Limosilactobacillus fermentum SWP-AFFS02 Improves the Growth and Survival Rate of White Shrimp via Regulating Immunity and Intestinal Microbiota

: White shrimp Litopenaeus vannamei is an important species of farmed shrimp. Intestinal bacterial composition and immune activity play important roles in regulating the health condi-tion of shrimp. Lactic acid bacteria Limosilactobacillus fermentum SWP-AFFS02 was isolated from the intestine of sea ﬁsh Rachycentron canadum , and the potential of its effect on growth, immunity, and intestinal microbiota of L. vannamei shrimp was investigated. Shrimps received feed with or without the addition of 8 log CFU/g L. fermentum SWP-AFFS02 thrice a day for 8 weeks. After 8-week treatment, weight gain, feed conversion rate, and survival rate of shrimp were greater in the L. fermentum SWP-AFFS02-feed group than in the control group. L. fermentum SWP-AFFS02 treatment increased the number of granular cells and semi-granular cells and decreased hyaline cell number when compared to the control group. L. fermentum SWP-AFFS02 promoted prophenoloxidase (PO) activity through increasing immune-associated gene expression in the hepatopancreas of shrimp. In addition, administration of feed containing L. fermentum SWP-AFFS02 regulated intestinal microbiota via decreasing the ratio of pathogenic bacteria, such as Vibrionaceae and Enterobacteriaceae , in the intestine of shrimp. This study demonstrated that administration of L. fermentum SWP-AFFS02 effectively prevented infection of L. vannamei shrimp by regulating intestinal microbiota and enhancing immunity in shrimp to increase the growth and improve their health status, which acted as a probiotic and provided beneﬁcial effects on shrimp.


Introduction
Litopenaeus vannamei, commonly known as white shrimp, is native to Central and South America, which has strong tolerance to the environment, disease resistance, and osmotic pressure regulation in low-salinity waters. It is characterized by its feed conversion rate, fast growth rate, low residual food rate, and suitability for high breeding density, which have made it an internationally important species of farmed shrimp. Developing an intensive farming model to enable an increase in stocking density per unit area has caused problems, such as water contamination. Chemicals, growth of pathogenic bacteria,

Husbandry Conditions and Feeding Trial
Shrimps used for this study were collected from an aquafarm at Kaohsiung, Taiwan. A total of 180 L. vannamei shrimps (initial body weight: 8.22 ± 0.38 g and 8.27 ± 0.53 g in blank and L. fermentum SWP-AFFS02 feed, respectively) were evenly distributed into 3 groups (60 shrimps/group), and they were stocked in a fiber reinforced plastics (FRP) bucket with a flow-through sea water system and continuous aeration. The indoor cement pool was maintained under natural light/dark regime. During the experimental period, the monitored water quality parameters (mean ± S.D.) were as follows: water temperature was 28.3 ± 2.4 • C, pH was 8.4 ± 0.5, and the salinity was 33.6 ± 0.2 ppt. All shrimp groups were fed feed with or without LAB to apparent satiation thrice a day (at 6:00, 17:00, and 22:00) for 8 weeks. The remaining feed residues after feeding were removed by siphoning and dried by using a freeze drier system to subtract from total feed intake. Body weight was also recorded once every two weeks.

Detection of Total Bacteria and Vibrio spp. Counts in Seawater Surrounding
The 1 mL of aquaculture breeding water was diluted to the appropriate ratio by a 10fold serial dilution method, and the 100 µL dilution solutions were smeared on 3% tryptone soy agar (TSA) plate medium (BD, Franklin Lakes, NJ, USA) and thiosulfate-citrate-bile salts-sucrose (TCBS) agar (Sigma-Aldrich, St. Louis, MO, USA) for total bacteria and Vibrio spp. cultivation, respectively. After cultivation for 24 h at 30 • C, the bacterial numbers between 25-250 colonies were counted [12]. The 0.8 mL of hemolymph was withdrawn from the ventral sinus in the first abdominal segment using a 26-gauge hypodermic needle. Each syringe was pre-filled with 0.2 mL of anticoagulant (10 mM Tris-HCl, 250 mM sucrose, 100 mM sodium citrate, pH 7.6). More anticoagulant was added to make an equal volume ratio of hemolymph to anticoagulant. The hemolymph samples of 180 L. vannamei shrimps (60 shrimps/group) from each treatment were analyzed individually [13].

Total Hemocyte Count (THC) and Differential Hemocyte Count (DHC)
A volume of 50 µL anticoagulated hemolymph was fixed with an equal volume of neutral buffered formalin (10%) for 30 min to measure the THC. Fixed hemolymph was smeared on a slide and stained with Giemsa solution (10%) for 10 min. The DHC was then characterized according to Tsing et al. (1989) [14], and 250-300 cells from each smear were counted. The remaining anticoagulated hemolymph was centrifuged at 300× g for 10 min at 4 • C to separate the hemocytes from plasma. Hemocytes were suspended and adjusted to 5 × 10 6 cells/mL in ice-cold cacodylate buffer (0.01 M sodium cacodylate, 0.45 M NaCl, 10 mM CaCl 2 , 26 mM MgCl 2 , pH 7.0). A volume of 2.5 mL (5 × 10 6 cells/mL) hemocyte suspension was respectively separated for superoxide anion and phenoloxidase activity [13].

Measurement of Superoxide Anion Respiratory Burst
According to recent study [15], after adding 100 µL (5 × 10 6 cells/mL) of diluted hemolymph to the wells of a 96-well microplate, the plate was centrifuged at 800× g at 4 • C for 20 min. The supernatant was discarded, and the pellets were rinsed with MCHBSS buffer (Sigma-Aldrich, St. Louis, MO, USA). Subsequently, he mocytes were mixed with 100 µL of DCFH-DA (Sigma-Aldrich, St. Louis, MO, USA) and zymosan (1 mg/mL) (Sigma-Aldrich, St. Louis, MO, USA) in the 96-well plate. After a 30-min incubation in the dark, cells were washed twice with MCHBSS buffer. The supernatant was discarded 30 min post-incubation and the pellets were rinsed once using MCHBSS buffer. Finally, the fluorescent intensity was measured using a spectrofluorometer with an excitation wavelength of 480 nm and an emission wavelength of 523 nm.

Assay for Phenoloxidase (PO) Activity
The PO activity was assayed, as described by Sung et al. (1994) [16]. using L-3,4dihydroxyphenyl-alanine (L-dopa) (Sigma-Aldrich) as a substrate. To obtain lysates from hemolymph (200 µL; 5 × 10 6 cells/mL), frozen hemocytes were thawed and frozen several times. Hemocyte lysate supernatants (HLS) and plasma were assayed for PO activity using a spectrophotometer (Hitachi U-2000) to measure the OD 490 nm . The concentration of total plasma protein was determined by the Bradford method [17] using bovine serum albumin (Bio-Rad Protein assay Kit II) as a standard. One unit of PO activity was defined as an increase in absorbance of 0.001 min/mg protein [18].

Microbiota Sequencing
Each intestine of shrimp was collected and immediately soaked in liquid nitrogen and stored at −80 • C for subsequent use. Total genomic DNA from samples was extracted using QIAamp PowerFecal DNA Kit (Qiagen) for 16S rRNA gene sequencing. V3-V4 regions (515F-806R) were amplified, and PCR reactions were carried out with Phusion ® High-Fidelity PCR Master Mix (New England Biolabs). The sequencing library was prepared and sequenced on an Illumina MiSeq platform.

Statistical Analysis
Experimental results were analyzed in five repeats and expressed as the mean ± standard error of mean. The results were examined by using one-way analysis of variance (ANOVA) and Duncan's multiple range tests, and the significance of differences between sample means was calculated. A p value ≤ 0.05 was considered significant.

Variation of LAB Feed-Contained L. fermentum SWP-AFFS02 after Stroage
L. fermentum SWP-AFFS02 was collected after cultivation for 48 h (approximately 10 9 CFU/mL) in MRS medium at 37 • C. The standard feed (sterilized) was mixed with a culture solution or deionized water (1/1: w/v) under laminar airflow and stored at 4 • C or 25 • C for 7 days after freeze-drying and vacuum-packaging. Finally, the LAB numbers in the L. fermentum SWP-AFFS02-rich feed were measured and calculated on MRS agar under anaerobic conditions. As shown in Table 1, the bacterial counts were 4.47 × 10 8 CFU/g on day 0. However, a decrease in bacterial levels was observed with increasing storage days (during days 0-7) in both 25 • C and 4 • C treatments. However, maintaining a low temperature could effectively retain the number of viable bacteria. The initial body weights of L. vannamei shrimp were 8.22 and 8.27 g in the control and LAB feed groups, respectively. The weight gain (WG), feed conversion rate (FCR), and survival of L. vannamei shrimp in the two groups were evaluated after 8 weeks. We found that the final weight of L. vannamei shrimp in the LAB group was higher than that in the control group. WG, FCR, and survival rate were greater in the LAB feed group than in the control group (Table 2).

Investigation for Immunity Index
Total hemocyte count (THC) is a marker to indicate shrimp immunity, and the THC ratio approximately reaching 2.0-4.0 × 10 7 cells/mL was reported [22]. Another study demonstrated that THC levels were affected by internal factors (including sex, shelling cycle, age, food intake, season, and body temperature), as well as external environmental factors (including pollutants, temperature, pH, salinity, and dissolved oxygen) in crustaceans [23]. Moreover, the THC level was markedly reduced in shrimp, thereby leading to infections caused by pathogenic bacteria or viruses [24]. We found that the LAB feed markedly elevated THC levels in shrimp when compared to the control group ( Figure 1A). SWP-AFFS02 could be developed as a probiotic to improve immunity and growth in L. vannamei shrimp, thereby elevating FCR efficacy.
The lipopolysaccharide and beta-1,3-glucan binding protein (LGBP) and the cascade via PO-activating factor (PPAF) both activate the PO system, thereby promoting immune function in shrimp [28]. In addition, heat-shock protein 70 (HSP70) and glutathione peroxidase (GPx) are defense agents associated with antioxidant activity against stress [29]. The LAB feed containing L. fermentum SWP-AAS02 significantly increased LGBP and PPAF expression in the hepatopancreas of L. vannamei shrimp ( Figure 3A, B), but the levels of HSP70 and GPx were not elevated by LAB feed administration ( Figure  3C, D). These results indicate that L. fermentum SWP-AFFS02 markedly elevated immune activity in L. vannamei shrimp.   Differential hemocyte count (DHC) was divided into three types of immunocytes, including granular cells, semi-granular cells, and hyaline cells, and these cells constituted THC. In addition, DHC plays different roles in regulating immune function. Granular cells and semi-granular cells exert cytotoxicity and activate the prophenoloxidase system (PO system), but hyaline cells do not activate the PO system. Phagocytosis and respiratory burst (production of superoxide anions) are the major immune abilities of hyaline cells [25]. We found that the LAB feed markedly increased the number of granular cells ( Figure 1B) and semi-granular cells ( Figure 1C) after 8 weeks of administration. However, the number of hyaline cells treated with the LAB feed was reduced both in week 4 and week 8, compared to the control group ( Figure 1D).
Respiratory burst was monitored to identify the body defense ability mediated by superoxide anion production; respiratory burst activation has been found to improve the shrimp defense system [26]. The results indicated that LAB feed administration did not promote respiratory burst (Figure 2A), which may be because the LAB feed could not promote hyaline cells. The activation of the PO system led to quinonoid production mediated by tyrosine oxidation, and quinonoid compounds exerted cytotoxicity [27]. As shown in Figure 2B, the LAB feed markedly promoted the PO activity of L. vannamei shrimp in both week 4 and week 8. These results demonstrate that L. fermentum SWP-AFFS02 could be developed as a probiotic to improve immunity and growth in L. vannamei shrimp, thereby elevating FCR efficacy.

Figure 1. The variations of THC (A), granular cells (B), semi-granular cells (C), and hyaline cells (D)
in shrimp feed with control or LAB feed. Significant difference is shown by different letters (a, b) (p < 0.05) between control and LAB feed groups.

Protection of Intestinal Microbiota in L. Vannamei Shrimp by LAB Feed Containing L. Fermentum SWP-AFFS02
There are 391 different types of bacteria within the intestinal stool of L. vannamei shrimp fed a control feed or LAB feed. However, the operational taxonomic unit (OTU) numbers were 126 and 201 in the intestinal stool of L. vannamei shrimp fed a control feed or a LAB feed, respectively ( Figure 4A). These results suggested that the LAB feed containing L. fermentum SWP-AFFS02 may potentially regulate microbial beta diversity in The lipopolysaccharide and beta-1,3-glucan binding protein (LGBP) and the cascade via PO-activating factor (PPAF) both activate the PO system, thereby promoting immune function in shrimp [28]. In addition, heat-shock protein 70 (HSP70) and glutathione peroxidase (GPx) are defense agents associated with antioxidant activity against stress [29]. The LAB feed containing L. fermentum SWP-AAS02 significantly increased LGBP and PPAF expression in the hepatopancreas of L. vannamei shrimp ( Figure 3A,B), but the levels of HSP70 and GPx were not elevated by LAB feed administration ( Figure 3C,D). These results indicate that L. fermentum SWP-AFFS02 markedly elevated immune activity in L. vannamei shrimp.    Figure 4A). These results suggested that the LAB feed containing L. fermentum SWP-AFFS02 may potentially regulate microbial beta diversity in the intestine, compared to the control group, as shown by PCA. Furthermore, the microbial beta-diversity of seawater was different from that of the stools of L. vannamei shrimp fed with control and LAB feed ( Figure 4B).   Figure 5B) in seawater were higher than in the stool of L. vannamei shrimp fed with control or LAB feed; however, other bacteria in seawater were markedly lower than those in the stool of L. vannamei shrimp, suggesting that some mechanisms may regulate special microbes remaining in the shrimp intestine, We analyzed the most abundant microbes (top 35) at the genus and family levels in the stool of L. vannamei shrimp fed a control feed or a LAB feed. The results indicated that Pseudomonas, Bacillus, Idiomarina, NS5_marine_group, Aliidiomarina, and Alcanivorax at genus level ( Figure 5A) and Flavobacteriaceae, Pseudomonadaceae, Bacillaceae, Idiomarinaceae, Rhodobacteraceae, Legionellaceae, Halieaceae, Alcanivoracaceae, Paenibacillaceae, Halomonadaceae, and Shingomonadaceae at family level ( Figure 5B) in seawater were higher than in the stool of L. vannamei shrimp fed with control or LAB feed; however, other bacteria in seawater were markedly lower than those in the stool of L. vannamei shrimp, suggesting that some mechanisms may regulate special microbes remaining in the shrimp intestine, compared to seawater bacteria staying in shrimp. In addition, the levels of Paraprevotella, Diallister, Oscillibacter, Lachnoclostridium, and Sutterella at the genus level were elevated by LAB feed-containing L. fermentum SWP-AFFS02, compared to the control group; however, the levels of Acinetobacter, Ruminococcaceae_UCG_002, and Escherichia_Shigella were reduced by LAB feed-containing L. fermentum SWP-AFFS02, compared to the control group. Moreover, we found that the levels of Vibrionaceae, Corynebacteriaceae, Marinifilaceae, Tannerellaceae, and Enterobacteriaceae were suppressed by the LAB feed. These results indicate that the LAB feed potentially regulates the intestinal microbiota to improve host health in L. vannamei shrimp.
Moreover, short-chain fatty acids, including acetic acid, propionic acid, and butyric acid in the stools of L. vannamei shrimp fed the LAB feed were measured, as shown in Table 5. The results showed that acetic acid and butyric acid were elevated in the stools of L. vannamei shrimp fed the LAB feed, compared to the control group. Table 5. The level of intestinal short chain fatty acid in L. vannamei shrimp fed with LAB feedcontained L. fermentum SWP-AFFS02.

Short-chain Fatty Acids
Acetic Acid Propionic Acid Butyric Acid

Discussion
A study has found that a Bacillus subtilis-fermented soybean meal could improve intestinal microbiota and disease prevention in shrimp [30]. Xie et al., (2017) indicated that piglets fed feed fermented by Lactobacillus, Bacillus, and Saccharomyces cerevisiae potentially improved growth performance [31]. In a similar study in fish, Lactobacillus and Pediococcus were abundant bacteria in the intestines of Atlantic salmon fed the fermented feed, and the expression of mucin 2 (muc2) and aquaporin 8ab (aqp8ab) significantly increased in the fermented feed group, compared to the control group [32].
Vibrio spp. are widely distributed in marine environments and are among the most abundant flora in shrimp digestive systems, and vibriosis is the most serious disease in

Discussion
A study has found that a Bacillus subtilis-fermented soybean meal could improve intestinal microbiota and disease prevention in shrimp [30]. Xie et al. (2017) indicated that piglets fed feed fermented by Lactobacillus, Bacillus, and Saccharomyces cerevisiae potentially improved growth performance [31]. In a similar study in fish, Lactobacillus and Pediococcus were abundant bacteria in the intestines of Atlantic salmon fed the fermented feed, and the expression of mucin 2 (muc2) and aquaporin 8ab (aqp8ab) significantly increased in the fermented feed group, compared to the control group [32].
Vibrio spp. are widely distributed in marine environments and are among the most abundant flora in shrimp digestive systems, and vibriosis is the most serious disease in worldwide aquaculture that causes shrimp death [33,34]. Huang et al. (2016) demonstrated that pathogenic Photobacterium and Vibrio were the predominant bacteria at the post-larval stage (80%) and juvenile stage (89.1%~94.2%), resulting in shrimp infection [35]. Similar to this study, Vibrio spp. were predominantly found in the intestines of shrimp, but these pathogenic bacteria were reduced in shrimp fed a synbiotic feed [34], thereby reducing disease arising from Vibrio infection by promoting immunity [36,37], as well as by avoiding hepatopancreatic damage [38]. A previous study suggested that activation of the antioxidant system avoided these negative effects caused by Vibrio infection in shrimp [39]. The immune and antioxidant systems are two primary physiological mechanisms that protect aquatic animal health against Vibrio spp. infection [40]. Non-specific immunity is the major defense system for stress resistance and protection against pathogen infection in crustaceans, and the PO system is also one of the primary response pathways in shrimp [41]. PPAF and LGBP are the key regulators of the immune response in shrimp [28]. In our study, we found that the LAB feed containing L. fermentum SWP-AFFS02 potentially suppressed intestinal Vibrionaceae in shrimp ( Figure 5B) and promoted immune-associated gene expression in the hepatopancreas of L. vannamei shrimp, including LGBP and PPAF ( Figure 3A,B). Acinetobacter spp. is a pathogenic bacterium of L. vannamei shrimp [42]. Corynebacterium is a gram-positive (non-spore forming) and aerobic bacterium that opportunistically reduces immunity and defense systems in hosts associated with systemic infection [43]. Escherichia_Shigella (Enterobacteriaceae family) is a pathogenic bacterium that causes gut leakiness and infection [44]. We found that administration of LAB feed that containing L. fermentum SWP-AFFS02 markedly inhibited these pathogenic intestinal bacteria in L. vannamei shrimp.
The animal intestine plays an important role in nutrient absorption and metabolism and acts as the first line of defense against pathogen infection and environmental stress [45]. Intestinal bacteria and bacterial metabolites directly influence the physiology of the host, including absorption and metabolism, intestinal barrier integrity, and immune function [46]. We believe that the LAB feed containing L. fermentum SWP-AFFS02 could limit pathogenic bacteria in the intestine of L. vannamei shrimp, thereby improving growth, FCR, and survival (Table 2).
In a recent study, feeding L. vannamei with feed supplemented with bacteria (Bacillus subtilis and L. rhamnosus) enhanced the antioxidative status by increasing the activity of antioxidant enzymes in shrimp [47]. Thus, shrimp feed with bio-friendly agents. such as Lactobacillus spp., could be used as effective alternatives to antibiotics for treating bacterial infections, especially V. harveyi infection in shrimp aquaculture [48]. Pooljun et al. (2020) confirmed that, after feeding juvenile shrimp (Penaeus vannamei) with L. acidophilus and Saccharomyces cerevisiae mixture (1:1, at 10 8 and 10 9 CFU/kg feed), the adverse effects of acute hepatopancreas necrosis disease induced by V. parahaemolyticus infection were ameliorated by improving hemocyte parameters, including the THC, granular hemocyte percentage, and phenoloxidase activity, as well as by highly upregulating hemocyte genes (crustin and penaeidin), which lead to high shrimp survival rate [49].

Conclusions
This study aimed to evaluate the effects of feeding L. vannamei shrimp with a feed rich in the LAB L. fermentum SWP-AFFS02. The effect of L. fermentum SWP-AFFS02 on the immunity and microbiota, as well as the changes in size of white shrimp, were investigated. Our study makes a significant contribution to the literature due to L. vannamei being an economically important species of shrimp, which are often plagued by pathogenic bacteria that cause their death. We noticed a marked inhibition of pathogenic bacteria in the shrimp intestine, which could in turn promote growth and improve immunity in L. vannamei. This study provided useful insights for the use of L. fermentum SWP-AFFS02 in shrimp. Enhancing the health status of cultured shrimp by using beneficial microbes as a probiotic is a good approach to control pathogens.  Ethical Approval: It is not required to get an 'Affidavit of Approval of Animal use Protocol' when using invertebrates, including shrimp, as an experimental animal in Taiwan.

Institutional Review Board Statement: Not applicable.
Data Availability Statement: Not applicable.