/). Paraoxonases (PONs) from mammalian sera also have lactonase-like activities in addition to their involvement in the hydrolysis of organophosphates (55). In contrast to AHL acylase and PONs, which have variable substrate spectra, as reported previously, AHL lactonase is by far the most specific AHL-degrading enzyme, with both short- and long-chain AHLs as substrates and no or little activity with other chemicals (18, 52).

Many Bacillus AHL lactonase-like enzymes have been found, and they are closely related (∼90% sequence identity) (14, 28). They contain the conserved motif HXHXDH and a zinc binding motif, and they are classified as metallo-β-lactamases (MBLs) (16). The active sites of autoinducer inactivation (AiiA) AHL lactonases contain a dinuclear zinc binding center bridged by an aspartate and an oxygen species (29, 30, 35). The AHL lactonase AiiB from Agrobacterium tumefaciens also has the same active sites (31). Given their quorum-quenching abilities, AHL lactonases provide a new tool that may inhibit bacterial infection. For example, a plant engineered to express AHL lactonase had substantially enhanced resistance to E. carotovora infection (16, 17). AHL lactonase expression in the pathogens Erwinia amylovora, Pseudomonas aeruginosa PAO1, and Burkholderia cepacia reduced their virulence by degrading AHLs (34, 47, 54). In addition, AHL lactonases have also been expressed in Escherichia coli and Pichia pastoris (7, 52).

For the study reported here, we isolated a Bacillus strain from pond sediment in China and showed that it has AHL lactonase activity. The gene responsible for this activity, Bacillus sp. strain AI96 aiiA (aiiA_AI96), was cloned and expressed in Escherichia coli. The physical properties that dictate the level of activity for the purified recombinant enzyme, AiiA_AI96, suggest that it could be used as an aquatic food additive, and indeed, we found that it attenuated the virulence of Aeromonas hydrophila in zebrafish.

MATERIALS AND METHODS

Strains, plasmids, enzymes, and chemicals.

The AHL-sensing bacterium Agrobacterium tumefaciens KYC55(pJZ372)(pJZ384)(pJZ410), which has a β-galactosidase reporter gene and is most sensitive to N-(3-oxooctanoyl)-l-homoserine lactone (3-oxo-C₈-HSL), was donated by Jun Zhu (Nanjing Agricultural University, Nanjing, People's Republic of China) (57). The biosensor Chromobacterium violaceum CV026, obtained from Shuishang Song (Hebei Academy of Sciences, Shijiazhuang, People's Republic of China), with the purple pigment violacein as a reporter, is most sensitive to C₆-HSL and more sensitive to C₄-HSL than strain KYC55 (32). A. hydrophila NJ-1, which had been isolated from a crucian carp (Carassius carassius), was donated by Yongjie Liu (Nanjing Agricultural University, Nanjing, People's Republic of China). E. coli Trans-I and the vector pEASY-T3 were used for gene cloning and sequencing (both from TransGen, Beijing, People's Republic of China). E. coli BL21(DE3) and the vector pET-28a(+) were used for gene expression (both from Novagen, Darmstadt, Germany). All strains were cultured in Luria-Bertani (LB) medium. Pfu Taq DNA polymerase and restriction endonucleases were obtained from TaKaRa (Otsu, Japan). T4 DNA ligase was obtained from Invitrogen (Carlsbad, CA). C₄-HSL, C₆-HSL, N-heptanoyl-dl-homoserine lactone (C₇-HSL), N-octanoyl-l-homoserine lactone (C₈-HSL), N-decanoyl-l-homoserine lactone (C₁₀-HSL), N-dodecanoyl-l-homoserine lactone (C₁₂-HSL), N-tetradecanoyl-l-homoserine lactone (C₁₄-HSL), N-(3-oxohexanoyl)-d-homoserine lactone (3-oxo-C₆-HSL), 3-oxo-C₈-HSL, N-(3-oxodecanoyl)-l-homoserine lactone (3-oxo-C₁₀-HSL), N-(3-oxododecanoyl)-l-homoserine lactone (3-oxo-C₁₂-HSL), N-(3-oxotetradecanoyl)-l-homoserine lactone (3-oxo-C₁₄-HSL), N-(3-hydroxydodecanoyl)-dl-homoserine lactone (3-hydroxy-C₁₂-HSL), N-(3-hydroxytetradecanoyl)-dl-homoserine lactone (3-hydroxy-C₁₄-HSL), and proteases (including trypsin, α-chymotrypsin, subtilisin A, collagenase, and proteinase K) were obtained from Sigma-Aldrich (St. Louis, MO). Isopropyl-β-d-1-thiogalactopyranoside (IPTG) and 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-gal) were obtained from Promega (Madison, WI). The PageRuler prestained protein ladder was obtained from Fermentas, a subsidiary of Thermo Fisher Scientific (Waltham, MA). All other chemicals were analytical grade and commercially available.

AHL lactonase assay.

AHL lactonase activity was determined as described previously (46), with some modifications. Reaction mixtures containing 179 μl phosphate-buffered saline (PBS) (pH 8.0), 20 μl of AiiA_AI96 solution, and 1 μl of 3-oxo-C₈-HSL (l mg/ml) were incubated at 30°C for 30 min, and the reactions were then terminated by the addition of 10% SDS (50 μl) to the mixture. Agar (1.5% [wt/vol]) strips with dimensions of 6 by 8 mm were prepared, which contained AT minimal salt medium (50) and 20 μg/ml X-gal. After the termination of a reaction, 10 μl of each mixture was added into a punched circular area (r = 2 mm) at one end of the agar strip, and the AHL-sensing strain KYC55, on a toothpick, was placed onto the strip at 4-mm intervals away from the punched area. After the agar strip was cultivated at 30°C for 24 h, the colony point that turned blue was calculated to detect the 3-oxo-C₈-HSL diffusion distance. For the standard curve (y = 6.52 × e^{0.35 × x}/10⁶; R² = 0.995 [where y, in nmol, is the amount of 3-oxo-C₈-HSL and x, in mm, is the distance of diffusion of 3-oxo-C₈-HSL]), the initial concentrations of pure 3-oxo-C₈-HSL were varied and added directly into the punched area of individual strips, and at the end of the assay, concentrations and the lengths over which the 3-oxo-C₈-HSL samples had diffused were correlated. One unit of AHL lactonase activity was defined as the amount of enzyme that hydrolyzed 1 nmol 3-oxo-C₈-HSL per minute.

Isolation and identification of Bacillus sp. AI96.

A sediment sample was collected from a pond in Wuqing, Tianjin, People's Republic of China. One gram of pond sediment was suspended in 10 ml of sterile water. The suspension was 10-fold serially diluted six times, and 100 μl from each dilution was individually cultivated on LB plates at 20°C for 5 days. Pure cultures were obtained by repeatedly streaking and cultivating the bacteria in LB medium at 20°C for 1 to 3 days. Cells were collected by centrifugation at 10,000 × g at 4°C for 5 min, suspended in PBS (pH 7.4), and sonicated in an ice bath. Cell lysates were centrifuged at 12,500 × g at 4°C for 15 min, and the supernatant were assessed for AHL lactonase activity by the agar strip assay. The genera of the strains with AHL lactonase activity were identified by the sequencing of their 16S rRNA genes (36). Bacillus sp. AI96 was chosen for further study for the greatest AHL lactonase activity.

Cloning of aiiA_AI96.

Two primers, AI96-up (5′-CTGATATGAGAAGGTGGATA-3′) and AI96-down (5′-AACAGCATTATGATTTCCC-3′), with sequences derived from the upstream and downstream sequences of genes encoding known AiiA-like enzymes, and genomic DNA from Bacillus sp. AI96 were used to amplify the full-length aiiA_AI96 gene. The PCR program was 95°C for 5 min, followed by 30 cycles of 94°C for 30 s, 45°C for 30 s, and 72°C for 1 min, with a final step of 72°C for 5 min. The PCR product was purified and ligated into pEASY-T3 for sequencing. The nucleotide sequence and open reading frame were determined by using Vector NTI 10 software and the NCBI open reading frame finder (http://www.ncbi.nlm.nih.gov/gorf/gorf.html), respectively. SignalP (http://www.cbs.dtu.dk/services/SignalP/) was used to search for a signal sequence. Sequence alignment was performed with BLAST at the NCBI website and with Vector NTI 10 software. The tertiary structure of AiiA_AI96 was predicated by using SWISS-MODEL (http://swissmodel.expasy.org/) and Swiss-PdbViewer DeepView v4.0 software with the AHL lactonase from Bacillus thuringiensis (Protein Data Bank [PDB] accession number 2A7M) (79% sequence identity) as the template (30).

Expression and purification of AiiA_AI96.

A gene encoding AiiA_AI96 was amplified by using primers AI96aiia-EF (5′-CCGGAATTCATGACCGTTAAAAAACTTTAC-3′ [the EcoRI restriction site is underlined]) and AI96aiia-ER (5′-ATTGCGGCCGCTTATAAAAATTCAGGAAATG-3′ [the NotI restriction site is underlined]). The amplification program was 95°C for 5 min, followed by 35 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 1 min, with a final step of 72°C for 10 min. The PCR product was gel purified, digested with EcoRI and NotI, and cloned into the EcoRI and NotI sites of pET-28a(+) using T4 DNA ligase. The recombinant plasmid, pET-aiiA_AI96, was transformed by electroporation into E. coli BL21(DE3) competent cells that were then cultured overnight at 37°C in LB medium with 50 μg/ml kanamycin. Positive transformants were individually grown in LB medium containing 50 μg/ml kanamycin at 37°C for 2 to 3 h to an optical density at 600 nm (OD₆₀₀) of ∼0.6. AHL lactonase expression was induced by the addition of 1 mM IPTG to the mixture, and the cells were cultured for an additional 12 h at 18°C with constant agitation at 180 rpm.

To purify AiiA_AI96, cells were harvested by centrifugation, washed with Ni-nitrilotriacetic acid (NTA) elution buffer (20 mM Tris-HCl [pH 7.6], 500 mM NaCl, 10% [wt/vol] glycerol), and resuspended in the same buffer. The cells, on ice, were disrupted by sonication, and the cell lysate was collected by centrifugation at 12,500 × g for 15 min. The cell lysate (5 ml) was loaded onto Ni-NTA resin (1 ml; Qiagen, Hilden, Germany) that had been equilibrated with elution buffer. The protein was eluted with a linear gradient of 20 to 300 mM imidazole in elution buffer. Fractions with AHL lactonase activity were pooled and concentrated with a Nanosep centrifugal device (10 kDa; Pall, East Hills, NY).

The apparent molecular mass of the purified protein was determined by SDS-PAGE (12% [wt/vol]) after staining it and molecular mass standards with Coomassie brilliant blue R-250. After isolation from the gel, AiiA_AI96 was sequenced by liquid chromatography-electrospray ionization-tandem mass spectroscopy (LC-ESI-MS/MS) at Tianjin Biochip Corporation Co. Ltd. (Tianjin, People's Republic of China). The peptide sequences determined by tandem mass spectrometry were compared with those found in the deduced AiiA_AI96 sequence. Protein concentrations were routinely determined by the Bradford method (5).

Physical parameters that affect AiiA_AI96 activity and measurement of catalytic constants.

3-Oxo-C₈-HSL was routinely used as the substrate in all assays except for the substrate specificity assay. The activity-pH profile was determined at 30°C with buffers with pH values that ranged from 4.0 to 11.0. To determine the stability of the enzyme at various pH values, it was incubated at 30°C in various buffers for 30 min without the substrate, and the residual activity was then measured under standard conditions (pH 8.0 at 30°C for 30 min). The buffers used were McIlvaine buffer (pH 3.0 to 6.0), PBS (pH 6.0 to 8.0), 0.05 M Tris-HCl (pH 8.0 to 9.0), and 0.1 M glycine-NaOH (pH 9.0 to 12.0).

The activity-temperature profile was determined at pH 8.0 at temperatures between 0°C and 70°C. The thermal stability of AiiA_AI96 was determined by measuring its residual activity under standard conditions after incubating the enzyme at 70°C or 80°C for 2, 5, 10, 15, 20, 30, and 60 min and at 90°C for 1, 3, and 5 min. The effects of metal ions and chemical reagents on activity were determined at concentrations of 1 or 10 mM. A sample without an additive served as the control.

Values of the kinetic constants K_m and V_max were determined by using assay systems that consisted of mixtures of PBS (pH 8.0) and 0.017 to 16.578 nM 3-oxo-C₈-HSL and an incubation time of 5 min at 30°C. To determine the kinetic values, the data were plotted as a Lineweaver-Burk curve.

To determine its resistance to proteolysis, AiiA_AI96 was individually incubated with trypsin, α-chymotrypsin, proteinase K, subtilisin A, or collagenase at a protease-to-protein (wt/wt) ratio of 1:10 (20) at 30°C for 30 and 60 min, respectively. The residual activities were detected under standard conditions with no protease-treated enzyme as the control.

The resistance of AiiA_AI96 to digestion with intestinal juice was also evaluated. Intestinal juice was prepared as described previously (44). In brief, the full-length intestine of a carp (Cyprinus carpio) that had been anesthetized with the anesthetic MS-222 (tricaine methanesulfonate) 1 h after feeding was gently washed three times with PBS (pH 7.4) and then dissected with a scalpel. The intestinal mucus was washed out with PBS (pH 7.4) and centrifuged at 8,000 × g for 20 min at 4°C. The supernatant (carp intestinal juice) was retained and was incubated with AiiA_AI96 at a ratio of 1:10 (vol/vol) for 30 and 60 min at 30°C. Residual activity was determined under standard conditions.

The protease sensitivity experiment was carried out as described above, with casein as the positive control, and the degree of degradation of casein was determined by the Folin-phenol method (27). The stability of proteases and intestinal juice in the absence of the target protein was measured. After being treated by using the same protocol, residual protease activities were estimated by the Folin-phenol method (27).

Substrate specificity assays.

The ability of AiiA_AI96 to hydrolyze C₄-HSL, C₆-HSL, C₇-HSL, C₈-HSL, C₁₀-HSL, C₁₂-HSL, C₁₄-HSL, 3-oxo-C₆-HSL, 3-oxo-C₁₀-HSL, 3-oxo-C₁₂-HSL, 3-oxo-C₁₄-HSL, 3-hydroxy-C₈-HSL, and 3-hydroxy-C₁₄-HSL, which were individually reacted in PBS (pH 8.0), was measured as described above for 3-oxo-C₈-HSL under otherwise standard assay conditions, with A. tumefaciens KYC55 or C. violaceum CV026 as the biosensor. All hydrolytic reactions were carried out for 30 min under conditions (pH 8.0 at 30°C) in which A. hydrophila is apt to grow.

Preparation of A. hydrophila NJ-1 cells and fish diet.

After cultivation at 30°C for 12 h, A. hydrophila NJ-1 cells were collected by centrifugation at 3,000 × g for 15 min, washed three times with sterile PBS (pH 7.4), and suspended in sterile distilled water at an OD₆₀₀ of ∼0.15. This bacterial suspension was used to grow fish. To ensure the validity of pathogens, a fresh A. hydrophila NJ-1 cell suspension was prepared and exchanged every 2 days. The CK diet (control diet) was supplied by Zhejiang Xinxin Feed Co. Ltd., Jiaxing, People's Republic of China, and contained 47% fishmeal, 24% soybean meal, 24% wheat flour, 2% soybean oil, 3% premix, and no antibiotics (42.0% crude protein and 7.29% crude lipid).

Effect of AiiA_AI96 on A. hydrophila NJ-1 infection.

Wild-type zebrafish (4 months of age, with an average weight of ∼200 mg and an average length of ∼2.5 cm) were randomly divided into six test groups: group A, which was fed the CK diet and immersed in clean water; group B, which was fed the CK diet and immersed in A. hydrophila NJ-1-containing water; group C, which was fed the CK diet supplemented with 3.74 U AiiA_B546 per gram of feed and immersed in clean water (7); group D, which was fed the CK diet supplemented with 3.74 U AiiA_AI96 per gram of feed and immersed in clean water; group E, which was fed the CK diet supplemented with 3.74 U AiiA_B546 per gram of feed and immersed in A. hydrophila NJ-1-containing water; and group F, which was fed the CK diet supplemented with 3.74 U AiiA_AI96 per gram of feed and immersed in A. hydrophila NJ-1-containing water. Each group had four replicate tanks, and each tank had 30 fish. The zebrafish were allowed free access to food (see below) at 09:30 and 16:30. Mortality during the 25-day experimental period was recorded each day. Dead fish of groups B, E, and F were removed daily and examined for bacteriological contamination, as was the water containing A. hydrophila NJ-1. The fish and water of groups A, C, and D, which served as controls, were also examined at 25 days. To test bacterial contamination, the fish body was sterilized with 75% ethanol, and the body fluid was extracted with a syringe under sterile conditions and streaked onto an ampicillin blood agar plate (24). The water sample was streaked onto the same plate directly.

The amount of AiiA_AI96 needed to protect the fish against A. hydrophila was assessed by adding various amounts of AiiA_AI96 to the feed. The zebrafish were divided into six groups: group A, which was fed the CK diet (the regular diet) and immersed in clean water; group B, which was fed the CK diet and immersed in A. hydrophila NJ-1-containing water; and groups C, D, E, and F, which were fed the CK diet supplemented with 3.74, 3.74 × 10⁻¹, 3.74 × 10⁻², and 3.74 × 10⁻³ U AiiA_AI96 per gram of feed, respectively, and immersed in A. hydrophila NJ-1-containing water. Each group had four replicate tanks, and each tank had 30 fish. Mortality during the 35-day experimental period was recorded each day.

The stability of AiiA_AI96 in the experimental diet was assessed by storing AiiA_AI96-supplemented feed at 25°C for 2, 4, and 8 weeks and then assaying the feed for activity with 3-oxo-C₈-HSL as the substrate.

Nucleotide sequence accession numbers.

The aiiA_AI96 and the 16S rRNA gene nucleotide sequences of Bacillus sp. AI96 have been deposited in the GenBank database under accession numbers HM750247 and HQ680867, respectively.

RESULTS

Identification of Bacillus sp. AI96.

Significant AHL lactonase activity was detected only in the intracellular protein of strain AI96, and this strain was classified as a Bacillus strain based on its 16S rRNA gene sequence, which was identical to those of Bacillus pseudomycoides J14 (GenBank accession number GU826151.1) (100% identity) and B. mycoides (accession number GU171377.1) (100% identity). A sample of Bacillus sp. AI96 was deposited in the China General Microbiological Culture Collection Center (Beijing, People's Republic of China) under registration number CGMCC 4164.

Cloning and sequence analysis of aiiA_AI96.

An aiiA_AI96 genomic gene fragment of ∼900 bp was amplified by using AI96 genomic DNA and primers AI96-up and AI96-down. The aiiA_AI96 open reading frame is 753 bp and encodes a protein of 250 residues and a stop codon, with a calculated molecular mass of 28.0 kDa and a pI of 5.08. No signal peptide was identified by using SignalP. The deduced AiiA_AI96 sequence is most similar to a hypothetical protein of B. pseudomycoides DSM 12442 (GenBank accession number ZP_04152085) (98% sequence identity), is ∼80% identical to the sequences of AHL lactonases from other Bacillus spp., and is ∼32 to 43% identical to those from other sources (Fig. 1). Homology modeling revealed that AiiA_AI96 has a putative structure typical of metallo-β-lactamase (MBL) family proteins. Compared to the AHL lactonase from B. thuringiensis (PDB accession number 2A7M), AiiA_AI96 tended to have variation in the location of the helical or loop-rich regions (Fig. 2).

Fig 1.

Amino acid sequence alignment of AiiA_AI96 with Bacillus AiiA-like proteins and with AHL lactonases from other species. The sequences are those of Bacillus sp. AI96 CGMCC 4164 (GenBank accession number HM750247) (B. sp. AI96), Bacillus sp. strain 240B1 (accession number AAF62398) (B. sp. 240B1), Bacillus sp. strain B546 (accession number FJ816104) (B. sp. B546), Bacillus thuringiensis serovar shandongiensis (accession number AAR85481) (B. thu sha) (referred to as AiiA SS10 in this paper), Bacillus thuringiensis serovar kurstaki (accession number AAM92140) (B. thu kur) (referred to as PDB accession number 2A7M in Fig. 2), Bacillus cereus (accession number ACR46836) (B. cer), Bacillus subtilis (accession number AAY51611) (B. sub), Arthrobacter sp. strain IBN110 (accession number AAP57766) (AhlD), Klebsiella pneumoniae (accession number AAO47340) (AhlK), Agrobacterium tumefaciens C58 (accession number AAK91031) (AiiB), and Agrobacterium tumefaciens C58 (accession number AAD43990) (AttM). The similar and identical residues are shaded in gray and black, respectively. The conserved motif HXHXDH is boxed. The active-site amino acids are indicated with arrows, and the amino acids involved in the modification of a helix to a random coil (Fig. 2) are indicated with diamonds.

Fig 2.

Putative tertiary structure of AiiA_AI96 modeled using the crystal structure of B. thuringiensis serovar kurstaki AiiA (GenBank accession number AAM92140 and PDB accession number 2A7M) (30) as the template. The AiiA_AI96 and AiiA (PDB accession number 2A7M) ribbon diagrams are shown in blue and red, respectively.

Expression and purification of AiiA_AI96.

In the expression study, AHL lactonase activity (2.57 U/ml) was found in the cell lysate at the end of the culture period. No activity was detected in the cell lysate supernatants of uninduced transformants or of induced transformants that harbored an empty plasmid.

AiiA_AI96 in the cell lysate was purified to electrophoretic homogeneity by Ni-NTA chromatography. In SDS-PAGE gels, the purified protein migrated as a single band with an apparent molecular mass of ∼31 kDa (including part of the His tag) (Fig. 3). The predicted and experimental molecular masses were nearly the same. Four interval peptides sequenced by LC-ESI-MS/MS (LYFLPAGR, GTFVEGQILPK, TEHDAALHR, and SGSVLLTIDASYTQENFEQGVPFAGFDSEMASQSINR) have the same sequences as those deduced from the aiiA_AI96 nucleotide sequence. The specific activity of purified AiiA_AI96 was 9.13 U/mg.

Fig 3.

SDS-PAGE gel of recombinant AiiA_AI96 from E. coli BL21(DE3). Lanes: M, protein molecular mass markers; 1, cell extract of E. coli BL21(DE3) that harbored an empty pET28a(+) vector; 2, cell extract of E. coli BL21(DE3) that harbored pET28a(+) carrying aiiA_AI96 and that was not treated with IPTG; 3, cell extract of E. coli BL21(DE3) that harbored pET28a(+) carrying aiiA_AI96 and that was treated with IPTG; 4, purified AiiA_AI96.

Physical parameters that affect AiiA_AI96 activity and kinetic constants.

AiiA_AI96 was fully active against 3-oxo-C₈-HSL at pH 8.0 when the temperature was 30°C, and it retained >95% of the maximal activity between pH 6.0 and 8.5 (Fig. 4A). The protein was stable between pH 6.0 and 11.0 at 30°C, as assessed by its complete retention of activity after preincubation in buffers of differing pHs (Fig. 4B). The optimal temperature at which AiiA_AI96 retained higher levels of activity at pH 8.0 was ∼10°C to 40°C (Fig. 4C). AiiA_AI96 retained >90% of its activity after incubation in PBS (pH 8.0) at 70°C for 60 min, and it retained ∼60% of its activity after incubation for 20 min at 80°C or for 3 min at 90°C (Fig. 4D).

Fig 4.

Physical parameters that affect the enzymatic activity of AiiA_AI96. (A) Effect of pH on AiiA_AI96 activity at 30°C. (B) pH stability assay. (C) Effect of temperature on AiiA_AI96 activity at pH 8.0. (D) Thermostability of AiiA_AI96. Each value is the mean ± standard deviation (SD) (n = 3).

The effects of various chemicals on AiiA_AI96 activity were determined under standard conditions (Table 1). None of the chemicals (1 mM) stimulated activity. At 10 mM, Co²⁺, Cr³⁺, Pb²⁺, Fe²⁺, and Mn²⁺ partially inactivated the enzyme. Ag⁺ and SDS at both 1 mM and 10 mM inhibited AiiA_AI96 activity.

Table 1.

Effects of metal ions and chemicals on AiiA_AI96 activity

Chemical	Mean relative activity (%) ± SD at concn (mM) ofa:
	1	10
None	100.00	100.00
Na+	98.65 ± 0.22	99.52 ± 0.00
K+	99.33 ± 0.00	99.52 ± 0.00
Li+	99.33 ± 0.00	99.52 ± 0.00
Ca2+	99.59 ± 0.11	99.52 ± 0.00
Mg2+	99.02 ± 0.22	83.58 ± 0.00
Co2+	99.33 ± 0.00	167.03 ± 0.22
Cr3+	99.02 ± 0.22	118.12 ± 0.22
Cu2+	99.33 ± 0.00	99.52 ± 0.00
Ni2+	99.33 ± 0.00	99.52 ± 0.00
Zn2+	99.33 ± 0.00	99.78 ± 0.11
Pb3+	99.33 ± 0.00	118.12 ± 0.22
Fe2+	99.81 ± 0.11	118.12 ± 0.22
Mn2+	99.02 ± 0.22	118.43 ± 0.00
Ag+	21.86 ± 1.10	21.90 ± 0.53
SDS	83.67 ± 0.00	40.61 ± 0.00
β-Mercaptoethanol	99.33 ± 0.00	99.52 ± 0.00
EDTA	99.59 ± 0.10	99.52 ± 0.00

^aEach value is the mean ± standard deviation (n = 3) and was normalized to the control value.

The K_m and V_max values were 8.87 mM and 1.65 mM/min, respectively, with 3-oxo-C₈-HSL as the substrate.

Resistance of AiiA_AI96 to proteases and intestinal juice.

AiiA_AI96 was very resistant to all proteases tested, retaining ∼100% of its activity after 30- and 60-min incubations with proteases. AiiA_AI96 was also resistant to carp intestinal juice and retained ∼90 and 80% of its activity after 30 and 60 min, respectively. In the protease sensitivity experiment, casein (the positive control) was almost completely degraded by all proteases used and intestinal juice (data not shown). The residual activities of trypsin, α-chymotrypsin, proteinase K, subtilisin A, collagenase, and intestinal juice were 98, 96, 97, 99, 106, and 98%, respectively, at the end of the protocol period. The results indicated that proteases and intestinal juice are highly active under the tested conditions.

Hydrolytic activities of AiiA_AI96 toward AHLs.

The hydrolytic activities of AiiA_AI96 toward various AHL substrates are shown in Table 2. AiiA_AI96 had a broad substrate spectrum, including C₄-HSL, C₆-HSL, C₇-HSL, C₈-HSL, C₁₀-HSL, C₁₂-HSL, C₁₄-HSL 3-oxo-C₈-HSL, 3-oxo-C₆-HSL, 3-oxo-C₁₀-HSL, 3-oxo-C₁₂-HSL, 3-oxo-C₁₄-HSL, 3-hydroxy-C₈-HSL, and 3-hydroxy-C₁₄-HSL. Due to the insensitivity of biosensor bacteria, activities toward several AHLs could not be quantified.

Table 2.

Hydrolytic activities of AiiA_AI96 against various AHL substrates^a

Substrate	Biosensor bacterium	Enzyme activity (U/ml)
C4-HSL	KYC55/CV026	D/44.97
C6-HSL	KYC55/CV026	D/2.28
C7-HSL	KYC55/CV026	1.90/0.39
C8-HSL	KYC55/CV026	1.29/0.33
C10-HSL	KYC55	11.19
C12-HSL	KYC55	L
C14-HSL	KYC55	D
3-Oxo-C6-HSL	KYC55/CV026	L/0.36
3-Oxo-C8-HSL	KYC55/CV026	9.13/8.54
3-Oxo-C10-HSL	KYC55	9.70
3-Oxo-C12-HSL	KYC55	3.08
3-Oxo-C14-HSL	KYC55	312.40
3-Hydroxy-C8-HSL	KYC55	5.22
3-Hydroxy-C14-HSL	KYC55	L

^aKYC55, A. tumefaciens KYC55(pJZ372)(pJZ384)(pJZ410); CV026, C. violaceum CV026; D, degraded but could not be quantified; L, little activity detected.

Assessment of AiiA_AI96 as a feed additive for protection against A. hydrophila infection.

In protection ability detection experiments, no fish died after being fed the CK diet (the regular diet) (group A) or the experimental diet (groups C and D) for 25 days in the absence of environmental A. hydrophila NJ-1 (Fig. 5), suggesting that the CK diet, AiiA_AI96, and AiiA_B546 are not toxic. AiiA_AI96 significantly attenuated A. hydrophila NJ-1 infection. Compared with group B, the percent survival of group F was significantly higher from day 4, increased about 20% at day 11, and maintained levels 20 to 38% higher until day 25. During the same period, significantly higher mortality rates were detected in group E than in group F, except for rates at days 10 to 14, suggesting the inability of AiiA_B546 to protect zebrafish by oral administration. AiiA_AI96 was stable in the experimental diet at 25°C. After a 2-month storage period at 25°C, AiiA_AI96 retained ∼100% of its activity. In the bacteriological contamination test, A. hydrophila NJ-1 was detected only in the dead fish of groups B, E, and F and in A. hydrophila NJ-1-containing water.

Fig 5.

Protection of zebrafish infected with A. hydrophila NJ-1 by AiiA_AI96. Groups were as follows: group A was fed the CK diet and immersed in clean water, group B was fed the CK diet and immersed in A. hydrophila NJ-1-containing water, group C was fed the CK diet supplemented with 3.74 U AiiA_B546 per gram of feed and immersed in clean water, group D was fed the CK diet supplemented with 3.74 U AiiA_AI96 per gram of feed and immersed in clean water, group E was fed the CK diet supplemented with 3.74 U AiiA_B546 per gram of feed and immersed in A. hydrophila NJ-1-containing water, and group F was fed the CK diet supplemented with 3.74 U AiiA_AI96 per gram of feed and immersed in A. hydrophila NJ-1-containing water. Each value is the mean ± SD (n = 4). Data marked with * (0.01 < P < 0.05) and ** (P < 0.01) are significantly different between groups B and F; data marked with # (0.01 < P < 0.05) and ## (P < 0.01) are significantly different between groups E and F.

Various amounts (on a feed-weight basis) of AiiA_AI96 were added to the CK diet to assess the minimum amount of AiiA_AI96 needed for protection. Compared with the percent survival of fish in group B (Fig. 6), groups C, D, E, and F all showed protection against A. hydrophila NJ-1 infection, with increased survival rates of 20, 10, 10, and 9%, respectively, at 35 days. Because the rates of survival of fish in groups D, E, and F did not show significant differences compared to those of group B, the smallest amount of AiiA_AI96 needed for protection was defined as 3.74 U/gram of feed.

Fig 6.

Determination of the smallest amount of AiiA_AI96 needed to protect zebrafish against A. hydrophila NJ-1 infection. Groups were as follows: group A was fed the CK diet and immersed in clean water; group B was fed the CK diet and immersed in A. hydrophila NJ-1-containing water; and groups C, D, E, and F were fed the CK diet supplemented with 3.74, 3.74 × 10⁻¹, 3.74 × 10⁻², and 3.74 × 10⁻³ U AiiA_AI96 per gram of feed, respectively, and immersed in A. hydrophila NJ-1-containing water. Each value is the mean ± SD (n = 4). Data marked with * (0.01 < P < 0.05) and ** (P < 0.01) are significantly different between groups B and F, and data marked with # (0.01 < P < 0.05) and & (0.01 < P < 0.05) are significantly different between groups B and C and between groups B and D, respectively.

DISCUSSION

Here we isolated AHL-degrading Bacillus sp. AI96, which is closely related to B. pseudomycoides and B. mycoides. Broad-spectrum AHL-degrading AiiA-type enzymes were found to be widespread in B. thuringiensis and Bacillus cereus strains (14, 28). Recently, Bacillus amyloliquefaciens, B. subtilis, and B. marcorestinctum were shown to have AHL-degrading activities, whereas AHL-degrading activity has not yet been reported for B. pseudomycoides (23, 41, 56). Although B. mycoides was reported previously to produce AHL-inactivating proteins, their inactivating mechanism is still unknown (14).

The deduced amino acid sequence of AiiA_AI96 is very similar to those of AiiA-like proteins from other Bacillus spp. (∼80% sequence identity) (Fig. 1). Given its sequence, AiiA_AI96 belongs to the MBL superfamily, which currently contains 20 families, including the glyoxalase II, cyclase, arylsulfatase, alkylsulfatase, flavoprotein, and RNase families (3). MBL superfamily members have structural features in common, including a tertiary structure, two active-site zinc ions, and the conserved HXHXDH motif (37). Zinc atoms Zn1 and Zn2 in the active sites of AiiA-like proteins are coordinated by His-104, His-106, and His-169 as well as Asp-108, His-109, and His-235, respectively. Asp-191 and an H₂O molecule or a hydroxide ion form two bridges between the two zinc ions (29). Moreover, Tyr-194 was also conserved, which plays a key role in hydrogen bonding with the substrate (AHLs) ester oxygen to form a complex (29). All these conserved amino acids can be identified in the corresponding sites of AiiA_AI96. Moreover, certain residues, located near or far from the active site, e.g., Gly-12, Pro-34, Leu-39, Asp-50, Gly-52, Gly-91, Pro-94, and Asp-96, are also conserved (Fig. 1). Their functions remain to be determined. As shown in Fig. 1, the amino acids of the modified helix are conserved in AiiA-like proteins. However, the residues up- and downstream of this conserved motif (including the active-site His-104, His-106, Asp-108, and His-109) in AiiA_AI96 are different from those of other AiiA-like proteins. This difference might lead to a subtle change from a helix to a random coil (Fig. 1, arrows) and consequently may affect the structure and function of AiiA_AI96.

To date, only the Bacillus AHL lactonases AiiA_SS10 (AiiA of Bacillus thuringiensis serovar shandongiensis), AiiA_B546, and AiiA_240B1 have been characterized (7, 45, 52). These AHL lactonases are most active at 20°C to 30°C (at pH 8.0), a temperature range that is lower than that for AiiA_AI96 (40°C). Moreover, AiiA_AI96 retained 20% of its activity at 60°C, a temperature at which AiiA_B564 and AiiA_SS10 are completely inactive (45, 52). The thermostability of AiiA_AI96 is exceptional for an AHL lactonase. All other tested AHL lactonases, which had been expressed in E. coli, are not thermostable. For example, AiiA_240B1 activity decreased sharply after incubation at 45°C for 2 h (16); AiiA_SS10 was inactive after incubation for 2 h at 40°C (45), and AiiA_B564 retained 21.3% of its initial activity after incubation at 37°C for 30 min (7). Conversely, AiiA_AI96 was stable at 70°C for 2 h and retained 80% of its activity at 80°C for 10 min and ∼67% of its activity at 90°C for 3 min.

Distinct from other MBL superfamily members, AHL lactonases do not require Zn²⁺ for activity (7, 45, 52). In addition, inhibitors of MBL activity, e.g., EDTA and β-mercaptoethanol, have no effect on AHL lactonase activity (7, 45, 52). Although AHL lactonases have the typical MBL superfamily structure and contain the same conserved residues and Zn ions, the effects caused by metal ions that promote activity and inhibitors differ between the AHL lactonases and other MBLs. Therefore, whether the Zn ions function in AHL lactonase catalysis and whether AHL lactonases are truly metalloenzymes remain to be determined. AiiA_AI96 is more resistant to the effects of heavy metal ions (Pb²⁺, Mn²⁺, Ni²⁺, Fe²⁺, Zn²⁺, and Cr³⁺) than are other AHL lactonases (7, 45, 52). Notably, Cu²⁺, which is an AHL lactonase inhibitor, had no effect on AiiA_AI96 activity. To stimulate fish growth, metal ions, e.g., the macrominerals Ca²⁺, Mg²⁺, K⁺, and Na⁺ and the microminerals Cu²⁺, Fe²⁺, Zn²⁺, and Mn²⁺, are routinely added into fish feed (38). To be a useful fish feed additive, an additive enzyme must not be affected by these metal ions or by proteases. AiiA_AI96 was strongly resistant to all the tested proteases and to carp intestinal juice.

Currently, only two groups of enzymes that are specific for AHL degradation are known, namely, AHL lactonases and AHL acylases (10, 14, 15). Only AHL lactonases hydrolyze many different AHL-signaling molecules. We found that AiiA_AI96 degrades C₄-HSL, C₆-HSL, C₇-HSL, C₈-HSL, C₁₀-HSL, C₁₄-HSL, 3-oxo-C₆-HSL, 3-oxo-C₈-HSL, 3-oxo-C₁₀-HSL, 3-oxo-C₁₂-HSL, 3-oxo-C₁₄-HSL, and 3-hydroxy-C₈-HSL. Conversely, most AHL acylases hydrolyze only long-chain AHLs and penicillin G (42). PONs that have AHL lactonase activity hydrolyze many different lactones (19). Given the different substrate specificities of the above-mentioned AHL-degrading enzymes, AHL lactonases are more suitable for the control of AHL-producing bacteria.

Aquatic pathogenic and opportunistic bacteria can completely kill mollusk, fish, and shrimp populations that are reared in aquaculture facilities (2, 9, 12). The addition of antibiotics and disinfectants has led to bacterial resistance and has also caused aquaculture environmental pollution and food safety issues (12). Various strategies have been used, e.g., microbial-matured water, green-water systems, bacteriophage therapy, immunostimulants, and vaccines, to combat such problems. However, bacterial quorum quenching has not been attempted (12). The important Gram-negative fish pathogens Vibrio spp., Aeromonas spp., and Yersinia ruckeri use AHLs as signaling molecules (6, 12), and a quorum-quenching strategy might efficiently control these pathogens (10–12, 22, 40). A. hydrophila is an opportunistic, important zoonotic fish pathogen associated with hemorrhagic septicemia as well as fin and tail rot (2, 9). Previous studies of the AhyR/AhyI system, which is analogous to the LuxR/LuxI system, in A. hydrophila indicated that C₄-HSL is the major signaling molecule and regulates the expression of virulence factors, biofilm maturation, extracellular protease expression, and so on (4, 25, 26, 48, 49, 51). The mutation of the A. hydrophila quorum-sensing system increased the rates of survival of infected Artemia shrimp (12). We have shown that the coinjection of recombinant AiiA_B546 and A. hydrophila into common carp decreased carp mortality rates and delayed the time to death in comparison with the results for carp infected only with A. hydrophila (7). However, the oral administration of AHL lactonases is more viable for application in aquaculture. In a previous study, the addition of N-acyl homoserine lactone-degrading bacterial enrichment cultures (ECs) into water or feed increased the survival rate of Macrobrachium rosenbergii larvae with Vibrio harveyi infection (40). In our protection ability experiments (Fig. 5), two AHL lactonases were supplemented into experimental diets to control A. hydrophila infection. The results showed that the oral administration of AiiA_AI96 significantly decreased the mortality of infected zebrafish, but AiiA_B546 lost its protection ability by direct feeding. Therefore, AiiA_AI96 might be used directly for the aquatic control of microbial pathogens. Compared with other AHL lactonases, AiiA_AI96 has superior properties as a feed supplement; i.e., it is optimal over wide ranges of pHs and temperatures (pH 6.0 to 8.5 and 10°C to 40°C), is thermostable, is resistant to proteases, is C₄-HSL hydrolyzed, and attenuates infection. These properties make AiiA_AI96 an outstanding quorum-quenching tool for aquaculture. Supplementation with AiiA_AI96 appears to be a feasible and economical way to decrease A. hydrophila infection.