First insights into the Aurelia aurita transcriptome response upon manipulation of its microbiome

Introduction The associated diverse microbiome contributes to the overall fitness of Aurelia aurita, particularly to asexual reproduction. However, how A. aurita maintains this specific microbiome or reacts to manipulations is unknown. Methods In this report, the response of A. aurita to manipulations of its native microbiome was studied by a transcriptomics approach. Microbiome-manipulated polyps were generated by antibiotic treatment and challenging polyps with a non-native, native, and potentially pathogenic bacterium. Total RNA extraction followed by RNAseq resulted in over 155 million reads used for a de novo assembly. Results The transcriptome analysis showed that the antibiotic-induced change and resulting reduction of the microbiome significantly affected the host transcriptome, e.g., genes involved in processes related to immune response and defense mechanisms were highly upregulated. Similarly, manipulating the microbiome by challenging the polyp with a high load of bacteria (2 × 107 cells/polyp) resulted in induced transcription of apoptosis-, defense-, and immune response genes. A second focus was on host-derived quorum sensing interference as a potential defense strategy. Quorum Quenching (QQ) activities and the respective encoding QQ-ORFs of A. aurita were identified by functional screening a cDNA-based expression library generated in Escherichia coli. Corresponding sequences were identified in the transcriptome assembly. Moreover, gene expression analysis revealed differential expression of QQ genes depending on the treatment, strongly suggesting QQ as an additional defense strategy. Discussion Overall, this study allows first insights into A. aurita’s response to manipulating its microbiome, thus paving the way for an in-depth analysis of the basal immune system and additional fundamental defense strategies.


Introduction
Cnidarians, such as the moon jellyfish Aurelia aurita, are distributed worldwide and play essential roles in shaping marine ecosystems (Brekhman et al., 2015). Cnidaria are dated back to about 700 million years and are considered a sister group to the Bilateria (Putnam et al., 2007;Park et al., 2012). Thus, they are among the simplest animals at the tissue level organization possessing two germ layers (ectoderm and endoderm) separated by the mesoglea (Ball et al., 2004). In addition to their morphological simplicity, many Cnidaria, particularly Scyphozoa, have a high level of developmental plasticity, allowing for an enormous tolerance, regeneration the Cnidarian Hydra, the autoinducer signaling molecule 3-oxo-homoserine lactone has been shown to be converted into the inactive 3-hydroxy counterpart by a host-derived oxidoreductase allowing host colonization of the main colonizer Curvibacter sp. (Pietschke et al., 2017).
The Cnidarian A. aurita harbors a highly diverse and dynamic microbiota specific to the animal, the different sub-populations, and life stages (Weiland-Bräuer et al., 2015a). In the absence of the specific microbial community, the fitness of A. aurita was significantly compromised, and notably, asexual reproduction was almost halted (Weiland-Bräuer et al., 2020a). This microbial impact is crucial at the polyp life stage before entering the process of asexual offspring production (strobilation) to ensure a normal progeny output (Jensen et al., 2023). Moreover, in A. aurita, three proteins interfering with bacterial QS were identified (Weiland-Bräuer et al., 2019). Incubation of native animals with potentially pathogenic bacteria induced the expression of the identified QQ-ORFs, strongly suggesting a host defense strategy.
Despite the growing knowledge about the impact of microbes on the host and the fundamental strategies of the host to respond, research to understand how microbiomes influence host gene expression is still in its infancy (Nichols and Davenport, 2021). Many studies of model organisms and humans demonstrated an interlinkage between the microbiome and the host's gene expression. However, the direction of causality mainly remained unanswered (Nichols and Davenport, 2021). Comparing conventional (microbiome-containing) to germ-free systems is one way to assess whether the microbiome plays a causative role in regulating gene expression (Bäckhed et al., 2012;Al-Asmakh and Zadjali, 2015;Fu et al., 2017;Pierre, 2022). Genome-wide transcriptomic analyses are now routinely used to quantify the changing levels of each transcript under different conditions (Conesa et al., 2016).
In the present study, we aimed to determine the influence of the associated microbiota on A. aurita's gene expression. After microbiome manipulation (by antibiotic treatment or bacterial challenge), RNA was extracted from polyps, followed by RNA-Seq and a de novo transcriptome assembly. Gene ontology categories and gene expression patterns were analyzed to elucidate how A. aurita recognizes and responds to the manipulation of its native microbiome and the presence of potential pathogens. A particular focus was on hostderived QQ activities as an additional potential defense strategy.

Reduction of the native Aurelia aurita polyp microbiota by antibiotics
Single native polyps were placed in 48 well multiwell plates in 1 mL ASW supplemented with an antibiotic mixture (Provasoli's antibiotic mixture with final concentrations of 360,000 U/L penicillin G, 1.5 mg/L chloramphenicol, 1.8 mg/L neomycin, and 9,000 U/L polymyxin B; all components from Carl Roth, Karlsruhe, Germany). No food was provided during the antibiotic treatment. The reduction and consequent change of the microbiota were tested by plating a single homogenized polyp (10 replicates) on Marine Bouillon agar plates (Carl Roth, Karlsruhe, Germany). Plates were incubated for 5 days at 20°C. Colony forming units (cfu) were calculated, and an 87 ± 9% reduction per polyp was determined.

Experimental design for transcriptome analysis
The North Atlantic sub-population polyps were generated from a single mother polyp by clonal budding. Pools of 20 daughter polyps were separated into 6-well plates in 4 mL 3% ASW. Polyps were kept in five conditions without food supply: 1. native polyps without treatment, 2. antibiotic (AB)-treated polyps, 3. native polyps challenged with 10 8 cells/mL Klebsiella oxytoca M5aI, 4. native polyps challenged with 10 8 cells/mL Pseudoalteromonas espejiana, and 5. native polyps challenged with 10 8 cells/mL Vibrio anguillarum ( Figure 1). Three replicates (3 × 20 polyps) were used for each condition. For the bacterial challenges, 20 polyps in 4 mL ASW were supplemented with 4×10 8 cells and incubated for 30 min (Figure 1). The total RNA was extracted from each pool of polyps.

RNA isolation
Total RNA of a pool of 20 A. aurita polyps was isolated with an adapted protocol of Gold et al. (2019). In more detail, polyps were washed three times with sterile ASW (to remove antibiotic residues) and homogenized with a motorized pestle. RiboLock RNase Inhibitor (40 U/μL, Thermo Fisher Scientific, Waltham/Massachusetts, United States), 200 μL lysis solution (100 mM Tris/HCl, pH 5.5, 10 mM disodium EDTA, 0.1 M NaCl, 1% SDS, 1% ß-mercaptoethanol), and 2 μL Proteinase K (25 mg/mL, Thermo Fisher Scientific, Waltham/Massachusetts, United States) were added to the homogenate and incubated for 10 min at 55°C. Chilled solutions were added with 5 μL of 3 M sodium acetate (pH 5.2) and 250 μL phenol-chloroform-isoamyl alcohol (25:24:1) and incubated for 15 min on ice prior to 15 min centrifugation at 12,000 x g at 4°C. The upper phase was mixed with 1 volume 2-propanol, and precipitation occurred at-80°C overnight. The precipitate was centrifuged for 15 min at 12,000 x g at 4°C. The pellet was washed twice with 70% ethanol before the air-dried pellet was dissolved in 25 μL RNase-free water. DNA contaminations were removed with Turbo DNA-free DNase (Thermo Fisher Scientific, Waltham/ Massachusetts, United States). The RNA quality and quantity were assessed by NanoDrop1000 (Thermo Fisher Scientific, Waltham/ Massachusetts, United States) and 1.5% agarose gel electrophoresis, and the cDNA library was prepared with DNA-free host RNA (500 ng) using the TruSeq Stranded mRNA Library Preparation kit. The library was sequenced with the NextSeq 500 System (Illumina, San Diego/ California, United States).

Transcriptome analysis
The raw reads were trimmed with Trimmomatic (Bolger et al., 2014) to remove bad-quality reads and the adapters. A de novo assembly with the trimmed sequences was conducted with the Trinity package v2.8.4 (Grabherr et al., 2011a,b). The quality of the de novo assembly was assessed with several parameters, such as the N50 value and the percentage of the assembly-mapped reads, which was calculated with Bowtie2. Finally, Benchmarking Universal Single-Copy Orthologs (BUSCO V2/3) against metazoan cassettes (Simão et al., 2015) was used to evaluate the completeness of the assembly regarding the core genes found in metazoans. Experimental setup. RNA extraction with subsequent RNAseq and analysis was conducted on a pool of 20 polyps (3 biological replicates). Polyps were kept under native and microbiome-manipulated conditions. Created with BioRenders.
Frontiers in Microbiology 04 frontiersin.org A Blastx of the transcriptome assembly was executed against the Swiss-Prot database for metazoans. For the gene expression analysis, reads were mapped to the reference assembly with Bowtie2 (Langmead and Salzberg, 2012). Transcript quantification was performed with RSEM (Li and Dewey, 2011). The differential gene expression (DGE) analysis was done with edgeR (Robinson et al., 2010;McCarthy et al., 2012). Pairwise comparisons between all experimental conditions were performed using FDR ≤ 0.001 and 2-fold changes as statistical parameters. Furthermore, GO enrichment analysis was conducted using a Fisher's Exact Test in Blast2GOPRO (Conesa et al., 2005) with a value of p threshold of ≤0.05. Here, the total annotation file of the reference transcriptome was used as the "reference dataset," whereas the upregulated genes in each condition served as the "test dataset." Based on sequence depth and quality, only two out of three biological replicates were used for the gene expression analysis for the conditions of native polyps and native polyps under the bacterial challenges. Transcriptome data is deposited under the BioProject ID PRJNA938117. Gene expression patterns in the transcriptome data set were verified with qRT-PCRs of the housekeeping gene elongation factor 1 (EF1) (Weiland-Bräuer et al., Jensen et al., 2023). EF1 showed similar trends in gene expression for all treatments when comparing transcriptome data and qRT-PCR (Table 1). Moreover, the previously published QQ genes aaqq1, aaqq2, and aaqq3 of A. aurita (Weiland-Bräuer et al., 2019) were identified in the transcriptome data set. Their expression profiles have been studied using qRT-PCR in native, antibiotic-treated, and K. oxytoca-treated polyps [see study (Weiland-Bräuer et al., 2019)]. Gene expression profiles were similar in transcriptome data and qRT-PCR, which we rated and considered as internal control for our newly obtained data set and its validity (Table 1).

Quorum quenching assay
Quorum quenching (QQ) assays using the reporter strains AI1-QQ.1 and AI2-QQ.1 were performed with cell-free supernatants and cell extracts of EST clones from the A. aurita EST library as described (Weiland-Bräuer et al., 2015b). Briefly, QQ screening plates were prepared with LB agar containing 0.8% agar at 50°C supplemented with final concentrations of 100 μM N-(β-ketocaproyl)-l-homoserine lactone (3-oxo-C6-HSL) (Sigma-Aldrich, Munich, Germany), 100 μg/mL ampicillin, 30 μg/ mL kanamycin, and 10% (vol/vol) exponentially growing culture of the reporter strain AI1-QQ.1. Similarly, AI-2 QQ screening plates were prepared with final concentrations of 50 mM 4-hydroxy-5methyl-3-furanone (Sigma-Aldrich, Munich, Germany), 100 μg/mL ampicillin, 30 μg/mL kanamycin, and 5% (vol/vol) exponentially growing culture of the reporter strain AI2-QQ.1. LB agar plates were coated with the agar mixture. After 10 min, 5 μL of the test substances were applied, followed by overnight incubation at 37°C. QQ activities were visualized by the growth of the respective reporter strain. Preparation of cell extracts and cell-free culture supernatants was conducted with 5 mL overnight cultures of EST clones grown at 37°C and 120 rpm. Cells were harvested by centrifugation at 7,000 × g, and the culture supernatant was subsequently filtered using 0.2-μm centrifugal filter units (Carl Roth, Karlsruhe, Germany). The cell extract was prepared from the cell pellet using the Geno/Grinder 2000 (BT&C/OPS Diagnostics, Bridgewater, NJ). Cell extracts were filtered through 0.2-μm filter units.
Following the manufacturer's protocol, the plasmids of identified QQ-active single EST clones were purified using the Presto Mini Plasmid kit (GeneAid, New Taipeh City, Taiwan). The respective inserts were Sanger sequenced at the Institute of Clinical Molecular Biology in Kiel with the primer set T7_Promoter (5′-TAATACGACTCACTATAGGG-3′) and T7_Reverse (5'-TAGTT ATTGCTCAGCGGTGG-3′). Sequences are deposited at NCBI GenBank under Accession Nos. OQ581004 -OQ581041.

Results
A transcriptomics approach was applied to gain insights into the response of A. aurita polyps to manipulation of its native, associated microbiota, with a particular focus on quorum sensing interference as a potential host defense strategy. In general, pools of 20 polyps were treated and processed together in the experiments, each with three biological replicates using the following treatments: native, antibiotic-treated polyps resulting in 87 ± 9% reduced microbial cells per polyp (cfu/polyp), and native polyps challenged with Klebsiella oxytoca M5aI, Pseudoalteromonas espejiana, or Vibrio anguillarum (2 × 10 7 cells/polyp, Figure 1). The

Statistics of the de novo transcriptome assembly
The RNAseq approach overall resulted in 167,269,257 raw reads. After quality trimming, 145,746,530 reads (87.1% of the total reads) remained for further analysis and were used for the de novo transcriptome assembly (Supplementary Table S1A

Manipulation of its microbiome affects A. aurita's transcriptome
Transcriptome analysis identified five separated clusters corresponding to the different conditions, while biological replicates of a condition cluster together (Figure 2). It should be noted that three replicates were only analyzed for antibiotic (AB)-treated polyps, as all other conditions resulted in an unacceptable low sequence depth for one replicate. Notably, two clusters were identified within the hierarchical clustering of transcripts. Here, native and AB-treated conditions yielded a tremendous difference. AB-treated polyps showed a massive reduction of the microbial load (by 87 ± 9%), assuming a rigorous change in the abundance and diversity of microbial colonizers. Bacteria-challenged polyps showed a mixture of native and AB-treated expression patterns (Figure 2). In more detail, the comparison of native and AB-treated polyps resulted in the highest number of differentially expressed genes (22,073 genes), with 10,451 upregulated and 11,622 down-regulated genes in AB-treated compared to native polyps (Supplementary Table S3A). According to  the GO enrichment analysis (Conesa et al., 2005), the 10,451 upregulated genes were overrepresented in processes related to inflammation and immune response (e.g., acute inflammatory response to antigenic stimuli, MAP-kinase activity, cytokine production, and neutrophil cell activation) ( Figure 3A; Supplementary Table S4A). Genes involved in immune response and apoptosis included, e.g., Caspases, Apoptosis regulators, Interferones, and Toll-like receptors ( Figure 3B; Supplementary Tables S3A, S4A). On the contrary, down-regulated genes in AB-treated polyps were enriched for processes related to development, morphogenesis, reproduction, stimuli response, and signaling (Figures 3C,D;  Supplementary Tables S3A, S4A).
When comparing expression patterns between native and bacteria-challenged polyps, differences based on the bacterial species used in each challenge were revealed (Figure 4). The comparison of native polyps compared to those challenged with V. anguillarum gave the highest number of differentially expressed (DE) genes (7,193 genes, Figure 4A; Supplementary Table S3B), followed by polyps challenged with K. oxytoca (6,974 DE genes, Figure 4A; Supplementary Genes related to biological processes, such as defense, immune and inflammatory responses, and the regulation of apoptotic processes, were upregulated in bacteria-challenged polyps ( Figure 4B; Supplementary Table S3E). In contrast, genes related to organism development, cell cycle processes, and cytoskeleton organization were down-regulated in challenged polyps ( Figure  4B; Supplementary Table S3E). Moreover, exclusively upregulated and down-regulated genes were identified for each bacterial challenge ( Figure 4). 1,319 genes were exclusively upregulated in V. anguillarumchallenged polyps ( Figure 4A Table S3H). Similarly, exclusively down-regulated genes were identified ( Figure 4A, right panel). Analyzing the GO-enriched categories for each species (Supplementary Table S4B-D), we found categories common among all treatments (Supplementary Table S4E) as well as categories present in two different treatments or only in one ( Figure 4B; Supplementary Table S4B-D). Focusing on genes exclusively upregulated after each species-specific challenge, we indeed observed that the same GO categories, e.g., immune response, MAPK signal transduction, autophagy, and DNA repair, were enriched to different degrees, though represented by various genes (depicted in Figure 5).

Bacterial challenge of polyps affects host quorum quenching
Recent studies revealed that interfering with Qurom sensing, so-called quorum quenching (QQ), might be a fundamental, additional inter-phylum interaction to maintain metaorganismal homeostasis. Consequently, we aimed to identify host-derived QQ activities. An expressed sequence tag (EST) library from A. aurita  Exclusively upregulated genes in native vs. bacteria-challenged polyps. Bubble plot showing the degree of enrichment per category of exclusively upregulated genes in each bacteria-challenged compared to native conditions. polyps-derived mRNA was constructed in E. coli SoluBL21 (Ladewig et al., 2023). The library consisted of 29,952 clones with an insertion efficiency of approx. 98% and an average insert size of 1.46 kbp, resulting in 43 Mbps cloned host transcriptome, corresponding to an estimated 11.4% coverage [calculated A. aurita genome size 376 Mbps (Gold et al., 2019)]. The A. aurita EST library was successively screened for QQ activities toward acyl-homoserine lactones (AHL) and autoinducer-2 (AI-2) in cell-free cell extracts and culture supernatants of the EST clones using established E. coli-based reporter systems (Weiland-Bräuer et al., 2015b). Overall, 37 out of 29,952 EST clones were identified as QQ active (Table 2). Predominantly, QQ activities against the Gram-negative signaling molecule AHL were detected. Although those host-derived QQ activities were identified in a functional screen, their biological activities in vivo and their functions in A. aurita have to be explored. In a first attempt, plasmid insertions of QQ-active single EST clones were sequenced, and sequence data subsequently checked for homologies in the A. aurita transcriptome assembly. All QQ-conferring sequences were identified in the transcriptome assembly with homologies ranging from 83 to 100% (Supplementary Table S5). Furthermore, public databases (NCBI, UniProt, PFAM) were used for homology predictions and annotation (Supplementary Table S5). Unexpectedly, several QQ-ORFs showed homologies to highly conserved ribosomal proteins, suggesting a moonlighting function of those proteins (Jeffery, 2003;Singh and Bhalla, 2020). Secondly, analyzing transcription levels of those identified QQ-ORFs in the generated RNAseq data set (see above) allowed first insights into their transcriptional regulation in response to the respective treatments, which might support predicting their biological role. Relative expression of the 37 QQ-conferring genes was calculated for microbiome-manipulated versus native polyps. Hierarchical clustering observed four clusters of expression profiles. The first cluster represents QQ genes similarly expressed in microbiome-manipulated polyps compared to native ones ( Figure 6, right column highlighted in grey). The second cluster of genes showed decreased expression in all treatments, except when challenged with P. espejiana ( Figure 6, right column light orange cluster). The third Frontiers in Microbiology 09 frontiersin.org cluster summarizes those genes with increased expression regardless of the type of manipulation ( Figure 6, right column dark orange cluster). The fourth cluster includes those genes with an increased expression after adding potential pathogens but reduced expression when polyps were AB-treated ( Figure 6, right column green cluster). Notably, P. espejiana primarily showed increased expression of QQ-ORFs; thus, hierarchal clustering showed the highest dissimilarity compared to the other treatments.

Discussion
In the present study, we obtained the first insights into the transcriptomic response of A. aurita to microbiome manipulation. The manipulations of the polyp microbiome included a massive reduction of viable bacterial cells (87% reduction) due to applying a broad-spectrum antibiotics mixture. The mixture contained Penicillin, Polymyxin B, Chloramphenicol, and Neomycin (Provasoli and Pintner, 1980;Liu et al., 2017a;KleinJan et al., 2022), and has been shown to only partially eliminate an entire bacterial community (Azma et al., 2010;Liu et al., 2017a). We assume that by drastically reducing the colony-forming units by 87%, the diversity and abundance of community members on the polyp are crucially changed. Consequently, the polyp was confronted with the loss of certain community members and a drastic change in the relative abundance of remaining ones. This extreme change in microbiome composition resulted in the up-regulation of various immune response mechanisms. Pattern recognition receptors (PRRs) like Tolllike receptors and Ficolins were upregulated, inducing the first line of defense ( Figure 3). Further, Caspases were enriched in AB-treated polyps, potentially maintaining homeostasis through regulating cell death and inflammation (McIlwain et al., 2013). Contrarily, all processes involved in morphogenesis and development, thus not essential for survival, were down-regulated in AB-treated polyps ( Figure 3). Furthermore, the native microbiome was affected by challenging the polyps with three different bacteria in high cell numbers: (i) K. oxytoca has not been detected in any life stage of A. aurita and thus represents a non-native bacterium (Weiland-Bräuer et al., 2015a; (ii) V. anguillarum, an opportunistic pathogen of various invertebrates and vertebrates, has been isolated from an A. aurita polyp (Austin, 2010;Frans et al., 2011;Weiland-Bräuer et al., 2015a,b); and (iii) P. espejiana was initially isolated from seawater but was also found in high abundance associated with all life stages of A. aurita, thus representing a native bacterium (Isnansetyo and Kamei, 2009;Weiland-Bräuer et al., 2015a,b). Due to bacterial challenges, we observed the up-regulation of defense, immune, and inflammatory responses, and apoptosis regardless of the bacterial species. Based on this finding, we hypothesize that the host recognized the non-native and native bacteria as a potential threat, at least when present in such high cell numbers. The immune system of A. aurita likely responded with the four primary innate immune system functions as demonstrated for other Cnidarians and already partly observed for AB-treated polyps (Miller et al., 2007;Parisi et al., 2020). First, immune recognition occurred by PRRs, like Fucolectins, Ficolins, and NACHT-containing domain proteins, binding to bacterial MAMPs/ PAMPs (Figures 4, 5). Subsequently, various transcription factors from the NF-κB family were activated (Figures 4, 5). Intracellular signaling cascades (MAPK, interleukin, cytokines) led to target gene transcription to eliminate the threat and mitigate self-harm (Figures 4,  5). Lastly, autophagy, DNA repair, and programmed cell death were upregulated after bacterial challenge (Figures 4, 5). Our observations Functionally identified QQ-ORFs are listed with their QQ activity in the cell-free cell extract (CE) and culture supernatant (SN) against acyl-homoserine lactones (AHL) and autoinducer-2 (AI-2) with their potential annotations; x, activity; −, no activity.
Frontiers in Microbiology 10 frontiersin.org are consistent with previous results showing that A. aurita's general fitness and, in particular, its asexual reproduction were drastically affected in the absence of microbes and due to the manipulation of the microbiota by challenging with non-native colonizers, e.g., P. espejiana and V. anguillarum (Weiland-Bräuer et al., 2020a). The acquisition, establishment, and maintenance of a specific microbiota are advantageous for the host, and its disturbance can contribute to developing diseases (Zheng et al., 2020). The host strives to maintain this homeostasis and presumably uses other defense mechanisms besides the innate immune system (Nichols and Davenport, 2021). The use of bacterial communication and its interference by the host has been recently regarded as an additional interaction mechanism within the complex interplay of the host and its microbiome (White et al., 2020;Weiland-Bräuer, 2021). Several studies evaluated the implication of highly conserved paraoxonases (PON1, PON2, and PON3) as QQ enzymes of the host defense against the pathogen Pseudomonas aeruginosa (Mochizuki et al., 1998;Draganov et al., 2005;Stoltz et al., 2007). In the present study, we functionally detected QQ activities of A. aurita in a cDNA expression library and verified the respective transcripts within the de novo assembled host transcriptome. Their native expression in response to microbiome manipulation resulted in four different expression profiles ( Figure 6). A first cluster represents QQ-ORFs similarly low expressed in treated and native polyps. Those findings argue against a biological function in the defense against pathogens. The second cluster showed decreased expression in all treatments except when challenged with P. espejiana. Here, a specific defense against P. espejiana can be assumed. This assumption aligns with previous studies showing massive declined survival rates in the presence of this bacterium (Weiland-Bräuer et al., 2020a). The multitude of initiated defense mechanisms in the presence of the ubiquitous, native P. espejiana (various QQ-ORFs and acute inflammatory response, Figures 5, 6) might indicate that this bacterium represents a threat to A. aurita, at least in the high cell numbers, and must be rapidly defended. The third type of expression profile showed increased expression regardless of the type of microbiome manipulation, assuming a general defense. Among those is Ferritin, generally regarded as an intracellular iron storage protein and suggested to be involved in response to infection in different organisms, including fish and marine invertebrates (Moreira et al., 2020). The expression of Ferritin or Ferritin-homologs was upregulated in different tissues in response to bacterial infections or stimulation with LPS (Neves et al., 2009;He et al., 2013;Ren et al., 2014;Chen et al., 2016;Sun et al., 2016;Martínez et al., 2017; Frontiers in Microbiology 11 frontiersin.org 2017b). Furthermore, Ferritin was shown to protect shrimp and fish from viral infections (Ye et al., 2015;Chen et al., 2018). The fourth cluster was expressed in the presence of the non-native and native bacteria, regarded as potential pathogens, but not after antibiotic treatment, assuming a fast response to potential pathogens' colonization possessing similar PAMPs. Homologies to already described genes within the NCBI database of those QQ-ORFs to predict their function were rare. QQ-ORFs Aa_QQ_15 and Aa_ QQ_30, belonging to the fourth cluster, showed non-significant homologies to a chitinase and acetyl-transferase, respectively. Here, enzymatic modification (hydrolysis or acetylation) of the bacterial signaling molecule can be speculated. The four other QQ-ORFs within this expression profile showed the best homologies with highly conserved proteins, including actin, heat shock, and ribosomal proteins. Particularly, ribosomal proteins crucially involved in translation have been shown to comprise other functions (Hurtado-Rios et al., 2022), thus recognized as moonlighting proteins.
Moonlighting proteins are capable of performing more than one biochemical function within the same polypeptide chain, including inhibiting infectious bacteria, viruses, parasites, fungi, and tumor cells (Jeffery, 2003;Gao and Hardwidge, 2011;Wang et al., 2015;Singh and Bhalla, 2020). Thus, they have been considered antimicrobial peptides (AMPs) (Hurtado-Rios et al., 2022). Consequently, we hypothesize that the identified QQ activity, besides their canonical function, is promiscuous. Future studies have to be conducted to verify the biological function of the identified A. aurita QQ-ORFs, e.g., by genetic and biochemical approaches.
In conclusion, this study provides the first A. aurita transcriptome analysis focusing on the impact of microbiome disturbance on host gene expression. Overall it highlights the importance of the native microbiome since processes like morphogenesis, development, and response reactions are down-regulated when it is disturbed due to antibiotics. Microbiome disturbance further induced apoptosis and immune responses, indicating the microbiome's protective function. Similarly, challenging A. aurita with high loads of bacteria resulted in the up-regulation of defense, immune, and inflammatory responses to maintain metaorganismal homeostasis. We have further received indications that quorum quenching is likely an additional mechanism for maintaining the specific microbiota besides the innate immune system, particularly acting as a fast response to potential pathogens.

Data availability statement
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: PRJNA938117 (SRA), OQ581004-OQ581041 in the NCBI database.

Author contributions
RS and NW-B: conceptualization and supervision. DL: methodology. DL and NW-B: investigation. NW-B: formal analysis. VK: bioinformatics analysis and data curation. NW-B, VK, and RS: writing-original draft preparation, writing-review, and editing. RS: project administration and funding acquisition. All authors have read and agreed to the published version of the manuscript.

Funding
This work was conducted with the financial support of the DFG-funded Collaborative Research Center CRC1182 "Origin and Function of Metaorganisms" (B2) and the DFG-funded ImmuBase project. Project number 261376515 to the CRC1182 fund, and Project number 417981041 to the Immubase fund.