US National Research Council guidelines for the care and use of laboratory animals were strictly followed during this research (35). Adult Argopecten purpuratus (n = 123; 60 mm length) were obtained from the aquaculture concession of the Universidad Católica del Norte (UCN) located at Tongoy Bay, Coquimbo, Chile. Scallops were transferred to the Central Laboratory of Marine Cultures on the premises of UCN, acclimatized and kept in 200-L tanks for two weeks prior to each experiment. Animals were constantly maintained with aeration and replacement of filtered water at 16°C. Animals were fed daily with a mixture of microalgae (50% Isochrysis galbana and 50% Nannochloris spp.; 6x10⁶ cells/ml/day). Vibrio splendidus bacterial strain VPAP18 was used in challenge experiments (36). The bacteria were cultivated in Trypticase Soy Broth (Difco) supplemented with 2% NaCl at 18°C for 24 h. Subsequently, the bacterial suspension was inactivated for 2 h at 90°C and centrifuged at 12,000 × g for 10 min. The obtained pellet was washed twice with 0.22 µm microfiltered sterile seawater, and bacterial solution concentration was adjusted to 1 ×10⁸ cfu/ml. V. splendidus VPAP18 was heat inactivated and washed to eliminate microbe-microbe interaction between V. splendidus VPAP18 and scallop microbiota.

Two independent scallop immune challenges were performed, both including the following experimental conditions: (i) Vibrio-injected scallops (5×10⁶ cells in 50 µl), (ii) seawater-injected (SW) scallops (50 µl) and (iii) non-stimulated scallops. Samples were obtained after 12, 24, 48, 72 and 168 h. For the RNAi experiment [RNAi, ribonucleic acid (RNA) interference], experimental conditions considered scallops injected with 20 µg in 100 µl of: (i) ApBD1-dsRNA (dsRNA, double-stranded ribonucleic acid), (ii) ApLBP/BPI1-dsRNA and (iii) GFP-dsRNA (green fluorescent protein). Samples were obtained after 48 h. Scallop sample size for each experiment is shown in Figure S1. All animals were injected with sterile syringes (25G × ⅝) into the adductor muscle. To obtain hemocytes, 2 ml of hemolymph per individual were extracted and centrifuged at 800 × g for 10 min at 4°C to separate cells from plasma. Then, cells were lysed in Trizol^® reagent (ThermoFisher) at 4°C and stored at -80°C until total RNA purification. To obtain hemolymph bacteria, 2 ml of hemolymph were deep-frozen in liquid nitrogen and stored at -80°C for subsequent extraction of genomic DNA (gDNA).

Total hemocyte RNA was extracted using Trizol^® reagent (ThermoFisher) according to manufacturer’s instructions. Genomic DNA was removed with DNase I (Ambion). The integrity of RNA was verified on agarose gels, and purity and concentration were determined on an Epoch™ microplate spectrophotometer (BioTek). The cDNA was synthesized by reverse transcription from 1 µg of total RNA, using the RevertAid First Strand cDNA Synthesis Kit (ThermoFisher), according to manufacturer’s instructions. RT-qPCR (quantitative polymerase chain reaction prior to reverse transcription) assays were performed in triplicate on a Stratagene Mx3005p Real Time PCR System thermocycler (Agilent Technologies), using the specific gene primers (Table S1). Amplification was performed in a final reaction volume of 20 µl composed of: 10 µl of Takyon ™ Rox SYBR^® MasterMix dTTP Blue (Eurogentec), 0.6 µl of each primer (10 mmol L^-1) and 2 µl of cDNA (1:5). The amplification program consisted of an initial denaturation at 3 min at 95°C, followed by 40 cycles of 15 s at 95°C and 30 s at 60°C. Through serial dilutions of cDNA, the efficiencies of RT-qPCR were verified in the range between 95%–110% (E = 10 ^(–1/slope)). The relative expression of the immune response genes was calculated by the Pfaffl method (37). Differences in gene expression with the control groups (SW) were verified with two-way ANOVA test and Tukey´s posterior test (P < 0.05).

Genomic DNA (gDNA) present in A. purpuratus hemolymph was extracted with the Wizard Genomic DNA purification kit (Promega) according to manufacturer’s instructions. The integrity, concentration and purity of the gDNA was verified by 1% agarose gel, Qubit 3.0 (Life Technologies) and Epoch™ microplate spectrophotometer (BioTek), respectively. The 16S rDNA gene of bacterial communities was amplified and sequenced, targeting the variable regions V3-V4 (341F: 5’-CCTACGGGNGGCWGCAG-3’; 805R: 5’-159 GACTACHVGGGTATCTAATCC-3’) and using ~12.5 ng of gDNA from 5 biological replicates from each experimental condition: Vibrio-injected and SW-injected scallops at 48 h and 168 h, and non-stimulated scallops at 0 h. Paired-end multiplex sequencing (2×300 bp read length) was performed from individual samples by Macrogen Inc. on a MiSeq system (Illumina^®) using the MiSeq Reagent Kit v3 according to manufacturer’s instruction. Raw sequence data are available in the SRA database BioProject PRJNA639911.

Raw reads were processed using QIIME2 version 2019.1 (http://qiime2.org) and developed based on standards described by (38, 39) for microbiota community evaluation. Demultiplexed paired-end reads were imported as artifacts and denoised using DADA2 (40). In this step of the analysis, quality control as well as phiX reads (commonly present in marker gene Illumina sequence data) and chimera sequence filtering were applied to ensure the retainment of only high-quality reads. Amplicon sequence variants (ASVs), a higher resolution analog to operative taxonomic units (OTUs) (41), were obtained and further processed for taxonomic assignment and diversity analysis. Taxonomy was assigned based on the Green Genes database (gg-13_8 99%) (42, 43). Features assigned to the class Chloroplast (Phylum Cyanobacteria) and Family Mitochondria (Phylum Proteobacteria) were filtered out due to their contaminant character. Samples were rarefied to the maximum depth of the sample with less sequencing depth, and rarefaction curves were plotted. Alpha diversity was evaluated using the Shannon and Simpson Diversity Index, the richness index was evaluated by Chao1 y Faith PD (Faith’s Phylogenetic Diversity) and beta diversity was evaluated by assessing weighted and unweighted UniFrac distances (44, 45). Differences in alpha diversity between groups were analyzed using the Kruskal-Wallis (pairwise) test in QIIME2. A principal coordinate analysis was performed to visualize phylogenetic beta diversity and differences between experimental groups based on the distances were assessed by PERMANOVA considering 999 permutations in QIIME2. Significant differences between relative abundances among groups were evaluated using two-way ANOVA and Tukey-Kramer tests on STAMP (Statistical Analysis of Metagenomic Profiles) (46), significant differences were only considered when P ≤ 0.05 and a q-value < 0.3. Bacterial groups showing abundances greater than 1% were considered for analysis.

The 16S rDNA gene of four bacterial strains belonging to the Gammaproteobacteria, Epsilonproteobacteria, Betaproteobacteria, and Firmicutes taxa (Table S2), plus the 16S rDNA gene from Vibrio splendidus VPAP18 were amplified by PCR using the 16S rDNA universal primers 16SEUBAC (Table S1). The 16S rDNA fragments were cloned into the pCR2.1 vector included in the TOPO TA kit following manufacturer’s instructions. Specific cloning of each 16S rDNA bacterial strain was verified by plasmid sequencing. Then, seven-serial dilutions of each plasmid, ranging from 10¹⁰ to 10³ plasmid copies/µl were prepared to construct standard curves and to obtain absolute quantifications of the 16S rDNA gene copy number of each bacterial group by qPCR. Plasmid copies were calculated by the following formula: Number of copies/µl = 6.022×10²³ (molecules/mole) × DNA concentrations (g/µl)/Number of bases pairs × 660 Da, were 6.022×10²³ (molecules/mole) is Avogadro´s number and 660 Da is the average weight for a single base pair.

Specific primers for the 16S rDNA of each bacterial group were previously designed and validated (Table S1). qPCR assays were performed in triplicate on a Stratagene Mx3005p Real Time PCR System thermocycler (Agilent Technologies), using a final reaction volume of 20 µl composed of: 10 µl of Takyon ™ Rox SYBR^® MasterMix dTTP Blue (Eurogentec), 0.6 µl of each primer (10mmol * L-1) and 2 µl of plasmid DNA. The amplification program consisted of an initial denaturation at 10 min at 95°C, followed by 40 cycles of 30 s at 95°C and 1 min at 60°C and dissociation curve detection. Standard curves were obtained by plotting the threshold cycle (C_q) on the Y-axis and the natural log of concentration (copies/µl) on the X-axis. We considered the following criteria: PCR efficiency (95-110%) and correlation coefficient R² (0.99), which validates the linear relation between the threshold cycle and the natural log of concentration (copies/µl). Next, 50 ng of gDNA from scallop hemolymph was analyzed by qPCR using the same settings. The copy numbers of 16S rDNA gene of each bacterial group in the sample were obtained by relating the C_q value to the respective standard curve. Eubacteria universal primers were used to obtain the total copies of bacterial 16S rDNA in each sample and it was considered as the 100% to calculate the relative abundance of each bacterial group. Differences in the relative abundance of the bacterial groups between treatments were verified with ANOVA two-way and the Sidak´s posterior test (P <0.05).

Sequence specific primers were designed to amplify fragments of ApBD1 (401 bp) and ApLBP/BPI1 (543 bp) from hemocyte cDNA as template and to add the T7 promoter sequence TAATACGACTCACTATAGGG by PCR (Table S1). A plasmid containing the sequence for the green fluorescent protein (GFP) gene (GenBank no. HM640279) which is not present in A. purpuratus was amplified with specific primers and included as a dsRNA control of specific gene silencing. The PCR products were verified on agarose gel and then purified using E.Z.N.A.^® Gel Extraction Kit (Omega Biotek). PCR products were confirmed by sequencing. T7 RiboMAX^® Express RNAi System was used to synthesize the dsRNA in accordance with manufacturer’s instructions. The quantity and quality of the RNAi were determined using a Qubit 3.0 (Life Technologies) and agarose gel electrophoresis, respectively.

Thiosulfate-citrate-bile salts-sucrose (TCBS) selective agar plates for Vibrio spp. (Difco) were prepared in sterile sea water (50%): distilled water 1: 1 following the manufacturer’s instructions. Aliquots of 0.1 ml and their dilutions from scallop hemolymph from the RNA interference experiment were plated separately in the selective agar plates and incubated at 25°C. Three independent aliquots from each scallop hemolymph were plated (5 scallops per experimental condition). After 24 h of plate incubation, visible colonies were counted. Then, average colony forming units (c.f.u.) number with the corresponding standard deviation were calculated for each sample. Differences in the c.f.u. number of Vibrio spp. between treatments were verified with Welch’s t-test (P <0.05).

The gDNA extracted from bacteria present in the hemolymph of i) heat-inactivated Vibrio-injected scallops, ii) SW-injected scallops, and iii) non-stimulated control scallops was analyzed by 16S rDNA deep amplicon sequencing. As a result, a total of 2,546 million bases were sequenced complying 4,229,314 joined sequences, with an average length of 300 bp, and 92.7% presented a quality score above Q20 whereas 84.6% were above Q30. After filtering, denoising and chimera removal, 3,159,967 high quality reads were obtained, ranging per individual between 108,842 and 150,048 reads, representing all the phylotypes present in the A. purpuratus hemolymph bacterial microbiota (Table S3). After filtering the rare OTUs, 478 taxonomic groups were identified in the A. purpuratus hemolymph. Analysis of rarefaction curves indicated that the sequencing depth was sufficient to identify all unique OTUs among all experimental groups (Figure S2).

We first focused on the analysis of bacterial groups present in non-stimulated scallops to determine how variable was the scallop microbiota between control individuals. Common bacterial groups among five non-stimulated scallops were determined by Venn diagrams (Figure 1). From the 37 phyla identified, 35% (13 phyla) were common to the 5 individuals (Figure 1A). Within these 13 phyla, a 54% were predominant groups, displaying relative abundances around or higher than 1% of the total (Figure 1B). Those phyla were identified as Proteobacteria (71.04%), Spirochaetes (12.58%), Bacteroidetes (9.86%), Gracilibacteria (1.95%), and Firmicutes (0.92). Venn diagram analysis indicated that the number of phyla shared among the five individuals was higher than the specific phyla number found for each individual. Accordingly, two scallops lacked specific phyla, two scallops presented two specific phyla and one scallop (namely N3) displayed six particular phyla (Figure 1A).

From the 90 bacterial classes identified, 23% (21 classes) corresponded to common groups among all scallops (Figure 1C). Within these classes, 43% were predominant, displaying relative abundances higher than 1% of the total, such as: Gammaproteobacteria (18.16%), Brachyspirae (10.32%), Alphaproteobacteria (6.54%), Deltaproteobacteria (6.52%), Flavoproteobacteria (6.36%), Epsilonproteobacteria (4.28%), and Betaproteobacteria (1.16%) (Figure 1D). Venn diagram analysis also indicated that the number of classes shared among the five individuals was higher than the specific classes found for each individual. Therefore, four scallops presented between one to four distinct classes, and N3 scallop displayed a higher number of 16 particular classes (Figure 1C).

At the genus level, from the 329 bacterial genera identified in from the 5 non-stimulated scallops, 7.5% (25 genera) corresponded to common groups (Figure 1E). From these common genera, 88% had relative abundances higher than 1%. The most relative abundant were Winogradskyella (5.18%), Halomonas (5.17%), Pseudoalteromonas (4.87%), Uruburuella (4.32%), Glaciecola (4.31%), Streptococcus (3.19%), and Pseudomonas (1.71%) (Figure 1F). Venn diagram analysis indicated again that four scallops presented more common than particular genera, presenting between five to twelve specific genera. At this level, N3 scallop differed from the others in terms of variable groups, displaying 39 particular genera (Figure 1E). Overall, results suggested that control scallops share more common taxa compared to specific taxa, except for one individual that harbored an elevated number of particular groups (Figures 1A, C, E). In agreement, N3 individual displayed higher species richness compared to other control scallops, determined by Chao1 and Faith PD indices (Figures S3C and D). Similarly, the Unweighted UniFrac distance analysis confirmed that the bacterial communities from N3 individual differed from the rest of the non-stimulated scallops (Figure S4A).

A previous study identified bacterial shifts in structure and composition from whole scallop microbiota at 48 h after immune challenge (34). To determine whether imbalances in bacterial communities are sustained or restored over time after the immune response, we analyzed the bacterial composition present in the scallop hemolymph at 48 h and 168 h after challenge them. Then, the bacterial relative abundances at class and genus taxonomic levels obtained by deep amplicon sequencing analysis were compared between Vibrio- and SW-injected scallops (Figure 2). We found that the most noticeable changes on bacterial abundances between experimental conditions were observed at 48 h post challenge, at both class (Figure 2A) and genus (Figure 2B) levels. As expected, the relative abundance of Gammaproteobacteria was significantly higher in Vibrio-injected scallops (46%) compared to seawater scallops (31.2%) (P<0.05). Concomitantly, the classes Delta- and Epsilonproteobacteria displayed a lower abundance in Vibrio-injected (9.2% and 5.4%, respectively) when compared to SW-injected scallops (20.6% and 15.6%, respectively) (P<0.05). The class Mollicutes, which were found at 0.023% in SW-injected scallops, displayed an 10.2% of relative abundance in Vibrio-injected scallops at 48 h after immune challenge. At 168 h after immune challenge, relative abundances of Gamma-, Delta-, and Epsilonproteobacteria in Vibrio-injected scallops were statistically comparable to those from SW-injected scallops, suggesting a restoration of those communities after that time (Figure 2A). Betaproteobacteria showed no significant variation at 48 h among injected groups, though a slightly increase of this class was observed in the Vibrio-injected group (8%) compared to the SW-injected group (3.1%) at 168 h.

When we compared the relative abundances of bacteria at genera level, we found that Mycoplasma (from Mollicutes class) displayed the principal shift, although slightly non-significant (P = 0.065). Specifically, Mycoplasma increased from 0.2% in SW-injected scallops up to 26% in Vibrio-injected scallops at 48 h. As expected, Vibrio displayed significant increase in its relative abundance in Vibrio-injected scallops (5.4%) when compared to SW-injected scallops at 48 h (1.6%) (P<0.05). Remarkably, Vibrio genus abundance was only a fraction (11%) of total increase of Gammaproteobacteria abundance in Vibrio-injected scallops, revealing that the higher abundance found in this group was not only due to the immunostimulation approach. Other bacterial genera displayed shifts on their relative abundances among groups at 48 h, but still non-significant. For instance, Streptococcus and Halomonas were found in lower abundance in Vibrio-injected scallops (0.6% and 2.2%, respectively) when compared to SW-injected scallops at 48 h (2.4% and 6.2%, respectively). Pseudomonas displayed the opposite pattern at the same time point, showing a relative abundance of 2.6% in Vibrio-injected and 0.4% SW-injected scallops (Figure 2B). After 168 h of challenge, bacterial abundances at the genus level in Vibrio-injected scallops were also statistically similar to those from SW-injected scallops, although a higher variability in bacterial abundances were observed at this taxonomic level (Figure 2B). For instance, Mycoplasma relative abundance decreased in Vibrio-injected scallops after 168 h (1.2%) when compared to 48 h (26%) but it was still higher when compared to the SW-injected group which did not show any presence of this genera. Analogous variation in their relative abundances were found for Streptococcus, Halomonas, Pseudomonas, and Vibrio genera at 168 h (Figure 2B). Overall, our data revealed the existence of dynamic shifts of bacterial communities in the scallop hemolymph along immune response progression. Certain bacterial group shift abundances were sustained over time, while a greater proportion of bacterial groups were restored after 168 h.

Alpha diversity was evaluated using the Simpson and Shannon Diversity indices, while the species richness was evaluated by Chao1 y Faith PD indices (Figure S3). The Simpson index determined that Vibrio-injected scallops at 48 h presented more diverse bacterial communities than the other experimental conditions, specifically compared to non-stimulated scallops and SW-injected scallops at 168 h (P<0.05) (Figure S3). Shannon index presented a similar trend than the Simpson index but differences were non-significant. Similarly, the Chao1 and Faith PD richness indices suggested that Vibrio-injected scallops at 48 h and 168 h, respectively, tended to display a higher species richness compared to other injected scallops, although the differences were non-significant (P>0.05) (Figure S3). PCoA based on weighted (quantitative) and unweighted (qualitative) UniFrac distances suggested not distinct microbiota community among experimental conditions (Figure S4), with the exception of the weighted UniFrac distance between bacterial communities from SW-injected scallops at 168 h (SW168) and Vibrio-injected scallops at the same time (VS168) that was significant (PERMANOVA pseudo-F=2.43, P=0.023). The variance explained in the PCoA using the unweighted UniFrac was 19.08% (PC1 10.94% and PC2 8.14%) (Figure S4A) and the explained variance obtained using the weighted UniFrac was 54.19% (PC1 37.29% and PC2 16.91%) (Figure S4B).

We next investigated the expression of antimicrobial effectors that could be involved in host-microbiota interaction during scallop immune response. For that, we focused on the transcription patterns of four antimicrobial effectors expressed by hemocytes from the same analyzed scallops. The relative expression of the antimicrobial peptide big defensin ApBD1, the antimicrobial proteins ApLBP/BPI1 and ApLBP/BPI2, and the hydrolytic enzyme lysozyme ApGlys were assessed in all experimental conditions at 12, 24, 48, 72, and 168 h by RT-qPCR (Figure 3). Results showed that all antimicrobials were first overexpressed in Vibrio-injected scallops within the first 24 h compared to the sea water-injected control group. Thus, ApLBP/BPI1 was overexpressed 15- and 5.7- folds at 15 and 24 h, respectively whereas ApLBP/BPI2 was slightly overexpressed 2.3- fold at 15 h. ApBD1 was overexpressed 9.1- and 6.3- folds s at 24 and 72 h, respectively. The ApGlys gene was overexpressed 17- fold at 24 h (Figure 3). No significant overexpression was detected for the four immune genes after 72 h post injection. Also, all genes were expressed at basal level by hemocytes of non-stimulated scallops (Figure S5). Overall, the significant changes observed in the scallop bacterial communities occurred after the activation of the immune response and the expression of antimicrobial effectors.

In an attempt to assess the changes of target bacterial groups in scallop hemolymph during the expression of immune effectors, we determined their relative abundances by qPCR and absolute quantification of 16S rDNA. To do this, a second immune challenge was performed and the significant overexpression of antimicrobial effectors in scallop hemocytes was confirmed at 24 h post challenge (Figure S6). Then, we examined the relative abundances of Gamma-, Epsilon-, and Betaproteobacteria, Firmicutes, as well as the specific quantification of Vibrio spp. from scallop hemolymph at 24 and 48 h after challenge (Figure 4). Results showed that significant shifts in bacterial relative abundances could be detected in scallop hemolymph as early as 24 h after challenge (Figure 4A). Furthermore, the relative abundances of Gamma-, Epsilon- and Betaproteobacteria, and Firmicutes were similar in Vibrio-injected scallops compared to SW-injected scallops at both 24 h and 48 h (Figures 4A, B). We observed a significant increase in the Gammaproteobacteria (P<0.001) and a significant decrease in undetermined bacterial groups (others) (P<0.05) in Vibrio-injected scallops at both time points. Still, Gammaproteobacteria exhibited a higher relative abundance at 48 h (65.8%) compared to 24 h (59%) after Vibrio injection. Betaproteobacteria and Firmicutes relative abundances did not vary between groups or time points. Epsilonproteobacteria abundance decreased in Vibrio-injected scallops at 24 h and 48 h, still those changes were non-significant (P>0.05). When comparing data obtained by absolute quantification of 16S rDNA and the deep amplicon sequencing from samples at 48 h, results showed a consistent shift trend for Epsilon-, Beta-, and Gammaproteobacteria but not for Firmicutes.

We further determined the number of total copies of specific bacterial 16S rDNA gene of the genus Vibrio to explain the contribution of the Vibrio injection to the Gammaproteobacteria abundance observed in scallop hemolymph (Figures 4C, D). As determined by the deep amplicon sequencing analysis, the increase Vibrio spp. was only a fraction of total Gammaproteobacteria relative abundance. Indeed, in terms of relative abundance, Vibrio spp. correspond to the 50% and 16% of total Gammaproteobacteria at 24 and 48 h respectively. Thus, other bacteria belonging to the Gammaproteobacteria class increased their abundances in Vibrio-injected scallops.

Since we detected significant shifts of bacterial groups in scallop hemolymph together with the overexpression of antimicrobial effectors, we focused on the role of ApBD1 and ApLBP/BPI1 in shaping the scallop microbiota. Thus, we performed a RNA interference silencing approach by injecting sequence-specific dsRNA in non-immune stimulated scallops. Subsequently, we quantified the 16S rDNA of target bacterial groups present in scallop hemolymph by qPCR (Figure 5). Results showed that transcript expressions of ApBD1 and ApLBP/BPI1 in hemocytes from scallops injected with ApBD1-dsRNA and ApLBP/BPI1-dsRNA were significant suppressed by ~94% and ~45%, respectively, compared to scallops injected with GFP-dsRNA as control after 48 h (Figure 5A). In addition, the expressions of ApBD1 and ApLBP/BPI1 in scallops injected with GFP-dsRNA were consistent with the basal expression levels observed in hemocytes from non-stimulated scallops (Figure S6). Then, we assessed the relative abundances of Gamma-, Epsilon-, and Betaproteobacteria, and Firmicutes from every experimental group. Particularly, Gammaproteobacteria significantly increased when the expression of antimicrobial effectors was suppressed compared to the control group (Figure 5B). We observed the highest relative abundance of Gammaproteobacteria in ApLBP/BPI1-dsRNA injected scallops (81.8%) followed by ApBD1-dsRNA (69.3%) and GFP-dsRNA (35.9%) injected scallops. Concomitantly, the increase in Gammaproteobacteria was associated with a significant decrease in undetermined bacterial groups (named as others). Specifically, ApLBP/BPI1-dsRNA injected scallops displayed the lowest abundance of undetermined groups (9.3%) followed by ApBD1-dsRNA (18.4%) and GFP-dsRNA (35.9%) injected scallops. Epsilonproteobacteria also displayed a significant decrease in ApLBP/BPI1-dsRNA injected (5.7%) compared to control scallops (28.7%). Firmicutes and Betaproteobacteria did not show any significant shift in their relative abundances among experimental conditions. Moreover, relative abundances of bacterial groups from scallops injected with GFP-dsRNA were non-significantly different from SW-injected scallops.

In parallel, we detected a significant increase of the number of total copies of specific bacterial 16S rDNA gene from the genus Vibrio in scallops silenced for ApBD1 or ApLBP/BPI1 (Figure 5C). In terms of the relative abundance, Vibrio spp. corresponded to 29% and 22% of total bacteria in ApLBP/BPI1- and ApBD1-dsRNA injected scallops, respectively, while Vibrio spp. only represented 2% of total bacteria in GFP-dsRNA injected scallops. Finally, to confirm that the increase in Vibrio was due to cultivable Vibrio spp., hemolymph from all dsRNA-injected scallops were plated onto TCBS selective media. Results showed a significant increased number of Vibrio spp. c.f.u. in ApLBP/BPI1- and ApBD1-dsRNA injected scallops compared to the GFP-control group (P<0.05) (Figure 5D).