EFFECT OF PROTEIN SOURCE AND PROBIOTIC ON THE INTESTINAL TRACT OF PACIFIC WHITE SHRIMP Litopenaeus vannamei

This study aimed to evaluate the viability of supplementing two diets for the shrimp Litopenaeus vannamei with Lactobacillus plantarum. One using fish meal as a protein source and another using soy protein concentrate, as well as the effect of these formulations on shrimp intestinal microbiota. To assay probiotic viability in the formulated diets, the number of CFU g-1 was observed weekly over the course of four weeks. The viability of Lactobacillus plantarum in relation to the physical parameters of the diets, including stability, buoyancy, and expansion, was quantified. The effect of the diets on microbiota and intestinal tract morphology was determined by performing a 2x2 factorial experiment (two diets, with or without supplementation) in triplicate, totaling 12 experimental units, with five animals per unit, fed with 3.5% of biomass for 17 days. The concentration of lactic acid bacteria decreased over time, irrespective of protein source. The diet with fishmeal as a protein source, independent of probiotic supplementation, presented good stability and did not disintegrate after four hours. In contrast, the diet with soy protein concentrate, supplemented or not, disintegrated between 2.5 and 3 hours, presenting low stability. All diets presented 0% buoyancy. The expansion rate was higher in diets with soy protein concentrate, but without the influence probiotic supplementation or interaction between the factors. In the in vivo assay, both supplemented diets showed greater total heterotrophic bacteria count than without probiotic; however, no difference in count was noted in diets with different protein source. Lactic acid bacteria were only observed in the shrimp fed diets supplemented with probiotic. Histology of the intestinal tract showed that all intestines had intact, well-developed and well-organized cells, irrespective of diet. Thus, L. plantarum, when combined with different protein sources, produced similar effects on the structure and microbiota of the marine shrimp Litopenaeus vannamei.


INTRODUCTION
Feed represents the highest cost of production in shrimp farming, and protein sources are the most expensive ingredients (SOOKYING et al., 2013;JATOBÁ et al., 2014). Among the protein sources used to make diets for marine shrimps, fishmeal (FM) is highlighted because of its amino acid composition, vitamins, minerals and palatability (SOOKYING et al., 2013;CHAVEZ et al., 2016). However, FM is not considered a renewable or sustainable ingredient, owing to the practice of extractive industrial fishing (ZHAO et al., 2017). This calls for the development of alternative protein sources for shrimp farming with similar nutritional value, but lower cost, thus allowing for greater profit margin (CHIU et al., 2016;SUN et al., 2016).
In addition to nutritional challenges facing shrimp farmers, disease is a predominant problem, especially those of bacterial or viral origin that arise from inadequate management and low-quality feed (MOSS et al., 2012;ZOKAEIFAR et al., 2012). To minimize the losses caused by these diseases, new technologies that reduce environmental impacts and production costs need to be evaluated (CORREIA et al., 2014;WANG et al., 2016).
The use of preventive measures, such as supplementation with probiotics capable of increasing the immunocompetence of aquatic organisms, have proved to be an efficient alternative (LUIS-VILLASENOR et al., 2013;MERRIFIELD et al., 2014), in particular the use of lactic acid bacteria (LAB) with satisfactory results, both in vitro  and in vivo (VIEIRA et al., 2007JATOBÁ et al., 2008JATOBÁ et al., , 2011JATOBÁ et al., , 2017JATOBÁ et al., , 2018. However, little is known about the synergistic effects of the different feed ingredients used in combination with probiotics.
Therefore, this study aimed to evaluate the viability of supplementing two diets of the shrimp Litopenaeus vannamei with the lactic acid bacterium Lactobacillus plantarum, one using fishmeal as a protein source and another using soy protein concentrate, as well as the effect of these formulations on the shrimp intestinal microbiota.

METHODS
The research was carried out at the Laboratório de Aquicultura do Instituto Federal Catarinense, campus Araquari.

Formulation and preparation of diets
Two iso-energetic diets were formulated (Table 1)   Diets were produced at the Instituto de Ciências do Mar (LABOMAR), Universidade Federal do Ceará (UFC), Eusébio, Ceará, Brazil. The dry ingredients of the diet were previously ground and sieved (500 µL).Subsequently, the micro-ingredients were homogenized in a Y-mixer for 10 min and were added to the macroingredients to be homogenized in a food mixer for 10 min. Soon after, the oils and soy lecithin were added, and then 40% of hot water. Theresulting mixture was pelletized through a meat grinder and dried for approximately 18 h at 40 °C.

Incorporation of Lactobacillus plantarum in the feed
Lactobacillus plantarum is a probiotic strain isolated from healthy shrimp. It was selected and approved by in vitro tests against pathogens and identified as Lactobacillus plantarum at the Centro Pluridisciplinar de Pesquisas Químicas, Biológicas e Agrícolas (CPQBA), Universidade Estadual de Campinas, with accession number CPQBA 007-07 DRM01 (VIEIRA et al., 2007. Experimental diets were inoculated with 100 mL Kg -1 of probiotic, while a sterile medium was inserted in control diets, according to the protocol established by JATOBÁ et al. (2011).

Experimental design
After producing diets with different protein sources and incorporation of probiotic, a two-factor experimental design (2x2), including factor 1, protein source, FM or SPC, and factor 2, diet supplemented or not with probiotic, was established and resulted in four treatments.

Assay of physical and microbiological parameters of diets
Three 1.0 g samples of each treatment were submerged in a beaker with 1 L salinized water (3% NaCl) to evaluate the stability and buoyancy of the diets. The assay was carried out during four hours at 30-minute intervals. The floating pellets were collected and counted to estimate buoyancy (adapted from ABCC, 2005).
The pellet expansion rate was determined by measuring the initial and final diameter (digital caliper), as samples of the stability and buoyancy test. After four hours, the pellets were removed from the water, and the final diameter was measured.
All physical tests of the diets occurred weekly over the course of four weeks.
After 24 hours of dietary supplementation, 1.0 g samples were macerated and serially diluted (factor 1:10) in sterile saline solution (SSE) 0.65%. Samples of each dilution were seeded in MRS agar medium and incubated for 48 h at 30 °C for LAB count. This procedure was repeated weekly for four weeks.

In vivo assay
Sixty Pacific white shrimp (average weight of 6.85 ± 0.04 g) were distributed in 12 experimental units (22 L), five in each (stocking density of 227 shrimps per m 3 ), equipped with canister filter and constant temperature (28.1 -29.2 °C). During the experiment, dissolved oxygen, temperature and pH were measured two times per day. The experimental units were divided into four treatments with three replicas. Shrimp were fed three times a day (8:30, 1:30 and 5:00 p.m.), offering 3.5% of biomass, a schedule maintained throughout the 17-day experiment.
Following 12h of starvation at the end of the experimental period, guts of three shrimp from each tank were removed and pooled for microbiological and histological analysis. The pooled shrimp guts were homogenized and serially diluted 1:10 in 0.65% of NaCl sterile saline. Samples from each dilution were cultured in PCA (Plate Count Agar), TCBS (thiosulfate citrate bile salts sucrose) agar, cetrimide agar, and MRS agar media and incubated for 48h at 30 °C for viable culturable heterotrophic bacterial counts, including Vibrio spp., Pseudomonas spp., and LAB, respectively.
The cephalothorax and the last abdominal segments of two different shrimp used in the microbiological evaluation were sectioned with a razor, and the abdominal portion was fixed in standard marine Davidson solution for 24-48 h and then transferred to alcohol solution 70%. Afterwards, the samples were submitted to the standard histology procedure with dehydration, diaphanization and paraffin inclusion.
Paraffin blocks 5-μm in thickness were cut with a microtome for subsequent staining with Harris Haematoxylin and Eosin (HOWARD et al., 2004;KIM et al., 2006) and observed in a Differential Interference Contrast (DIC) AxioImager A2 (Zeiss).
At the end of the experiment, the apparent feed efficiency, protein efficiency ratio, final body weight, weekly gain and survival were assessed.

Statistical analysis
Data were first subjected to Bartlett's analysis to verify the homogeneity of variance, and microbiological data were log(x + 1) transformed. All data were subjected to bifactorial (two-way) ANOVA, and significant differences among treatments were analyzed using the Student-Newman-Keuls (SNK) test. All tests were conducted at a 5% level of significance (ZAR, 2010).

Physical and microbiological parameters of diets
The diet with FM as a protein source, independent of probiotic supplementation, did not disintegrate during the evaluated time (4 hours), whereas the diet with SPC disintegrated between 2.5 and 3 hours, supplemented or not. All diets showed 0% buoyancy, an essential characteristic for any benthic aquatic organism.
The expansion rate of the diets differed in the ingredient factor. Diets elaborated with SPC expanded more than those formulated with FM; however, probiotic supplementation did not influence this process (Table 2).
In the second and third week, the SPC and supplemented diet had higher concentrations of LAB, whereas at weeks 0, 1 and 5, no differences were observed between the supplemented diets ( Figure 1). LAB was not observed in diets without probiotic supplementation. 25012 Different uppercase letters indicate significant differences between ingredients, and different lowercase letters indicate significant differences between probiotic and control diet in the bifactorial (two-way) ANOVA and SNK tests. Not detectable (ND); Different uppercase letters indicate significant differences between ingredients, and different lowercase letters indicate significant differences between probiotic and control diet in bifactorial (two-way) ANOVA and the SNK tests. Figure 1. LAB count in diets formulated with different protein sources supplemented with probiotic over time. *Indicates significant differences between diets.

5/8
In vivo assay Lactobacillus plantarum changed the microbiota of the intestinal tract of shrimp fed with the probiotic diets, irrespective of protein source, whereas LAB was not detected in shrimp fed with control diets, and total heterotrophic bacteria decreased in shrimp fed with the probiotic diet. Neither protein source nor probiotic interfered with the bacterial count of Vibrio spp. (Table 3).
The qualitative analysis of the histological sections of the middle portion of the intestinal tract of the shrimp presented structures similar to those described by BELL and LIGHTNER (1988). In the histological sections of the intestinal tract, no qualitative differences between the treatments were recorded (Figure 2). The intestinal epithelium presented intact, developed, organized and well-defined cells, as well as absence of vacuoles and intercellular spaces (Figure 2).
The dissolved oxygen remained above 4.5 mg L -1 and pH at 7.9 throughout the experiment, in all treatments, not diverging among them.
Zootechnical variables did not diverge among treatments, regardless of the factor evaluated, supplementation or protein source (Table 4) and the survival rate of 100% were recorded in all treatments.  AFE (Apparent feed efficiency); PER (Protein efficiency ratio); FW (Final weight); WG (Weekly gain); Different uppercase letters indicate significant differences between ingredients, and different lowercase letters indicate significant differences between probiotic and control diet in bifactorial (two-way) ANOVA and the SNK tests. 6/8

DISCUSSION
The ABCC (2005) suggests a minimum disintegration time of 3.0 to 3.5 hours for marine shrimp diets. In the present study, the diet with SPC disintegrated between 2.5 and 3 hours, supplemented or not, which demonstrated lower stability when compared to the diet with FM, independent of probiotic supplementation. Few studies evaluate the rate of pellet expansion. However, different rates of diet expansion may be related to the potential for agglutination of the ingredients used as protein source (LIM and DOMINY, 1990).
To be nutritionally useful, a probiotic microorganism must colonize and stay alive in the digestive tract of its host (GATESOUPE, 1999). Therefore, it is necessary to offer the probiotic microorganism in concentrations higher than those normally found in host intestinal tract, and/or even in concentrations higher than other bacterial genera in the host intestinal tract. In the present study, the decrease of LAB agrees with the finding of NAVIN CHANDRAN et al. (2014) who observed a decrease in the amount of probiotics over time. Bearing in mind 105 CFU per g of diet, the minimum concentration of LAB offered in probiotic diets, SPC was able to sustain this concentration for one week more than FM. This could be explained by the fact that soybean has a large amount of proteins and fibers, as well as inulin and oligosaccharides with prebiotic potential, all of which increase the viability of microorganisms (FUCHS et al., 2005).
The use of LAB is also advantageous in conserving food since it reduces the presence of bacteria with pathogenic potential capable of deteriorating the quality of the diet (HOOVER and STEENSON, 2014;O'BRYAN et al., 2015), thus guaranteeing longer storage time, irrespective of protein source. According to VIEIRA et al. (2008), L. plantarum, the same strain used in the present study, remained at high concentrations for four days before the supply of probiotic was exhausted, however, the probiotic effects were maintained for two days, suggesting its continuous use. However, the authors did not consider the possibility of using prebiotics or other ingredients that might favor the colonization of these probiotic bacteria in the diet, with a view to increasing the viability and the residence time of these microorganisms. In the present study, shrimp fed with the diet supplemented with probiotic presented a reduction in the concentration of total heterotrophic bacteria, corroborating similar findings in other studies JATOBÁ et al., 2011JATOBÁ et al., , 2017JATOBÁ et al., , 2018BUGLIONE-NETO et al., 2013), most likely owing to the production of extracellular compounds, such as organic acids and bacteriocins (FULLER, 2012).
Neither protein source nor probiotic interfered with the count of Vibrio spp. This result diverged from that observed by VIEIRA et al. (2008VIEIRA et al. ( , 2016 who observed a reduction in the vibrionaceae population in the intestinal tract of L. vannamei fed with the same probiotic strain. It is common to report the decrease of Vibrio spp. in the intestinal tract of aquatic animals fed with diets supplemented with lactic acid bacteria JATOBÁ et al., 2008JATOBÁ et al., , 2011JATOBÁ et al., , 2018ZHANG et al., 2011;STANDEN et al., 2015;HAI, 2015;VIEIRA et al., 2016). However, the short experimental time, only 17 days, may have contributed to the absence of inhibition of the Vibrio spp. in this trial.
The absence of changes in shrimp epithelium suggests that neither protein sources damaged the tissue. JATOBÁ et al. (2017) observed that L. vannamei fed diets with SPC and FM, the same sources as those used in this experiment, presented the same food efficiency, demonstrating that shrimp have the ability to digest both protein sources. In the histology of the intestinal tract of Japanese eel (Anguilla japonica) fed diets supplemented with L. plantarum at concentrations below 107 CFU per g of diet, no differences were observed relative to animals fed with control diet (LEE et al., 2017). The same work shows that concentrations higher than 108 CFU per g of feed could modulate the intestinal microbiota of the Japanese eel (A. japonica). This result corroborates the results of other studies, demonstrating the need to offer probiotics consistently and in high concentrations (HAI, 2015;JATOBÁ et al., 2018).
Both factors, protein source and probiotic, did not influence the growth performance, unlike the observed by JATOBÁ et al. (2017) which observed a decreased of apparent feed efficiency, protein efficiency ratio and weekly gain for L. vannamei reared in biofloc, during 40 days, fed with same diets (without supplementation), the data of these authors were higher than observed in this study, probably due to the presence of the biofloc. During In vivo assay water variables values were considered adequate for the species, according to (BOYD and GAUTIER, 2000), suggesting that all changes are related to treatments.
The use of this probiotic has already shown beneficial effects on shrimp farms under outdoors conditions (VIEIRA et al., 2016;GAINZA et al., 2018), however in this study the probiotic did not show benefits in the zootechnical variables, this fact may be related to the duration of the experiment that was designed to evaluate changes in the microbiota.
The L. plantarum strain used in the present study has already demonstrated its probiotic potential to shrimp (L. vannamei), increasing survival after experimental infections against Vibrio spp. in post-larvae  and juveniles (VIEIRA et al., , 2010, as well as shrimp reared on commercial farms (VIEIRA et al., 2016), increasing the digestibility of diets (BUGLIONE-NETO et al., 2013) and improving food conversion (VIEIRA et al., 2016), justifying its application regardless of the protein source used.

CONCLUSION
The inclusion of L. plantarum did not interfere with the physical properties of the two dietary formulations. The protein source did interfere with the viability of L. plantarum in diets, but soy protein concentrate kept it in higher concentrations for longer. L. plantarum combined with different protein sources, fishmeal or soy protein concentrate, produced similar effects on the histology and microbiota of the marine shrimp (Litopenaeus vannamei).