Aposymbiotic and symbiotic Aiptasia of the clonal strain CC7 were generated and reared as described previously (Baumgarten et al., 2015) with some modifications. Aposymbiotic animals were generated through repeated 4 h cold−shocks at 4°C and treatment with 50 μM of the photosynthetic inhibitor diuron (Sigma−Aldrich, United States). Aposymbiotic animals were maintained in the dark for >3 years and are routinely inspected by fluorescence microscopy (Leica DMI3000 B) to confirm the absence of Symbiodiniaceae. Symbiotic anemones were generated by infecting aposymbiotic animals with strain SSB01 (Breviolum minutum, former Clade B) (Xiang et al., 2013). Of note, the algal symbiont strain SSB01 is not the native symbiont of Aiptasia CC7 but has been previously used as a stable and characterized Aiptasia host-algal symbiont combination (Röthig et al., 2016a; Wolfowicz et al., 2016; Simona et al., 2019). To obtain symbiotic CC7-SSB01 Aiptasia, individual anemones were exposed to 10⁵ Symbiodiniaceae cells/ml for 24 h, fed with Artemia nauplii after 48 h, and seawater was exchanged thereafter. All Aiptasia anemones were kept in autoclaved natural seawater and reared in 1 L tanks at a 12 h:12 h light:dark cycle at 20–40 μmol photons m² s¹ of photosynthetically active radiation at 25°C in an I-36LLVL incubator (Percival Scientific Inc., United States). Nematostella vectensis polyps were reared in half liter tanks in half strength artificial seawater at room temperature. Polyps of Hydra magnipapillata strain 105 were reared in half liter tanks in commercial spring water at room temperature. Nematostella and Hydra animals were fed Artemia nauplii twice a week.

We compared ectoderm surface topography of different cnidarian classes, namely Anthozoa (corals, Aiptasia, Nematostella vectensis) with Hydrozoa (Hydra magnipapillata), using Scanning Electron Microscopy (SEM). Coral colony fragments of Stylophora pistillata, Acropora humilis, and Porites sp. were collected from the central Red Sea at 6 m depth at the Al Fahal forereef (22°15.100′N, 38°57.386′E) by SCUBA and maintained in an open water aquaria system until use (CMOR, KAUST), whereas Aiptasia, N. vectensis, and H. magnipapillata animals were available from cultured lab strains. Three fragments (in the case of corals) or three polyps (in the case of Aiptasia, Nematostella, Hydra) were transferred to 24-well plates for processing. Coral fragments and model system cnidarians were left to expand in ∼2% magnesium chloride in artificial seawater (ASW) or fresh water for Hydra, or half strength ASW for Nematostella for 15 min, respectively. After that, specimens were fixed in ∼2.5% glutaraldehyde in 0.1 M cacodylate buffer at 4°C overnight. Samples were then washed in 0.1 M cacodylate buffer (pH 7.2–7.4) thrice for 15 min each and post fixed in 1% osmium tetroxide solution in 0.1 M cacodylate buffer for 1 h in the dark. Samples were further washed thrice in ddH2O for 15 min each and proceeded for dehydration in an EtOH gradient: 30, 50, 70, 90, and 100%, 15 min each step, with two final incubations in 100% EtOH, also for 15 min. The drying process was initiated by transferring the samples to a 1:2 solution of hexamethyldisilazane (HMDS):100% EtOH (v/v) for 20 min, then to a fresh solution of 2:1 HMDS:100% EtOH for 20 min, 100% HMDS for 20 min, and a final incubation in 100% HMDS. Samples were left to dry loosely covered in the chemical hood for the HMDS to slowly evaporate overnight. Each fragment was then mounted on an SEM specimen mount head pin and sputter coated with 4 nm of platinum/palladium or Iridium. All samples were imaged using a Teneo Volume Scope electron microscope (FEI, United States) operating at 1–3 kV.

Complementary to the SEM analyses, we imaged Aiptasia specimens by Transmission Electron Microscopy (TEM). Briefly, Aiptasia polyps were fixed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) for 12 h at 4°C and post fixed with 1% osmium tetroxide in 0.1 M cacodylate buffer for 1 h in the dark at 4°C. After three washes in ddH2O, the sample blocks were dehydrated through EtOH and acetone, infiltrated with a mixture of epoxy resin (Electron Microscopy Sciences, United States) and acetone, followed by a final embedding in pure resin. A Leica EM UC6 Ultramicrotome (Leica Microsystems, Germany) was used to cut 150 nm thin sections from the resin block. Finally, thin sections were collected on a 200 mesh copper grid, stained with 1% uranyl acetate (Electron Microscopy Sciences, United States) and Reynold’s lead citrate (Electron Microscopy Sciences, United States). Images were acquired using a Titan CT transmission electron microscope (Thermo Fisher Scientific, United States) operating at 300 kV.

To investigate the presence of bacteria inside the mucus layer using TEM, we modified the standard resin embedding protocol by introducing an agarose embedding step with subsequent cryosectioning of the specimens. Aiptasia anemones were prepared as follows: Individual polyps were chemically fixed in phosphate buffer (0.1 M pH 7.4) with 9% sucrose, containing 4% formaldehyde and 2.5% glutaraldehyde. Fixation was done at room temperature for 2 h after which samples were embedded in 4% aqueous agarose to ease the manipulation. Then, polyps were cut into ∼4 mm³ pieces and post stained in 2% osmium aqueous solution for 1 h in the dark. After rinsing in water, samples were dehydrated in a series of ethanol concentrations, ranging from 10 to 100%, then infiltrated with EPON resin (Electron Microscopy Sciences, United States), before polymerization at 60°C for 48 h. Thin sections of 80 nm were cut and mounted onto a Formvar film-coated, carbon-stabilized 100 mesh copper finder grid (Electron Microscopy Sciences, United States). Sections were post stained with UranyLess (Electron Microscopy Sciences, United States) and lead citrate (Electron Microscopy Sciences, United States) before being imaged with a Tecnai-12 transmission electron microscope (Thermo Fisher Scientific, United States) operating at 100 kV with a FEI eagle camera (Thermo Fisher Scientific, United States) using TIA software (Thermo Fisher Scientific, United States). Contrast, brightness, and sharpness of acquired images were adjusted using Adobe Photoshop.

Bacteria-depleted Aiptasia polyps were generated using a previously developed protocol consisting of a depletion priming step, followed by an antibiotic treatment, and subsequent recovery from the antibiotic cocktail (Costa et al., 2019). For the depletion priming step, polyps were transferred to 500 ml plastic tanks, reared in 0.22 μm filtered ASW (same light and temperature regime as described above) and fed with sterile decapsulated Artemia nauplii for 4 weeks. For the antibiotic treatment, anemones were transferred to petri dishes and washed repeatedly with ASW individually and incubated in antibiotic solution (50 μg/ml of Carbenicillin, Chloramphenicol, Rifampicin, and Nalidixic acid in ASW) for 15 min. Polyps were then transferred to 24-well tissue culture plates (Corning Costar, United States), one polyp per well, under sterile conditions, and incubated with antibiotic solution for 7 days, with daily media exchange, at a 12 h:12 h light/dark cycle in an incubator (20–40 μmol photons m^–2 s^–1 of photosynthetically active radiation) at 25°C. After 7 days, anemones were given 24 h for recovery in ASW before microbiome inoculation (see below). Effective bacterial depletion using this protocol was validated using culture-dependent and -independent techniques (Costa et al., 2019): bacterial depletion of treated Aiptasia polyps was confirmed by absence of colony forming units (CFUs) after plating anemone lysates on Marine Agar and subsequent incubation for at least 5 days. In addition, DNA extracted using the DNeasy Blood & Tissue Kit (Qiagen, Germany) from treated Aiptasia anemone lysates were used for PCR amplification of the 16S rRNA gene (95°C for 15 min, followed by 30 cycles of 30 s at 95°C, 90 s at 55°C, and 90 s at 72°C, followed by a final extension for 10 min at 72°C using the 16S universal primer pair 27F-1492R) and subsequent confirmation of absence of a PCR product by means of electrophoresis on an agarose gel.

We determined microbial carrying capacity of apo- and symbiotic control Aiptasia polyps (i.e., untreated) as well as of anemones after antibiotic treatment and microbiome transplants. To do this, single polyps from all experimental conditions were placed in 1.5 ml tubes under sterile conditions and 300 μl of ASW were added before the polyps were lysed using a motorized pestle and mortar. Animal lysates and ASW (negative control) were diluted 10-, 100-, and 1000-fold, and 50 μl were plated on Marine Agar (Difco Marine Agar 2216, BD Biosciences, United States) and incubated at 25°C for 24–48 h or 5 days in the case of antibiotic treated anemones, with subsequent counting of bacterial colonies. For statistical analysis, colony counts were log-transformed, normality was tested using the Shapiro–Wilk test, and a one-way ANOVA was conducted. In case of statistical significance, a Dunnett post hoc test was conducted. An unpaired t-test was used to assess for significant differences between aposymbiotic and symbiotic control conditions. In order to account for putative polyp size differences, CFU counts per polyp were normalized to the host total protein for each polyp, determined using a Micro BCA protein assay kit (Pierce, United States), and the same statistical analysis as above was conducted.

Fragments of Acropora humilis and Porites sp. were collected from the nearshore reef Tahala (22°15.7812′N, 39°3.099′E) (central Red Sea, Saudi Arabia) and processed on the same day. Fragments from three colonies per coral species were collected and combined prior to inoculation (see below). Each coral fragment was placed in a sterile Ziploc bag with 10 ml of ASW, and coral tissue was air blasted off the skeleton using a sterile 1 ml barrier tip inserted to a rubber hose connected to a bench air-pressure outlet. The slurry obtained from fragments from a given coral species were combined and transferred to a 50 ml polypropylene tube and the volume was adjusted to 50 ml with ASW to reduce viscosity. The slurry was homogenized for 30 s using an Ultra Turrax T18 homogenizer (IKA, Germany) and split in 25 ml preparations per tube. Control inocula were prepared using between 10 and 15 apo- and symbiotic Aiptasia polyps, respectively, in 30 ml of ASW, homogenized as described above, and homogenized a second time with a MicroDisTec homogenizer 125 (Thermo Fisher Scientific, United States) to ensure complete maceration. The final volume of the lysate was adjusted to 50 ml with ASW and split in 25 ml preparations.

Microbiome inocula were further processed in a biosafety cabinet, using sterile work practices. Lysates were centrifuged using a swing-bucket rotor at 500 g for 5 min to pellet zooxanthellae. The supernatant was collected and an aliquot was taken for visual inspection on an inverted epifluorescence microscope. All lysates were centrifuged once, except for the Porites sp. inoculum, which had one extra centrifugation step to completely remove visible traces of the zooxanthellae. The supernatants were pooled for each inoculum type and centrifuged at 3220 g for 30 min to pellet bacteria. The resulting pellet was resuspended in 25 ml of ASW and centrifuged twice. Pellets were resuspended in 15 ml of ASW and a 1 ml aliquot was set aside for quantification of bacteria.

Bacteria were quantified using BacLight Red Bacterial stain (Thermo Fisher Scientific, United States). Aliquots from each inoculum were stained using 1 μM of dye for 10 min and counted using a BD FACSCanto II flow cytometer (BD Biosciences, United States). Gates for bacterial counts were first defined using 1 μm and 2 μm reference beads (Thermo Fisher Scientific, United States) in forward scatter (FSC) vs. side scatter (SSC) and then by using Aiptasia bacterial cultures stained with BacLight Red and acquired in the PerCP-PI channel. Bacterial numbers were calculated after gravimetric calibration of the flow rates and using the positive events acquired using the defined gating strategy. Based on determined counts, inocula were diluted to 5 × 10⁵ bacterial cells/ml and 1 ml was used per polyp. In parallel to the flow cytometry-based bacterial counts, an aliquot of all final inocula was plated on Marine Agar (Difco Marine Agar 2216, BD Biosciences, United States) for CFU counts. Bacterial densities of inocula are provided in Supplementary Table 1.

For the microbiome transplants, we assessed 30 apo- and symbiotic anemones each across five experimental conditions using six biological replicates (60 anemones in total). Experimental conditions were as follows: (1) untreated control anemones from rearing tanks (APO and SYM), (2) gnotobiotic anemones after 1 day of recovery from antibiotic treatment (APO_AB and SYM_AB), (3) Acropora microbiome inoculum (APO+ACRinoc and SYM+ACRinoc), (4) Porites microbiome inoculum (APO+PORinoc and SYM+PORinoc), (5) Aiptasia microbiome inoculum (APO+APOinoc and SYM+SYMinoc). Anemones were kept in 24-well plates, 1 polyp per well, and gnotobiotic anemones were inoculated by adding 1 ml of the respective microbiome inoculum to the well (final volume of 1 ml, 5 × 10⁵ bacterial cells/ml) and subsequent incubation for 3 days. After that, anemones were washed twice with ASW and kept in ASW for 4 additional days before being collected (7 days after microbiome transplantation).

RNA-based 16S rRNA amplicon sequencing was conducted on five biological replicates from each experimental condition and symbiotic state of the microbiome transplant experiment (see above), in addition to no template DNA extraction and no template PCR negative controls. For RNA isolation, the Qiagen AllPrep DNA/RNA kit (Qiagen, Germany) was used. Briefly, 600 μl of RLT Plus buffer were added to 150 μl of fresh lysate (or artificial seawater for the negative control) followed by snap freezing of the tubes in liquid nitrogen and storage at −80°C until extraction, following the manufacturer’s instructions. RNA quantity and integrity were determined using Qubit (Thermo Fisher Scientific, United States) and BioAnalyzer (Agilent Technologies, United States), respectively. Total RNA was DNase-treated prior to reverse-transcription using the SuperScript First-Strand Synthesis System (Invitrogen, United States), according to the manufacturer’s instructions. For amplification of 16S rRNA amplicons from cDNA, the primers 784F and 1061R (Andersson et al., 2008; Bayer et al., 2013) with MiSeq 16S adapter sequences were used (forward: 5′-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGAGGA TTAGATACCCTGGTA-3′; reverse: 5′-GTCTCGTGGGCTCGG A GA TGT GTA TA AG A G A C A G CR R C A C GA G CTGACGAC -3′; Illumina overhang adaptor sequences are underlined). PCR reactions were performed in triplicate using the Qiagen Multiplex PCR kit (Qiagen, Germany) with 2 μl of cDNA and a primer concentration of 0.5 μM in a reaction volume of 20 μL. PCRs were performed as follows: 1 cycle at 95°C for 15 min, 27 cycles each at 95°C for 30 s, 55°C for 90 s, and 72°C for 30 s, followed by a final extension step at 72°C for 10 min. Triplicate PCRs for each sample were pooled and cleaned with Illustra ExoProStar (GE Healthcare, United States). Samples were subsequently indexed (eight PCR cycles) using Nextera XT barcode sequencing adapters (Illumina, United States). Indexed PCR products were normalized using Invitrogen SequalPrep normalization plates (Thermo Fisher Scientific, United States) and pooled in equimolar ratios. Pooled samples were checked for the presence of primer dimers on a BioAnalyzer (Agilent Technologies, United States) before sequencing. The library was sequenced at the KAUST Bioscience Core Lab using 2 × 300 bp at 6 pM with 20% phiX on the Illumina MiSeq (version 3 chemistry) according to the manufacturer’s specifications.

Sequence reads were demultiplexed and adapters and barcodes were removed. Resulting sequences were then processed using mothur v.1.39.5 (Schloss et al., 2009). Briefly, paired-end sequences were assembled using the “make.contigs” command, subsequently trimmed to exclude sequences <200 bp, and rare sequences (appearing once across the entire sequencing dataset) were removed. The remaining sequences were then aligned to the SILVA database (version 132) using “align.seqs,” then pre-clustered allowing a 2 nt difference, and finally chimeric sequences were removed using VSEARCH (Rognes et al., 2016). Taxonomical classification was done using the Greengenes (release gg_13_5_99, May 2013) and SILVA (release 138, December 2016) databases. Eukaryotic, archaeal, mitochondrial, and chloroplast sequences were removed prior to OTU clustering using a 97% similarity cutoff. Putative contaminants were determined based on their abundance in negative controls and removed from all samples. An OTU was considered a contaminant if: [Σ relative abundance OTU_i in negative controls]/[Σ relative abundance OTU_i in all samples] > 0.1. Beta diversity was examined via principal coordinate analysis (PCoA) of Bray-Curtis dissimilarity distances of log₁₀(x + 1) OTU abundances (Supplementary Data 1) using the phyloseq package (McMurdie and Holmes, 2013). A permutational multivariate analysis of variance (PERMANOVA) was carried out using the “adonis” function on dissimilarity Bray-Curtis distances of relative OTU abundances to test for differences between conditions and across symbiotic states. Pairwise PERMANOVA tests were conducted using an R wrapper function for multilevel pairwise comparisons (Martinez Arbizu, 2019). To determine overlap of bacterial taxa across microbiome transplants, we determined OTUs that were present across all samples for a given experimental treatment (i.e., APO, APO+APOinoc, ACR+ACRinoc, APO+PORinoc, SYM, SYM+SYMinoc, SYM+ACRinoc, and SYM+PORinoc) and overlapping taxa were represented using the package VennDiagram (Chen and Boutros, 2011).

We compared surface topographies using scanning electron microscopy (SEM) of three model system cnidarians (Hydra, Nematostella, and Aiptasia) and three coral species (Stylophora pistillata, Acropora humilis, Porites sp.) as a first proxy to determine microbial association (Figure 1). The ectodermal epithelium of the hydrozoan Hydra was composed of a smooth surface with few cilia (i.e., 5–10 μm long hair-like plasma membrane projections, made of microtubules in a 9 + 2 ultrastructure arrangement) (Figure 1A). By comparison, Hydra tentacles showed increased ciliation (and a higher density of villi), coinciding with increased mucus and detritus retention, but also exhibited smooth(er) tissue patches scattered across the tentacles, where some bacteria were found to be attached (Figure 1B). The ectodermal epithelium surfaces of column and tentacle of N. vectensis and Aiptasia were highly similar (Figures 1C–F). We observed extensive villi (i.e., smaller, numerous plasma membrane projections, lacking microtubules) coverage in the body column of both polyps, with villi protruding from ectodermal cells ranging between 1 and 2 μm in length (Figures 1C,E) and bigger cilia ranging from 4 to 6 μm in the tentacle region (Figures 1D,F). In contrast to Hydra, we did not detect any bacteria on the column surface or tentacle regions. This suggests two things: (1) bacteria are rather rare on the surface ectoderm and may be largely constrained to the surface mucus layer in anthozoans, and (2) bacteria may be more abundant on the surface ectoderm in the hydrozoan Hydra than in anthozoans. This is corroborated by a complementary TEM analysis of the Aiptasia ectoderm, showing bacteria above the ciliary/villi band, inside the electron dense mucus layer (Figure 2 and Supplementary Figure 1). The preserved surface mucus layer was estimated around 5–15 μm, extending beyond the villi. Of further note, the surface ectoderms of apo- and symbiotic Aiptasia were indiscriminate (Supplementary Figure 2).

We also analyzed the ectodermal epithelium topography of the tentacles and coenenchyme (i.e., coral tissue between the polyps) of the three scleractinian corals Stylophora pistillata, Acropora humilis, and Porites sp. (Figures 1G–I). We observed extensive villi coverage with longer cilia ranging from 4 to 6 μm extending from the epidermis in all corals. Surface topographies were similar to N. vectensis and Aiptasia, but distinct from Hydra. In the case of Porites sp., a thick mucus sheet was visible on top of the coenenchyme, intercalated with fouled areas (Supplementary Figure 3). Further, the surface mucus layer of Porites sp. seemed distinct from S. pistillata and A. humilis, creating a more compact mucus sheet that persisted chemical fixation and several washing steps. Such sheet-like mucus appearances are described for Porites compressa, as well as the observation of the presence of fouled regions (Johnston and Rohwer, 2007). Contrary to the coral surface ectoderm that seemed devoid of bacteria, many bacteria were found attached to the smooth epithelial surfaces of other eukaryotes that aggregated in fouled regions of Porites sp. (Figure 1L), corroborating the notion that the ciliated surface of anthozoans plays a role in preventing bacterial adhesion.

To determine the putative carrying capacity of Aiptasia anemones, we obtained CFU counts from apo- and symbiotic control anemones that were subsequently compared to the different experimental conditions, i.e., antibiotic treated and microbiome inoculated animals (Figure 3 and Supplementary Tables 2, 3). The average carrying capacity of aposymbiotic Aiptasia anemones was determined as 4.25 × 10⁴ CFUs/polyp (control animals). When normalized to protein biomass, we obtained 1.59 × 10⁵ CFUs/mg host protein. The carrying capacity of symbiotic anemones was estimated at 2.02 × 10⁵ CFUs/polyp and 1.10 × 10⁶ CFUs/mg host protein, respectively (control animals). Thus, the number of CFUs was higher in symbiotic polyps by about an order of magnitude compared to aposymbiotic polyps (unpaired t-test, P = 0.025, Supplementary Table 4). We did not obtain CFUs from antibiotic-treated anemones, confirming successful bacterial depletion (Figure 3B). Interestingly, in some cases (re-)infection of gnotobiotic Aiptasia with microbiome inocula of either aposymbiotic Aiptasia or corals increased the carrying capacity of aposymbiotic animals. For instance, aposymbiotic anemones exposed to Porites sp. inoculum exhibited a significant increase in CFU counts to an average of 1.85 × 10⁵ CFUs/polyp (Dunnett’s test P < 0.05, Figure 3B and Supplementary Tables 2, 4). Conversely, inoculation of symbiotic anemones (SYM+ACRinoc and SYM+PORinoc) did not result in a significant increase in CFU counts (Dunnett’s test P > 0.3, Figure 3B and Supplementary Tables 2, 4). Overall, the carrying capacity of Aiptasia polyps with ∼5 mm of oral disk was between 10⁴ and 10⁵ CFUs for apo- and symbiotic anemones, respectively. After microbiome inoculation the carrying capacity was at about 10⁵ CFUs/polyp, irrespective of the symbiotic condition.

We employed RNA-based 16S amplicon sequencing to assess active bacterial community composition and dynamics (in contrast to the resident community based on DNA-based 16S sequencing) associated with Aiptasia under the various experimental treatments (Figure 4). Bacterial community composition differed significantly between treatments (PERMANOVA, P = 0.001) (Supplementary Table 5). As previously reported based on DNA-based 16S marker gene sequencing (Röthig et al., 2016a), bacterial communities of apo- and symbiotic Aiptasia anemones were different (PERMANOVA, P = 0.015, Figure 4B). For this reason, we clustered apo- and symbiotic samples separate to resolve differences within apo- and symbiotic groups as a result of the treatments (Figure 4A). For both groups, antibiotic-treated anemones (APO_AB and SYM_AB) were clearly separated from control anemones (APO and SYM) and both were different from gnotobiotic anemones re-inoculated with microbial communities (with the exception of SYM+SYMinoc that closely clustered with SYM anemones) (Figure 4A).

The bacterial assemblage of antibiotic-treated anemones (APO_AB and SYM_AB) was markedly different from the microbiomes of other anemones in that it was considerably less complex. This was highlighted by the notion that >90% of the active community in apo- and symbiotic anemones was comprised of the same three bacterial taxa (OTU0002, OTU0003, and OTU0006) affiliated to the order Rhizobiales and the genera Marinobacter and Alcanivorax (Figure 4B). Conversely, 75% of the active community of aposymbiotic control anemones was comprised of only six bacterial taxa (OTUs 0005, 0012, 0013, 0014, 0015, 0018, Figure 4B), which were largely absent in the antibiotic treated Aiptasia. Notably, >70% of the bacterial communities of all inoculated aposymbiotic anemones were dominated by the same taxa regardless of the inoculum. The dominant bacterial taxa were representatives of the order Alteromonadales (OTU0001, OTU0004, OTU0007, OTU0008), which were barely detected in aposymbiotic control anemones. For the symbiotic anemones, the active bacterial community of control polyps was dominated by a few bacterial taxa of the order Alteromonadales and the genera Glaciecola and Thalassomonas (Figure 4B). Similarly, microbiome transplanted symbiotic Aiptasia resembled each other and were dominated by bacteria of the order Alteromonadales (OTU0001, OTU0004, OTU0007, OTU0008, OTU0009), just as the microbiome transplanted aposymbiotic Aiptasia. Thus, despite the differences in microbial community composition of apo- and symbiotic control anemones, microbiome transplanted gnotobiotic apo- and symbiotic anemones look much more alike with regard to their microbiome.

To get better insight into the bacterial taxa that were present in Aiptasia after microbiome transplantation, we compared the active bacterial community of control conditions (APO/SYM) and treatment conditions (Figure 5 and Supplementary Table 6). This comparison shows that all inoculated anemones (+APOinoc, +SYMinoc, +ACRinoc, and +PORinoc) harbored bacterial taxa that were already present (active) in control anemones as well as “novel” (i.e., previously undetected) bacteria. This may point to potential carryover from gnotobiotic anemones (not all bacteria were depleted). In the case of coral microbiome inoculations, we observed association with novel bacteria from the coral microbiomes (as determined by the comparison to APO and SYM anemones, Figure 5 and Supplementary Table 6). In particular, two bacteria taxa, of the genus Thalassomonas (OTU0016) and an unclassified Alteromonadales (OTU0021), are worthwhile mentioning because they were both found to be abundant (i.e., active) in coral microbiome transplanted Aiptasia (i.e., APO+ACRinoc, APO+PORinoc, SYM+ACRinoc, and SYM+PORinoc), but absent from Aiptasia transplanted with their own microbiomes (i.e., APO+APOinoc and SYM+SYMinoc) (Supplementary Table 6). Both OTUs are putative coral microbiome bacteria that were able to colonize apo- and symbiotic Aiptasia. On the contrary, a putative coral microbiome taxon from the genus Glaciecola (OTU0195) was only found associated with symbiotic anemones. Further, we found two Gammaproteobacteria (OTU0067 and OTU0099) to be specifically prevalent in anemones inoculated with Porites sp. microbiome preparations.