A potential alternative to traditional antibiotics in aquaculture: Yeast glycoprotein exhibits antimicrobial effect in vivo and in vitro on Aeromonas caviae isolated from Carassius auratus gibelio

Abstract In aquaculture, antibiotics are commonly used to provide protection against pathogens; however, this practice has become controversial due to increased occurrences of microbial resistance, and alternatives are needed. This study aimed to investigate the antimicrobial activity of yeast glycoprotein (YG) against Aeromonas caviae. Pathogens were isolated from liver of diseased Carassius auratus gibelio. Based on morphological and biochemical analysis, together with 16S rRNA gene sequencing, the isolated strains were identified as A. caviae and concluded as clones of a single strain and named L2. Further pathogenicity analysis revealed that A. caviae possessed β‐haemolysis, and its median lethal dose for C. gibelio was 1.33 × 106 CFU/ml. Hepatic adenylate kinase and pyruvate kinase activities of C. gibelio were inhibited post–A. caviae infection. Antimicrobial drug test suggested that A. caviae was a multidrug‐resistant organism but could be inhibited by YG in vitro. Minimum inhibitory and bactericidal concentration of YG was 83.3 mg/ml and 166.7 mg/ml, respectively. Microbiota sequencing results showed that YG supplement could obviously decrease the relative abundance of Aeromonas and increase the microbial diversity. Our study revealed that A. caviae from C. gibelio was a multidrug‐resistant bacteria strain, and could be significantly inhibited by YG in vivo and in vitro, thus providing important insights into ecological control and pathogenesis of A. caviae in aquaculture.


| INTRODUC TI ON
In past decades, antibiotics have been used in animal feed to reduce bacterial infections. However, this relative practice has been banned in the European Union and United States of America (Pradella, 2006), and become increasingly controversial in the world considering the possible adverse effects of antibiotics, such as developing of resistance strains, contaminating the environment, posing threat to food safety and predisposing to later intestinal disease for reduction in microbial diversity (Cotten et al., 2009;Phillips et al., 2004;Young & Schmidt, 2004). China Ministry of Agriculture has gradually reduced types and doses of antibiotics allowed in feed since 2017 with the intent to ban the usage of antibiotics as animal feed additives by 2020. Thus, finding safe and effective alternatives to traditional antibiotics in aquaculture are needed. Although a lot of attempts have been made over the years, such as using vaccine, probiotics and immunostimulants (Munir, Hashim, Nor, & Marsh, 2018;Wang et al., 2019), the effective alternatives to antibiotics in aquaculture are still to be developed. Yeast glycoprotein (YG) containing glycoprotein and cell wall polysaccharides may increase resistance to infections by Gram-negative bacteria, and serve as an immunomodulator because the internal β-dextran can modify immunosuppression (Amorim et al., 2012;Qin et al., 2019). The antimicrobial mechanisms might involve two aspects: (a) glycoprotein could highly gather on the surface of bacterial cell wall by binding to glycoprotein receptor, then interfere with the normal metabolism of bacteria (Amorim et al., 2012); and (b) cell wall polysaccharides activate some cell wall-degrading enzymes, including chitinase and glucanase, resulting the degradation of bacterial cell wall (Fisher, Meroueh, & Mobashery, 2005). Therefore, it might be explored as a natural alternative to prophylactic antibiotics. Besides the role in the digestion and absorption of nutrients, the microbiota in the gut play a critical role in prevention of pathogen colonization. Therefore, research concerning the effect of YG on microbial community will facilitate to illustrate its antimicrobial role. A basal diet containing 800 mg/kg YG has proven effective in improving the gut microbiota in weaned piglets (Qin et al., 2019). However, it is not clear whether YG can play an important role in modifying intestinal microbiota of aquatic animals.

Aeromonas caviae, a mesophilic species belonging to genus
Aeromonas, is one of the most prominent pathogens occurring ubiquitously in the aquaculture industry (Di Ianni et al., 2015;Yi et al., 2013).
The diseases caused by A. caviae in fish are characterized by bacterial septicaemia or surface ulcer with high mortality (Thomas et al., 2013).
So far, great advances have been made with respect to the pathology, epidemiology and diagnostic methods of A. caviae (Baldissera et al., 2018;Janda & Abbott, 2010;Schmidt, Bruun, Dalsgaard, & Larsen, 2001), the liver is one of the organs most affected by A. caviae (Igbinosa, Igumbor, Aghdasi, Tom, & Okoh, 2012). However, the ecological control and pathogenesis of A. caviae in aquaculture are largely unknown.
The detailed information of YG was shown in Figure S1. Based on the assumption that the synergistic effect of YG might strengthen the immune function and intestinal microbiota, we hypothesized that YG may be a suitable alternative to antibiotics and reach the similar effects without negative impact when added in the basal diet of aquatic animals. Different from Aeromonas hydrophila and Aeromonas veronii, research on the pathogenesis, prevention and control of A. caviae in fish is limited. To test the in vivo antimicrobial effect of YG, intestine microbial community of Carassius auratus gibelio feeding with a basal diet containing antibiotics and YG was determined by 16S rRNA sequencing. Thus, the objective of this study was to evaluate the antibacterial effects of YG in vivo and in vitro on A. caviae isolated from C. gibelio, the results will provide important insights into disease control and pathogenesis of A. caviae in aquaculture.

| Fish handling and diet
Carassius auratus gibelio, an important aquatic species widely cultured in China, was chosen to test the pathogenicity of the aetiological agent isolated from naturally infected C. gibelio and in vivo antimicrobial effect of YG. A total of 150 healthy C. gibelio with average weight 50 ± 5 g were purchased from a farm in Beibei District of Chongqing. Fish were allowed to acclimate to laboratory conditions for about 2 weeks before artificial infection. Fish were kept in rearing tanks supplied with running water at 25 ± 2°C and fed daily with basal diet (Xu, Wang, Li, & Lin, 2009). The temperature was kept constant for the experimental period.

| Aetiological isolation and identification
A total of three naturally infected C. gibelio characterized by bacterial septicaemia were obtained from a C. gibelio breeding farm in Beibei District of Chongqing municipality, China. Pathogenic bacteria were isolated from liver and inoculated on lysogeny broth (LB) agar medium at 28°C for 24 hr. The typical colonies from three individuals were subcultured thrice to obtain dominant single colonies. A total of three bacterial strains were obtained and named strain L1 to strain L3.
The three bacterial colonies were subjected to Gram stain, biochemical test and 16S rRNA sequencing for identification. Biochemical identification was carried out according to protocols in Bergey's Manual of Determinative Bacteriology (Buchanan & Gibbons, 1974).
The purified bacteria and A. caviae standard strain were inoculated on indole, VP (voges proskauer), MR (methyl red) and other 29 biochemical identification reaction tubes at 28°C for 12-48 hr.
For molecular identification, bacterial genomic DNA was extracted for 16S ribosomal RNA (16S rRNA) gene amplification following the standard protocol of extraction kit (TIANGEN). The universal primers of 16S rRNA (forward, AGAGTTTGATCCTGGCTCAG; and reverse, ACGGCTACCTTGTTACGACTT) were synthesized by Genomics Institute, Beijing, China. The PCR reaction (50 μl) contained 2 × Taq MasterMix (25 μl), DNA template (100 ng/μl, 3 μl), primers (10 mmol/L, 1 μl) and 20 μl of sterile water. After the initial denaturation (95°C for 5 min), all reactions went through 30 cycles of 94°C for 1 min, 55°C for 1 min and 72°C for 1 min, then 74°C for 10 min. After confirmed by 1% agarose gel electrophoresis, the PCR products were purified using QIAquick PCR purification kit (Qiagen) following the manufacturer's instructions, followed by sequencing in Beijing Genomics Institute, China. The result was compared with the corresponding sequences in NCBI GenBank for homology analysis, and phylogenetic tree was generated by neighbour-joining method (Saitou & Nei, 1987). Finally, to make it available to other researchers for reproducing our study,

A. caviae was submitted to China General Microbiological Culture
Collection Center (CGMCC) for preservation, which will take about 3 months to give a CGMCC number specific for A. caviae. Before the CGMCC number for A. caviae is given, other researchers could also get the isolated A. caviae strain by contacting the author via email.

| Pathogenicity analysis
In vivo, to confirm the pathogenicity of aetiological agent, reproduction of infection was conducted in C. gibelio by intramuscularly injecting with serially diluted A. caviae (strain L2). Aetiological agent was re-isolated and identified according to the above-mentioned methods, and mortality was recorded for calculating median lethal dose (LD50).
Briefly, rapidly growing strain L2 was collected from culture medium and enriched by centrifugation at 500 g for 10 min at 4°C, then suspended in 0.01-M phosphate buffered saline (PBS, pH 7.4). For five experimental groups, fish were injected intramuscularly with 500 μl of serially diluted bacteria suspension (1 × 10 5 , 1 × 10 6 , 1 × 10 7 , 1 × 10 8 and 1 × 10 9 CFU/ml), respectively. As the negative control group, fish were injected intramuscularly with equal volume of sterile PBS. Signs were observed twice a day and mortality was recorded for 2 weeks, then LD50 was calculated based on Reed method (Reed & Muench, 1938).
The extracellular products (ECPs) of strain L2 were prepared according to the methods described by Zhang, Pridgeon, and Klesius (2014) with minor modifications. Briefly, strain L2 was inoculated in liquid medium, tryptic soy broth and shake incubated at 28°C. The bacterial suspension was centrifugated at 8,049 g for 30 min and the supernatant recovered was filtered through a 0.22-μm PES membrane (Millipore), which was referred to ECP. Haemolytic tests of ECPs were carried out in accordance with protocol described in Wang, Xu, Cao, Wang, and Wang (2013). Haemolytic test was conducted in 10% rabbit blood agar medium. The medium plates were inoculated at 28°C for about 36 hr. The occurrence of transparent circle in rabbit blood agar medium indicated positive results of haemolytic activity.
Hepatic adenylate kinase (AK), pyruvate kinase (PK) and adenosine triphosphate (ATP) activities in both infected and uninfected liver of C. gibelio were measured according to the published methods (Dzeja, Vitkevicius, Redfield, Burnettm, & Terzic, 1999;Leong, Lai, Lim, & Clark, 1981;Wen et al., 2015). According to the methods, the results of AK and PK activities were expressed in pmol ATP formed/min/mg of protein and μmol pyruvate formed/min/mg of protein, respectively.
For ATP levels, the results were reported as pmol/mg of protein.

| Antimicrobial activity of antibiotics and YG in vitro
Antibiotic sensitivity was tested by Disc Diffusion Method (Bakht, Islam, Ali, Tayyab, & Shafi, 2011). Pathogenic strains (100 μl, 1 × 10 7 CFU/ml) were inoculated by spreading method (Elguindi et al., 2011), and the test paper of drug sensitivity was pasted on the LB culture medium at 28°C for 24 hr. According to the America The antibacterial activity of YG was evaluated according to the Oxford cup method (Wang, Zheng, Li, Wu, & Xiao, 2017). Briefly, the culture dish, Oxford cups (stainless tube with 6 mm in inner diameter, 8 mm in outer diameter, and 10 mm in height), LB medium and other reagents were sterilized in an autoclave before use. The bacterial suspension was adjusted to 10 6 CFU/ml and coated evenly on culture dish with LB agar medium, then the Oxford cups were placed on the surfaces of the medium and sample solutions of YG (2,000 mg/ml, diluted in sterile distilled water [DW]) were added by dripping. The negative controls were added equal volume of sterile DW. Finally, the culture dishes with the Oxford cups were incubated at 28 ℃ for 24 hr, and the diameter of the bacteria inhibiting loop was measured to evaluate the antibacterial activity of samples. The procedure was performed in triplicate.
The MIC of YG was determined by broth dilution methods according to CLSI (Cockerill et al., 2012). The stock solutions (2,000 mg/ml) of YG were twofold serially diluted in DW to obtain the nine following concentrations: 1,000, 500, 250, 125, 62.5, 31.3, 15.6, 7.8 and 3.9 mg/ml. Nine sterilized glass tubes were added 2 ml of rapid growing A. caviae (1.5 × 10 6 CFU/ml) in LB medium, For the positive control, additional sterilized glass tube was added 3 ml of 1.0 × 10 6 CFU/ml A. caviae. All the tubes were subjected to shake-tube inoculation at 28°C for 24 hr, after which all tubes were observed for growth. MBC of YG was derived by sub-culturing the mixture in LB agar medium at 28°C for 24 hr from tubes that showed no growth (determined by turbidimetric method) in the test of MIC, The minimum concentration of YG that produces total cell death (determined by plate culture) is taken as the MBC.

| In vivo antimicrobial test of YG and sample collection
A pre-experiment lasting for 1 month was carried out to confirm the safety of YG on C. gibelio when fed with a basal diet containing 800 mg/ kg YG, and we found that the survival rate was 100%, no other symptoms were observed, along with its ingredients ( Figure S1) and effects on growth performance of piglet [Qin et al., 2019]), we concluded that YG is safe to C. gibelio when fed with a basal diet containing 800 mg/kg YG.
Healthy C. gibelio were randomly allocated to four equal groups, including YG group, antibiotic group, negative control group and blank control group. Each group contained three replicates with 15 C. gibelio, respectively, and the trial lasted 14 days. All fish of YG group were fed with a basal diet containing 800 mg/kg (we referred to the addition in piglet, performed by Qin et al., 2019) YG (YG group), antibiotic group were fed with a basal diet containing 200 mg/kg Rifampicin serving as a positive control (PC group). The negative control group (NC group) were fed with basal diet. After 2 weeks, each fish in YG group, PC group and NC group was injected intramuscularly with 2 × LD50 viable A. caviae. Fish in blank control (BC group) was injected with equal volume of sterilized saline water. After 3 days post-infection, the intestine of three individuals in each group was collected for microbial diversity analysis.

| Microbial diversity analysis
Microbial DNA was extracted from intestine samples using

Ltd. Operational taxonomic units (OTUs) were clustered with 97%
similarity cut-off using UPARSE and chimeric sequences were identified and removed using UCHIME (Edgar, 2010). Representative sequences defined by abundance from each OTU were aligned to the SILVA bacterial database using PyNAST (Caporaso, Bittinger, et al., 2010;Koetschan et al., 2014). The taxonomy of each 16S rRNA gene sequence was analysed by RDP Classifier algorithm against the Silva (SSU123) 16S rRNA database using confidence threshold of 70% (Caporaso, Kuczynski, et al., 2010). Shannon and Simpson indices were employed for diversity analyses, and the Ace and Chao1 indices were employed for richness analysis.

| Statistical analysis
All data were sorted by excel 2016 after the initial order and analysed using SPSS 21.0 (2015, IBMSPSS Inc.). Variability in all the data was expressed as means ± SEM. Differences between mean values were compared using independent samples T test, and considered significant when p < .05.

| Identification of A. caviae based on morphological, biochemical and 16S rRNA sequencing analysis
Based on morphological, biochemical and 16S rRNA sequencing analysis, the three dominating bacterial colonies (L1, L2 and L3) were identified as A. caviae and concluded as clones of a single strain.
Strain L2 was randomly selected to illustrate its morphological and biochemical characters of A. caviae. It had round and smooth circled shape on the agar plates, examination under microscope showed the bacteria were Gram negative (red) and rod shaped (Figure 1a).
Based on physiological and biochemical identification, a total of 32 characteristics of strain L2 were highly consistent with those in A.

| Pathogenicity analysis of A. caviae
Cumulative mortality of C. gibelio in each group was recorded during 2 weeks of artificial infection, as shown in rabbit blood agar (Figure 2a), showing that ECPs of A. caviae had β-haemolysis.
Liver of three randomly chosen diseased C. gibelio was sampled for analysis of AK and PK activities and ATP levels. The results showed that AK and PK activities significantly decreased (p < .05) by 33.7% and 28.6% in the liver of infected animals compared with liver in uninfected animals, respectively (Figure 2b,c). ATP levels in livers also showed an obvious decrease (42.8%, p < .05) post-infection (Figure 2d).

| In vitro antimicrobial activity of antibiotics and YG
A total of 30 different antibacterial drugs were used in the tests, the results showed that A. caviae was sensitive to seven drugs, including ceftazidime, cefotaxime, vancomycin, nitrofurantoin, aztreonam, rifampicin and amikacin (Table 2). After incubating at 28℃ for 24 hr, an obvious bacteria-inhibiting zone appeared around Oxford cups adding YG in three culture dishes (Figure 3a). The assay was performed in triplicate, the average diameters of inhibiting zone was 25 mm (Figure 3b). No inhibiting zones were observed around negative controls adding DW.

| MIC and MBC of YG to A. caviae
After inoculation at 28°C for 24 hr, bacterial turbidity significantly decreased in tube 1-3 compared with tube 10 (positive control), and turbidity of tube 4-9 was similar to tube 10 ( Figure 4a). To further determine the MBC of YG, the products in tube 1, 2, 3, 4 and 10 were sub-cultured in LB agar plates at 28°C for 24 hr.
The results showed no bacteria growth was detected in products from tube 1 and 2, some bacterial growth was observed in tube 3 and nearly equally actively growing bacterium with positive control plates (tube 10) was detected in products of tube 4

| YG-associated changes in Aeromonas in intestinal microbiota
The influence of YG on the intestinal microbiota at the genus level is shown in Figure 5. Community barplot analysis demonstrated

| D ISCUSS I ON
The bacterial colonies isolated from liver of diseased fish were identified as A. caviae according to colony morphology, biochemical and physiological characteristics. A. caviae was confirmed as a causative pathogen because satisfying Koch's postulates in normal C. gibelio.
PCR was performed to identify the bacterial isolate at molecular level using primers specific to 16S rRNA gene of A. caviae. To our knowledge, this is the first pathogenic diagnosis of A. caviae in C.
gibelio, and there is limited research on ecological prevention and control of A. caviae in aquaculture even though it occurs ubiquitously in aquaculture industry (Ringø & Vadstein, 1998).
Reproduction of infection demonstrated that normal C. gibelio was susceptible to A. caviae infection, the LD50 of A. caviae for C. gibelio is 1.33 × 10 6 CFU/ml, indicating that A. caviae was highly virulent to C. gibelio. The production of haemolytic toxins has been regarded as indication of pathogenic potential, and all Aeromonas isolates with haemolysin-positive genotype were virulent in the suckling mouse assay model (Wong et al., 1996). In the present pathogenicity analysis, haemolytic test of ECPs showed that A. caviae possessed β-haemolysis, which was potentially associated with the signs of infected A. caviae, such as haemorrhage in the liver. AK and PK are two key enzymes linked to the communication between sites of ATP generation and ATP utilization, and a recent study found that experimental infection by A. caviae could decrease AK and PK activities in experimentally infected silver catfish (Baldissera et al., 2018). In this study, we observed that hepatic AK and PK activities in C. gibelio were inhibited by A. caviae, which resulted in decreased availability of hepatic ATP. Therefore, the inhibition of hepatic AK and PK activities by A. caviae infection might contribute to disease pathogenesis in C. gibelio.
The antibiotic sensitivity test showed that A. caviae was multi-antibiotic, and its drug-resistant characteristics were different from A. caviae isolated from other species (Huang, Wang, Zheng, & Huang, 2010;Yu & Ma, 1997). As A. caviae infecting C. gibelio has multiple antibiotic resistance, development of a microecological preparation is an alternative and promising prophylactic method to provide protection against A. caviae. A previous study showed that YG supplement could obviously increase the relative abundance of Lactobacillus in weaned piglets (Qin et al., 2019), it was not clear about its effect on pathogen and intestinal microbiota of aquatic animals. This study showed that YG could effectively inhibit A. caviae replication in vitro. Community barplot analysis demonstrated that YG supplement could more significantly increase the microbial diversity in the gut and decrease the relative abundance of Aeromonas than antibiotic did in vivo. Interestingly, we found that bacterial abundance of Erysipelotrichaceae and Porphyromonadaceae was highest in the YG supplement group.
According to the previous studies, these two key bacterial families belong to indigenous microbiota and are related to4 the digestion of protein and energy (Bermingham, Maclean, Thomas, Cave, & Young, 2017;Sakamoto, 2014). Therefore, YG supplement could efficiently inhibit A. caviae both in vivo and in vitro, and contribute F I G U R E 5 The influence of YG on the composition of intestinal microbiota at the genus level. BC: fish group feeding with the basal diet and injecting with sterilized saline water; NC, YG and PC group were injected with viable Aeromonas caviae. NC: artificial infected fish group feeding with the basal diet; YG: fish were fed with a basal diet containing 800 mg/kg YG. PC: fish were fed with a basal diet containing 200 mg/kg Rifampicin. Sequences that could not be classified into any known group were designated as "Unclassified". Statistical comparisons were performed by the Mann-Whitney test. YG, yeast glycoprotein to intestine health. Further study concerning optimizing the adding level of YG supplement is needed.
To conclude, this is the first report that C. gibelio is susceptible to A. caviae infection. Its β-haemolysis and decrease of AK and PK activities might be potentially associated with virulence and disease pathogenesis. Aeromonas caviae isolated from C. gibelio was a type of multiple antibiotic-resistant bacteria strain and highly virulent to C. gibelio. Fortunately, it could be effectively inhibited by YG in vivo and in vitro, providing a potentially safe and effective alternative method to traditional antibiotics. These conclusions shed light on ecological control and pathogenesis of A. caviae in aquaculture.

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest.

AUTH O R S' CO NTR I B UTI O N S
Yun Li, corresponding author, involved in the study design and concept, statistical analysis and manuscript authorship. Ronghua Wu and Junyu Shen involved in the project implementation, data acquisition, editing and major revisions of manuscript. Dandan Tianand Jiaqian Yu involved in the fish handling and data sorting. Tao He and Jianhua Yi involved in analysis of report and technical support. All authors have read and approved the final version of this manuscript.

E TH I C A L A PPROVA L
This study was conducted in strict accordance with the recommendations in the Guide for the Use of Experimental Animals of Southwest University, China. Before sacrificing and handling, experimental C. gibelio were anesthetized with ethyl 3-aminobenzoate methanesulfonic acid (MS222, Sigma, USA), and all efforts were made to minimize suffering.