Diversity, enumeration, and isolation of Arcobacter spp. in the giant abalone, Haliotis gigantea

Abstract Arcobacter have been frequently detected in and isolated from bivalves, but there is very little information on the genus Arcobacter in the abalone, an important fishery resource. This study aimed to investigate the genetic diversity and abundance of bacteria from the genus Arcobacter in the Japanese giant abalone, Haliotis gigantea, using molecular methods such as Arcobacter‐specific clone libraries and fluorescence in situ hybridization (FISH). Furthermore, we attempted to isolate the Arcobacter species detected. Twelve genotypes of clones were obtained from Arcobacter‐specific clone libraries. These sequences are not classified with any other known Arcobacter species including pathogenic Arcobacter spp., A. butzleri, A. skirrowii, and A. cryaerophilus, commonly isolated or detected from bivalves. From the FISH analysis, we observed that ARC94F‐positive cells, presumed to be Arcobacter, accounted for 6.96 ± 0.72% of all EUB338‐positive cells. In the culture method, three genotypes of Arcobacter were isolated from abalones. One genotype had a similarity of 99.2%–100.0% to the 16S rRNA gene of Arcobacter marinus, while the others showed only 93.3%–94.3% similarity to other Arcobacter species. These data indicate that abalones carry Arcobacter as a common bacterial genus which includes uncultured species.

still unknown (Houf et al., 2009). In contrast, other species have not been directly associated with animal or human diseases.
These results suggest that Arcobacter spp. are widely distributed in marine invertebrates, and potentially indigenous bacteria may play some important role in the host. However, knowledge on the presence and diversity of Arcobacter associated with marine invertebrates including abalone is still lacking compared to pathogenic Arcobacter.
Therefore, in order to gain knowledge on Arcobacter spp. in marine invertebrates, we tried to explore the diversity and abundance of the genus Arcobacter in the giant abalone Haliotis gigantea, an important fishery resource inhabiting shallow water environments. We used cultivation-independent methods, such as Arcobacter-specific clone libraries and fluorescence in situ hybridization (FISH). We also attempted to isolate Arcobacter strains using selective cultivation, and report here the genetic relationships between successfully isolated strains.

| Sample collection and DNA extraction
Nine cultivated giant abalones, H. gigantea, (sample code: CA) and rearing water samples from two tanks (sample code: RW) were collected from the Owase Farming Fishery Center (Owase, Mie, Japan) in February 2012. These samples were supplied for construction The internal organs, including the gut and gills, were collected from the abalones followed by the previously described method (Tanaka et al., 2004). To tear off bacterial cells from their host tissue, we used a beads beater on the condition of slower stroke and shorter time (We state that the method will not allow obtaining the whole bacterial diversity present in the host tissue as compared to an approach based on completely grinding the host tissue). Abalone specimens were pooled into each tube and homogenized using a beads beater (4,200 rpm, 30 s; Tietech Co., Nagoya, Japan). Host tissues were removed from CA samples by quick centrifugation (1 s, 8,000 g), and the supernatant was transferred to new tubes and centrifuged for 20 min at 15,000 g to recover bacterial cells.

| Fluorescent in situ hybridization
The total number of Arcobacter cells in abalone samples was counted using the FISH method (Kepner & Pratt, 1994). Aliquots of bacterial pellets obtained from CA and RW samples were rinsed by 600 µl of sterile PBS and centrifuged at 12,000 g for 5 min, An epifluorescence light microscope (Eclipse 400; Nikon Corp., Tokyo, Japan), was used for observing the stained cells. Due to a technical error during sampling, we were unable to detect the mean ± SE from RW samples.

| Isolation of Arcobacter spp.
For the isolation of Arcobacter species, the procedure described by Salas-Massó et al. (2016) was followed, but the media was slightly modified by changing 2.5% NaCl to artificial seawater. The bacterial mixture from three abalones was diluted 10 times using sterile Daigo's Artificial Seawater SP (Nihon Pharmaceutical Co., Tokyo, Japan) and 100 µl of the diluted mixture was inoculated into Arcobacter Broth (Oxoid Ltd., Hampshire, UK, USA) with CAT supplement [cefoperazone at 8 mg/L, amphotericin B at 10 mg/L and teicoplanin at 4 mg/L] (Oxoid Ltd., Atabay & Corry, 1997) (Atabay & Corry, 1997). Next, the filters were carefully removed and the flow-through was spread on Marine Agar 2216. The media was incubated at the same temperature and time as the primary culture.
After cultivation, presumed Arcobacter colonies (tiny and beige to off-white in color) were selected and applied to colony PCR in the same way as Arcobacter-specific clone libraries using ARC94F and ARC1446R primers. Positive PCR products were sequenced using standard Sanger sequencing.

| Cluster analysis of the bacterial community structure
For each sample, sequences were aligned and grouped in Operational Taxonomic Units (OTUs) with >97% sequence identity (Stackebrandt & Goebel, 1994 (Thompson, Higgins, & Gibson, 1994), using MEGA 7.0 (Kumar, Stecher, & Tamura, 2016). A phylogenetic tree was constructed using the maximum-likelihood method of MEGA 7.0, with 1,000 replicates in the bootstrap analysis and Kimura's two-parameter model (Kimura, 1980). Distances were estimated with the Jukes-Cantor correction.

| Arcobacter-specific clone libraries
To perform a comprehensive search for Arcobacter spp. from abalone and their surrounding seawater, we established Arcobacterspecific 16S rRNA gene clone libraries. A total of 120 and 30 clones were obtained in our study from abalone (CA) and rearing water (RW) samples, respectively (Table 2). Among these clones, Arcobacter sequences were observed as 12 OTUs from CA and seven OTUs from RW.
TA B L E 2 16S rRNA gene sequences identified in the clone library and isolation from abalone or seawater Arcobacter marinus strain CL-S1 (EU512920) 100 The 16S rRNA genes identified from abalone (CA) using the

| Isolation of Arcobacter Spp.
When incubation of samples from natural abalones collected in Mie,
Various methods including enterobacterial repetitive intergenic consensus PCR (ERIC-PCR), randomly amplified polymorphic DNA-PCR (Houf et al., 2002), and amplified fragment length polymorphism (On, Harrington, & Atabay, 2003 have been used to detect and elucidate the transmission routes or to trace the sources of Arcobacter outbreaks . Multiplex-PCR that can detect multiple species simultaneously has also been used (Brightwell et al., 2007;Houf et al., 2000;Khan et al., 2017). ERIC-PCR in particular has been successfully applied to outbreak investigations (Vandamme et al., 1993) in food. Although these methods have advantages due to its simplicity and cost, they detect only a specific species from isolated strains or from mixed cultures based on specific culture conditions (González, Bayas Morejón, & Ferrús, 2017).
To prevent bias resulting from culture-dependent methods in this study, we used Arcobacter-specific clone libraries to directly identify 16S rRNA gene sequences. Twelve OTUs relating to Arcobacter were detected from abalones using Arcobacter-specific clone libraries (Table 2), all clustered with previously reported Arcobacter sequences ( Figure 2). All the OTUs showed similarity to 16S rRNA genes detected from marine environments such as marine invertebrates or seawater, but not from those identified from terrestrial sources such as poultry.
Furthermore, these sequences are not classified with any other known Arcobacter species. Interestingly, using our analytical approach, the samples from abalones also did not show the presence of pathogenic Arcobacter spp., A. butzleri, A. skirrowii, and A. cryaerophilus, commonly isolated or detected from bivalves. Bivalves such as mussels and clams are filter feeders that feed plankton using gills, while abalones feed on brown algae. Thus, they have a more developed digestive system compared to bivalves. The result suggests that gastropods such as abalone may not be host or harbor pathogenic Arcobacter species, perhaps due to their different feeding habits and digestive system.
In this study, we employed the ARC94F Arcobacter-specific probe for FISH against cells isolated from abalone for detection and quantification. This ARC94F probe has been used for specific counting of genus Arcobacter in seawater (Fera et al., 2008(Fera et al., , 2010Moreno et al., 2003). The ratio of ARC94F-positive cells suggested that Arcobacter might be a common bacterial genus in abalones (6.96%, Table 3).
Haliotis gigantea appears to be a habitat for Arcobacter species, but the role and effect on their hosts are still unclear.
Regarding the cultivation of Arcobacter spp., several conditions have been introduced, such as altering the NaCl concentration (Salas-Massó et al., 2016) or the requirement of sea salts (Diéguez, Balboa, Magnesen, & Romalde, 2017). Hence, we used artificial sea salt instead of NaCl and added more than 2.5% sea salt during isolation. In addition, the incubation temperature used for Arcobacter isolation was set to 15 or 25°C, which are closer to the seawater temperatures of the natural habitat of the abalones at Ise Bay, Mie prefecture, against common methods (30 to 37°C: Vandamme et al., 1991;Houf et al., 2000;Collado, Guarro, et al., 2009;Merga et al., 2011;Salas-Massó et al., 2016;Salas-Massó, Figueras, Andree, & Furones, 2018;Laishram et al., 2016;González et al., 2017). As a result, 10 Arcobacter isolates (five known and five novels) were recovered from samples 15T96H and 25T96H. These were classified into three genotypes based on 97% sequence similarity (15T96H-1, 15T96H-2, and 25T96H-1). The 16S rRNA gene of 25T96H-1 had F I G U R E 2 16S rRNA gene-based phylogenetic tree of Arcobacter spp. from abalone and environmental samples. Circles colors indicate origins or pathogenicity of Arcobacter spp. as follows: blue, marine habitats, orange, terrestrial environments and red, pathogenic species. The tree was generated using the maximum likelihood (ML) method with 1,000 replicates in the bootstrap analysis. The distances were estimated with the Jukes-Cantor correction. The tree was rooted with Campylobacter fetus subsp. fetus ATCC 27374, and gene sequences are followed by GenBank accession numbers in parentheses. Scale bar represents 2% sequence divergence   a high similarity of 99.2%-100% to A. marinas, which has been isolated from a mixture of seawater and starfish (Kim, Hwang, & Cho, 2010). In contrast, the isolates 15T96H-1 and 15T96H-2 had no closely related sequences in all other known Arcobacter isolates.
Including comparisons with uncultured clones, the 16S rRNA genes of 15T96H-1 or 1596H-2 have related to an uncultured clone detected from the Pacific oyster Crassostrea gigas, an invertebrate living in shallow water (Madigan et al., 2014) and marine bulk water (Teeling et al., 2012). Both 15T96H-1 and 15T96H-2 were isolated only at the 15°C incubation temperature. From these observations, we believe that 15T96H-1 and 15T96H-2 will be able to be isolated from various marine invertebrates with sea salt medium at lower temperatures. In terms of incubation time, no Arcobacter species were isolated within 2 days of incubation. This implies that when incubation temperature is lower than 37°C, Arcobacter requires more than 96 hr to grow before colonies can be detected. There are still many species of Arcobacter detectable by molecular methods in abalones that are not cultivable. We feel that isolation methods should be improved to obtain these uncultured Arcobacter species.
In conclusion, we succeeded in detecting several new Arcobacter genotypes from abalone using Arcobacter-specific 16S rRNA gene libraries. Furthermore, since most of the clones showed low similarity with other known Arcobacter spp. and no pathogenic Arcobacter were detected or isolated from abalone, we need further investigations for uncultured Arcobacter spp. which remains to be determined in H. gigantea.

ACK N OWLED G M ENTS
This work was supported by a JSPS Research Fellowship (no. 18J14216) for Young Scientists from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

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

AUTH O R CO NTR I B UTI O N S
YM, SI, and RO performed the experiments; RT, TM, and SF designed and supervised the study; YM wrote the paper. All authors read and approved the final manuscript.

E TH I C S S TATEM ENT
None required.

DATA ACCE SS I B I LIT Y
Raw sequencing data are available at the NCBI website (http://www. ncbi.nlm.nih.gov/). Accession numbers of 16S rRNA gene sequence data from Arcobacter-specific clone libraries are LC133140-LC133161, and from bacteria isolations, LC457972-LC457974.