Modulation of the gut microbiota of Pacific white shrimp ( Penaeus vannamei Boone, 1931) by dietary inclusion of a functional yeast cell wall- based additive

Several eco- friendly natural substances can enhance the shrimp immune defence system therby acting as a prophylactic agent in feed additives. Agents such as (1, 3)- (1, 6)- D β - glucan and complex mannan- oligosaccharides located in yeast cell walls present such immunomodulatory and potential prebiotic properties. The aim of our study was to evaluate the effect of a commercial yeast cell wall extract (YCW) on shrimp performance and health status, and influence on gut microbiota. Juvenile Penaeus vannamei (Boone, 1931) were raised at an intensive shrimp farm and fed with two different diet inclusions levels of YCW, that is, 0.5% and 1.0%, in addition to a yeast free control group. After 102 days, animals were sampled, and standard nutrition performance parameters were measured. Additionally, the phylogenetic profile and composition of shrimp gut microbiota were evaluated. Animal performance, including growth and survival, was significantly better on animals fed with YCW than the control group. Furthermore, beneficial bacteria phylotypes were stimulated by the presence of YCW, positively modulating the gut microbiota, with emphasis on 1.0% YCW treatment. Therefore, YCW can be re-garded as a prophylactic functional agent in the intensive rearing of juvenile P. vannamei thus improving animal performance and contributing to a healthy intestinal microbiota.

| 1115 SERVIN ARCE Et Al. major studies directed to this field (Davies et al., 2019). An optimal nutritional condition leads to a satisfactory immune response when animals face an infectious or non-infectious challenge such as stress and environmental pressures. Nowadays, aiming at an eco-friendly and sustainable shrimp farming industry, researchers and farmers are searching for new prophylactic treatments to prevent and protect animals against pathogens and increase resistance during stressful situations (e.g. disease conditions, environmental disturbances, inadequate nutrition, handling) (Thitamadee et al., 2016).
The concept is to develop nutritional strategies to enhance animal health and production, without the use of pharmaceuticals and antibiotics, the latter being of particular concern due to antimicrobial resistance and other emerging pathogens. Natural compounds (nutritional or non-nutritional) can positively modulate the animal immune system, resulting in increased resistance against diseases and/or pathogens . Also, they promote better growth and production efficiency. Many of the natural compounds are termed as prebiotics and consist of defined carbohydrate cell wall structures mostly originating from fungi, bacteria, plants, microalgae and macroalgae (Ringø et al., 2010;Song et al., 2014). These have proved to be most attractive for aquaculture as they are derived from natural sources and meet consumer expectations for safety and environmental concerns. The yeast cell wall, for example, is composed by different polysaccharides (Ringø et al., 2010), which may have prebiotic effects. Polysaccharides are polymers of simple sugars with two main biological functions: energy storage and extracellular structural integrity (Nelson & Cox, 2004). The mechanism of action of polysaccharides is related to specific binding proteins, opsonins and other defence proteins that activate the cellular function when reacting with β-glucans or LPS, that is, when the β-glucans bind to the receptors on the haemocytes the stimulation of the immune response may be initiated (Karunasagar et al., 2014;Meena et al., 2013;Smith et al., 2014).
The gut microbiota has been described as the 'new organ' in the animal imparting major systemic effects (Baquero & Nombela, 2012;McFall-Ngai et al., 2013), with an important role in digestion (physiology), nutritional metabolism and influencing immune components in disease and inflammatory responses (Akhter et al., 2015).
The factors to be considered in conducting microbiome research were described comprehensively by Goodrich et al., (2014) in terms of design, execution and interpretation of data analysis for animal studies. To date, less work has been undertaken on fish and shrimp compared to other species and humans.
The relevance of the microbiome for the Pacific white shrimp has been evaluated by Cornejo-Granados et al., (2017) showing differential bacterial community composition between wild, aquaculture raised and shrimp with AHPND/Early Mortality Syndrome outbreak conditions. Wild shrimp presented larger biodiversity gut microbiota than shrimp raised under controlled conditions. Likewise, AHPND/ Early Mortality Syndrome led to a loss of hepatopancreas microbiota diversity. These authors have stated that little is known about shrimp microbiota remodelling. With all the different scenarios that shrimp will encounter, there is a need to relate the microbiota to the diet composition and potential to modulate favourably the defence barrier mechanisms in shrimp. There is now much importance given to the effect of functional feed additives in aquaculture and particularly in fish but also gaining momentum for shrimp (Gainza & Romero, 2020). Thus, there is much interest in comprehending the links between diet immune stimulants and the microbial ecology of the gastrointestinal tract.
It has been stated by Dai et al., (2020) that the gut keystone taxa disproportionally affect the function and stability of their resident community and thus are candidate targets for improving host health.
However, exactly how disease progression changes the assembly of gut bacterial community remains unclear requiring more elucidation. Their influence on shrimp gut microbiota can be a great asset to understand patterns in bacteria and environmental changes. Thus, the search to include sustainable feed supplements into shrimp diets could lead to the stabilization of potential beneficial bacteria in the digestive system and act as effective control agents to mitigate against disease and infection. This has been recently comprehensively reviewed by Holt at al. (2020). These authors have stated how gut microbiome manipulation may offer an attractive option for aquaculture (e.g. improved digestion, immunity and health as well as an alternative to antibiotics). It has been suggested as a possible alternative to the use of broad-spectrum antibiotics in the management of disease processes in shrimp.
Therefore, the purpose of our study was to identify the biological potential and environmental and microbial ecological aspects relating to the composition of gut microbiota in cultured shrimp when fed with a commercial complex of brewer's yeast cell wall extract with supportive proteic by-products (enzymes) from fungus (powder from Trichoderma longibrachiatum spp) (PAQ-Gro ™ Phibro, Teaneck, USA). PAQ-Gro ™ is a unique, patented propriety feed additive product for fish and shrimp diets. It has been stated that it can improve growth performance, FCR, survival rates and the overall health status of shrimp and fish during the critical hatchery, nursery and growout phase. We examine this potential in juvenile Pacific white shrimp (Penaeus vannamei Boone, 1931) under typical intensive farming conditions employing a commercial propriety feed.

| Experimental design
A total of 1,200 juvenile Pacific white shrimp (P. vannamei), 3 g ± 0.25 g, were raised for 102 days in an intensive system with seawater with minimum water exchange (<10% cycle), in the experimental station of Laguna Aquaculture, Colima city, Mexico. Animals were kept in twelve floating cages randomly allocated in a 1500 m 2 shrimp pond with HDPE liner with sand bottom. Floating cages dimensions were as follows: a) length: 1.60 m; b) width: 1.65 m; depth: 0.60 m; bottom area: 2.56 m 2 ; volume: 1.58 m 3 . Floating cages were preferred rather than free on the pond due to better control of the trial conditions. A single anchoring kept the cages flowing around the ponds. No solid mud or soil was inside the cages, as they were 0.6 m depth. Cages were covered with 5 mm mess to keep shrimp inside the cages and to avoid predation, especially from birds. Shrimp were randomly distributed in twelve cages, 100 animals per cage and four cages per treatment.
Initial shrimp density was 63 shrimp per m 3 . The same stocking density was used outside the cages; thus, the standard farm protocol and commercial conditions were maintained. Water was pumped from an artesian well with oceanic salinity. Constant aeration was provided by two horsepower (hp) paddlewheel aerators located outside the floating cages, with total capacity of 16 hp per hectare.
The water quality parameters were monitored daily throughout the duration of the trial. The minimum and maximum values observed during the study period were as follows: salinity (33-35 g L −1 ), ammonia (0.1-0.3 mg L −1 ), nitrite (0.2-0.5 mg L −1 ), pH (8.5-9.6), water temperature (29-34°C) and dissolved oxygen (4-10 mg L −1 ). Biomass measurements occurred weekly, with 20 shrimp per cage, that is, 20% of the total population were randomly selected with the use of nets and individually weighed at the site.
Growth performance and feed efficiency were determined with the indexes described below. Statistical analyses were calculated using one-way ANOVA, followed by Tukey post hoc test, with a pvalue < .05 considered statistically significant.

| Experimental diets
Two different doses of a commercial shrimp feed additive, containing primarily yeast cell wall components, that is, β-glucans and mannans from Saccharomyces cerevisiae (YCW, i.e., PAQ-Gro TM Phibro, Teaneck, USA), were added 'on top' of the shrimp feed formula mixture, and compared to a control group (without YCW supplement).
Four cages were fed with 5 kg of YCW per tonnes of feed (0.5%), four cages were fed with 10 kg of YCW per tonnes of feed (1.0%), while the remaining four cages were fed with commercial diet without supplement (control group). A typical shrimp commercial diet, with standard commercial nutrient levels, was used as basal diet (Table 1). The proximate composition of each protein ingredient is presented in Table S1.
In Mexico, the current use of wheat flour in shrimp feed ranges from 30% to 50%. Particularly due to established milling techniques (e.g. double preconditioner plus conditioner), it is possible to use a die of 1.5 mm and achieve a good pellet stabilization. Similarly, kelp hydrolysate was used as a binder as it helps in the stabilization of the feed pellet with the presence of alginates. Also, the starch from wheat flour gelatinization supports the feed stabilization for a minimum of two hours under the water. The YCW supplement was added 'on top' of the commercial formula mixture. It was included with micro ingredients in the mash, prepared by mixing the feed in a mixer with the YCW prior to standard pelleting. As we have used a standard shrimp commercial diet rather than an experimental diet, the tested additive was included in addition to the feed blend. It is frequent practice under realistic commercial feed mill conditions for testing a new product without altering the primary specification.

| Sample collection
After the 102 days of the feeding trial, the cages were harvested.
A total of 21 shrimp were randomly sampled for intestine collection, that is, 7 shrimp per treatment (2 shrimp from the first three replicas and 1 shrimp from the fourth replica). Shrimp were euthanized by thermal shock (33-9°C), carcass surfaces were cleaned with 70% (v/v) ethanol and the midgut with its content was collected, using sterilized forceps and scissors. As shrimp were constantly fed, with continuous feeding activity, at the moment of gut collection, the midgut was filled with the provided feed without natural food. Samples were immediately fixed in 70% (v/v) molecular grade

| Analysis of high throughput sequencing results
The main focus of our study was to address the modulation of the gut microbiota with regards to the stabilization of a desirable, symbiotic microbiota profile. We adopted the protocols for conducting a microbiome study by Goodrich et al., (2014) (Segata et al., 2012), with a significant p-value < .05 and effect size threshold of 2. Finally, the Venn diagram was built to identify the core microbiota, as well as unique and shared OTUs between treatments, using Venny 2.1 software (http://bioin fogp.

| Data availability
The 16S rRNA gene raw sequence data are deposited and pub-

| Ethics statement
The

| Growth performance and feed efficiency
Growth performance results are summarized in Table 3

| High throughput sequencing results
The microbiota profile of P. vannamei intestine was analysed based on the sequencing of the 16S rRNA target gene, using Ion Torrent™ technology, in two different feed doses of YCW, that is, 0.5% and 1.0%, and a control group. Sequencing resulted in a total of 1,983,557 raw sequences. After removing singletons (OTUs observed fewer than two times) and Streptophyta phylum (chloroplast diet associated), 1,238,780 reads were qualified as high quality. Good´s estimator of coverage presented values above 0.992, demonstrating that almost the entire bacterial diversity was identified. The rarefaction curves revealed that a satisfactory sequencing coverage was achieved, with signs of saturation for all experimental groups ( Figure   S1). Table 4 summarizes the High Throughput Sequencing results.

| Relative abundance at phylum and genus levels
Among the five most abundant phyla in the taxonomic analysis  of shrimp that received 0.5% of YCW and in those from the control group, the latter group being statistically significantly higher than 1.0% treatment (p = 0.0029). The intestinal microbiota of animals that received 1.0% of YCW presented a high relative abundance of phyla Proteobacteria and Fusobacteria, being Proteobacteria relative abundance significantly higher than 0.5% and control groups (p = 0.0073). These two bacterial phyla comprised had relative abundances greater than 80% in all the three analysed groups, that is, 80%, 82% and 89% in 0.5% and 1% treatment, and control group, respectively. Finally, the relative abundance of Bacteroidetes was statistically lower in 1% treatment than in control group (p = 0.001).

| Similarities and dissimilarities
Concerning the similarities and dissimilarities of the bacterial population, the PCoA revealed a spatial separation between the categories, mainly between control and 1.0% treatment. Weighted UniFrac distance ( Figure 2a) showed that 1.0% treatment samples clustered all together, with one exception, in the opposite direction of all control samples, while 0.5% treatment samples were dispersed. Unweighted UniFrac distance (Figure 2b) displayed similar results, with spatial differentiation between control and 1.0% treatment, and with five of seven 0.5% samples clustering close to control.

| Linear discriminant analysis Effect Size.
Regarding possible distinct taxa with statistical significance and biological relevance, the linear discriminant analysis effect size (LEfSe) identified 25 distinct taxa in control group, three on 0.5% YCW treatment and 22 on 1.0% YCW treatment that could explain differences among treatments (Figure 3). The logarithmic LDA score revealed the effect size of each factor, with positive scores ranging between 2.0 and 5.0. The gut microbiota of 1.0% YCW treatment showed the highest LDA score (5.0) with Sphingobium genus, also notable on the relative abundance at the genus level. Further, this treatment showed the smallest LDA score, though always above 2.0.
Control group displayed the greatest quantity of distinct taxa, 25 in total, all above or close to 3.0 LDA score, highlighting Shewanella and Clostridium genera. Finally, 0.5% YCW treatment revealed only three distinct taxa, with LDA score between 3.0 and 4.0.

| Venn diagram and shared OTUs
In order to define the intersection list of OTUs between treatments, as the core gut microbiota of P. vannamei, as well as to determine the unique OTUs of each group, a Venn diagram was constructed, at the genus level ( Figure 4)

| DISCUSS ION
A high priority for contemporary aquaculture is to find efficient and safe technologies for prophylaxis and stimulation of the animal immune system with the associated mechanism of the gut mucosal interface and the role of the gut microbiota. This complex relationship All the growth performance and feed utilization parameters analysed (growth rate, feed conversion efficiency and survival) were deemed to be adequate for the established standards for shrimp farming. Currently, the standard growth parameters in Mexican shrimp farm are growth rate 1 g week −1, FCR 1.6 to 2.0, and survival rates 40% to 65%. Moreover, animals that received YCW inclusion, especially those that were fed with 1.0% YCW, presented higher overall performance than the control group. Both treatments groups fed diets containing YCW displayed significantly better survival compared to the control diet. Additionally, when analysing the growth parameters, that is, the growth rate and the final weight, animals that received 1.0% YCW feed showed improved performance than those that received 0.5% YCW or from the control group. It is recognized that the dose and the frequency that various prebiotics and immunostimulants are administrated can notably influence animal response , and a better performance in our investigation was observed in animals fed with 1.0% YCW.
It is worth noting that when doing evaluations under commercial conditions, ideal results are rare. Moreover, the economic relevance should be the main objective. In the trial described in this article, the survival rates and animal performance were significantly superior compared to the control group and the farm inventory record F I G U R E 3 Distinct enriched taxa in the gut microbiota of Penaeus vannamei, with two different doses of Yeast Cell Wall (YCW), that is, 0.5% and 1.0%, and control group, at genus level. (a) LDA score indicating the scale of difference among taxa. (b) Relative abundance of the five most abundant bacteria, at genus level, in order to support LefSe results. p < 0.05; LDA threshold of 2 logs. Commercial aquaculture is a multi-diverse activity where each farm looks for profit or cost benefit rather than simply performance, adjusting the parameters according to their market, carrying capacity and productivity. Additionally, in regard to shrimp survival and compensatory growth, a typical scenario is when the shrimp population reduces, shrimp size increases. This is related to carrying capacity. At our site, typical capacity was 600 g m −3 . During our feeding trial, final biomass in each tank was maintained above the capacity (control group 624 g m 3 ; 0.5% YCW 965 g m −3 ; 1.0% YCW 978 g m −3 ). Our hypothesis in this regard is that the immune robustness, due to inclusion of the tested product, aids the shrimp to maintain adequate survival in both treatment groups, but not in the control group. Additionally, shrimp survivals are widely interpreted from the technical point of view, that is, economic standards and health management. The farm records survival rates fluctuate from 40 to 65% (exempting the trial, survivals range from 40% to 70%), which is considered to be a good trend. Our goal in this experiment was to link the performance results with the inclusion of the functional feed additive and the gut microbiota profile. Thus, we aimed at building a scientific case based on the hypothesis that beneficial bacterial can increase stress tolerance as well nutrition absorption increasing the robustness of the shrimp.
Regarding the growth parameters, the superior performance observed for shrimps that received the YCW may have resulted from the direct induction of the immune system by the polysaccharide-rich feed, that is, the β-glucan from the yeast cell wall along with other constituents such as proteoglycans and mannan-oligosaccharides (MOS). Particularly, in our study, the natural compounds found within the yeast cell wall matrix appeared to maximize the shrimp production in terms of weight gain, feed efficiency and survival.
It should be mentioned that Sajeevan et al., (2009)  The addition of 1.0% YCW was the treatment that had the most influence on the shrimp intestinal microbiota composition. This finding was also corroborated with the PCoA results that revealed a distinct separation and dissimilarity between 1.0% YCW treatment and control group fed shrimp. Animals that received this inclusion level of the commercial YCW product presented the phylum Proteobacteria as the most prevalent constituent. This phylum, in fact, has been reported by other studies as the most dominant in shrimp (Li et al., 2018;Xiong et al., 2017). Meanwhile, the Fusobacteria phylum was the most prevalent in the digestive tract among the animals that received 0.5% YCW and those from the control group. This phylum has also been described as one of the most ubiquitous phyla

F I G U R E 4 Venn diagram showing unique and shared OTUs
(Operational Taxonomic Units) in the gut microbiome of Penaeus vannamei, with two different doses of Yeast Cell Wall (YCW), that is, 0.5% and 1.0%, and control group diets, at genus level in the intestine of the Pacific white shrimp (Souza Valente et al., 2020). In fact, these two phyla, among others, were described to be part of the autochthonous gut microbiota of P. vannamei, since larval until adult stage (Zeng et al., 2017). Moreover, these two phyla were reported to be upregulated by WSSV infection (Wang et al., 2019). Luis-Villaseñor et al., (2013) addressed that the phyla Proteobacteria, Fusobacteria, Sphingobacteria and Flavobacteria were the most prevalent in the gut of P. vannamei after being fed with a Bacillus probiotic mix.
Those findings can also be observed in the most dominant microbial genera. The genus Cetobacterium was the most predominant in all treatments, although significantly lower in 1% YCW. This genus, and specifically C. somerae species, is related to the production of vitamin B 12 (cobalamin) in fish (Rodiles et al., 2018;Tsuchiya et al., 2008). In Chinese crabs, cobalamin is associated with the nonspecific immune responses (Wei et al., 2014). In shrimp, vitamin B 12 is commonly supplemented as a form of cyanocobalamin, in optimal doses of 0.1 -0.2 mg kg −1 in complete diets (Koshio, 2014). Although there is recent research showing the relation between Cetobacterium and cobalamin in fish, there is a lack of studies on this bacterium and its role in shrimp. As the most prevalent genera in the three analysed group of the present study, certainly this genus is worthy of future attention.
Similarly, the genus Sphingobium was higher in the 1.0% YCW group in comparison to the control group, indeed reaching a relevant percentage of relative abundance for that treatment.
Sphingobium was isolated from fresh and treated water (Corre et al., 2019;Sheu et al.,2013) and, as some bacteria from this genus may degrade polycyclic aromatic hydrocarbons, they can be used for soil bioremediation (Chen et al., 2016). In the LEfSe analysis, this genus was enriched and presented the biggest LDA score, showing that this genus had great relevance, and could explain part of differences between this treatment and the other groups. It is a relatively new described genus, first proposed by Takeuchi et al., (2001) and, at the moment, it is only mentioned to be part of some intestinal shrimp microbiota (Hu et al., 2017), but its relevance is rarely discussed. The genus counts as the 40 most common taxa found in the arthropod gut microbiota, from soil and the aquatic environment (Esposti & Romero, 2017). It was isolated from the rhizosphere of an aquaponics system (Schmautz et al., 2017) and associated with an antibiotic resistance (glycopeptide resistance gene) in an experimental aquaculture facility (Colombo et al., 2016). The role of this genus on the gut microbiota of shrimp from the present study remains unclear.
Noteworthy, the inclusion of YCW within the experimental shrimp diets, both at 0.5%, and 1.0%, significantly increased the relative abundance of the genus Bacillus in the gut microbiota. This result is remarkable due to the significant probiotic importance of this genus. Bacillus is considered as an autochthonous member of crustacean's environment and is among the widely used probiotic bacteria for crustaceans (Castex et al., 2014), mainly due to its capacity to activate both cellular and humoral shrimp immune responses (Rengpipat et al., 2000) and to naturally produce antibiotic compounds (Van Hai & Fotedar, 2009 Exiguobacterium has been proposed to be a potential probiotic for P. vannamei (Cong et al., 2017) and may increase shrimp survival and growth (Sombatjinda et al., 2014). Moreover, this genus has potential biotechnological use to industry and agriculture (Kasana & Pandey, 2018). Equally, although Vibrio genus also encompasses some opportunistic pathogens, others may act as probiotics for crustaceans (Castex et al., 2014), such as Vibrio alginolyticus (Austin et al., 1995). Furthermore, Vibrio may play a relevant role in enhancing shrimp digestion, due to its genes related to digestive enzymes (Gao et al., 2019). Therefore, shrimp aquafeed with YCW not only preserved relevant bacteria genera in the gut microbiota but also selected for two other beneficial ones (i.e. Exiguobacterium and Vibrio).
Lastly, even with changes observed in the gut microbiota composition due to the addition of YCW especially on 1.0% treatment that showed dissimilarity to the control group, we noticed a stable core microbiota. A stable and permanent bacterial community was preserved, regardless the addition of YCW in the shrimp diets tested in this study. The core microbiota, composed by LAB and recognizable or promising probiotic strains in this study, is intimately associated with healthy and diseased animals, being paramount on the host-bacteria interaction. A healthy shrimp gut microbiota is characterized by a high diversity with cooperative interactions, while diseased animals tend to have less diversity and simple gut microbiota (Yao et al., 2018). Moreover, when elucidating the core microbiota composition, it is easier to manipulate it in order to develop effective strategies to promote animal health and growth (Steinberg, 2018). Thus, the preservation of a healthy and balanced gut microbiota, as observed in the inclusion of YCW, may result in more resilient and stronger shrimp, likely to better respond to stressful situations.
In conclusion, the present study has implied that the dietary inclusion of a yeast cell wall extract complex (PAQ-Gro™ Phibro, Teaneck, USA) resulted in beneficial attributes for shrimp farming.
Particularly, for juvenile P. vannamei raised on an intensive floating cage system. YCW can be supplied to shrimp as a prophylactic agent in order to promote animal performance and gut health, especially at a 1.0% inclusion level in a formulated diet.
The product we evaluated (PAQ-Gro™ Phibro, Teaneck, USA) is not a pure β-glucan but also contains other important constituents of the yeast cell wall such as MOS and proteoglycans and enzymes from fungal sources that could influence the gut health but confounding our understanding of the direct causative effect. As a complex bioactive agent, it may maximize shrimp yield farming profitability, likely providing economic advantages to the producer.
Future studies should also be directed to assessing the cost benefit analyses of YCW under various production scenarios and within the context of disease challenge studies with specific pathogenic agents encountered in the shrimp farming industry (e.g. White Spot Disease and Acute Hepatopancreatic Necrosis Disease). These collectively offer much scope for the future of applying such yeast cell wall products as feed supplements in formulated diets for shrimp to achieve superior health and resilience. Further work should be directed towards metagenomic techniques to establish the functionality of the various microbial taxa and specific roles of commensal bacteria in relation to the shrimp immune competence.
Additionally, it was shown that supplementation of a propriety shrimp feed with a functional feed additive under practical farming conditions resulted in positive outcomes for production. This investigation will therefore allow us to better understand their applications under varying environmental and rearing conditions for more effective control and use of prebiotic type additives for attaining superior health status. This will be transformational to achieve a more sustainable industry and meet with consumer expectations for a superior quality end product.

| DATA S TOR AG E AND D O CU M E NTATI O N
All data that support the findings presented in the manuscript are available within the manuscript and supplemental material.