Effect of Streptomyces probiotics on the gut microbiota of Litopenaeus vannamei challenged with Vibrio parahaemolyticus

Abstract This study assessed the intestinal microbiota of juveniles of the White shrimp Litopenaus vannamei, whose feed was enriched with three probiotic formulations: Streptomyces sp. RL8 (RL8); a mix of Lactobacillus graminis and Streptomyces spp. RL8 and N7 (Lac‐Strep); and a mix of Bacillus spp. and Streptomyces spp. RL8 and N7 (Bac‐Strep). The analysis was performed by sequencing the V3 region of the 16S rRNA gene of treated animals and the control group before and after Vibrio parahaemolyticus challenge. After challenge, the highest Shannon diversity indexes corresponded to RL8 and Bac‐Strep (3.94 ± 0.11 and 3.39 ± 0.3, respectively) and the lowest to the control group (2.58 ± 0.26). The most abundant phyla before and after challenge were Proteobacteria, Actinobacteria, and Bacteroidetes. The principal component analysis and Statistical Analysis of Metagenomic Profiles (STAMP) showed that the gut microbiota of the groups RL8 and Bac‐Strep after challenge was different from the other experimental groups, which was characterized by a higher bacterial diversity, as well as a significant stimulation of the Bacteriovorax population and other antimicrobial producing genera that protected shrimp from infection.

. Therefore, probiotics have shown to be a promising and environmentally friendly alternative for disease prevention, especially in crustacean aquaculture of high commercial value (Lobo et al., 2014).
Several studies have indicated that probiotics could contribute to enzymatic digestion, inhibit pathogenic microorganisms, promote growth factors, and increase the immune response of aquatic organisms (Krummenauer et al., 2014). Consequently, new beneficial microorganisms that could be used as probiotics in aquaculture are constantly explored (Lazado, Caipang, & Estante, 2015). Marine actinomycetes are among those promising candidates by virtue of their ability to produce a wide variety of antibiotics and extracellular enzymes (Barka et al., 2016;Prakash et al., 2013). In fact, some studies have shown that marine strains of the genus Streptomyces increased growth, survival and resistance to disease in the shrimp Penaeus monodon (Augustine, Jacob, & Philip, 2016;Das, Lyla, & Ajmal Khan, 2006;Das, Ward, & Burke, 2010).
Previous experiments have also shown the in vitro probiotic effect of Streptomyces spp. isolated from marine sediments of Cuba (García-Bernal et al., 2015), as well as the increased resistance to infection and survival of L. vannamei juveniles treated with those strains and challenged with V. parahaemolyticus CAIM 170 (García-Bernal, Medina-Marrero, Campa-Córdova, & Mazón-Suástegui, 2017). Therefore, the objective of this research was to determine the effect of Streptomyces strains on the intestinal bacterial community in juveniles of the White shrimp L. vannamei, as part of a previous study revealing the probiotic effect of Streptomyces strains alone or combined with Bacillus and Lactobacillus (García-Bernal, .

| Test organisms
The Streptomyces spp. RL8 and N7 isolated from marine sediments of Cuba (García-Bernal et al., 2015), a Bacillus (Bac) mixture com-   (Abasolo-Pacheco et al., 2016;Luis-Villaseñor et al., 2011), whereas Streptomyces strains were added at a ratio of 1 × 10 8 CFU/g of feed, which is the mean of the dose range used for most of the probiotics (Newaj-Fyzul & Austin, 2015). Treated shrimp were fed ad libitum three times a day during 30 days with the probiotic-sprayed commercial diet, whereas the control group was fed with the commercial diet sprayed with sterile seawater (García-Bernal, García-Bernal et al., 2018). The bacterial load in the food was confirmed by plate count; particulate matter was daily removed by siphon during the probiotic feeding period followed by the addition of the same amount of discarded water (25%), as reported in the preceding paper ( García-Bernal, . No water exchange was performed during challenge, and dead animals were regularly removed from tanks throughout the daylight hours. Intestine samples for metagenomic studies were taken after the probiotic treatment (day 30) and at the end of V. parahaemolyticus CAIM 170 challenge (day 35, 5 days postchallenge) ( Figure 1). Samples from probiotic-fed shrimps were taken from the same amount of surviving animals as in the control (30% survival).

| DNA extraction and sequencing
The DNA was extracted using the method of Sambrook, Fritsch, and Maniatis (1989). The complete intestinal tissue was homogenized in a lysis buffer containing Tris-EDTA-sodium dodecyl sulfate (SDS) (100 mmol/L); NaCl, 50 mmol/L; Tris (pH 8), 100 mmol/L; EDTA (pH 8); SDS (1%); and 100 µl of lysozyme (50 mg/ml; Sigma) at 37°C for one hour. Once homogenized, the tissue was incubated overnight at 65ºC with 20 µl of Proteinase K (20 mg/ml; Sigma), followed by the addition of 200 µl of 6 mol/L of NaCl, incubation on ice (20 min) and centrifugation (13,000 g, 4ºC, 10 min). The DNA was precipitated from the supernatant with absolute ethanol, left to settle overnight at 4°C and collected by centrifugation (8,000 g, 4ºC, 5 min). Extracted DNA was washed with 70% ethanol, dried, and resuspended in 50 μl of molecular grade water. DNA purity and concentration was determined with a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific™). Finally, DNA samples were stored until sequencing in the Laboratory for Microbial Genomics (Centro de Investigación en Alimentación y Desarrollo). To determine the microbiota present in the samples, the 16S rDNA variable region V3 was amplified by PCR (338F and 533R with Illumina adaptors) and barcoded following the protocol recommended by Illumina. Three 16S-amplification replicas were performed to each DNA sample. Samples were quantified through Qubit Fluorometer and mixed together as an equimolar pool before sequencing in an Illumina Miniseq machine using standard conditions (300 cycles, 2 × 150).

| Statistical analysis
Sequencing reads of the 16S rRNA gene were processed with QIIME software (Caporaso et al., 2010). Read preparation was performed with the pair-end_cleaner v0.9.7 (https ://github.com/Genom icaMi crob/pair-end_cleaner) program. The minimum sequence length was 170 bp, and singletons were discarded. Chimeric sequences were detected and eliminated with the program chimera_detector version 1.3.3 (https ://github.com/Genom icaMi crob/). Metagenomic analysis was performed with the Microbiomal Helper (Comeau, Douglas, & Langille, 2017) program, using QIIME1 (Caporaso et al., 2010). The free chimera sequences were grouped in operational taxonomic units (OTUs) (97% identity). To assign OTUs, the script "pick_open_reference_otus.py" was used. The taxonomic data for each OTU were obtained from the reference bases using the script "assign_taxonomy.py." Low confidence (0.1%) OTUs were removed with the script "remove_low_confidence_otus.py." Rarefaction was performed with the script "single_rarefaction.py" utilizing the read count obtained as the lowest limit. Postrarefaction data allowed to calculate the relative abundance of the IM composition.
Alpha diversity was calculated through richness (Chao-1) estimations as well as Shannon and Simpson indexes, using the script "alfa_ diversity.py." Comparisons among estimations were calculated with the software Past (Hammer, Harper, & Ryan, 2001). Diversity among groups (beta diversity) was estimated with weighted UniFrac implemented in the "beta_diversity.py" script and visualized graphically in a principal component analysis (PCA) plotted with EMPeror (Vazquez-Baeza, Pirrung, Gonzalez, & Knight, 2013). Significant differences of the beta diversity estimates among and within groups were assessed with the nonparametric tests ANOSIM and PERMANOVA, using the average rank dissimilarity and the sum of squares of the distances between diversities, respectively, as implemented in the script "com-pare_categories.py" with 999 permutations and p < .05. The statistical differences of beta diversity were observed and plotted with the Statistical Analysis of Metagenomic Profiles (STAMP) (Parks, Tyson, Hugenholtz, & Beiko, 2014), using the Welch's test with correction of Benjamin Honchberg FDR (q value < 0.05).

| Obtained sequences
To determine the bacterial microbiota composition of the gastrointestinal tract of shrimp fed with different Streptomyces-based probiotics, a total of 25,000 valid sequences with an average read length of 170 bp were obtained by sequencing the V3 region of the 16S rRNA gene using Illumina platform. In general, bacterial OTUs from these sequences were assigned to 14 phyla, 49 families and 46 genera.

| Richness and diversity analysis
Bacterial richness and diversity were estimated by Simpson, Shannon, and Chao-1 indexes. Bacterial diversity in the groups RL8_ACH and Bac-Strep_ACH (after challenge) with Shannon indexes of 3.94 ± 0.11 and 3.39 ± 0.3, respectively, was greater than the control groups.
The rarefaction curve of the experimental groups reached the saturation plateau (Appendix Figure A1), which indicated that sampling captured the most representative bacterial richness present in shrimp intestine. In general, microbiota diversity in the majority of the experimental groups was higher after than before challenge with V. parahaemolyticus.

| Composition of bacterial microbiota
The composition and abundance of the bacterial community of different experimental groups is shown in Figure 3 (Figure 3a,b).

| D ISCUSS I ON
The animals' intestine is a vital organ for food storage, nutrient digestion, and absorption besides playing an important role in immunity (Ringø et al., 2016;Tzuc et al., 2014). Several intestinal functions are achieved through bacterial metabolism, which may also benefit the host by improving the immune response, nutrient absorption, and homeostasis maintenance (Hooper & Macpherson, 2010).
Consequently, modulation of the IM, through optimization of diet formulation or supplementation with pre-and probiotics, is important to improve the general physiological development and increase the productivity and economic revenues during shrimp farming.
Bacterial diversity was estimated with the Shannon index whose However, this was not the case in this study where both treated and control groups had similar dominant classes (Figure 4a,b).
One of the most significant results that derived from this study was the detection of members of the family Bacteriovoracaceae, mostly of the genus Bacteriovorax from the order Bdellovibrionales in the groups treated with RL8 and Bac-Strep after V. parahaemolyticus challenge. Bacteriovorax is a small mobile predator bacteria that invades the periplasmic space of certain gram-negative bacteria, including Vibrio species, altering the cellular wall of its prey, consuming the cytoplasmic content, and lysing the cell until it releases  Crossman et al., 2013). These predator bacteria are often isolated from estuarine seawater  where they can attack and lyse a great variety of gram-negative bacteria (Chen et al., 2012).
As a consequence of its capacity to limit the proliferation of bacterial   & Zhang, 2006;Frans et al., 2011), and others, such as V. coralliilyticus and V. shiloi, are coral pathogens, most species from this genus are benign (Thompson & Swings, 2006;Thompson & Polz, 2006 The change in the microbiota composition of the control group with respect to probiotic-fed groups cannot be attributed exclusively to V. parahaemolyticus, even though they were under the same experimental conditions after challenge with this agent. Instead, some indirect and stochastic effects which did not arise from the direct shrimp V. parahaemolyticus interaction may account for this variation (Zaneveld, McMinds, & Vega Thurber, 2017). By contrast, groups fed with the probiotics Rl8 and Bac-Strept gained a more diverse and resilient microbial community composition that help them cope with the infection and any other detrimental stochastic effect.
Microbial colonization and survival in the intestines of targeted organisms are usually claimed as crucial prerequisites for potential probiotics (Lakshmi, Viswanath, & Sai Gopal, 2013). However, these conditions do not seem to be strictly required for shellfish organisms which can benefit from their continuous interaction with beneficial microorganisms thriving in the water and sediment (Li et al., 2018).

| CON CLUS ION
This study revealed that Proteobacteria, Actinobacteria, and Ormart, and Diana Fischer for editorial services.

CO N FLI C T O F I NTE R E S T S
None declared.