Effects of Dietary Soy Isoflavone and Soy Saponin on Growth Performance, Intestinal Structure, Intestinal Immunity and Gut Microbiota Community on Rice Field Eel (Monopterus Albus)


 Background: Soy isoflavone and soy saponin are heat-stable glycosides present in soybean meal, which is the major factor restrict soy protein using in aquatic feed. This study is aimed to conduct that dietary soy isoflavone and soy saponin on growth performance, intestinal structure, intestinal immunity and gut microbiota community on rice field eel (Monopterus albus), and revealed its mechanism.Results: In current study, dietary 1g/kg soy isoflavone, 2g/kg soy saponin and their mixed feed decreased growth of M. albus. Dietary 1g/kg soy isoflavone significantly increased serum alkaline phosphatase (P < 0.05), dietary 2g/kg soy saponin remarkably declined intestinal immunoglobulin M (P < 0.05). Dietary 1g/kg soy isoflavone and 2g/kg soy saponin mixed feed significantly decreased intestinal complement 3, complement 4 and immunoglobulin M, amounts of intestinal goblet cells per root and shannon of gut microbiota (P < 0.05). Dietary 1g/kg soy isoflavone, 2g/kg soy saponin and their mixed feed down regulated intestinal tight junction protein (Interleukin-12, Interleukin-15, Tight junction protein ZO-2-like) and anti-inflammatory (Interleukin-10, Transforming growth factor beta 2) mRNA expression, up regulated pri-inflammatory factors (Interleukin-1 β, Interleukin-6, Interleukin-8, Interleukin-12 and Interleukin-15) mRNA expression.Conclusion: Based on the current results, dietary soy isoflavone and soy saponin could damage the intestinal structure and changes gut bacteria community of M. albus, and destroyed the intestinal barrier, eventually induced intestinal inflammatory occurring, soy isoflavone and soy saponin mixed feed has more serious effect than they single supplemented.


Introduction
Fish meal (FM) is a high-quality protein source for aquatic feed because of its balance of amino acids, high fatty acid content and low carbohydrate content. With the continuous development of aquaculture industry in recent years, the demand for FM has increased rapidly. At the same time, due to climate change and with the increase of global shing, the global sh resources have declined sharply and the prices have soared. Nevertheless, the probability of hunting for wild marine sh is gradually di cult, conventionally used proteins has considered to be no longer sustainable from both economic and ecologic points of reality view [1] .
Soybean protein is a kind of high-quality plant protein, it's also the largest production plant protein in the world, and its short growth cycle. Recently, the aquatic feed industry rapidly expanded, soybean protein is widely used to replace FM in aquatic feed [2] . However, various anti-nutritional factors like trypsin inhibitors, antivitamins, lectins, soy antigens, iso avone, saponins, phytic acid, phytoestrogens and oligosaccharides are major shortcoming of soybean products used in aquatic feeds [3] . Lectins, trypsin inhibitors and antivitamins are heat labile factors could be destroyed by high temperature and extruded treatment, and phytates and antigenic compounds in aquatic feeds containing commercially available soybean protein sources did not affect growth performance and immunity of sh [4] . The reasons is the presence of some anti-nutritional factors (ANFs) in variable amounts in the raw soybean grain.
The ANFs in soybean mainly consist of heat labile and heat stable factors [5] . The process of extruding for soybean is acknowledged to be very useful to remove heat labile ANFs, however, there are thermal stability ANFs in extruded soybean meal has been reported that not only has the negative effect on growth performance, but also damage the intestinal function [6,7] .
Soy iso avone are heat-stable glycosides present in soybean meal, also widely known to exert oestrogen like biological effects on animals because of the similarity of their structure to the natural oestrogen structure [8] . There has many physiological function about soy iso avone, especially for immune system regulation [9] , include effects of antioxidant, anti-in ammatory on white shrimp (Litopenaeus vannamei) [10] , and improve the immune response and against Aeromonas hydrophila challenge on juvenile grass carp (Ctenopharyngodon idella) [11] . However, soy iso avone affected growth performance and altered the physiological processes of on Atlantic salmon (Salmo salar L.) [12] . Obviously, effect of soy iso avone aquatic animals is positive or negative? It's hard to say.
Soy saponin is the main heat-stable glycosides present in soybean meal, even extruded soybean meal [13] , and it's also a kind of membrane permeabilising activity [14] . Soy saponin is considered to have negative effects for feed intake, weight gain, ability uptake of nutrients including minerals and vitamins in intestine, and declining the digestibility of protein [6,7,15] . Generally, soya saponin could increase the intestinal epithelial permeability, made the intestinal barrier weaker, and be considered to be the factor responsible for distal intestine enteritis induced by soybean meal used in Atlantic salmon (Salmo salar) [15] , juvenile Japanese ounder (Paralichthys olivaceus) [7] and Zebra sh (Brachydanio rerio var ) [16] . Soy saponin may the main factor to restrict the capacity of using soybean meal in aquatic feed.
Rice eld eel (Monopterus albus, M. albus) is subtropical freshwater benthic sh, mostly inhabit ponds, paddy elds, ditches and other water bodies, cave dwelling, day and night emergence, can endure low oxygen, feeding way for phagocytosis and swallowing, mainly by the way of phagocytosis [17] . M. albus is an economic species in central and southern of China, and popularly cage raised owing to its farming scale and commercial value [18] . M. albus requires high quality of protein level in diet referenced our previous studies [19,20] . Our published study reported that soybean meal [20] and soy protein concentrate [21] replaced FM decreased the growth performance on M. albus, and soy protein concentrate was better than soybean meal for M. albus. The digestive tract of M. albus is a long and coiled tube, and divided into four parts: mucosa, lamina propria-submucosa, muscularis and serosa. What's more, histology and cellular structures of epithelia and mucosa of the digestive tract of M. albus is alike other carnivorous sh [22] . Based on these characteristics of M. albus, in current study, we aimed to study the effects of soy iso avone and soy saponin on intestinal structure and immunity in M. albus.

Ingredients and experimental feeds
Six isonitrogenous and isolipidic experimental feeds were made. A normal sh meal diet (550 g/kg sh meal), a soybean meal diet (286 g/kg sh meal; 372 g/kg soybean meal) and a soy protein concentrate diet (286 g/kg sh meal; 226 g/kg soy protein concentrate) were set up according to our previous article [21] . Soy iso avone (40%) and soy saponin (≥98%) and their mixed were supplemented with 2.5 g / kg and 2 g / kg in soy protein concentrate diet, respectively, purchased from Shanghai yuanye Bio-Technology Co.,Ltd. The levels of soy iso avone and soy saponin were correspond to the soybean meal from previous study [23] , we have determined the level of soy iso avone and soy saponin of soybean meal before making the experimental feeds. The composition of these diets and nutrition level were showed in Table 1. After crushing and sieving through 80 meshes, the feed material is evenly mixed according to the experimental formula. The feed material by air-dried at 25℃ until the content of moisture is less than 10%. Then it is stored in a refrigerator at -20℃ for reserve. Water was added before feeding and using as dough.  [24] .
Feed samples analyzed according to AOAC [25] . Moisture was measured by drying to a constant weight at 105 °C until it achieved a constant weight. Crude protein (N×6.25) was analyzed by the Kjeldahl method with an auto digestor (FOSS, Tecator, Hoganos, Sweden) after acid digestion. Crude lipid was determined by petroleum ether using the Soxtec System HT (Soxtec System HT6, Tecator, Sweden).
Ash determined by a mu e furnace at 550 °C until complete ashing. Soy iso avone and soy saponin of experimental diets were measured by high performance liquid chromatography (Agilent 1260, Agilent Technologies Co., Ltd., Santa Clara, CA, USA).

Fish and experimental management
M. Albus gained from Changde, China, raised in oat cages (2.0×1.5×1.5 m) for 2 weeks to adapt. The sh were fed with earthworm a week, after that, the experimental feed of FM treatment was gradually added, the content of earthworm gradually decreased until the sh could completely consume the feed. The cages were cover by 95% of the fresh alternanthera philoxeroides (Mart.) Griseb to imitate the realistic living environment of M. albus.
Uniform size juveniles M. albus (initial weight 20.12 ± 0.15 g) randomized to 18 oat cages (2.0 ×1.5 ×1.5 m) after fasting 24 h. Every treatment consisted of triplicate groups with 80 sh, and fed apparent satiation once per day at 18:00 for 8 weeks. The feeding rate was 3-4% of body weight and adjusted according as the intake diet during the feeding trial (water temperature 28 ± 4 °C, dissolved O 2 ≥ 6.0 mg/L, NH 4+ -N 0.5 mg/L, respectively).

Sample collection and analyses
According to the guidelines established by the National Institutes of Health. All experimental sh anesthetized with eugenol (1:12,000) (Shanghai Reagent Corporation, Shanghai, China) before sampling to minimize suffering.
After feeding trial, M. albus were weighed and counted to calculate survival, weight gain and feed conversion ratio after fasting 24 h. Hepatosomatic index, visceral somatic index and condition factor of M. albus were calculated based on the weight of liver, viscera, body and body length.  [26][27][28] .

Activities of intestinal enzyme
There have no signi cantly difference in amylase and lipase among all treatment (P > 0.05). Compared with FM, trypsin remarkably declined in SBM (P < 0.05). Compared with SPC, dietary soy iso avone and soy saponin group (SI, SS, SI+SS) of trypsin were lower than SPC, not signi cantly (P > 0.05) ( Table 6).   were increased if sequences number is under 2609, the OTU Number in SS was increased the fastest among all groups, which in SBM was the lowest (Figure 3).

Bacterial diversity index based on 16S rDNA gene sequence
Good's estimated sample coverage (ESC) in all groups were higher than 99 %, which indicated that the data covered most of bacterial species. Compared with FM, shannon, simpson, chao1, abundance based coverage estimator (ACE) and operational taxonomic units (OTUs) remarkably decreased in SBM (P < 0.05), these indices in SPC also decreased, not signi cantly. Compared in SPC, shannon signi cantly decreased in SI+SS (P < 0.05) ( Table 8). As gure 4 showed, rmicutes were the most abundant groups of gut ora of M. albus. For the level of phylum, compared with FM, abundance of rmicutes were trended to increase in SBM and SPC, rmicutes in SBM was the highest, while proteobacteria were trended to decrease in SBM and SPC, proteobacteria in SBM was the lowest. Compared with SPC, abundance of rmicutes were trended to increase in SS and SI+SS, proteobacteria were trended to decrease in dietary soy iso avone and soy saponin groups (SI, SS, SI+SS). As gure 5 showed, compared with FM, ratio of gut stenotrophomons in SBM and SPC decreased, and SBM was the lowest; ratio of gut lactococcus in SBM and SPC increased, ratio of unidenti ed_clostridiales was the highest in SBM. Compared with SPC, romboutsia, unidenti ed_ clostridiales and stenotrophomons increased in dietary soy iso avone and soy saponin groups (SI, SS, SI+SS), while cetobacterium decreased in dietary soy iso avone and soy saponin groups (SI, SS, SI+SS). Dietary soy iso avone and soy saponin group were different from the SPC group, and SI+SS group was quietly ( Figure 6).

Discussion
This study was continuously explore soy protein replace sh meal on M. albus, and referenced our previous studies. In this study, the growth performance of M. albus in SBM remarkably lower than FM, while SPC group not remarkably decreased, and growth performance of M. albus in dietary soy iso avone and soy saponin group (SI, SS, SI+SS) were lower than SPC. Here, dietary soy iso avone and soy saponin also decreased the hepatopancreas index and viscerosomatic index of M. albus, in particular soy iso avone and soy saponin exist simultaneously. This phenomenon was similar to study that high level of soy saponin decreased growth performance on juvenile Japanese ounder (Paralichthys olivaceus) [7] . While dietary soy iso avone has the negatively affect growth performance on Atlantic salmon (Salmo salar L.) [12] , dietary soy iso avone not signi cantly in uence growth performance on rainbow trout (Oncorhynchus mykiss) [8] . However, dietary higher soy iso avone showed better growth than without supplemented soy iso avone on juvenile golden pompano (Trachinotus ovatus) [31] and juvenile grass carp (Ctenopharyngodon idella) [11] . We speculated that different aquatic animals have different tolerance to different level of soy iso avone and soy saponin. In our study, dietary level of soy iso avone and soy saponin which corresponded to the soybean meal were all decreased the growth performance of M. albus, and these two heat-stable anti-nutritional factors have negative stacking effect for growth performance on M. albus.
Acid phosphatase (ACP) enzyme is involved in protein pinocytosis and intracellular digestion [32] . Alkaline phosphatase (AKP) is key enzyme with protective role in sh under stress, parasitic infection and wound healing [33] , and it also play a main role in dephosphorylating and detoxifying the endotoxin components of lipopolysaccharides in the intestinal mucosa [34] . Transaminases are produced by liver, if the liver damaged, the permeability of the cell membrane increases, and transaminases been secreted from cells, causing thus transaminase higher in blood. Glutamic oxalacetic transaminase (GOT) and glutamic pyruvic transaminase (GPT) used to evaluate liver whether injury [35] . Complement system is considered that defense system for resistance against microbial infection, mainly included C3 and C4 [36] . Superoxide dismutase (SOD) can eliminate free radicals, like hydroxyl radicals and superoxide anion radicals, improve antioxidant capacity [37] . Immunoglobulin M (IgM) is the essential immunoglobulin and sensitively react xenobiotics, used as an important biomarker for humoral adaptive immune response in sh [38] . Total antioxidant capacity (T-AOC) is an important indicator to measured the antioxidant ability of the body, previous studies hold that soy iso avone and soy saponin are antioxidant factor [10,14] . In present study, compared with SPC, serum AKP was remarkably increased in SI, serum GOT in dietary soy iso avone and saponin group were signi cantly increased, but serum SOD were showed the decreasing trend in dietary soy iso avone and soy saponin groups. On the one hand, dietary soy iso avone and soy saponin have antioxidant capacity for M. albus, this phenomenon was similar to juvenile grass carp (Ctenopharyngodon idella) [11] . On the other hand, these anti-nutrition factors also damaged the liver of M. albus. What's more, we also determined intestinal non-speci c immune indices. Here, compared with SPC, intestinal IgM in SS group was remarkably declined, intestinal C3, C4 and IgM also signi cantly decreased in SI+SS group. We hold that soy saponin decreased the intestinal immunity, and soy iso avone and soy saponin had the stacking effect on M. albus. While there have dietary soy iso avone improved the immune response on juvenile golden pompano (Trachinotus ovatus) [31] and juvenile grass carp (Ctenopharyngodon idella) [11] . May be different types of sh have varying degrees of tolerance by soy iso avone and soy saponin.
In this work, activity of intestinal trypsin enzyme in SBM group were remarkably lower than FM group, dietary soy iso avone and soy saponin and its mixed decreased trypsin. M. albus is a kind of carnivorous aquatic animal [39] , higher activity of trypsin is bene t for protein digestion [24] . These results may explain that dietary soy iso avone and soy saponin decreased growth of M. albus. While our results was opposite to the study that dietary soy iso avone increased intestinal digestive of trypsin on Atlantic salmon (Salmo salar L.) [12] . The speci c reasons need further exploring.
In order to observe dietary soy iso avone and soy saponin on intestinal tract more intuitively, we had made the intestinal H & E staining. Based on these intestinal morphology photos, we have counted intestinal villus height, crypt depth, intestinal muscular thickness and amount of intestinal goblet cells. In this study, compared with FM, intestinal villus height and intestinal muscular thickness in SBM remarkably decreased, intestinal crypt depth in SBM was increased, and these indices had the same phenomenon in SPC, but not remarkably. Compared with SPC, the trend showed that intestinal crypt depth increased, while intestinal villus height in dietary soy iso avone and soy saponin (SI, SS, SI+SS) were decreased, amounts of intestinal goblet cells per root in SI+SS was remarkably declined. It well explain that dietary soy iso avone and soy saponin decreased intestinal trypsin and immunity, also decreased the abilities of intestinal digestion and absorption, thus made the negative effect on performance of M. albus. These results were similar to the study on juvenile Japanese ounder (Paralichthys olivaceus) [7] , Zebra sh (Brachydanio rerio var) [4] , Atlantic salmon (Salmo salar L.) [12] and juvenile turbot (Scophthalmus maximus) [6] . Soy saponin could increase the intestinal epithelial permeability [40] , and made the intestinal cells easily damaged, and soy iso avone promote this process, thus the function of intestinal barrier decreasing. In contrast to our results, dietary soy iso avone didn't affected intestinal structure of Rainbow trout (Oncorhynchus mykiss) [8] .
To further exploring effects of dietary soy iso avone and soy saponin intestinal immune and its mechanism, we have intestinal barrier and immune mRNA expression. The intestinal tight junction protein, include occludens, claudin and zonula, which are indispensable in protecting barrier integrity and function, they play an epithelial barrier and prevented in ltration by various viruses and antigens [41] . Tight junction proteins would damage the intestinal permeability if they mRNA expression declined. Much research proved that IL-1β, IL-12, IL-15,IL-8 and IFNγ were the key pro-in ammatory cytokines [42][43][44] , and IL-10 and Tgfb mainly prevent in ammatory occuring [45] . When intestinal tight junction damaged, the risk of intestinal in ammation elevates, intestinal in ammation mediated by antiin ammatory cytokines and pro-in ammatory cytokines, and pro-in ammatory cytokines would promote the differentiation of T cells and B cells and make in ammatory occuring [46] . Then, IL-1β, IL-12 and IFNγ mRNA expression up-regulated mean that in ammation has occurred. On contrary to pro-in ammatory cytokines, the anti-in ammatory cytokines could reduce the risk of cytokines. For example, Tgfb could inhibit the lymphocyte proliferation [47] . NF-κB has been demonstrated that the most important regulators of in ammatory gene expression [48] . After the stimulation of pro-in ammatory pathways, the IκB would be degraded, and subsequently activate the translocation of NF-κB into the nucleus, which regulated the expression of pro-in ammatory cytokines and caused the in ammatory responses [49] . Soybean saponins induced NF-κB by blocking IκBα degradation. The current study was different with the published study that soybean saponins exhibit anti-in ammatory by suppressed in ammatory cytokine genes transcription through the NF-κB signaling pathway in mice [50] . In present study, compared with FM, intestinal tight junction protein mRNA (OCC, CL12, ZO-1, ZO-2, IL-10) down regulated in SBM, expression of anti-in ammatory factors mRNA (Tgfb1 and Tgfb2) and in ammatory signaling molecules mRNA (IκBa and IκBe) signi cantly down regulated in SBM, pro-in ammatory factors mRNA (IL-1β, IL-8, IL-12, IL-15, IFNγ) remarkably up regulated in SBM, ZO-1, Tgfb1, NF-κB mRNA expression signi cantly down regulated in SPC. Dietary soy iso avone and soy saponin down regulated CL-12, IL-15, ZO-2, IL-10, Tgfb2 mRNA expression, while IL-1β, IL-6, IL-8, IL-12 and IL-15 mRNA expression up regulated in these groups. Our ndings were similar to study in Atlantic salmon (Salmo salar L.) [40] . It suggested that soy saponin increased the intestinal permeability, damaged the intestinal structure and caused the function of intestinal barrier decreasing, and soy iso avone promote this progress, eventually, made the intestinal in ammation of M. albus easily occurred. A paper also reported that dietary soy saponin made higher expression of intestinal IL-1β and IL-8 on juvenile turbot (Scophthalmus maximus), and damaged the intestinal barrier and induced enteritis [6] . While dietary soy iso avone improved immune performance (such as the responses of immune cytokines) of juvenile grass carps [11] . We think that herbivorous sh may have higher adaptability to plant anti-nutritional factors than carnivorous sh.
As we know, healthy gut microbiota is essential to promote host intestinal structure and maintain it healthy statement [51] . Our previous studies have demonstrated that diet could change the composition of gut bacterial communities of M. albus [17,24,52] . In current study, OTU numbers ranked as SS > SPC > FM > SI > SS+SI > SBM from sequence 0 to 60000 until stability, Good's estimated sample coverage (ESC) in all treatments were higher than 99 %, which indicated that our data is credible. Compared with FM, shannon, simpson, chao1, abundance based coverage estimator (ACE) and operational taxonomic units (OTUs) were remarkably decreased in SBM, these indices in SPC also decreased, not signi cantly. And shannon signi cantly lower in SI+SS than SPC. This phenomenon may indicate that dietary soy iso avone and soy saponin made the gut ora imbalance. For the level of phylum, rmicutes were the most abundant group of M. albus, compared with FM, abundance of rmicutes were trended to increase in SBM and SPC, rmicutes in SBM was the highest, while proteobacteria were trended to decrease in SBM and SPC, proteobacteria in SBM was the lowest. Compared with SPC, gut rmicutes were trended to increase in SS and SI+SS, proteobacteria were trended to decrease in dietary soy iso avone and soy saponin groups (SI, SS, SI+SS). For the level of genus, compared with FM, ratio of gut stenotrophomons in SBM and SPC decreased, and SBM was the lowest; ratio of gut lactococcus in SBM and SPC increased, ratio of unidenti ed_clostridiales was the highest in SBM. Compared with SPC, romboutsia, unidenti ed_ clostridiales and stenotrophomons increased in dietary soy iso avone and soy saponin groups (SI, SS, SI+SS), while cetobacterium decreased in dietary soy iso avone and soy saponin groups (SI, SS, SI+SS). Firmicutes were the most abundant groups of phylum in the gut bacteria of M.albus, which was similar to our previous results [24] , it is also key microbiota to affect immune [53] , it's an good evidence to explain that dietary soy iso avone and soy saponin mixed feed caused immunity declining.

Conclusion
Soy iso avone and soy saponin could damage the intestinal structure and change gut bacteria community of M. albus, thus destroyed intestinal barrier, eventually induced intestinal in ammatory occurring, also made the negative effect on growth performance. Dietary soy iso avone and soy saponin mixed feed has more serious effect than they single supplemented.

Declarations
Ethics approval and consent to participate All experiments involving animals were supported by the Animal Care Committee of Hunan Agricultural University (Changsha, Hunan Province, China) and were conducted according to the Chinese guidelines for animal welfare.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This study was nancially supported by National Natural Science Foundation of China (Grant No. 31572626).

Authors' contributions
Yajun Hu determined the index in of the study and wrote this article.
Wuying Chu was in charge of designing gene in this trial.
Junzhi Zhang participated in experimental designing for this study.
Yi Hu participated in experimental designing and revised this manuscript.     Effects of dietary soy iso avone and soy saponin on OUT number of bacterial diversity index after 8 weeks (n =3).  Effects of dietary soy iso avone and soy saponin on OUT number of bacterial diversity index after 8 weeks (n =3).

Figure 4
Effects of dietary soy iso avone and soy saponin on OUT number of bacterial diversity index after 8 weeks (n =3).

Figure 5
Effects of dietary soy iso avone and soy saponin on phylum of bacterial diversity index after 8 weeks (n =3).

Figure 5
Effects of dietary soy iso avone and soy saponin on phylum of bacterial diversity index after 8 weeks (n =3).

Figure 5
Effects of dietary soy iso avone and soy saponin on genus of bacterial diversity index after 8 weeks (n =3). Effects of dietary soy iso avone and soy saponin on phylum of bacterial diversity index after 8 weeks (n =3).

Figure 7
Effects of dietary soy iso avone and soy saponin on genus of bacterial diversity index after 8 weeks (n =3).

Figure 8
Effects of dietary soy iso avone and soy saponin on genus of bacterial diversity index after 8 weeks (n =3).

Figure 9
Effects of dietary soy iso avone and soy saponin on gut principal component analysis (PCA) after 8 weeks (n =3). Note: SISS expressed SI+SS because the analysis software can't distinguish "+".