Prior to sequencing, stock populations of the anemones were maintained in replicate 10-gallon tanks at 25 degrees in 34 ppt Instant Ocean artificial seawater (ASW). Each tank contained one genet (H2, CC7, VWA, or VWB) in one symbiotic state (symbiont-rich or symbiont-depleted). Symbiont-depleted anemones were subjected to a bleaching protocol as previously described (Lehnert et al., 2012) and maintained in dark conditions for approximately 6–9 months before nucleic acid isolation. Symbiotic states were defined based upon pale appearance to the naked eye, although the symbiont-depleted anemones from all genets retained some algal cells ( Figure 1 ). The H2 and VWB clonal strains originate from Coconut Island, HI (Xiang et al., 2013). VWA is a multi-origin strain (Poole et al., 2016). CC7 was originally obtained from Carolina Biological Supply (Sunagawa et al., 2009). Symbiont-rich anemones were kept on a 12:12 light:dark cycle using white LED lights at about 12µmol photons/m/s. All anemones were fed freshly hatched Artemia shrimp nauplii 2–3x weekly and were subject to regular salinity checks and water changes.

Anemones were transferred to a 6-well plate for each symbiotic state and clonal line (genet), with one or two individuals per well in 3mL ASW. The anemones in the 6-well plates were kept in an incubator (Percival AL30L2) for a week of acclimation under optimal conditions including white light on a 12:12 schedule and twice weekly feedings with freshly hatched Artemia shrimp nauplii.

After the acclimation period, anemones were anesthetized for approximately 10 minutes and rinsed in a solution of sterile 0.5% MgCl₂ in ASW in order to relax their bodies for dissection. Once the anemones were anesthetized, they were dissected using a dissecting microscope, fine forceps, and a needle. The anemones were sliced just under the oral disc, resulting in the separation of the oral disc and tentacle region, and the stem and peduncle region. Anemones were kept separate by genet and symbiotic state combination during the dissection process, and the tools were sterilized and wiped clean with ethanol rinses and clean lint-free disposable cloths between uses.

Once the anemones in each group (3–6 replicates each) were dissected into top (oral disc and tentacles) and bottom (stem and peduncle), they were placed into 1.5mL Eppendorf tubes. Each tube was then labeled with a 4-character code indicating the genet, symbiotic state, replicate, and body part, and placed in a -80°C freezer for storage.

Nucleic acid isolation was performed using a modified SlimeAway protocol (Vargas et al., 2021). Anemone samples were first homogenized using a motorized pestle inside their Eppendorf tubes on dry ice. Next, a CTAB extraction buffer was prepared by mixing 17µl molecular grade water, 28µl 0.5M EDTA, 56µl 1M TRIS, 224µl 5M NaCl solution, 112µl 0.1M PVP (Polyvinylpyrrolidone K30) solution, and 112µl 0.1M CTAB (Cetyltrimethylammoniumbromide) solution per sample. The buffer was incubated in a water bath at 60°C for 10 minutes, after which 12µl β-Mercaptoethanol per sample was added to the buffer. 550µl CTAB buffer was then added to each sample tube, followed by 5µl Proteinase K, inverting the tube to mix. Samples were incubated at 56°C for between 2 hours to overnight, inverting the tubes regularly to avoid tissue clumping, until there were no visible tissue clumps in solution.

Once the solution was free of clumps, 150µl 5M KoAc solution was added to each tube, and they were incubated again at 56°C for 10 minutes. The samples were then centrifuged for 15 minutes at 4°C at 15000xg, ~600µl supernatant was removed avoiding any pellet or viscous precipitate, and 600µl (1 volume) chloroform was added. After this addition, samples were shaken vigorously for 20 seconds then centrifuged again at the same parameters. Finally, ~400µl of the supernatant was transferred, avoiding the interphase, to a new, labeled 1.5mL Eppendorf tube.

DNA was precipitated using 400µl (1 volume) isopropanol added to each sample, then sample tubes were shaken for 15 seconds before being incubated at room temperature for 10 minutes. They were placed in the -20°C freezer overnight. The next day, the samples were centrifuged at 4°C for 30 minutes at 15000xg. The isopropanol supernatant was discarded without disturbing the pellet, and 62.5µl cold (-20°C) 100% ethanol, 1µl 20mg/mL glycogen, and 2.5µl 3M NaAc solution were added to each sample and mixed. The samples were incubated at -20°C between 2 hours and overnight, then centrifuged at 4°C at 15000xg for 30 minutes. Next the supernatant was carefully removed, the pellet was washed with 500µl cold (-20°C) 75% ethanol and vortexed, and the pellet was centrifuged again with the same settings for 15 minutes. This ethanol removal, wash, and spin step was repeated one more time. The supernatant was once again removed, and the pellet was allowed to dry for 10 minutes, followed by resuspension in 30µl molecular-grade water and stored at -80°C. Samples were treated with Monarch RNase A and subsequently purified with the Monarch PCR & DNA Clean Up Kit (New England Biolabs, Massachusetts). All samples were eluted in 30 µL 1X TE buffer after cleanup. DNA was then quantified using the Qubit HS ssDNA kit and assessed for quality on the Nanodrop 8000 spectrophotometer (ThermoFisher, California).

16S ribosomal RNA libraries were generated from the V3-V4 variable region according to the Illumina 16S Metagenomic Sequencing Library Preparation guide. Samples were normalized to 5 ng/µL in 1X TE buffer. 550 bp amplicons from the V3-V4 region of the rRNA gene were generated for each sample The primers included the V3-V4 locus specific sequences underlined below based on Klindworth et al., 2013, as well as an Illumina overhang adapter sequence.

Forward 5’ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG

Reverse 5’ GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC

Amplicons were cleaned using HighPrep beads (MagBio, Maryland) and target size was verified using Fragment Analyzer (Agilent, California). Illumina Nextera DNA CD unique dual indexes were added to the amplicons, then cleaned using HighPrep beads (MagBio, Maryland). Final libraries were quantified using the Qubit HS DNA assay (Thermo Scientific, California) and library size was verified using Fragment Analyzer (Agilent, California).

52 amplicon libraries were normalized to 2 nM, then pooled together in equal volumes. The library pool was then denatured and diluted following the MiSeq dilute and denature guide. Sequencing was performed on the Illumina MiSeq using a paired end 600 cycle v3 sequencing kit (Illumina, California).

The DADA2 (Callahan et al., 2016) pipeline was followed in R v. 4.2.1 (R Core Team, 2022) for filtering, trimming, denoising sequences, merging reads, and performing chimeric checks. All statistical analysis and data visualization was performed using R. Silva ver. 132 was used to assign taxonomy. Non-bacterial sequences were removed from downstream analysis. Sequence data was compiled into a Phyloseq object to perform diversity and abundance analyses (McMurdie and Holmes, 2013). Data were not normalized prior to diversity analyses. Alpha-diversity was measured using the observed number of ASVs, the Shannon index, and Pielou’s evenness. Beta-diversity distances were calculated using the Bray-Curtis method. Vegan (Oksanen et al., 2022) was used to create rarefaction curves. Data was visualized using GGplot2 (Wickham, 2016) and Vegan (Oksanen et al., 2022). DESeq2 (Love et al., 2014) was used for differential abundance analysis. The results of the differential abundance analysis were further filtered using a threshold of ⅔ presence across samples belonging to the differentially abundant category of interest to focus on ASVs that were commonly observed. Core microbiome analysis was performed by making lists of ASVs present in 100% of samples. We also searched for a core set of ASVs present in each category of analysis by sub-setting samples by clonal line, body part, and symbiotic state, similar to Röthig et al. (2016). Results were considered significant at an alpha level of 0.05.

Alpha-diversity metrics (observed ASVs, Shannon’s index, Pielou’s evenness) did not indicate any significant differences between symbiotic states and body parts ( Figure 2 ). Pairwise Wilcox rank-sum tests using FDR correction showed significant differences between observed alpha-diversity between four genet pairings, however only one genet pairing showed significant difference using the Shannon metric for alpha-diversity. Anemones from the VWA and VWB clonal lines displayed greater evenness compared to anemones belonging to CC7 and H2 ( Figure 2C ). Beta-diversity was visualized via principal coordinates analysis (PCoA) based upon the Bray-Curtis dissimilarity index reveals some clustering by symbiotic state and genet ( Figure 3 ). Results of permutational multivariate analysis of variance, where the R² value represents the proportion of variation in the distance matrix explained by each variable, confirmed substantial variation driven by genet (R² = 0.31; p = 0.001) and symbiotic state (R² = 0.03, p = 0.02), but not body part (R² = 0.02, p = 0.146). Additional pairwise comparisons were performed between each individual genet. Differences between each pairing were significant, further cementing our finding that genets have distinct microbial communities.

The mean relative abundance of bacteria was used to provide an overview of dominant ASVs at multiple taxonomic levels across sample types. ASVs originating from Proteobacteria had the highest relative abundance (74%), followed by Plancomyetes (11%), Actinobacteria (4.5%) and Lentisphaerae (3.5%) ( Figure 4 ). The abundance profile of phyla remained similar between genets, body parts, and symbiotic states.

At the family level, ASVs belonging to Rhodobacteraceae (21%), Phycisphaeraceae (8.3%), Teraskiellaceae (5.7%), Beijerinckiaceae (5.2%), Caulobacteraceae (4.4%), and Stappiaceae (3.9%) were common. Samples from the VWA clonal line abundances revealed greater evenness between families, with Beijerinckiaceae (12.4%) and Caulobacteraceae (11.2%) being particularly common. In contrast, Stappiaceae (10.2%) was particularly abundant within the VWB genet. Abundances were much more variable across genets at the genus level, however Marivita, Methylobacterium, Phycisphaera, Brevundimonas, Labrenzia, and Thalassobius were common. A core microbiome, defined by the presence of an ASV in 100% of samples, was not identified by our data. Similarly, no ASVs were found to be universal within each body part, symbiotic state, nor clonal line.

To further investigate taxonomic differences between anemone body part and symbiotic state, differential abundance analysis was performed using DESeq2. We identified a total of 91 differentially abundant ASVs between symbiotic states (Supplementary File S2) and 32 differentially abundant ASVs between anemone body parts ( Supplementary File S3 ; Figure 5 ). To account for the compositional dissimilarity between samples, we considered ASVs found in at least ⅔ of samples belonging to the category of interest as part of a “common” group. 12 ASVs in this common group were shown to be significantly more abundant in symbiont-depleted anemones than symbiont-rich anemones, including ASVs belonging to Marivita, Magnetospira, Nitrospina, Kordiimonas and Mycobacterium. In contrast, only 2 ASVs, belonging to Mesorhizobium and Maritalea, were found to be significantly more abundant in symbiont-rich anemones with modest L₂FC difference. Only 7 common and differentially abundant ASVs between anemone body parts were identified, all of which were more abundant in the bottom half of the anemone. Visualization of the differential abundance analysis results reveal greater consistency in abundance of abundant ASVs in symbiont-depleted anemones compared to symbiont-rich anemones, implicating algal-symbiont identity as a possible driver of variation ( Supplementary Figure S2 ).