The research described herein was performed by using Ciona specimens collected in the Mar Piccolo of Taranto (Taranto, Italy, ~40°29’29”N, 17°17’55”E) and in the Fusaro lagoon (Naples, Italy, ~40°49’10”N, 14°03’28”E), in locations that are not privately-owned nor protected in any way, according to the authorization of Marina Mercantile (Decree of the President of the Republic (DPR) 1639/68, Sep. 19, 1980, confirmed on Jan. 10, 2000). The study did not involve mammalian or vertebrate subjects, or endangered or protected species, and was carried out in strict accordance with European (Directive 2010/63/EU) and Italian (Legislative Decree n. 26/2014) legislation for the care and use of animals for scientific purposes. C. robusta is considered an introduced species and is not regulated or protected by environmental agencies in Italy. The animal collection services contracted in this study maintain current permits and licenses for collection and distribution of marine invertebrates to academic institutions. No special permission was required to collect ascidians, and animal handling was in accordance with the guidelines of our academic institutions. Animals were recovered and brought to the laboratory alive and maintained in clean water with aeration, temperature control and properly fed. In accordance with general animal protocols, the least number of specimens required per experiment were utilized. Animal waste products were disposed appropriately.

To obtain stage 4 Ciona juveniles, in vitro fertilization was performed using eggs and spermatozoa surgically collected from the gonoducts of different animals that had been exposed to constant light to elicit gamete maturation. In vitro egg fertilization follows the procedure described previously (40), with the exception that 0.22 μM filtered seawater (FSW) was used and no sterilization step was performed. After the time required for egg fertilization (about 10 minutes), fertilized eggs were diluted in FSW in 150 mm Petri dishes and raised at 18°C. Eighteen hours (hr) later, when animal have reached the non-feeding swimming tadpole larval stage, batches containing at least 90% normally developed larvae were selected for inflammatory treatments. Then, larvae were gently transferred to 6-well plates containing 4 ml FSW and 600-700 larvae per well. Animals were let to grow O/N at 18°C, and the day after the FSW was replaced with fresh one to eliminate post-metamorphic stages not attached to the dish, or unsettled larvae not properly developed. When juveniles have reached stage 4 (33), 3-4 days post fertilization, they were ready to be treated with different inflammatory stimuli, such as LPS (Sigma #L2880), Pam2CSK4 (InvivoGen #tlrl-pm2s-1) and zymosan (InvivoGen #tlrl-zyn). We referred to literature for the concentration of each inflammatory agent that best suits the activation of the immune system. LPS is usually injected in adults in the quantity of 100 μg dissolved in marine solution (27–29), or resuspended in a cell line-based heterologous system at the concentration of 2.5 and 5 μg/ml (17). Here, we used LPS at an intermediate concentration of 10 μg/ml. The triacylated form Pam3CSK4 was tested on cell lines in a heterologous system at the concentration of 0.5 and 2 μg/ml (17). Here, we used Pam2CSK4 at 1 and 10 μg/ml concentrations. The third inflammatory agent, zymosan, was reported in Ciona as a modulator of TLR activation in an in vitro heterologous system at the concentration of 100 μg/ml (17). Here, we used it at 10 and 100 ug/ml concentration. Each compound was resuspended at the appropriate concentration in 4 ml of 0.22 μm FSW, and animals were treated with these solutions for 30 minutes (min), 2 hr and 4 hr, while control animals were exposed to normal FSW. At the end of each time treatment, samples for gene expression analysis were collected in RNAlater by replacing the FSW supplemented with the microbial stimulus with 1 ml RNAlater solution in each well. Sample storage followed the manufacturer’s instruction; briefly, after O/N incubation in RNAlater at room temperature (RT), RNAlater solution was removed and samples (whole animals still attached to the bottom of 6-well plates) were frozen at -20°C for long term storage until RNA extraction. In turn, samples for proteomic analyses were processed by removing FSW supplemented with the microbial stimulus followed by fast freezing in dry ice and storage at -80°C until protein extraction.

To increase knowledge about the landscape of molecular dynamics triggered by inflammation, a protein-protein interaction map was generated using the STRING database (54). Multiple sequence queries were followed by evaluation of protein sequence identity by means of Multiple Sequence Alignment (MSA) in Jalview (46) and BlastP (44, 45). The resulting interaction networks were modified and merged with the Cytoscape v.3.9.0 software (55) in order to build a complete protein-protein interactome in which to connect all molecules whose gene expression was analyzed in this study. Among the applications available in Cytoscape “App Manager” for functional analyses of immune-related pathways, we installed and used the STRING Enrichment application, that is based mainly on KEGG (56), SMART, COMPARTMENTS (57), Panther (58, 59), AmiGO 2 (60) and QuickGO (61), Pfam (62), InterPRO and Reactome (63) databases. This application was used to retrieve annotations about compartmentation, Gene Ontology (GO), pathway analysis and domain analysis of the selected nodes of the map. The main findings are reported in the protein-protein interaction network.

In this study, the effect of 30 min, 2 hr and 4 hr exposure to microbial stimuli in C. robusta juveniles has been investigated by evaluating changes in the expression levels of immune genes and proteins in the whole-body animal ( Figure 1A ). Also, juveniles were exposed for 4 hr at all selected concentrations (10 μg/ml LPS, 1 and 10 μg/ml Pam2CSK4, 10 and 100 μg/ml zymosan), raised in FSW for 24 hr, and then analyzed for survival and altered gross phenotypes. In all experimental conditions, juveniles were alive and morphologically normal, suggesting that the concentrations of the microbial stimuli did not trigger a strong immune response that could be lethal for the animals ( Figure 1B ).

Previous studies have identified the main molecular actors of the innate immune system of C. robusta, revealing high-rate conservation of PRRs, cytokines and complement system molecules, to name a few. In our work, to widen the spectrum of molecules that could be responsive to the inflammatory stimuli here tested, more counterparts of genes involved in the inflammatory response of vertebrates were included. The genes encoding the tyrosine-protein kinase SYK and Nuclear factor of activated T-cells 5 (NFAT5), respectively co-interactor and transcription factor of PRR pathways, were identified by orthologous gene search and had already been annotated (http://www.aniseed.cnrs.fr). Then, more immune genes were uncovered by homology search analysis of C. robusta genome and proteome ( Table 1 ). Specifically, we searched for molecules containing the CLECT and the Ig domains, since they are present in proteins with immune functions (64–67), and the IRF domain contained in the homonymous transcription factor involved in PRR pathways (68).

The CLECT domain allowed to identify the Dectin-1 CLR family 4 members M (CLEC4M) and F (CLEC4F), and Macrophage mannose receptor 1-like (MR). Ig- or Ig-like domains were found in FAM187A, fibronectin (FN), fibronectin like (FN-like), and tyrosine-protein kinase receptor 3 (TYRO3). TYRO3 is a receptor which is also characterized by FN and Tyrosine kinase catalytic domains and is involved in inflammation resolution (69, 70). FAM187A (Cirobu.g00001161) groups together with the ascidian Botryllus schlosseri Triggering receptor expressed on myeloid cells (TREM)-like 2 gene (TREML2) (Boschl.g00005277) in the “Gene Phylogeny” section of the ANISEED database. In human, the TREM2 pathway is expressed in various tissue macrophages, such as the microglia of the central nervous system, and increasing evidence suggests that TREM2 is involved in neuroinflammatory responses and neurodegenerative diseases like Alzheimer and Parkinson (71, 72) and contributes to mucosal inflammation in the digestive tract during the development of colitis in mice (73). The IRF domain was present in IRF-like gene. Furthermore, search for C. robusta NFAT5 led to the identification of FAM136A, a mitochondrial protein conserved across metazoans (74) whose human orthologue shows a correlation with Meniere’s disease, an inner ear problem with an autoimmune condition (75).

In detail, the effect of LPS, Pam2CSK4 and zymosan has been tested on the activity of the following Ciona factors: i) immune receptors such as TLR1, TLR2, CLEC4M, CLEC4F and MR; ii) Ig-like V-type domain-containing proteins as FAM187A, FN, FN-like and TYRO3; iii) the co-interactor SYK; iv) transcription factor coding genes NF-κB, IRF-like and NFAT5; v) FAM136A; vi) cytokines IL17-1, IL17-2, IL17-3, and their receptor IL17R, TNFα, TGFβ and MIF; vii) defensins like mamA and viii) complement system genes C3 and C3aR. The results obtained for each stimulus are described below. For sake of simplicity in the presentation, FAM136A and IL17R were included in the categories of transcription factors and cytokine signaling, respectively, besides their different roles.

To explore the effect of each stimulus treatment at protein level, the targeted proteomic approach has been conceived for detecting changes in protein amount of the immune molecules here investigated. Indeed, A LC-MRM/MS method was developed for the detection of changes in protein abundance for the immune molecules. We first selected a panel of 1-3 proteotypic peptides for each protein and then recorded from 3 to 6 transitions (a combination of each precursor ion to several fragment ions) for each peptide. A total number of 32 peptides belonging to 16 proteins (listed in the Supplementary Table 1 and in the Figure 7B ) was selected by monitoring 139 transitions displaying a good instrumental response (the same retention time for all transitions and a peak area higher than 1,000). An example of MRM chromatogram was reported for one peptide of the NFAT5 protein ( Supplementary Figure 6 ). After selecting the best instrumental response (see instrumental parameters in Supplementary Table 2 ), the relative quantification was performed by comparing the areas underlying the extracted ion chromatogram (EIC) peaks reflecting the differential expression of selected 16 proteins following the various stimulation versus the control sample. As an example, the EIC peak areas of two proteins, e.g. FAM136A and FN, were reported in a histogram representation ( Supplementary Figure 7 ), displaying a similar trend in the samples. A significant downregulation was observed for the FAM136A and fibronectin protein along the Pam2CSK4 and zymosan treatment (independently from the stimulus kinetics) in comparison to the other samples, included the control.

The EIC peak areas were then uploaded on Perseus software to discover the response of C. robusta juveniles to stimulation with 10 μg/ml LPS, 1 μg/ml Pam2CSK4 and 100 μg/ml zymosan at 30 min, 2 hr and 4 hr of treatment compared to the control samples. The PCA analysis allowed to summarize and visualize the overall protein response to the microbial stimulation within a biplot where the control and treated samples were graphically represented. Among all revealed components, the first two ( Supplementary Table 3 ) resulted in a two-dimensional PCA biplot ( Figure 7A ). PCA reduced the dimensionality of the multivariate data to two principal components explaining almost 90% variance (79.6% for Component 1 and 9.7% for Component 2) with minimal loss of information. The statistical analysis revealed a clear separation between a large cluster that included control, LPS (30 min, 2 hr, 4 hr), Pam2CSK4 (30 min) and zymosan (30 min, 2 hr) treatments, a small cluster consisting of 2 hr Pam2CSK4 and 4 hr zymosan treatments, and the 4 hr Pam2CSK4 treatment that was different from all other analyzed conditions ( Figure 7A ). This finding suggested that the stimulation of C. robusta juveniles with zymosan for 4 hr and Pam2CSK4 for 2 hr, and with Pam2CSK4 for 4hr induced a significant dysregulation of protein abundance as a consequence of microbial treatment response. Even the heatmap representation enabled the visualization of hierarchical clustering where the aforementioned microbial stimuli displayed the greatest response of immune molecules ( Figure 7B ).

A protein-protein interaction network, based on STRING output and then modified and merged in Cytoscape, has been constructed to include all the immune molecules whose gene expression has been investigated in this study ( Figure 8 ). The interactome shows that both receptors TLR1 and TLR2 are connected with the three transcription factors NF-κB, IRF-like, NFAT5, and with the interleukin receptor IL17R. Notably, TLR2 is also connected with cofactor SYK, receptors MR and TYRO3, complement molecule C3, cytokine MIF and the molecule FAM136A. Although the lack of STRING annotations for IL17 gene products, possibly due to difficulties in investigating such proteins in C. robusta, the presence of connections with the receptor, IL17R, could provide clues about the interactions of these interleukins ( Figure 8 ).

As reported for IL17s, lack of annotation information also affects other proteins, such as CLEC4M, CLEC4F, C3aR, mamA and TNFα. However, we found connection of similar domains belonging to other proteins, as in the case of CLEC17A-like that reveals a connection with the protein SYK ( Figure 8 ). The interaction map shows that other main players of TLR signaling are connected, including MyD88, IRAK4, IKK, TAK1, p38, TRAF3 (13, 76, 77) and MAPKs. The genes coding for these proteins were not investigated in this work but should be future objects of similar studies. All the molecules included in the interaction map that were not investigated by transcriptional and proteomics analyses in this study, are listed in the Supplementary Table 4 .

The enrichment analysis has been performed through STRING Enrichment app in Cytoscape to depict the Biological Processes, Molecular Functions and Cellular Components in which these molecules are involved. Some of the results are represented as Split Donut Chart (chosen option in the Network specific settings for STRING Enrichment table) in the protein-protein interaction network, using certain colors from the Enrichment color palette to underline specific outputs. To generate a map easier to understand, we have highlighted the biological processes that are more significant for this study. These comprise TLR signaling pathways (GO:0002224), innate immunity including inflammation and response to external stimulus (GO:0050896, GO:0002682, GO:0050794, GO:0050778), regulation and response to cytokine production (GO:0034097, GO:0001817, GO:0001818, GO:0001819), CTL or CRD domains (SM00034), and leukocyte-mediated immunity and adhesion to endothelial cells (GO:0045785, GO:0050900, GO:0002443) ( Figure 8 ).

To interpret the inflammatory response in ascidian juveniles, it is important to recall some basic immunological concepts, such as the definition of inflammation and which factors/conditions activate it. The term “inflammation” defines the process triggered by innate immune cells when the homeostatic state is altered due to microbial infection or tissue injury (11–13). Marine organisms are continuously exposed to, and challenged by, a multitude of microorganisms (e.g., bacteria, archaea, fungi, viruses, protozoans) inhabiting the surrounding environment (83). These microbes may be beneficial for the host by helping to shape the immune system and influencing developmental and physiological processes (84). The superorganism theory emphasizes how the concept of self and non-self has changed over time by incorporating host microbiota in the definition of self (84–86). During early development, the crosstalk between host and microbes is crucial to shape immune system maturation, that will allow to discriminate between self and non-self, and to establish a homeostatic state with beneficial components (86). These processes are mostly studied within the gastrointestinal tract where the symbiotic interactions mostly take place (87, 88). This equilibrium is broken when pathogenic microbes, or their components (as PAMPs), invade the host epithelial barrier and induce infection. Colonization and invasion by pathogens activate an inflammatory response that primarily involves PRRs (and their signaling pathways), the main players of the innate immune system (13, 68, 89, 90), with the aim to eliminate the infectious agents and to restore homeostasis (11, 12). PRRs have broad specificity and can recognize many PAMPs, which have a common structural motifs or patterns, thus representing a sort of pathogenicity markers (89, 91). In 2009, Vance and coauthors have proposed the “patterns of pathogenesis”, or POP hypothesis, according to which the immune system recognizes pathogens not only by virtue of the presence of PAMPs but also by their pathogenic behaviors (92). These include growth upon host invasion, cytosolic invasion and disruption of the normal functions in the host cell cytoskeleton (92). On these grounds, a microbe can be considered pathogenic or nonpathogenic depending on the site of infection and on the immune state of the host. Hence, POP do not define a pathogen, rather a pathogenic behavior (86, 92). These new concepts may help to better understand the effect of PAMPs on the immune response, and thus on the onset of the inflammation, in different organisms and physiological states.

Based on the POP hypothesis, it is important to consider the physiological conditions at which the immune challenge is encountered. Ciona stage 4 juveniles are immunological naïve and probably they still do not have a stable microbiota. As they start interacting with the surrounding environment by seawater filtration, ascidian juveniles are here exposed to resuspended microbial components (8, 33, 40) and an immune response is observed if microbial components interact with the epithelial barrier. In line with Ciona LPS injection-based studies (14, 25, 93), the resuspension approach used herein may act differently in terms of immune activation patterns on older juveniles, like stage 8 (2^nd ascidian stage) (33).

In this work we found that LPS, the most common inflammatory stimulus used in vertebrates (13, 68, 94, 95), ascidians and other marine invertebrates (96–99), does not significantly alter gene expression in the adopted experimental setup, that are naïve metamorphic stage 4 juveniles. In line with the transcriptional results, targeted proteomics data show that LPS treatment does not elicit protein level variations. The POP hypothesis may help to explain the lack of an immune response to LPS in C. robusta juveniles treated by LPS resuspension. Since these organisms inhabit a habitat rich in LPS-containing Gram-negative bacteria (78) that are continuously filtered by the organism and interact with host mucosal sites, ascidian juveniles may not recognize LPS as a PAMP and/or LPS may not cross the epithelial barrier of their gastrointestinal tract. Instead, the other two microbial components that we used as inflammatory stimuli, the bacterial Pam2CSK4 and the fungal zymosan, are apparently able to interact with the epithelial barrier and then be sensed by host immune surveillance as PAMPs, thus triggering an immune response.

Pam2CSK4, which is not commonly used as a microbial stimulus in vertebrates as well as marine organisms, has been shown to bind the heterodimer TLR1/TLR6 (38) and, in monocytes, to enhance expression and function of Fcγ, a receptor involved in phagocytosis and inflammatory cytokine production (100). In mice, in vitro and in vivo stimulation of macrophages induces the activation of MAP kinase and NK-κB pathways upon TLR2 binding (101). Also, Pam2CSK4 treatment of human platelets in vitro activates TLR2/TLR6 complex and initiates signaling events that stimulate increase of NF-kB protein level and interactions between platelets and endothelial cells (ECs). This event increases inflammatory cytokine production and reduces EC permeability (102). In our experiments, Pam2CSK4 was the most effective inflammatory stimulus in ascidian juveniles, featuring a concentration-dependent influence on both TLR and Dectin-1 pathways. Also, it is worth mentioning that 1 μg/ml is lower than the concentration of Pam2CSK4 used on human cell lines (10 ug/ml) (100, 102), suggesting that ascidian juveniles are highly responsive to this microbial stimulus. The immune response to Pam2CSK4 observed in Ciona juveniles at stage 4 include the modulation of the PRRs expression, TLRs and CLECs, and of the transcription factors just at the lower concentration tested. While expression data show that Pam2CSK4 exposure modulates both TLR and Dectin-1 pathways, the protein-protein interaction network could not confirm it, as long as CLEC4M and CLEC4F receptors are concerned.

The time of activation of each pathway depends on the amount of microbial component that interacts with host mucosal barrier. We may hypothesize a stronger, but vital, immune response induced by 10 μg/ml Pam2CSK4 that can penetrate mucus barrier (due to the higher amount of Pam2CSK4 molecules), interact with epithelial layer and affect gene expression of Ciona complement system components C3 and C3aR, that consequently can activate cell adhesion, chemotaxis and phagocytosis processes in order to fight the invading microbial molecules (103, 104). Moreover, as in human platelets (102), Pam2CSK4 is able to alter transcriptional levels of cytokines (e.g., IL17-3 and TGFβ) in Ciona. The activation of TGFβ following IL17-3 modulation could be explained as a possible resolution of the inflammation induced by this microbial stimulus (105), in agreement with the observation that juveniles after 24 hr treatment are healthy ( Figure 1B ). That 10 μg/ml Pam2CSK4 can induce a strong inflammatory response is corroborated by the evidence of expression changes of other cytokine coding genes like IL17-1 and MIF. In mammals, MIF has pro-inflammatory and immunoregulatory properties, and upregulates TLR4 expression (106). In LPS-injected Ciona adults, a major role of MIF signaling pathway has been described in regulating IL17s and TGFβ expression (93). A similar effect is observed also here in the inflammatory state induced by Pam2CSK4 treatment, where a downregulation of MIF expression is concurrent with the upregulation of IL17s and TGFβ (after 2 and 4 hr treatment with 10 μg/ml Pam2CSK4).

Compared to Pam2CSK4 infection, zymosan modulates a smaller and diverse set of molecules and pathways in Ciona juveniles ( Figure 9 ). As a cell wall preparation of yeast S. cerevisiae, zymosan is a mix of glucans, mannans, mannoproteins and chitin (107). These components are implicated in yeast recognition by innate immune cells, stimulating phagocytosis by macrophages (108) and cytokine production (i.e. TNFα, IL1β, IL-10 and TGFβ) by monocytes, macrophages and dendritic cells (39, 109–111). In vertebrates, zymosan activates both TLR and Dectin-1 pathways (39). Moreover, a cell signaling cascade activated by the binding of β-glucan to Dectin-1 receptor can initiate both Syk-dependent and Syk-independent cascades (68, 112). In mammals, Syk is involved in cytokine transcription upon recruitment by Dectin-1 (113). However, Syk factor is involved in both PRR pathways (68, 114).

In Ciona, the protein-protein interaction network here generated highlights a potential connection among TLR2, SYK and MR molecules. Although activation of Ciona TLRs was induced by zymosan in a heterologous cellular system (17), here we did not observe changes in the expression of the two TLR genes investigated, but we found an alteration of SYK expression that does not permit us to rule out the hypothesis of an involvement of TLR pathway. The high zymosan concentration seems to induce a stronger inflammatory response that affect mainly gene transcription of cofactors involved in both TLR and Dectin-1 pathways, highlighting again an interconnection between the two pathways as observed in juveniles exposed to Pam2CSK4, although we did not observe a direct effect on the expression of the PRRs investigated. The finding that high concentration induces upregulation of the gene coding for TYRO3, a coreceptor involved in the resolution of inflammation, is in line with the upregulation of TGFβ observed after the increase of the expression of the two pro-inflammatory cytokines IL17-3 and MIF. The role of TYRO3 in the negative regulation of cytokine production has been depicted also by GO analysis of the biological process of Ciona protein-protein interaction map.

In this study, we have observed an effect on the transcriptional levels of Ig-domain containing molecules (FAM187A, FN and FN-like) at both zymosan concentrations. Of note, the connection of these molecules with the Dectin-1 and TLR immune pathways was confirmed by the protein-protein interaction network. The immune role of these Ig-domain containing molecules represents an interesting starting point for future investigation in deciphering their role in the inflammatory response to fungal wall components. Moreover, the finding that Ciona FAM187A is phylogenetically related to B. schlosseri TREML2 gene and that homology search analysis of C. robusta genome and proteome revealed the presence of genes coding for SYK (a downstream effector of the TREM2 pathway) and TYRO3, the latter binding in mammals TYRO protein tyrosine kinase binding protein (TYROBP), also known as DNAX-activating protein of 12 kDa (DAP12) whose putative receptor is TREM2 (115), induce to hypothesize that the existence of the TREM2 signaling is an ancient trait whose origin dates back to the common chordate ancestor. It also prompts for a better understanding of the evolution of this pathway in immune function, and in particular in the response to molecules of fungal origin.

Finally, we report that zymosan affect the transcription of the Ciona complement system, suggesting the activation of a phagocytosis process. This has been observed also in mammals, where zymosan is phagocytosed by macrophages with or without opsonization and can activate alternative pathway of complement system (116–118).

Proteomic data reveal an effect on protein level modulation after 2 hr Pam2CSK4 and 4 hr zymosan treatments, and a major effect after 4 hr treatment with Pam2CSK4. Here, the targeted proteomic analysis performed on a small subset of proteins (encoded by the genes analyzed at the transcriptional level) and experimental conditions highlights that i) Pam2CSK4 has a major effect in the immune regulation respect to zymosan, and that ii) the immune response to these two PAMPs differs in the time of activation. As to the latter aspect, low concentration of Pam2CSK4 is sufficient to induce early expression changes (30 min and 2 hr) as suggested by the transcriptional modulation of a higher number of genes at these time points ( Figure 9 ). This evidence corresponds to a significant effect in a general protein modulation at 2 hr and 4 hr of treatment ( Figure 7 ). A similar delay is observed also in the case of zymosan, which induces a significant modulation of protein levels only at 4 hr, compared to the transcriptional response observed at 2 and 4 hr ( Figures 7 , 9 ). These analyses help draw a first consideration concerning the temporal delay observed in the synthesis of proteins with respect to mRNA expression. The initiation of protein translation occurs within minutes after mRNA export into the cytoplasm, thus justifying the lag between transcription and translation. However, we cannot exclude the involvement of mechanisms, such as translational control through RNA binding proteins, that tightly regulate the production of specific proteins, thereby helping to resolve the inflammatory response (119). In mouse dendritic cells treated with LPS, time delay between transcriptional induction and protein level increases was described, with rapid expression of immune response genes (5 hr post LPS treatment) followed by the best quantitative correlation of protein levels to the mRNA levels at 12 hr (120).

In our study, we have added a further tile in the use of the ascidian C. robusta as an experimental system in comparative immunology field. Specifically, we have i) developed Ciona as an in vivo inflammatory model for studying the activation of the immune response to selected microbial stimuli, showing an interconnection between different PRR pathways and indicating the upregulation of cytokines gene expression (IL17-3 and TGFβ) as markers of inflammation, ii) constructed a first protein-protein interaction map that can help to predict potential molecular interactions, and iii) correlated changes observed at transcriptional and translational levels. This new marine invertebrate inflammatory model represents the starting point for future studies, that include either large-scale sequencing or other “-omics” approaches for better defining the cellular pathways or biological processes affected by microbial treatments, but also for investigating host response to PAMPs in different physiological conditions and at different stages of maturation of the immune system. These advancements will contribute to our understanding of the crosstalk between host and microbiota and to test the POP hypothesis. As suggested by Newton and Dixit (2012), it is important to understand how, in a whole organism, innate immune cells exposed to multiple inflammatory stimuli can integrate signaling triggered by different receptors to identify critical components that can be targeted for therapeutic benefit in inflammatory disorder (13). In this framework we believe that, based on the results obtained in this study, this marine model organism could represent a proficient experimental system. Future studies can lead to the use of C. robusta experimental system in translation research or in any kind of approach (i.e., biotechnological or ecotoxicological) where the effect of molecules, either drugs or pollutants, on the activation and regulation of the innate immune system has to be investigated, like in large-scale screening of inflammatory modulators.