Environmental Microbiology Reports (2019) 11(3), 372–385 doi:10.1111/1758-2229.12686 Brief Report Coral microbiome database: Integration of sequences reveals high diversity and relatedness of coral- associated microbes Megan J. Huggett1,2* and Amy Apprill3* timely given the escalated need to understand the role 1School of Environmental and Life Sciences, University of the coral microbiome and its adaptability to changing of Newcastle, Ourimbah, NSW, 2258, Australia. ocean and reef conditions. 2School of Science, Edith Cowan University, Joondalup, WA, Australia. Introduction 3Marine Chemistry and Geochemistry Department, Woods Hole Oceanographic Institution, Woods Hole, Corals are complex ‘holobiont’ organisms (Knowlton and MA, USA. Rohwer, 2003), hosting multi-species microbial interac- tions which sustain their productivity and growth in oligo- trophic reef waters. It is well known that corals rely on Summary endosymbiotic microalgae (Symbiodinium)tosupply Coral-associated microorganisms are thought to play a sugars, amino acids, lipids and other nutrients (Trench, fundamental role in the health and ecology of corals, but 1971), but corals also host an assemblage of other micro- understanding of specificcoral–microbial interactions organisms including endolithic algae, bacteria, archaea, are lacking. In order to create a framework to examine fungi and viruses (Rohwer et al., 2002; Knowlton and coral–microbial specificity, we integrated and phyloge- Rohwer, 2003; Ainsworth et al., 2017). Of these microor- netically compared 21,100 SSU rRNA gene Sanger- ganisms, bacteria and archaea are thought to play a criti- produced sequences from bacteria and archaea associ- cal role in the cycling and regeneration of nutrients for the ated with corals from previous studies, and accompany- coral holobiont (Rohwer et al., 2001, 2002; Beman et al., ing host, location and publication metadata, to produce 2007; Kelly et al., 2014). There is also evidence that coral- the Coral Microbiome Database. From this database, we associated bacteria are capable of producing antibiotic identified 39 described and candidate phyla of Bacteria and other secondary metabolite compounds, which could and two Archaea phyla associated with corals, demon- provide protection to the coral (Kelman, 2004; Ritchie, strating that corals are one of the most phylogenetically 2006; Raina et al., 2009). Understanding the individual diverse animal microbiomes. Secondly, this new phylo- contribution of these disparate organisms to the coral holo- genetic resource shows that certain microorganisms are biont, and if and how these microorganisms enhance the indeed specific to corals, including evolutionary distinct resistance and resilience of corals to changing ocean and hosts. Specifically, we identified 2–37 putative monophy- reef conditions, are current challenges within this field. letic, coral-specific sequence clusters within bacterial Over the past 15 years, numerous review articles have genera associated with the greatest number of coral reflected upon the relationship between corals and bacte- species (Vibrio, Endozoicomonas and Ruegeria)aswell ria and archaea (Rosenberg and Ben-Haim, 2002; as functionally relevant microbial taxa (“Candidatus Rohwer and Kelley, 2004; Mouchka et al., 2010; Bourne Amoebophilus”, “Candidatus Nitrosopumilus” and under et al., 2016). These reviews repeatedly demonstrate that recognized cyanobacteria). This phylogenetic resource coral–microbes are taxonomically distinct from cells provides a framework for more targeted studies of residing in the seawater, and there is general recognition corals and their specific microbial associates, which is of shared microbial groups over coral species and reefs (i.e., Roseobacter clade, Vibrio spp.). However, although most of the coral–microbial data are publically available, Received 17 January, 2018; accepted 4 August, 2018. *For corre- these data from multiple investigators and study environ- spondence. E-mail [email protected]; Tel. +61 2 4348 fi 4025; Fax +61 2 4349 4404; E-mail [email protected]; Tel. 01-508-289- ments data are not integrated. In other elds of study, 2649; Fax 01-508-457-2164. combining existing data provides valuable resources for © 2018 The Authors. Environmental Microbiology Reports published by Society for Applied Microbiology and John Wiley & Sons Ltd. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. Coral microbiome database 373 the community and especially for students and new advance investigations of the functional role of the coral investigators to these fields (e.g., Simister et al., 2012). microbiome and specificcoral–microbial interactions. Integration of coral–microbial data is critical in the current framework of limited understanding about coral–microbial associations and during a time when many corals and reefs Results and discussion are experiencing disturbances from climate and anthropo- Overview of coral-specific microbial sequence database genic sources (Hoegh-Guldberg et al., 2007; Carpenter et al., 2008; Hughes et al., 2017, 2018). A growing commu- The Coral MD includes 21,100 sequences derived from nity of scientists is pursuing studies aimed at understanding coral-associated bacteria and archaea which originated the resistance, resilience and impact of these events on from public databases, and are compiled in the ARB soft- coral microbiomes (e.g., Lee et al., 2015; Webster et al., ware environment within the framework of the Silva refer- 2016). For example, studies have shown that coral- ence database (Quast et al., 2013) (Supporting associated microbial communities demonstrate taxonomic Information Methods). The database is available for shifts in composition in response to temperature, nutrient download at BCO-DMO (https://www.bco-dmo.org/ and pH-based stressors (Littman et al., 2010; Meron et al., dataset/724355). The mean and median lengths of 2011; Shaver et al., 2017). The current research trend is to sequences in the database are 893 and 750 bp respec- produce copious partial SSU rRNA gene sequences using tively. There are 6833 nearly full-length sequences next generation sequencing methods targeted at bacteria greater than or equal to 1200 bp, and 5811 sequences and archaea to inform questions about how microbial diver- are short reads, less than or equal to 600 bp. The sity, community structure and taxonomic composition sequences span 37 described and candidate phyla of changes under different environmental scenarios. However, Bacteria (Proteobacteria considered as single phylum) sequences generated by these approaches are typically and 2 phyla of Archaea (Euryarchaeota and Thaumarch- not long enough to allow for in-depth phylogenetic analyses aeota) (Fig. 1). Of the bacteria, sequences are most com- or primer and probe design. Further, taxonomic assignment monly affiliated with the Proteobacteria (especially is generally facilitated using reference databases such as Gammaproteobacteria and Alphaproteobacteria), Bacteroi- Silva (Quast et al., 2013) or Greengenes (DeSantis et al., detes, Cyanobacteria and Firmicutes. Fewer sequences 2006), and these databases include sequences from were associated with candidate divisions, including BRC1, biomedical-based studies and it is often difficult to identify OD1,OP11,OP3,OP8,SR1,TM7andWS3. sequences originating from corals, thus hindering the devel- In order to examine if the phyla-based distribution of opment of coral-specific analyses and resources, such as sequences in this database are representative of those specific primers and probes. available in next-generation sequencing studies, we In order to examine relatedness of microorganisms examined phyla-based diversity of 715,114 coral– detected by sequencing approaches across numerous microbial sequences spanning nine studies and 23 coral studies of corals, we followed an approach utilized by Sim- species from the Alcyonacea as well as complex and ister et al. (2012) to examine the phylogeny of sponge- robust Scleractinian lineages available within the Coral associated microbes. Specifically, we utilized publically Microbiome Portal (CMP) (https://vamps2.mbl.edu/portals/ available Sanger produced sequence data from worldwide CMP) (Sunagawa et al., 2010; Barott et al., 2011; Morrow studies of corals to produce the Coral Microbiome Data- et al., 2012; Apprill et al., 2013; Bayer et al., 2013a, b; base (Coral MD), a coral-specific SSU rRNA gene data- Bourne et al., 2013; Šlapeta and Linares, 2013; Lesser base hosting 21,100 coral–microbial sequences, housed and Jarett, 2014). We identified high consistency within the ARB software (Ludwig et al., 2004). We utilized between the two types of sequence data, with three line- the Coral MD to examine the phylogenetic diversity of ages represented in the next-generation studies that coral-associated bacteria and archaea and these findings were not present in the Sanger database (Candidate divi- were confirmed in a subset of next-generation sequencing sion OP10, TG-1 and WS6), and five lineages in the data available for corals. Secondly, we identified putative Sanger sequencing based Coral MD database (BD1-5, monophyletic, coral-specific sequence clusters and exam- Elusimicrobia, NPL-UPA2, SHA-109 and WCHB1-60) ined the distribution of these clusters among the Alcyona- which were absent from