Study of the microbiota associated to Ruditapes decussatus and Ruditapes philippinarum clams by 16S rRNA metabarcoding, dilution to extinction, and culture-based techniques

The study of the microbiota associated to clams is important not only to know their sanitary status but also to prevent pathobiology events. The use of different microbiological techniques can help to obtain a better picture of the bacterial diversity of clams as well as to isolate new bacterial taxa. In this study, two clam species, Ruditapes decussatus and R. philippinarum, were analyzed in two locations of Galicia (northwest of Spain) in April and October, by combining classic culturing, dilution-to-extinction approach, and 16S rRNA gene target sequencing. 16S rRNA gene target sequencing revealed a great diversity within the clam samples, shedding light into the vast microbial communities associated to these bivalves. All samples were dominated by the same bacterial genera in the different periods, namely Mycoplasma, Vibrio, and Cutibacterium. The α-diversity in the samples obtained during the month of October was lower and showed the dominance of rare bacterial taxa, such as Methylobacterium or Psychrobacter. Dilution-to-extinction technique demonstrated its usefulness to culture rare bacterial taxa that were not found in clams under the classic culturing techniques, including Rahnella, Brachybacterium, Micrococcus, Jantinobacter, and Lelliottia. Altogether, our study provides valuable information on the microbiota associated to R. decussatus and R. philippinarum, demonstrating the high complexity and dynamics of these microbial populations.

R. philippinarum (Adams & Reeve, 1850) (Manila clam), represents an important economic activity in the aquaculture of Galicia (NW Spain). Due to their filter-feeding habit, clams accumulate a rich and diverse bacterial microbiota (Balboa et al., 2016). The study of such microbiota is essential to know the sanitary status, as well as to determine the pathobiological bases of the periodic disease outbreaks affecting clam populations. On the other hand, the introduction of seeds and adult specimens from other countries, due to the overexploitation of natural beds, has increased the risk the introduction of foreign pathogens (Prado et al., 2005).
In recent years, bivalve-microbiota relationships have faced new challenges of crucial importance with the introduction on Next-Generation Sequencing (NGS) technologies. In this sense, symbiotic interactions and pathobiotic events have been extensively explored as well as the mollusc responses to environmental perturbations (Gootenberg & Turnbaugh, 2011;McFall-Ngai et al., 2013;Milan et al., 2018;Lasa et al., 2019). Culture-independent techniques have been introduced to describe microbial community populations associated with bivalves, including clams and oysters (Wegner et al., 2013;Lokmer et al., 2016;Milan et al., 2018;Vezzulli et al., 2018;King et al., 2020;Gerpe et al., 2021). The composition and establishment of the microbial association is influenced by several biotic and abiotic factors, including physiological, immune, and genetic characteristics of the host, the environment, and the interactions among the microorganisms. NGS-based studies of the microbiota of different bivalve molluscs have shown the overestimation using classical techniques of certain bacterial groups or genera, such as Vibrio (Trabal et al., 2014;Roterman et al., 2015;Lokmer et al., 2015;Lasa et al., 2016;Lokmer et al., 2016;Vezzulli et al., 2018). Moreover, these studies have revealed the presence of uncultured species and also some bacterial genera never described before in studies with culture-based techniques.
In the current work, we combined the use of different culture-based techniques, classic culturing, and dilution to extinction, together with a metataxonomic approach, in order to obtain a complete picture of the microbial diversity associated to clams. Thus, the study of the microbiota associated to whole specimens was carried out using 16S rRNA amplicon target sequencing. We have also included classic culturing techniques on marine agar (MA) and Thiosulfte-Citrate-Bile-Sucrose (TCBS) plates for comparative purposes. Finally, the dilution-to-extinction approach, reducing nutrient stress for oligotrophic and facultative oligotrophic bacteria adapted to the marine environment, was employed to attempt the isolation of rare taxa that can be masked in the media used with conventional cultivation techniques. Besides, two different sampling periods and sites were selected to include temporal and spatial variability in the analysis.
Each sample comprised 25 individuals. After washing with distilled water in order to eliminate epibiotic bacteria from the outer shell, clams were aseptically opened and whole specimens were homogenized for bacterial isolation and DNA extraction for 16S rRNA metabarcoding analysis.

DNA extraction
Bacterial cells of each homogenate were separated from eukaryotic cells and concentrated by gradient centrifugation using OptiPrep™ gradient density medium as previously described (Gerpe et al., 2021). The final pellet was resuspended in 1 ml of PBS and stored at − 20 °C until DNA extraction.
DNA was extracted using the MasterPure Complete DNA and RNA Purification kit (Epicentre Biotechnologies) following the manufacturer's instructions. DNA concentration and quality were determined by agarose gel electrophoresis (1% wt/ vol agarose in Tris-acetate-EDTA buffer) and using NanoDrop ND-1000 spectrophotometer (Thermo Scientific) and QUBIT 3 fluorometer (Invitrogen). Extracted DNA was stored at − 20 °C until use for PCR amplification.
Amplicons were sequenced at Sistemas Genómicos (Valencia, Spain) using Illumina MiSeq platform, generating paired-end 2 × 250-bp reads. Illumina reads were analyzed for quality control using FastQC software (Brabaham Bioinformatics). Quality trimming of reads was performed based on quality scores (Q < 30) and length trimming (200 base pairs bp), using Trimmomatic 0.32 (Bolger et al., 2014) program, as well as chimera detection and removal. The filtered paired-end reads were then merged using the command fastq-join (Quast et al., 2013) and clustered at 97% level of similarity into OTUs. Chimera detection and removal were performed.
Ribosomal RNA gene reads were classified against the non-redundant version of the SILVA SSU reference taxonomy (release 123; http:// www. arb-silva. de). For bacterial diversity estimation in the samples, the number of operational taxonomic units (OTUs) at 97% sequence identity was determined, and rarefaction analyses were carried out. Briefly, the reads were aligned against the 16S rRNA sequences of the SILVA database followed by a quality filtering, including length, ambiguity, and homopolymer checks. A de-replication step was performed to collapse identical reads into one single sequence and OTU's were clustered at 3% divergence threshold. The mitochondria, chloroplasts, and unassigned reads were deleted for the taxonomic analysis. Non-metric Multidimensional Scaling (NMDS) and Analysis of similarities (ANOSIM) were performed from the dissimilarity matrix using vegan package of R (Clarke 1993;Oksanen, 2017).

Bacterial isolation
Serial dilutions of the homogenized samples from the October sampling period were carried on with PBS and 100 μl of each dilution was cultured in marine agar (MA, Difco) and Thiosulphate-Citrate-Bile-Sucrose (TCBS, Oxoid) at 24 ± 1 °C for 48 h. Isolated colonies were culture in marine agar and preserved at − 80 °C in marine broth (MB, Difco) supplemented with 15% of glycerol.

Dilution-to-extinction approach
For the experiment (Fig. 1), 20 clams were selected in October, 5 specimens of R. decussatus and 5 of R. philippinarum from each sampling site (Carril and Redondela). As before, clams were aseptically opened and the soft tissues of each clam were homogenized. The number of viable bacteria in the clam homogenates was determined using Live/ Dead BacLight bacterial viability kit (Invitrogen) and were visualized and counted in a fluorescent microscope. Samples were diluted to 1-5 cells/ well in 96-well microplates containing marine broth diluted 10 or 100 times (1-5 cells/well). These cultures were incubated at 24 ± 1 °C in the absence of light for 60 days. Then, each well, containing bacterial growth, was streaked onto marine agar plates (MA, Difco) diluted 10 (MA1/10) and 100 times (MA1/100). Then, isolated colonies were cultured in marine agar 1/10 and 1/100 and stored at − 80 °C in marine broth (MB, Difco) diluted 1/10 and 1/100 supplemented with 15% of glycerol.

s rRNA gene sequencing
For all the obtained isolates, DNA was extracted using the InstaGene Matrix (Bio-Rad) following the manufacturer's instructions. The 16S rRNA gene sequence was amplified and sequenced using the universal primer pair 27F (5′ -AGA GTT TGA TCC TGG CTC AG-3′) and 1510R (5′-GGT TAC CTT GTT ACG ACT T-3′) as described by Hutson et al. (1993). Sequence data analyses were performed using DNASTAR Lasergene SEQMAN program and 16S rRNA sequence similarities were determined using the EzTaxon-e server (www. eztax on-e. ezbio cloud. net) and the BLASTN program.

16S rRNA sequence metataxonomy
After filtering raw sequences obtained from V3/V4 region of 16S rRNA, a total of 441,672 reads were obtained from samples of R. decussatus and R. philippinarum with an average length ranging from 453 to 461 pb and a total of 22,044 OTUs were obtained (Table 1). Rarefaction analysis (at 97% sequence identity level) of R. decussatus and R. philippinarum reflected higher α-diversity in April samples rather than October (Fig. 2).
Taxonomic assignment of the sequences using the non-redundant version of SILVA database identified Proteobacteria and Actinobacteria as the main phyla in all samples (Fig. 3). Other phyla with lower relative abundances included Tenericutes, Firmicutes, and Bacteroidetes. Samples taken in October showed an increase in the relative abundance of Proteobacteria group in both sites and clam species. On the other hand, Actinobacteria displayed a different pattern depending on the sampling site, with a reduction in Carril in October, while in Redondela the relative abundance was increased during this period. 16S rRNA amplicon analysis at the genus level revealed a great bacterial diversity associated to R. decussatus and R. philippinarum species (Fig. 4).
Mycoplasma genus was detected in all samples showing more stability in Carril than in Redondela, which showed a drop in their relative abundance from 17.8% at both locations were at high relative abundances (ranging from 11.5% to 18.2%). Interestingly, the NGS amplicon sequencing approach allowed the identification of several uncultured bacteria, such as Propionibacteriaceae or Microscillaceae, which in some cases reached relative abundances above 20%, as in R. decussatus samples from Carril in April (CD1) or from redondela in October (RD2). Microbial composition differences have been reflected when a Non-metric Multidimensional Scaling (NMDS), applied on ANOSIM distance matrix, analysis was performed. This analysis allowed to further investigate whether the clam species, sampling site, and sampling period affected the microbial composition of R. decussatus and R. philippinarum species (Fig. 5). This analysis demonstrated that samples from October clustered together, regardless their geographical origin or clam species. On the other hand, samples from April were separated further apart one to each other.

Classical culture-based technique
Bacterial isolation from specimen homogenates resulted in a total of 94 isolates from MA and TCBS culture media (Supplementary Tables S1 and  S2), which, in overall, consisted of 37 strains from R. decussatus and 13 from R. philippinarum in the localization of Redondela and another 21 belonged to R. decussatus and 23 R. philippinarum in the localization of Carril. In all samples, Vibrio was the most abundant genus with percentages from 42 to 57%, together with Pseudoalteromonas that was well represented in Carril (Fig. 6). Together with these two genera, Shewanella was one of the predominant genera in samples of R. philippinarum from Carril and  Tables S1 and S2).
16S rRNA gene sequence analysis of the isolated bacteria showed the predominance of Gammaproteobacteria group in all samples, followed by Actinobacteria, which was more abundant in samples from dilution-to-extinction approach than in samples cultivated on MA and TCBS, especially in Carril. At the genus level, the dominance of the genus Pseudomonas was observed in all samples of MA1/100, while in samples isolated from MA1/10 Pseudomonas and Shewanella were the most abundant genera (Fig. 6, Supplementary Tables S3 and S4). In general, samples from Carril showed higher number of different genera, especially samples obtained from MA1/100.
In Redondela samples, microbiota was represented by few genera. In samples cultivated in MA1/10, Pseudomonas and Shewanella were the most abundant genera that represented about 60% in RP1/10 and 80% in RD1/10 of the total of microbiota. Micrococcus and Rahnella represented 25% of the microbiota in RP1/10. In samples from this site cultivated in MA1/100 Pseudomonas was the most abundant bacteria accounting for 77% of total of microbiota in R. philippinarum and 42.67% in R. decussatus. Serratia genus presented percentages around 15-16% in both clam species and Rahnella genus was well represented in RD1/100 (25%).

Discussion
The present study combined culture-dependent techniques together with a metataxonomic approach in order to study the microbiota associated to R. decussatus and R. philippinarum clam species. The use of different techniques provided a complete overview of the microbial population associated to these two clam species, from the already well-defined culturable fraction to the non-culturable vast microbial community. Indeed, the dilution-to-extinction experiment made possible the assessment of the microbial population, by successfully cultivating 131 new bacterial isolates.
Previous study of microbiota of R. philippinarum and R. decussatus reported the predominance of Vibrio and Pseudoalteromonas genera (Romalde et al., 2013;balboa et al., 2016). These two genera represented more than 50% of the isolates on MA and TCBS media, followed by other bacteria frequently found in association with marine organisms or in the seawater, such as Shewanella, Photobacterium, Aliivibrio, or Bacillus. Furthermore, these two genera were not detected within the strains obtained by dilutionto-extinction culturing, in which Pseudomonas was the predominant genus in most samples. Some genera detected by dilution-to-extinction experiment were not detected with classic culturing technique in this study or in previous studies on R. philippinarum and R. decussatus clams (Balboa et al., 2016), including Rahnella, Micrococcus, and Lelliottia. These bacteria are frequently found in aquatic environments, but they have also been reported to cause bacteraemia in humans, as for R. aquatilis (Ruimy et al., 2010). Besides, Rahnella and Lelliottia isolates were found during the experiment in several samples, both in diluted 1/10 and 1/100 MA plates, although they were not detected when using neither 16S rRNA gene amplicon sequencing nor classic culturing techniques. Other genera, such as Janthinobacterium, Gordonia, Brachybacterium, or Microbacterium isolated in dilution-to-extinction experiment, were also found using 16S rRNA gene amplicon sequencing but at very low relative abundances (< 0.5%). As expected the restrictive selected culture conditions applied in the dilution-to-extinction effort, the number of isolates obtained was low in comparison to the number of OTUs detected with NGS technique, barely 0.6% of the total diversity. Different factors might be considered in future studies in order to culture a higher number of rare taxa, including quantification errors, culture media selection, nutrient availability, temperature, or time of incubation.
In general, seasonal and spatial variations were observed, pointing out the high complexity of these microbial populations. It is well known that the interaction between the host and their microbes provides them beneficial effects (Shapira, 2017;Torda et al., 2017;Offret et al., 2019), although, in bivalves, the characteristics and functionality of the associated microbiota are poorly understood (Desriac et al., 2014;Offret et al., 2019). Our work provides a new insight into the microbiota associated to R. decussatus and R. philippinarum, which, ultimately, demonstrates the complexity of such multicellular consortia. The results demonstrated that, although variability is observed, clam microbiota was dominated by few bacterial groups that are present in all samples in the different conditions, including Mycoplasma, Vibrio, and Cutibacterium, suggesting they might be considered as autochthonous bivalve bacteria.
Vibrio and bivalve mollusc interactions have been widely studied (Balboa et al., 2016;Destoumieux-Garzón et al., 2020) as pathogens or opportunistic pathogens but also as commensals or by establishing neutral associations. Vibrios are among the most common and widespread prokaryotes in temperate marine environments, representing the major culturable fraction of the marine microbial community (Ceccarelli et al., 2019); however, the abundance of Vibrio populations in molluscs can reach concentrations close to 100-fold higher than those in seawater (Shen et al., 2009). In clams, they represent the main taxa in the culturable fraction as presented above; however, when a NGS approach is applied their real abundance compared to the whole community is clearly reduced.
Mycoplasma genus has been found associated with healthy oysters (Lasa et al., 2019) and was also dominant in clam hepatopancreas (Milan et al., 2018). It has been suggested that members of this genus might play a beneficial role in bivalve fitness and health status, although their role within the host is still largely unknown (Romero et al., 2002;King et al., 2012). Additionally, Cutibacterium, which was found at high relative abundances, is a novel genus (Scholz & Kilian, 2016), after splitting Propionibacterium into three different genera, and they have been mostly associated to humans, especially with acne (Platsidaki et al., 2018). However, our results indicate that the habitat of bacteria belonging to this genus may be larger than previously known and we also found several OTUs assigned to uncultured Propionibacteriaceae, indicating that these bacteria might be frequently found in marine environments.
The analysis of α-diversity revealed lower microbial richness in R. decussatus and R. philippinarum in October samples, while NMDS analysis showed that microbial communities of samples obtained in autumn were more similar. Other studies revealed similar trends in oysters, where bacterial load and diversity were lower in winter (Pujalte et al., 1999;Zurel et al., 2011). Indeed, those samples showed the dominance of few bacterial taxa, besides Mycoplasma, Vibrio, and Cutibacterium, such as Methylobacterium and Sphingomonas in samples from Carril and Redondela in October or Psychrobacter in samples from Carril during October sampling period. These genera increased their relative abundances considerably respected to spring, when their concentration was residual. These changes in the microbiota composition may be related to a limited microbial diversity in the surrounding environment but also as a consequence of physiological changes that occur in clams during late spring/summer related to a great energy demand to complete gamete maturation and reproduction (Meneghetti et al., 2004). During this period, clams increase their filtration rates and increase the phytoplankton uptake, therefore the bacterial uptake is increased too.

Conclusion
The present study explored into the high complexity of the microbial populations associated to R. decussatus and R. philippinarum clam species. Dilutionto-extinction technique revealed its usefulness for culturing rare bacterial species that, otherwise, would not be feasible. Overall, metataxonomic results demonstrated that the composition and dynamics of the associated microbiota are affected by seasonal periods, although few bacterial taxa appear to be autochthonous to these clam species.
Acknowledgements This work was supported in part by grant AGL2013-4268-R and AGL2016-77539-R from the Ministerio de Economía y Competitividad (Spain).
Author contributions DG and AL contributed to investigation, formal analysis, and writing of the original draft; AL performed bioinformatic analysis; SB performed formal analysis and writing, reviewing, and editing of the manuscript; JLR contributed to conceptualization, writing, reviewing, and editing of the manuscript, funding acquisition, and supervision.
Funding Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This work was supported in part by grant AGL2013-42628-R and AGL2016-77539-R from the Ministerio de Economía y Competitividad (Spain).
Data availability All data produced from this study are provided in this manuscript. Sequence files for all samples used in this study have been deposited at NCBI SRA with accession: PRJNA428215.

Conflict of interest
The authors declare there is no conflict of interest.

Ethical approval
The research complies with ethical standards.

Consent for publication Not applicable.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.