Bioaugmentation of biofloc system with enzymatic bacterial strains for high health and production performance of Penaeus indicus

The beneficial effects of two probiotic bacterial strains Marinilactibacillus piezotolerans and Novosphingobium sp. during the culture of Indian white shrimp, Penaeus indicus, under biofloc and clear water system were evaluated. The experimental variation were CW1 (M. piezotolerans in clear water), BFT1 (biofloc + M. piezotolerans), CW2 (Novosphingobium sp. in clear water), BFT2 (biofloc + Novosphingobium sp.) and control (without bacterial strains and biofloc). Growth and survival considerably increased in probiotic bio-augmented treatments. Probiotic incorporation significantly improved water quality, especially ammonia reduction. Microbiota analysis from gut samples taken from different treatments revealed varied microbial population structure among clear water culture, biofloc culture and control. Proteobacteria and Firmicutes were the top phyla observed in the treatments which were significantly higher in bio-augmented systems than the control. Vibrio genera were predominantly observed in control and clear water system compared to that of biofloc systems. Immune genes were significantly altered in response to probiotic gut microbial supplementation than the control. Higher gene expression profile of important immune genes was observed in the biofloc reared shrimps. Expression of digestive enzyme related genes such as trypsin, chymotrypsin, cathepsin L, cathepsin B and alpha amylase were also upregulated significantly in probiotic supplementation especially in the biofloc treatments. Proteomic analysis of hepatopancreas of shrimps from different treatments was carried out by using 2D gel electrophoresis and MALDI-TOF analysis. The proteins were mostly related to growth and stress tolerance. Eukaryotic initiation factor 4E binding protein was expressed in all the groups and it was high in biofloc treated animals followed by animals treated solely with probiotics compared to those of control groups. The results concludes that biofloc already proved as an effective culture method for healthy shrimp production and supplementation of probiotic bacterial strains registered additional benefit for growth, survival, microbial, immunological status of P, indicus culture.


Results and discussion
The effect of bio-augmentation of Marinilactibacillus piezotolerans and Novosphingobium sp. were studied during the culture of P. indicus in biofloc system and the results compared with those from clear water system culture. The results revealed that biofloc culture system is an effective treatment system in terms of growth and survival. These two strains promoted the growth of shrimp significantly (P < 0.05) compared to the control (Fig. 1). Growth increased substantially in the BFT-augmented groups (BFT1 and BFT2) and a twofold higher growth was observed in combination with biofloc system. Amongst these two strains, M. piezotolerans bio-augmented BFT system was found to be more effective compared to the other treatments. Also, the survival was significantly higher (P < 0.05) in strains bio-augmented groups compared to control (82%). The survival was 85, 90, 86 and 91% respectively for CW1, BFT1, CW2 and BFT 2 groups.
By using biofloc technology, intensive shrimp culture can be carried out with reduced water usage, pathogen free and reduced waste output 23,24 . In shrimp, biofloc rearing has been extensively standardized with P. vannamei in terms of growth, water quality and health 8,9,13,[25][26][27] . This technology can be popularized further by adopting it to other species of shrimps. Our study revealed that the addition of live microbial supplements considerably improved the performance in terms of growth and survival. Scanty information is available on the use of probiotic with biofloc system. Krummenauer et al. 28 reported that the supplementation of probiotic mixture in biofloc system significantly improved the growth and survival of P. vannamei infected with V. parahaemolyticus. A similar trend was observed in an earlier study 14 conducted by bioaugmenting several probiotic strains in BFT system. The findings in the present study also corroborate the same. Llario et al. 29 reported that the addition of probiotic strain B. amyloliquefaciens directly into biofloc water enhanced the immunity of P. vannamei culture; however, there was no effect on growth. In the present study, growth and immunity significantly increased when BFT system was augmented with enzymatic gut microflora as probiotics supplement. Water quality parameters improved substantially in the biofloc groups compared to the control. Ammonia level was significantly (P < 0.05) reduced in strain supplemented BFT groups (BFT1 and 2) than strain supplemented clear water groups (CW1 and 2). Other parameters such as dissolved oxygen level were maintained better in BFT groups than the clear water system and control. Not much variation was observed in other parameters such as pH, salinity among the treatments ( Table 1). The carbonate, bicarbonate and hardness levels varied between experimental groups especially in BFT1, CW2 and BFT2 due to restricted water exchange and addition of probiotics and lime. Higher carbonate and bicarbonate in the biofloc treatments indicate a higher alkalinity level is maintained in BFT groups, which gives more buffering capacity compared to control. Higher buffering capacity will always lower diurnal fluctuations of pH and other parameters in the system. However, among the treatments depending on the nature of the probiotics used or presence of different heterotrophic bacterial community, there is non-significant variations in these parameters. Generally biofloc system is known for maintaining the water quality by continuous recycling of nitrogenous wastes generated from the culture system 6 . The strains selected as a supplement were observed to produce different kind hydrolytic enzymes 31 and this will give additional support for the nitrogen recycling and avoids the toxic deposition in the culture system. Analysis of gut microbiota. Microbial community is a nascent constituent of biofloc based aquaculture and the structure and function of microbial community highly influences the growth and survival rate. Culturable microbes can be identified with conventional methods but uncultured microbial communities are also taken into account and evaluated through metagenomics analysis of V3-V4 hypervariable region of 16 s rRNA by using high throughput sequencing approaches. The results of the high throughput sequencing analysis of gut samples taken from different treatments showed varied microbial population among clear water culture, biofloc culture and control. Many of the researchers have explored the microbial communities in biofloc system which influences the functional quality of the same. But in recent studies, it was inferred that gut of biofloc reared animals was also having important role in the host physiological activities by producing variety of enzymes 31 . So the microbial variation in gut samples of different treatments was analyzed and the major phyla of different treatments are depicted in Table 2. Proteobacteria and Firmicutes are the predominant phyla observed in the treat- Table 1. Water quality parameters observed in different treatments (mean ± SE (n)). Means in the same row having different superscript a,b,c,d differ significantly. *P < 0.05; Means in the same row having same superscript are not significant.  www.nature.com/scientificreports/ ments and they were higher in bio-augmented systems than the control. Bacteroidetes and Planctomycetes were the other dominant microphyla observed in the treatments. Figure 2 represents the dominant genus observed in the treatment and control. Vibrio dominance was high in control and clear water system compared to that of BFT system. This phenomenon of heterotrophic bacterial community inhibiting the vibrio population was successfully demonstrated through the 16 s high throughput sequencing approach in our previous study 25 . The genus Enterobacter was also observed to be high in biofloc system especially BFT 2. The next observed genera were Bacillus, which was greatly aggregated in biofloc treatments than clear water system of strain supplementation. Bacillus genera in lesser amount were observed in control culture. Ruegeria, a genus of the Rhodobacteraceae family is a marine Agrobacterium which was observed to be higher in BFT2 group followed by BFT 3. Halomonas genus was observed majorly in BFT 2 and CW2 than others (Fig. 2). Top 20 major bacillus genera were observed in different treatments and control (Fig. 3).
In the present study, 16 s High throughput sequencing approach was followed and the analysis of gut samples revealed considerable variation in control, clear water with probiotic supplement and biofloc with probiotic supplement. Rarefaction curves for species richness in all five samples plateau at the maximum depth, indicating an adequate sampling was conducted ( Supplementary Fig. S1). The bacterial diversity in each treatment, alpha diversity (ACE, Chao and Shannon) was analyzed. The richness index ACE varied from 375.03 to 604.40 and Chao1 varied from 363 to 605.92. The diversity index Shannon varied from 2.81 to 3.92. All the indices showed significantly ascending trend in BFT1 group (Supplementary Table S1). Microbial colonization and its function are usually attributed to the host nutrient and environmental status 25,30,31 . Among all the treatments, Proteobacteria was the major phylum observed followed by Firmicutes and Bacteroidetes; these populations are highly colonized in biofloc treatments. Many marine proteobacterial populations are found to produce valuable natural bioactive compounds like astaxanthin, haliangicin, ecotin, bryostatins etc. and these compounds benefit nutritionally and physiologically to the host animal 32 . Also, the colonization depends on the type of carbon sources supplied. Wei et al. 33 observed that biofloc produced by using glucose, starch and glycerol was majorly dominated by Proteobacteria and Bacteroidetes, an observation corroborated by the present study. Martínez-Córdova et al. 34 observed that Planctomycetes, Proteobacteria and Bacteroidetes are the dominant bacterial phyla observed in biofloc produced with amaranth meal whereas biofloc produced with wheat harbored Bacteroidetes as the dominant phyla.
Gene expression. Effect of the bio-augmentation of two strains on the immune status revealed an improved health status of the animals reared in these systems compared to that of the control. The effect was more significant in biofloc system when compared to clear water system. Also, the digestive system of the animal was highly influenced in the biofloc system revealing a positive effect as inferred by using expression of digestive enzyme related genes. A significant alteration of immune related genes was observed with supplementation of two strains in clear water and biofloc system. Peroxinectin and hemocyanin mRNA transcript upregulation was significantly higher (P < 0.01) in BFT2 and an eight-fold increase was observed compared to that in control. The expression of antimicrobial peptide crustin gene was observed to be upregulated in CW2 (four fold) than BFT 2 (three fold) (Fig. 4). The expression of SOD gene was observed to be high in CW2 and CW1 than biofloc treatments and www.nature.com/scientificreports/ control. The β-glucan binding protein was significantly upregulated (P < 0.01) in BFT2 and it was five-fold higher than in control. In other treatments a marginally higher expression was observed compared to that in control. The expression of prophenol oxidase (proPO) was also upregulated with these two strains added treatments and it was better in CW2 (Fig. 5).
One of the most advantageous benefits of the biofloc system is improving the immunity. In the present study, the expression of selected immune related genes was studied based on the effect of two different strains in conventional and biofloc systems. It has been reported that the transcriptional response of immune genes was   www.nature.com/scientificreports/ highly induced by bioflocs 25,35,36 and the results of our study revealed that the biofloc system combined with live microbial supplementation efficiently improved the immune status. Significant improvement of prophenoloxidase, β-glucan binding protein and peroxinectin expression was observed in both the strain supplied systems than the control. Prophenol oxidase system is a notable shrimp innate immune system associated with melanization. This can be activated by many factors including β-glucan binding protein and peroxinectin. Peroxinectin cell adhesion molecule plays a significant role in immunity 37,38 . The expression of these two confers the efficacy of live microbial supplementation in biofloc system. The other observed parameters like antimicrobial peptide, crustin and antioxidant enzyme SOD were also found to be significantly expressed in the system where probiotic strains were added. It is generally reported that the antimicrobial status and antimicrobial activity are important factors in building immunity. The higher expression of SOD, crustin and hemocyanin results in a higher antioxidant enzyme activity and antimicrobial status. Aguilera-Rivera et al. 39 reported that biofloc rearing of P. vannamei shrimp considerably increased the SOD and hemocyanin transcripts and provided protective response against Vibrio harveyi. Cardona et al. 40 reported that biofloc rearing was more beneficial in improving antioxidant and antimicrobial defense in Litopenaeus stylirostris. The efficacy of different strains supplemented on the digestive enzyme activities of P. indicus was analyzed using the expression of mRNA transcripts of different enzymes such as trypsin, chymotrypsin, cathepsin L, cathepsin B and alpha amylase. The results in general, suggested that mRNA transcript of the selected enzyme genes were significantly (P < 0.05) upregulated in treatments supplemented with strains and among these, probiotic strains with biofloc system had higher levels of these enzymes as evident from the upregulation of the concerned genes. A three-fold upregulation of trypsin was observed in BFT1 and BFT2 compared to CW1 and CW2. The chymotrypsin level was also substantially upregulated. In case of cathepsin L, it was expressed at a higher level in CW1 and CW2 than BFT1 and BFT2. Cathepsin B was upregulated in BFT 1 compared to CW1, but the same was downregulated in BFT 2 compared to CW2. The level of alpha amylase expression was very high in biofloc reared animals than the clearwater reared shrimps (Figs. 6 and 7).
As stated above, biofloc system continuously supplies viable nutrients throughout the culture period and these paves the way in stimulating the digestive status of the host animal. In the present study, the effect of two different bacterial strains on the digestive enzyme status of the shrimp was studied by qPCR. The level of selected enzymes trypsin, chymotrypsin, alpha amylase, cathepsin L and cathepsin B were upregulated significantly (P < 0.05) in biofloc system. Cardona et al. 40 reported that digestive enzyme activities of L. stylirostris was significantly influenced in transcript and enzymatic level with biofloc system compared to clear water system and without biofloc, which is consistent with our findings. Our previous study revealed the benefits of biofloc based system over the conventional system and the manner in which the microbial community plays a vital role in inducing nutritional immunomodulation 2,31 . Xu and Pan 41 opined that biofloc system stimulates the activity of digestive enzymes such as amylase and protease in P. vannamei which resulted in improved growth. www.nature.com/scientificreports/ Proteomic analysis. Proteomic analysis of hepatopancreas of different samples was carried out by using 2D gel electrophoresis and MALDI-TOF analysis. The results revealed that different proteins were expressed in different treatments that the biofloc reared animals were exhibiting less proteins spots than clear water group animals and control. The proteins mostly related to growth and stresses were observed in clear water and control group animals. Eukaryotic initiation factor 4E binding protein was observed to be expressed in all the groups and it was high in biofloc treated animals followed by probiotic alone group and control. It was reported that proper dietary protein supplementation would directly influence the protein synthesis and deposition. The protein synthesis included initiation, elongation and termination 42 . Hence, the high expression of Eukaryotic initiation factor 4E binding protein in BFT groups indicated the optimum nutrient availability through optimum C:N ratio that confers the growth. Kelch motif family protein was observed to be upregulated in BFT 1 and BFT 2  www.nature.com/scientificreports/ followed by CW2. It was not observed in control and CW1 ( Fig. 8 and Table 3, Supplementary Table S2; Supplementary Fig. S2). These proteins are identified as propellers of cell function and having diverse functions in the cells 43 . They also play an important role in maintaining the cell integrity during stress conditions and also in signal transduction protein during WSSV infection 44,45 . It is important that the expression of hypoxia inducible factor 1a was exclusively observed in BFT1 group animals. These transcription factors are important to regulate many genes in response to hypoxia 46 which increases the susceptibility of shrimp to WSSV infection 47 . Structural maintenance of chromosome protein is only expressed in BFT 1 animals which are responsible for maintenance of chromosome transmission during replication and segregation. Serine proteinase was expressed only in BFT2 group animals which is a major enzyme also responsible for immunity as observed in P. monodon during V. harveyi infection 48 . At the same time, Alpha-amylase OS are only observed in BFT1 animals. Vacuolar protein sorting-associated protein 29 OS was reported to have a relationship with the immune response of insects 49 which is observed only in CW2 group animals. Hemocyanin is a multifunctional oxygen-carrying protein immune defense factor reported to have antibacterial and antiviral defense properties 50,51 . This was considerably upregulated in CW2 group animals than control, whereas the animals in other groups do not show the expression of the same. Similarly, Serine proteinase inhibitor 6 was observed in CW2 group animals only and this protein plays a role in controlling many biological processes as well as host defense mechanism such as pathogen digestion 52 .
It is therefore concluded that biofloc technology is an efficient technology for aquaculture and additional supplementation of probiotic strains to the biofloc would enhance the efficiency. This has been inferred from the analysis of differentially expressed proteins as well as the expression of immune and digestive enzyme related genes, in addition to a well accounted advantage in terms of growth, immunity and survival.

Selection of strain. The bacterial strain-Marinilactibacillus piezotolerans strain AP BFT8 (MK934556) and
Novosphingobium sp. strains AP BFT4 (MK934552) were selected for this study. The strain was isolated from P. indicus previously reared on biofloc based system with different carbon sources. These strains were selected from a total of 94 isolated colonies based on its extracellular enzymatic (amylase, lipase, protease) profile and antibiofilm activity against pathogenic Vibrio sp. 31 . The candidate species P. indicus (average weight 1 g) was stocked in 500 L tanks at the rate of 100 no/per tank. The effect of these two strains were evaluated in a biofloc system and compared with shrimps reared in a clear water system. The experimental combinations were as follows: CW1 (Novosphingobium sp. in clear water), BFT 1 (Novosphingobium sp. + biofloc), CW2 (M. piezotolerans in clear water), BFT 2 (M. piezotolerans + biofloc) and a treatment devoid of these strains and biofloc were taken as control (C). The experiment was conducted in triplicate.
Prior to stocking, tanks were prepared for the floc development. Initially the tanks were filled with disinfected (Bleach@60 ppm) sea water, for developing autotrophic bloom the fertilizer such as (Urea (3 g/m 3 ), Triple super phosphate (0.5 g/m 3 ), and dolomite (5 g/m 3 ) were added in an alternate days for 3 days, then the aerobically fermented biofloc inoculum were added to appropriate tanks regularly till the floc level reaches upto 2-5 ml/L to convert the autotrophs into heterotrophic medium. Biofloc inoculum was prepared by fermenting a mixture of carbon sources and bacillus microbial consortium (CIBA-floc) and allowed to proliferate; the filtrate of this was added to the biofloc treatment tanks at the rate of 1 L/m 3 . Thereafter, the animals were stocked and fed with commercial nursery shrimp feed (Inve Aquaculture, Thailand) at 10% of the body weight in each tank. Biofloc was added to the respective tanks on alternate days. The water was added at 10% per week to compensate the evaporation loss in the experimental tanks whereas 50-60% water was exchanged in control and clear water system tanks per week. The probiotic strains were grown on liquid broth medium (peptone-0.5%, yeast extract-0.1%, KH 2 PO 4 -0.05%, sodium nitrate-0.1%, sodium chloride-2% and dextrose-0.1%). Thereafter, the respective probiotic strains were inoculated and kept in a shaker at 150 rpm at 32 °C. After reaching the OD at 600 nm, the cultures were inoculated to the respective tanks at the rate of 20 mL per tank (0.1%). The probiotic strains were added once in 2 days. Sampling of shrimps was done once in 15 days to monitor and record growth and survival. www.nature.com/scientificreports/ Analysis of water quality, growth and survival. Physico-chemical parameters such as salinity, pH, DO, ammonia nitrogen, nitrite, and phosphate were measured periodically by following standard methods described in APHA (1995). During sampling, the length, weight and survival were recorded once a fortnight, and the survival was calculated as the ratio of the number of shrimps survived to the number of shrimps stocked × 100, and weight gain was computed as the difference between the final and initial average weights. The amplicons were purified using Ampure beads. Additional 8 cycles of PCR were performed using Illumina barcoded adapters to prepare the sequencing libraries. The sequence data quality was checked using FastQC and MultiQC software. After sequencing, the degenerate primers were removed by trimming (20bp) from 5′ end and the adapter sequences were removed using Trimgalore (Babraham, 2004) and further imported to mothur software and aligned to form contigs. After screening for errors, 300 bp and 532 bp sequence lengths were retained. The nonspecific regions were aligned with 16s rRNA database. Thereafter, the gaps and overhang in the end region of the contigs were removed and the possible chimera sequences were removed by using UCHIME algorithm 53 . Further, it was classified into taxonomical outline and clustered into operational taxonomic unit using GREENGENES v.13.8-99 database 54 .

Gut microbial analysis.
Immune response. Expressions of immune related genes were analyzed in hepatopancreas samples and the selected genes were prophenoloxidase, ß-glucan binding protein, hemocyanin, crustin, SOD and peroxinectin.
The digestive enzyme related genes such as trypsin, chymotrypsin, cathepsin L, cathepsin B and alpha amylase were also analyzed. Total RNA from hepatopancreas was isolated using TRIzol reagent (RNAiso plus, Takara) and converted into cDNA using Prime Script 1st Strand cDNA Synthesis Kit (BioRad, USA). Thereafter, the cDNA was serially diluted for further relative quantification of target immune genes. The primers used are listed in Table 4. Total 20 μL of reaction constituted [10 μL of 2 × SYBR Green qPCR master mix (Bio-Rad, USA), 1 μL each of forward and reverse primers (10 pmol), 1 μL of template DNA (30-60 ng) and 7 μL of PCR-grade water]. PCR was performed in Real-time PCR (Applied Biosystem's Real-Time PCR system StepOne Plus). The following conditions were maintained; holding stage of 10 min at 95 °C (initial denaturation), followed by 45 cycles of 15 s at 95 °C (denaturation), and 1 min at 60 °C (annealing and extension). Expression of ß-actin gene was used as an internal control and the data were analyzed using cycle threshold (CT) method 55 . Table 4. List of primers used for real-time PCR (qPCR).

Gene
Primer sequence (5′-3′) Accession no./ Reference   Trypsin  F-TCC TCT CCA AGA TCA TCC AA  R-GGC ACA GAT CAT GGA GTC  Stephens et al., 2012   Chymotrypsin  F-GGC TCT CTT CAT CGACG  R-CGT GAG TGA AGA AGT CGG  Stephens et al., 2012   Cathepsin L  F-CTC AGG ACG GTA AGT GTC G  R-TTC TTG ACC AGC CAG TAG G  Stephens et al., 2012   Cathepsin B  F-GGA TGT AAC GGA GGC TTC  R-CTG TAT GCT TTG CCT CCA  Stephens et  www.nature.com/scientificreports/ Proteomic analysis. Hepatopancreas sample were taken from shrimps belonging to varied experimental groups and proteins were extracted by using lysis buffer (135 mM NaCl, 2.7 mM KCl, 1.5 mM KH2PO4, and 8 mM K2HPO4, 0.1% protease inhibitor cocktail mixture pH 7.2) in liquid nitrogen. For two-dimensional gel electrophoresis of processed proteins (150 µg), 13 cm IPG strips of pH 4-7 (GE Healthcare, Uppsala, Sweden) were employed in the first dimension 56,57 . Passive Rehydration process was performed and proteins were focused for a total of 50,000 Vhs at a constant temperature (20 °C) under linear voltage ramp after at 30 V in a PGPhor III (GE Healthcare, Uppsala, Sweden) apparatus with following IEF conditions, 100 V gradient for 1 h, 300 V gradient for 2 h, 1000 V gradient for 1 h, 5000 V gradient for 5 h and 5000 V step on hold for 7 h. Following IEF, each IPG strip was placed in the equilibration buffer containing 2% DTT first followed by incubation in another buffer in which the DTT was replaced by 2.5% iodoacetamide. The second dimension PAGE (12.5%) was carried out in an SE600 (GE Healthcare, Uppsala, Sweden) at 1 W/gel for 1 h and 13 W/gel for 3 h. Further, different spots from each sample were picked and analyzed by using MALDI-TOF and similarity search was made using Mascot search engine.
Statistical analysis. The data obtained from the experiment were analyzed by SPSS (Version-17). The comparison between all the treatment groups and between the treatments (P < 0.05) was made using one way ANOVA and Duncan's multiple range test (DMRT), respectively. The probability level was kept at 5% for the statistical analysis.