Antibiotic Resistance of Bacteria Isolated from the Internal Organs of Edible Snow Crabs

Antibiotic resistance and microbiota within edible snow crabs are important for the Chionoecetes (snow crab) fishing industry. We investigated these parameters using culture methods and antibiotic susceptibility tests with six internal organs from three species of Chionoecetes. Each sample revealed many unexpected microbial species within Chionoecetes internal organs. On the basis of 16S rRNA sequence analysis of 381 isolates, the most abundant genera identified in Chionoecetes opilio were Acinetobacter spp. (24%), Bacillus spp. (4%), Pseudomonas spp. (34%), Stenotrophomonas spp. (28%), and Agreia spp. (11%). In Chionoecetes sp. crabs, Acinetobacter spp. (23%), Bacillus spp. (12%), and Psychrobacter spp. (20%) were most prevalent, while Agreia spp. (11%), Bacillus spp. (31%), Microbacterium spp. (10%), Rhodococcus spp. (12%), and Agrococcus spp. (6%) were most abundant in C. japonicus. Our antibiotic resistance test found resistance to all nine antibiotics tested in 19, 14, and two of the isolates from C. opilio, Chionoecetes sp., and, C. japonicus respectively. Our results are the first to show that microbes with antibiotic resistance are widely distributed throughout the internal organs of natural snow crabs.


Introduction
Snow crabs belong to the subphylum Crustacea, Order Decapoda, Family Majidae, and Genus Chionoecetes. These Chionoecetes are found in colder water at depths less than 2000 m where it is muddy or sandy [1]. Three kinds of Chionoecetes are prominent on the east coast of Korea, including snow crab (Chionoecetes opilio), red-tanner crab (Chionoecetes japonicus), and the hybrid Neodo-Daege (Chionoecetes sp.) [2]. Chionoecetes fishing is a major industry and source of income in the area [3][4][5][6]. However, contamination of the East Sea of Korea from human activities has raised serious sanitation concerns that could potentially threaten this industry [7]. Research on antibiotic resistance has primarily focused on human disease; there is limited understanding of antibiotic resistance genes in natural environments. The relationship between environmental microorganisms and human pathogens is not clear; a recent report showed that soil bacteria and human pathogens shared an antibiotic resistome [8]. In this study, we analyzed the microbiota within parts of Chionoecetes using culturing methods, and the culturable microbial isolates were tested for antibiotic resistance. Here, we report the microbial populations and antibiotic resistance of isolates from the internal organs of Chionoecetes. These results may be used to monitor snow crab populations and to identify potentially dangerous changes in microbiota that could threaten the snow crab industry.

Collection of snow crabs
No specific permissions were required for these locations or activities because many snow crabs were sold daily at the market near the Harbor. Wild-caught, uncooked snow crabs (Chionoecetes spp.) were collected from a retail seafood shop at Jukbyeon Harbor and were placed into clean, resealable plastic bags. Samples were stored in a cooler during transfer from Jukbyeon Harbor to the laboratory where they were then stored at 4uC until processed. The morphological characteristics of snow crabs can be distinguished by their carapace color, the arrangement of granules on the lateral carapace, and the presence or absence of spines on the lateral carapace [4]. The crabs were originally caught in the East Sea of Korea [5,6]. There are no-take periods from June through November for snow crab (Chionoecetes opilio) and July through August for red-tanner crab (Chionoecetes japonicus), and no prohibition for Neodo-Daege (Chionoecetes sp.) for management of the snow crab industry.

Sample preparation
Samples were divided into the following six parts: guts (D), gills (G), heart (H), leg meat (LS), carapace meat (S), and carapace juices (J). Each sample was homogenized with 10 mL of 10 mM potassium phosphate buffer, and 100 mL of each sample was spread onto agar plates.

Enumeration of microbial populations
The serial dilutions were spread plated on various media to determine microbial counts of the Chionoecetes spp. in sterile water, and the dilutions were dispensed onto agar plates. For all experiments reported herein, we cultured aerobic microbes on the following media: YPD (yeast extract, peptone, dextrose, BD Bioscience, USA) with chloramphenicol (100 mg/L) and streptomycin (100 mg/L) for yeast; PDA (potato dextrose agar, BD Bioscience) with chloramphenicol (100 mg/L) and streptomycin (100 mg/L) for fungi; R2A (BD Bioscience); NA (nutrient agar, BD Bioscience); TSA (tryptic soy agar, BD Bioscience) for general bacteria; or MA (marine agar, BD Bioscience). The plates were incubated aerobically at 28uC for 1 wk, and average CFU (colonyforming units) values were obtained from triplicate plate counts.

Isolation of bacteria
All colonies from individual plates of one plate or two plates from the plates of colony counts were picked up and cultured separately. In total, 381 individual isolates were transferred to fresh plates three times and then processed for sequencing of 16S rRNA and ITS genes.

Sequencing
The primers used to amplify the 16S rRNA and ITS genes for bacteria and other microbes were 27F and 1492R [9] and ITS1 and ITS4 [10], respectively. The PCR reaction was performed with 20 ng of genomic DNA as the template in a 30-mL reaction mixture by using EF-Taq DNA polymerase (Solgent, Korea). The thermocycler conditions were 95uC for 5 min, followed by 35 cycles of 95uC for 2 min, 55uC for 60 s, and 72uC for 60 s, then a final extension step for 10 min at 72uC. Thereafter, the amplification products were purified using a multiscreen filter plate (Millipore Corp., Bedford, MA, USA). Sequencing reactions were performed using a PRISM BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Bio-systems, Foster City, Calif., USA). The DNA samples containing the extension products were added to Hi-Di formamide (Applied Biosystems, Foster City, CA). The mixture was incubated at 95uC for 5 min followed by 5 min on ice, and then analyzed using an ABI Prism 3730XL DNA analyzer (Applied Biosystems, Foster City, CA). DNA sequencing of isolates was performed by Macrogen Inc. (Seoul, Korea).

Phylogenetic trees
The 16S rRNA and ITS sequences were aligned using the Nearest Alignment Space Termination (NAST) aligner [11]. Aligned sequences were then compared to the Lane mask using the Greengenes website [12]. Sequence matching to the Ribosomal Database Project [13] was used to find GenBank sequences representing the most closely related type of strain for each isolate. These type strains were included as references in the phylogeny using the Greengenes Automatic Taxonomic Classification algorithm [12]. Phylogenetic trees were constructed using neighbor-joining [14] with MEGA5 for Windows [15]. Evolutionary distances were calculated with the Kimura 2-parameter method. Bootstrap analyses of the neighbor-joining data were conducted based on 1000 samples to assess the support for inferred phylogenetic relationships.

Antibiotic susceptibility test
Bacteria were considered susceptible to a particular antibiotic if the bacteria formed a clear zone around a disc on the media (disc diffusion susceptibility testing). After autoclaving the nutrient agar and cooling the agar to 50-55uC, bacterial colonies were mixed into the autoclave medium flask and poured onto petri dishes. Each antibiotic susceptibility testing disc (BD Bioscience) was placed onto a plate and incubated at 28uC for 24 or 48 h. The results indicated whether the isolates were resistant or susceptible to each antibiotic.
PCR assays for detection of resistance genes and sequencing of the PCR products Bacteria were tested through PCR method with the primers of antibiotic resistance genes as shown in Table 1. For the PCR, the reaction mixtures contained 10 ml 26 DNA polymerase enzyme (PowerAmp TM 26 premix), 4 ml primer mixtures, 5 ml template DNA and sterile distilled water to bring the final volume to 20 ml. The PCR was performed with TaKaRa PCR Thermal Cycler Dice TP600 (TaKaRa, Japan). The reaction was started with a 15min denaturation step at 95uC. In the PCRs, the temperature cycles consisted of 30 sec at 95uC, followed by 1 min at 58uC and 1 min at 72uC and each cycle was repeated 35 times. The final cycle was followed by incubation of the reaction mixture for 10 min at 72uC. Amplified PCR products were analyzed by gel electrophoresis in 2% agarose gels stained with ethidium bromide, visualized with ultraviolet illumination, and imaged with the Gel Doc 2000 documentation system (Bio-Rad, Hercules, CA, USA). DNA sequencing of antibiotic resistance genes was performed by Macrogen Inc. (Seoul, Korea).

Culturable microbiota
A total of 381 isolates, including 221 from C. opilio, 76 from C. japonicus, and 84 from Chionoecetes sp. were isolated using six different media ( Figure 1A). Figure 1 summarizes the phylogenetic distribution of the 16S rRNA gene sequences. The genera Pseudomonas, Stenotrophomonas, and Acinetobacter predominated the bacterial communities found in Chionoecetes, commonly representing more than 60% of the sequences isolated from the three crab species. The microbial isolates from C. japonicus and C. opilio included 19 and 20 genera, respectively, whereas the isolates of Chionoecetes sp. included 21 genera ( Figure 1A). The greatest diversity of microbiota among the six different internal organs investigated was found in the gills; we identified 14, 15, and 17 genera in gills of C. japonicus, C. opilio, and Chionoecetes sp. respectively. Nine or fewer genera were identified in the other organs ( Figures 1B and S1).
Isolates from snow crab were predominantly Acinetobacter spp.  Figure S1g). Bacillus spp. were common in all three species of snow crabs. Pseudomonas spp. (n = 68 isolates) were predominantly found in C. opilio, while only three and two bacterial species were isolated from the other two Chionoecetes spp. (Figures 1 and S1a). Acinetobacter spp. were predominant in C. opilio and C. japonicus (53 and 18 isolates, respectively), and only one bacterial isolate was found in Chionoecetes sp. (Figures 1 and S1b). Stenotrophomonas spp. dominated in C. opilio (62 isolates), while only one bacterial species from this genus was isolated from each of C. japonicus and Chionoecetes sp. (Figures 1 and S1c). Numbers of Bacillus spp. isolated included nine from C. opilio, nine from C. japonicus, and 26 from Chionoecetes sp. (Figures 1 and S1d).
In C. opilio, 67% (nine isolates) of Bacillus spp. were localized in the heart, while 33% (nine isolates) were found in the gill and carapace in C. japonicus. Bacillus spp. from Chionoecetes sp. were found in the carapace juices and heart (26% and 30% of 26 isolates respectively). Some bacterial isolates were confined to particular parts of Chionoecetes. Some isolates were not amplified using 16S rRNA gene sequencing; therefore, Rhodotorula (one strain) and Pichia (three strains) were identified from the gills of Chionoecetes sp. using ITS gene sequencing.

Enumeration of total cultivable bacteria
As shown in Figure 2, aerobic bacteria counts numbered from 10 3 to 10 5 cells/g in gills and from 10 2 to 10 3 cells/g in carapace meat in the three snow crab species on TSA, R2A, MA, and NA media, whereas no aerobic bacteria or other prokaryotes appeared on YPD or PDA media (Figures 2A and 2B). In comparison with the other organs, gills of the three crab species contained relatively high aerobic bacterial populations (up to 10 5 ) on TSA, R2A, MA and NA (Figures 2A and 2C).

Enumeration of antibiotic-resistant bacteria
More than 50% of the 221 C. opilio isolates were resistant to ampicillin, erythromycin, penicillin, ticarcillin, and vancomycin. Isolates from C. japonicus and Chionoecetes sp. were resistant to the nine antibiotics tested, representing a 15% and 30% resistance ratio respectively (Figure 3 and Table 2). In C. opilio, Pseudomonas, Acinetobacter, Enterobacter, Psychrobacter, Stenotrophomonas, and Lactobacillus spp. were resistant to the nine antibiotics tested (19 isolates); Pseudomonas, Acinetobacter, and Stenotrophomonas spp. were resistant to seven or eight of the nine antibiotics (12 isolates; Figures S1a, S1b, S1c and Table 2). The MDR bacteria identified in our study were affiliated with Agreia and Psychrobacter; two isolates from C. japonicus (red-tanner crab; Figure S1f and Table 2) and nine genera (14 isolates) from Chionoecetes sp. were affiliated with Shewanella, Rhodococcus, Agrococcus, Leifsonia, Deinococcus, Staphylococcus, and Agreia bacteria, and with Rhodotorula and Pichia yeasts ( Figure S1g and Table 2).

Phylogenetic distribution
Phylogenetic analysis revealed that isolates within Pseudomonas and Stenotrophomonas were grouped separately, representing 66 and 62 isolates from C. opilio (snow crab, C), while two and three isolates from C. japonicus (red-tanner crab, CJ) and Chionoecetes sp. (Neodo-Daege, B) were represented (Figures S1a and S1c). For Acinetobacter spp., 49, 14, and one isolate were found in C. opilio, C. japonicus, and Chionoecetes sp., respectively ( Figure S1b). In C. japonicus three clusters formed with high similarly to A. johnsonii DSM 6963, A. haemolyticus DSM 6962, and A. guillouiae ATCC 11171 ( Figure S1b). Up to 26 Bacillus spp. were isolated from Chionoecetes sp. and nine Bacillus spp. isolates were found in each of C. opilio and C. japonicus ( Figure S1d). These isolates showed similarity with several comparative groups, suggesting that the relationship between the species of Chionoecetes and the isolated Bacillus spp. was not strong ( Figure S1d). Other actinobacteria were found in Chionoecetes sp. and C. japonicus (34 and 21 isolates, respectively). Interestingly, 10 isolates of Salinibacterium from C. japonicus were clustered together (Figures S1f and S1g). Isolates from Chionoecetes sp. were affiliated with Microbacterium (eight isolates), Rhodococcus (nine isolates), and Deinococcus (two isolates; Figure S1g). Proteobacterial taxa were affiliated with 16 isolates of c-proteobacteria from a total of 17 isolates from C. japonicus ( Figure  S1f). Isolates from C. opilio were affiliated with a-(seven isolates), b-(five isolates), and c-proteobacteria (nine isolates; Figure S1e). Bacteroidetes taxa from C. japonicus were affiliated with Flavobacterium (two isolates) and Chryseobacterium (one isolate) in Chionoecetes sp. Bacilli, excluding the Bacillus, were isolated from C. opilio (five isolates), C. japonicas (five isolates), and Chionoecetes sp. (six isolates). Specifically, Exiguobacterium was isolated from C. japonicus ( Figure  S1f).

PCR detection of antibiotic resistance genes
We tested the antibiotic multiresistant bacteria, i.e., Acinetobacter spp., Leclercia sp., Pseudomonas spp., Stenotrophomonas spp., Lactobacillus spp., and Bacillus sp. for the detection of antibiotic resistance genes, i.e., bla SHV , bla OXA , and bla TEM as ampicillin resistance genes; catA1, catA2, catA3, and catB3 as chloramphenicol resistance genes; ermA, ermB, and mefA genes as erythromycin resistance genes; pbp2a as penicillin resistance gene; aphA-3 as kanamycin resistance genes; tetA, tetB, tetC, and tetG as tetracycline resistance genes; and vanA and vanB as vancomycin resistance genes. The catA1 gene was detected in all the tested bacteria. However, resistance genes against ampicillin, erythromycin, penicillin, and kanamycin were not found in the tested bacterial isolates. Interestingly, Leclercia sp. possessed catB and Pseudomonas sp. possessed tetB. Vancomycin resistance gene, vanB was detected in Pseudomonas spp. and Stenotrophomonas spp (Table 3). Sequencing analysis of the PCR products showed that the sequences of catA1 gene were identical, with 100% nucleotide homology in the tested isolates except for the gene of Leclercia spp. The catA1 sequence of Acinetobacter sp. C-G-MA6 (AB826491) and Pseudomonas sp. C-D-MA7 (AB826493) showed that the gene represented 99% and 100% nucleotide identity to an antibiotic resistance gene of Klebsiella pneumonia subsp. pneumonia KPX plasmid pKPX-1 DNA. Moreover, amino acid sequences translated from the nucleotide sequences of the PCR products showed 100% identity with the amino acid sequence of chloramphenicol acetyltransferase. This result indicates that  catA1 gene is derived from the chloramphenicol resistance gene detected in several pathogenic bacteria.

Discussion
Until recently, the study of Chionoecetes was conducted mainly by artificial cultivation for the examination of disease [23,24]. In contrast, the present study was performed to characterize the types of microbiota within Chionoecetes, as little is known about microbial dynamics within Chionoecetes [23,24]. In this study, we provided ratios of antibiotic resistance and microbial community distribution in three Chionoecetes species. Our results indicate that future microbial studies of Chionoecetes in their natural ecosystems are necessary to assess and monitor potential human risk.
This study revealed four genera prevalent in Chionoecetes: Pseudomonas, Acinetobacter, Stenotrophomonas, and Bacillus ( Figure S1). Microbial diversity was high in the gills of C. opilio ( Figure 2C). Schuwerack et al. [25] reported that bacterial colonies enmeshed in polysaccharide-like films produced indentations in the gill cuticular surfaces and dissociation of microvillus membranes at the basal zone of epithelial cells of gill lamellae of the fresh crab Potamonautes warren.
The yeasts Rhodotorula and Pichia, which were identified from Chionoecetes sp. Rhodotorula, Cryptococcus, Torulopsis, Candida, Trichosporon, and Aureobasidium, have previously been isolated from the meat of Dungeness (Cancer magister) and King crabs (Paralithodes camtschatica) [26]. The discovery of yeast in C. opilio in the present study, as well as in Dungeness and King crabs [26], suggests that future ecological studies of yeast populations will be necessary as well.
While several studies show conclusively that antibiotic resistance is a natural phenomenon that predates the modern selective pressures of clinical antibiotics and agricultural use of antibiotics [27][28][29][30][31][32][33], human activity has probably increased the prevalence of MDR bacteria in air, soil, and marine and freshwater ecosystems. Most antibiotic resistance genes are acquired through horizontal gene transfer [8]. In this study, MDR bacteria from Chionoecetes demonstrated antibiotic resistance in nonclinical environments, suggesting an ecological role for antibiotics that warrants additional investigation. Some Bacillus spp. (e.g., Bacillus cereus) are ubiquitous in nature and constitute a major portion of the microbial populations in contaminated food, causing food spoilage and poisoning to the detriment of the consumers [34]. Two Bacillus species are considered medically significant: B. anthracis, which causes anthrax, and B. cereus, which causes a food-borne illness [34]. Because we found Bacillus spp. in all Chionoecetes, the incidence and survival of Bacillus spp. is thought to be controlled by cooking Chionoecetes at high temperatures prior to consumption.
In terms of antibiotic-resistant Acinetobacter spp., the bacteria were found in C. opilio and C. japonicus; infections generally occur in hospitalized patients with weakened immune systems. Therefore, understanding antibiotic resistance in C. opilio isolates is clinically important for cases involving multidrug resistance (MDR) [35]. In this study, 54 isolates of Acinetobacter spp. were isolated from C. opilio, more than 60% of which were resistant to at least one of the antibiotics tested. Antibiotic resistance was high not only for the dominant Pseudomonas, Acinetobacter, and Stenotrophomonas (c-proteobacteria), but was also high regardless of whether bacterial strains were gram negative or gram positive. c-Proteobacteria demonstrated high antibiotic resistance to isolates from C. opilio, while the actinobacteria of Chionoecetes sp. were resistant to all of the nine antibiotics tested (Table 2). Here, we revealed that actinobacteria were commonly isolated from Chionoecetes sp., including many multidrug-resistant strains (Table 2). Moreover, clinical reports of secondary urinary or respiratory infections by Pseudomonas and Enterobacter spp. have been presented [36,37]. While isolates of snow crabs rarely infect respiratory organs or the skin, proper heating of food prepared from Chionoecetes must be ensured in order to protect against infection.
The results of the present genetic study showed that the catA1 gene is widespread in many bacteria (Table 3). These data indicated that this gene moved between species via horizontal gene transfer. However, whether the multiresistance of the bacteria could be derived from intrinsic characteristics of bacteria or from unknown mechanisms (e.g., uncharacterized specific genes and dissemination through unknown transposable elements) is an open question. Therefore, we suggest that further studies are necessary to elucidate whether the resistance gene of snow crabs is intrinsic or arises from horizontal gene transfer between the environmental and pathogenic resistomes. Additional research is required to determine how resistance genes become incorporated into a range of bacteria species. In the future, it is essential that the implications of MDR for human consumption of snow crabs be entirely understood and that the penetration of antibiotic resistance into natural environments be prevented.
In summary, we revealed for the first time a high level of microbial infiltration or inclusion in the internal organs of three Chionoecetes species. In addition, we isolated 381 microbial strains from three species of Chionoecetes spp.; unexpectedly, microbes with antibiotic resistance are widely distributed throughout the internal organs of wild, commercial snow crabs. In the future, additional research on antibiotic resistance and its mechanism and on microbial dynamics in the fishery industry will enhance further understanding of the clinical and ecological implications of these results.  GN : gram-negative bacteria, 12) GP : gram-positive bacteria. doi:10.1371/journal.pone.0070887.t002