The Genetic and Phenotypic Diversity of Bacillus spp. from the Mariculture System in China and Their Potential Function against Pathogenic Vibrio

Bacillus spp. could be one of the most suitable substitutes for the control and prevention of aquatic diseases. The occurrence of species population, antimicrobial character, and virulence diversity in Bacillus spp. recovered from the mariculture system in China between 2009 and 2021 were investigated, screening for probiotic Bacillus strains with good biological safety that can inhibit Vibrio parahaemolyticus, V. alginolyticus, V. harveyi, V. owensii, V. campbellii. The results showed that 116 Bacillus isolates were divided into 24 species, and the top three species were B. subtilis (37/116), B. velezensis (28/116), and B. amyloliquefaciens (10/116). Among the 116 Bacillus isolates, 32.8% were effective against V. parahaemolyticus, 30.1% for V. alginolyticus, 60.3% for V. harveyi, 69.8% for V. owensii and 74.1% for V. campbellii. More than 62% of Bacillus isolates were susceptible to florfenicol, doxycycline and tetracycline, etc., and 26/116 Bacillus isolates were found to be multiple-antibiotic-resistant (MAR), with MARI values ranging from 0 to 0.06. Eighteen kinds of antibiotic resistance genes were tested; only tetB, blaTEM, and blaZ were detected. And 9 isolates in 2 Bacillus species were excluded by 6/10 kinds of Bacillus-related toxin gene (hblA, hblC, nheB, nheC, entFM, cykK). Bio-safety testing indicated that three kinds of probiotics were good probiotic candidates to prevent Vibriosis. These results provide comprehensive genetic diversity, potential risks, and probiotic characteristics of Bacillus in the mariculture system in China, and provide basic support for green and healthy development of aquatic industry.


Introduction
Aquatic products play an important role as a component of healthy and indispensable supplements to nutritional diets. The aquaculture output reach to 80 million tons in recent years [1]. However, the occurrence of diseases is gradually becoming a major issue for the aquaculture industry [2]. For instance, approximately 60% of the reduced production in shrimp aquaculture can be attributed to diseases, 20% of which are caused by bacterial diseases [3]. Vibrio spp. is a Gram-negative bacteria commonly found in ocean and estuarine environments. Vibriosis is regarded as the bacterial disease caused by Vibrio spp. in shellfish and finfish aquaculture and is also a main cause of mortality in cultured shrimp throughout the world [4]. Pathogenic Vibrio spp. has been reported to cause foodborne diseases associated with seafood consumption. Meanwhile, Vibriosis in shrimps has been reported by many researchers, and at least 14 Vibrio species have been deemed responsible, Furthermore, in terms of Vibrio species, the antibacterial ability of Bacillus isolates showed differences among species and isolates. For the 38 isolates that were effective against V. parahaemolyticus, there were 12 B. subtilis isolates, 10 B. velezensis isolates, 7 B. The left is the strain number, and the right is the antibacterial diameter character. The colors from light to dark represent 0-11 mm, 11-16 mm, 16 For B. velezensis, among the 28 B. velezensis, 18 isolates had no effect on V. parahaemolyticus, and the inhibition zones of the remaining 10 isolates ranged from 13.8 to 21.7 mm; 15 strains had no effect on V. alginolyticus, and the diameter range of remaining 13 isolates ranged from 14.3 to 18.5 mm; for the V. harveyi, only 2 strains had no effective, remining isolates. The maximum diameter could reach 33.8 mm, while 14.7 is the minimum diameter. All 28 isolates were effective on V. owensii, from 11.7 to 33.4 mm; 4 isolates had no effect on V. campbellii, the diameter of the remaining 24 isolates ranged from 14.2 to 29.2 mm.
For B. amyloliquefaciens, among the 10 B. amyloliquefaciens, 3 strains had no effect on V. parahaemolyticus, and the inhibition zones of the remaining 7 isolates ranged from 13 to 19.7 mm; 2 strains had no effect on V. alginolyticus, and the inhibition zone of the remining 8 isolates was 14.7-19.7 mm; 10 strains had effective on V. harveyi, from 20.3-26.8 mm; 10 strains were effective on V. owensii, ranging from 22.2 to 27.9 mm; only 2 strains were effective for V. campbellii. The inhibition zone of the remining eight isolates was 20.3-27.1 mm.

Antibiotic Susceptibility Characterization of Different Kinds of Bacillus Isolates
Antimicrobial susceptibility tests indicated that high sulfadiazine resistance (67/116) was observed in 116 isolates. However, high susceptibility was observed to florfenicol Furthermore, 26/116 Bacillus isolates were found to be multiple antibiotic resistant (MAR) and displaying resistance to at least three of the antibiotics tested in this study. The MARI of all the isolates ranged from 0 to 0.06. None of the tested isolates were resistant to all antibiotics at the same time. Two isolates (strain 7, strain 23) were resistant to seven types of antibiotics. The MAR of isolates resistant to 3 types of antibiotics was the most concentrated, and the proportion of isolates resistant to 4-6 types of drugs was lower (Table 3). Furthermore, 26/116 Bacillus isolates were found to be multiple antibiotic resistant (MAR) and displaying resistance to at least three of the antibiotics tested in this study. The MARI of all the isolates ranged from 0 to 0.06. None of the tested isolates were resistant to all antibiotics at the same time. Two isolates (strain 7, strain 23) were resistant to seven types of antibiotics. The MAR of isolates resistant to 3 types of antibiotics was the most concentrated, and the proportion of isolates resistant to 4-6 types of drugs was lower (Table 3).
Under 67 sulfonamides resistant isolates, none of the selected sulfonamides antimicrobial resistance genes were identified. All isolates were sensitive to tetracycline, but tetB was detected in 116 isolates. The blaTEM and cfr was detected some isolates that susceptibility to cefalexin, sulbactam, thiamphenicol and florfenicol.

Distribution of Virulence Genes among Bacillus Isolates
Based on the reported Bacillus-associated virulence genes, 10 kinds of virulence genes were detected in the Bacillus isolates. Only 9 of 116 isolates (7.8%) carried virulenceassociated genes, of which 9/116 carried entFM, 7/116 carried hblA and hblC, 6/116 carried nheB, 5/116 carried nheC and cytK. The hblD, nheA, and ces genes were not detected in any strains. The positive, uniformly sized PCR products were sequenced and reconfirmed based on the NCBI database.

Holistic Assessment of Potential Probiotics
Among the 116 Bacillus isolates, 12 Bacillus isolates were shown to have an inhibitory effect on 5 Vibrio species and can be used as potential probiotics to inhibit Vibriosis (Table 4), including 3 B. subtilis isolates, 5 B. velezensis isolates, and 4 B. amyloliquefaciens isolates. The largest inhibition zone for V. parahaemolyticus was 20.1 mm, and those for V. alginolyticus, V. harveyi, V. owensii, V. campbellii were 19.7 mm, 28.7 mm, 33.4 mm, and 28.7 mm, respectively.
For antibiotic susceptibility characterization, B. subtilis strain 4, 6 was resistant to neomycin and azithromycin, and B. velezens strain 46, 61 was resistant to ampicillin and thiamphenicol. B. velezens strain 40, 46, 50, 61 was resistant to sulfonamides. Only B. subtilis strain 4 carried chloramphenicol gene cfr. The other 11 isolates contain 1 resistance gene tetB but were susceptible to tetracycline. All virulence genes were not detected in these strains ( Figure 3). For antibiotic susceptibility characterization, B. subtilis strain 4, 6 was resistant to neomycin and azithromycin, and B. velezens strain 46, 61 was resistant to ampicillin and thiamphenicol. B. velezens strain 40, 46, 50, 61 was resistant to sulfonamides. Only B. subtilis strain 4 carried chloramphenicol gene cfr. The other 11 isolates contain 1 resistance gene tetB but were susceptible to tetracycline. All virulence genes were not detected in these strains ( Figure 3).

Safety Testing of Potential Probiotics Bacillus Isolates in L. vannamei
The 12 potential probiotics Bacillus suspension of 10 8 CFU/mL was injected into the muscle of the second abdominal segment of L. vannamei and observed for 7 days to confirm the safety of the strains. The results showed that B. subtilis strain 4, 6, 40, B. velezens 46, 56, 61, and B. amyloliquefaciens 69, 72, 75 had pathogenic and lethal effects. B. velezens strain 46 mortality rates reached 85%, and the mortality rate of the remaining strains was between 15% and 55%. For the death experimental groups. Some minor mortality groups may be caused by the deterioration of the culture environment, rather than the pathogenicity of the strain. And the absolutely safety Bacillus isolates for L. vannamei were selected, 3 strains with obvious non-lethality were B. subtilis strain 8, B. velezens strain 57, and B. amyloliquefaciens strain 74 were a safe strain without death, and can be applied as a potential probiotic in L. vannamei (Figure 4). strain 46 mortality rates reached 85%, and the mortality rate of the remaining strains was between 15% and 55%. For the death experimental groups. Some minor mortality groups may be caused by the deterioration of the culture environment, rather than the pathogenicity of the strain. And the absolutely safety Bacillus isolates for L. vannamei were selected, 3 strains with obvious non-lethality were B. subtilis strain 8, B. velezens strain 57, and B. amyloliquefaciens strain 74 were a safe strain without death, and can be applied as a potential probiotic in L. vannamei (Figure 4).

Discussion
Biological antagonism aims to achieve the effect of "treating bacteria with bacteria" under natural conditions. Since Pseudomonas bromoutilis was isolated as antagonistic bacteria for the first time in 1966 [22], more and more antagonistic bacteria have been utilized. Probiotic bacteria isolated from the culture medium might exhibit antagonistic effects on present pathogens. Bacillus isolated from the marine environment can improve diseaseresistant activity against bacterial infection [23]. It is a priority and inherently advantageous to use B. pumilus in the marine environment as a biocontrol agent or probiotic in

Discussion
Biological antagonism aims to achieve the effect of "treating bacteria with bacteria" under natural conditions. Since Pseudomonas bromoutilis was isolated as antagonistic bacteria for the first time in 1966 [22], more and more antagonistic bacteria have been utilized. Probiotic bacteria isolated from the culture medium might exhibit antagonistic effects on present pathogens. Bacillus isolated from the marine environment can improve disease-resistant activity against bacterial infection [23]. It is a priority and inherently advantageous to use B. pumilus in the marine environment as a biocontrol agent or probiotic in aquaculture [24]. To form the basis of morphological characterizations and gene sequencing, the study reflected the diversity of Bacillus in our aquaculture systems. A total of 116 Bacillus isolates were categorized into 24 different species, including B. subtilis, B. velezensis, B. amyloliquefaciens, and B. stercoris, in this study. At present, Vibriosis has become an economically important disease in marine culture and adversely affected many cultured animals [25]. In the shrimp aquaculture system, V. harveyi, V. alginolyticus and V. parahaemolyticus are most frequently isolated [26]. V. campbellii, V. owensii were also pathogens of shrimp acute hepatopancreatic necrosis disease [27,28]. In this study, different Bacillus isolates had a variety of effects and also showed interspecific differences in pathogen inhibition effects. In addition, for the homogeneous Bacillus species, different effects were shown on the same type of Vibrio isolate. In this regard, the isolation of Bacillus from the mariculture system may have provided these isolates with an advantage in the inhibition of pathogens isolated from the marine environment.
According to the guidelines suggested by the European Food Safety Authority, this is essential for the absence of virulence factors in probiotic strains [29]. The presence of toxin genes in probiotic strains could potentially contribute to the prevalence of toxin genes among bacteria through horizontal transfer in some cases. To ensure the safe use of probiotic Bacillus, a safety evaluation must be carried out, common virulence genes must be screened, and the use of strains carrying virulence genes should be reduced as much as possible. In recent years, studies have been focused on the toxin genes character in probiotic Bacillus spp. Fu et al. [19] found that surveillance of anthrax toxin revealed that cya was detected in 8 of 31 farms. Almost half of the isolates could produce enterotoxins and various cytotoxic surfactin-like toxins in Cui et al. [30,31] report, and they summarized the state of current knowledge about the toxins of B, cereus sensu lato to be considered for safety assessment of probiotic candidates. Among the 116 Bacillus spp. isolates, very few isolates detected the presence of virulence genes, and 9 isolates of Bacillus can be excluded by virulence genes' testing. Five isolates (4.3%) were tested as positive for hemolytic cytotoxin K (cytK), which is responsible for severe food poisoning. None of the species were positive for the emetic toxin gene ces, and the prevalence of nheABC and hblCDA was shown to be rare in the comparison [30,32]. Hemolytic activity assays are considered to be an important screening process. Hemolysin is a very common virulence factor, which frequently causes anemia and edema in the host; hence, hemolytic strains are not recommended for use as feed additives. The non-hemolytic strains would be preferable for probiotic use [33]. Generally, nheABC and hblCDA genes were widely distributed among B. cereus isolates from various origins, except for probiotic origin [30]. Deng et al. [34] suggested that nheABC and hblCDA were gradually prevalent in probiotic Bacillus spp. isolates. The current observations suggest that hblC and nheA were detected in B. cereus and B. thuringiensis isolates. No virulence genes were detected in 12 potential probiotics isolates.
Antibiotic resistance is one of the imminent challenges to global health with the current trend of escalating environmental contaminants; furthermore, the natural environment is a potential repository for the spread of antibiotic resistance [35]. Antibiotic susceptibility is one of the most important features of probiotic bacteria [36]. Reacher Yaylaci et al. [37] reported that B. pumilus was susceptible to furazolidone, erythromycin, ampicillin, oxytetracycline, rifampicin, and ciprofoxacin. Sorokulova et al. [38] previously reported that Bacillus spp. may be resistant to chloramphenicol. Our study is not similar in terms of results: 116 strains of Bacillus isolates showed low resistance to antibiotics, and more than 62% isolates were susceptible to florfenicol, doxycycline, sulfamethoxazole-trimethoprim, tetracycline, ofloxacin, enrofloxacin, sulbactam, cefalexin, sulfamethoxazole, ampicillin, azithromycin, thiamphenicol, kanamycin and erythromycin. The MARI reflects the degree of the environmental pollution caused by antibiotics that may be dangerous to human health [39]. A value of higher than 0.2 indicates a high antibiotic exposure risk, while a value lower than 0.2 indicates a low antibiotic exposure risk [40]. The MARI values of the 116 Bacillus isolates ranged from 0 to 0.06, which indicates a low antibiotic exposure risk.
Furthermore, the antibiotic resistance burden has serious implications for human health owing to the potential transfer of antibiotic resistance genes between bacteria, thereby impairing the efficacy of antibiotic treatment and compromising public health [41]. A total of 18 kinds of antibiotic resistance genes were tested, and only tetB, blaTEM, and blaZ were detected in this study. The previous studies indicated that the majority of Bacillus spp. for probiotics exhibited resistance to tetracycline, which mainly resulted in mobile tetracyclineresistance genes such as tetB and tet45 [19,31,32]. In this study, 112 Bacillus isolates showed an antibiotics phenotype that was sensitive to tetracycline, but the tetB detection rate was 100%. Many researchers have observed discrepancies between antimicrobial resistance phenotypes and the prevalence of antimicrobial resistance genes [42,43]. This might be due to the antimicrobial phenotypes that can be expressed upon the stimulation of many different genetic determinants [44]. The macrolide-resistance genes present on extrachromosomal elements have been identified in mobile elements, such as the plasmid-encoded ermC from B. subtilis [45]. Tetracycline-resistance determinants have also been found in mobile elements, including the plasmid-encoded tetL gene from B. subtilis [46], and the tetM contained within the conjugative transposon Tn5397 of B. subtilis [47]. Other tetracycline resistance genes, such as tetK, have been observed in some Bacillus isolates [48]. The detection of chloramphenicol gene cfr (8.62%) and β-lactams gene blaTEM (2.59%) was low. Chloramphenicol resistance can be attributed to the enzymatic inhibition of the drug facilitated by chloramphenicol acetyl-transferases [49]. Beta lactams are the most broadly used class of antimicrobial agents, characterized by minimal toxicity and employed in the treatment of various bacterial ailments, including those attributed with different Vibrio species [50]. Apparently, resistance among the beta-lactam and chloramphenicol drugs was low.

Bacterial Strains and Growth Condition
A total of 116 Bacillus isolates were collected from a mariculture system in China with geographic and chronological differences between 2009 and 2021. All samples were cultured on tryptone soybean broth (TSB) agar medium at 28 • C for 36 h. Single colonies were picked and inoculated on TSB medium to obtain purified bacteria. Then single colony was picked up with a tip under sterile conditions, transferred to 100 µL ultrapure water and repeatedly pipetted and, after 12 min in a metal bath at 99 • C, centrifuged for 5 min at 4 • C and 12,000 rpm with a high-speed centrifuge, the supernatant was taken.
The DNA concentration was determined in the supernatant using a spectrophotometer (Nano Drop 1000) as template DNA in PCR. The PCR system consists of 1 µL of DNA template, 25 µL of Taq Master Mix (Vazyme, Nanjing, China), 1 µL of forward primers and 1 µL of reverse primers, and double-distilled water for a total volume of 50 µL. PCR amplification conditions were: 94 • C, 4 min; 30 cycles: 94 • C, 30 s, 55 • C, 30 s, 72 • C, 1.5 min; store at 72 • C, 10 min, 4 • C. PCR products were sequenced after detection by 1% agarose gel electrophoresis. Primer synthesis and gene sequencing were completed by Sangon Biotech (Shanghai, China). Colonies were preliminarily identified by morphology and 16S rDNA sequencing. BLAST was used for nucleotide comparison to obtain percentage similarity. The sequence analysis was carried out based on recA sequence to more accurately identify the dominant bacteria, as described by Mohkam. Five common pathogenic Vibrio spp. of Litopenaeus vannamei, including V. parahaemolyticus, V. alginolyticus, V. harvey, V. owensii, and V. campbellii, were tested to observe the antibacterial effect of Bacillus isolates. These strains were activated and cultured at 28 • C for 24 h in TSB for activation.

Antibacterial Ability of Bacillus Isolates against Vibrio spp.
Antibacterial activity was assessed using the agar well diffusion assay method. The Bacillus and Vibrio isolates were cultured in TSB media at 28 • C, 180 rpm for 24 h. Then, the bacteria suspension was adjusted to 1 × 10 9 CFU/mL for Bacillus, 1 × 10 6 CFU/mL for Vibrio. 100 µL Vibrio suspension was smeared on the surface of TSB medium and 100 µL of Bacillus suspension was added into the hole and cultured at 28 • C. After 36 h, checks were carried out for the appearance of inhibition haloes surrounding the putative antagonists' spots and the diameter of the inhibition zone was measured with Scan 1200 inhibition zone reader function (Interscience, Saint-Nom-la-Bretèche, France) to screen the beneficial strains with good antibacterial effect. All the experiments were performed in triplicate.

Antibiotic Susceptibility Test
The antimicrobial resistance of 116 Bacillus isolates was determined using the Kirby-Bauer disk diffusion method on TSB agar medium according to the guidelines by the National Committee for Clinical Laboratory Standards with minor changes [51]. A total of 16 kinds of antibiotics were used in this study, including thiamphenicol, erythromycin, kanamycin, ofloxacin, florfenicol, azithromycin, cefalexin, flumequine, sulfamethoxazoletrimethoprim, enrofloxacin, tetracycline, sulfadiazine, neomycin, sulbactam, ampicillin, and doxycycline. Escherichia coli ATCC 25922 was used for quality control. The plates were incubated at 28 • C for 24 h. All the experiments were performed in triplicate. Then, the antibiotic susceptibility was reported as resistant (R), intermediate (I) and sensitive (S) based on the inhibitory zone diameter according to the CLSI standards.
Furthermore, the isolates were considered multiple-antibiotic resistant (MAR) if the isolate was resistant to three or more separate antimicrobial classes [52]. The MAR and multiple antibiotic resistance index (MARI) were measured for all isolates according to the CLSI protocol [53].

Molecular Detection of Potential Virulence Gene among Bacillus Isolates
Based on Fu et al. [54], the presence of 10 kinds of Bacillus-related toxin genes was detected in 116 Bacillus isolates, including hblA, hblC, hblD, nheA, nheB, nheC, cytK, entFM, bceT and ces. The primers and the expected size of the DNA product for each of the investigated genes are shown in Table 5.  The primers were synthesized by Sangon Biotech (Qingdao, China). The DNA of the purified bacteria was extracted by the Beijing Tiangen biological bacterial genome DNA extraction kit. Extracted DNA material served as a template for PCR amplification for each tested sample. Each assay was composed of 1 µL of DNA material, 12.5 µL of Taq Master Mix (Vazyme, China), 1 µL of forward and 1 µL of reverse primers, and double-distilled water to obtain a total of 25 µL. The PCR product was analyzed by electrophoresis using a 1% (w/v) agarose gel and then visualized under a UV transilluminator. The positive bands' sequence information was determined and compared to verify the accuracy of the sequence information.

Bio-Safety Testing in Litopenaeus vannamei
To evaluate the in vivo security of 12 potential probiotics Bacillus isolates, L. vannamei was used in this study. Bacillus were cultured in TSB and incubated with shaking for 36 h at 28 • C. Shrimp (5-8 cm) were randomly divided into 13 groups (12 groups for Bacillus infection and 1 group for control), with each group containing 20 shrimp. The shrimp were injected with 1.0 × 10 8 CFU/mL Bacillus suspension, and PBS was injected as control. Then, mortality was monitored and recorded for 7 days.

Statistical Analysis
Data were preliminary analyzed by excel 2016 software, and all statistical analyses were carried out using SPSS software (version 26.0). The results are expressed as mean ± SE of mean (SEM) with One-way ANOVA at 5%. The colony photographs were adjusted with Adobe Illustrator 2019 to best represent the colony morphology. Survival curves were made by GraphPad.

Conclusions
Our study provides comprehensive diversity and shows the potential risks of the Bacillus species collected from geographically distinct regions in the mariculture system in China for the first time. The phenotype and genetic diversity of different kinds of Bacillus spp. were investigated. An integral evaluation of the virulence and antimicrobial resistance risk of Bacillus isolates was carried out. Then, the potential probiotics were screened with multiple antibacterial activities for different Vibrio species and evaluated the probiotic character. Finally, three potential probiotic strains were found. These results provide basic support for microecological preparations in aquatic disease preservation and promote the healthy and green development of aquatic industry.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest:
The authors declare no conflict of interest.