Fresh A. japonicus were purchased from Nanshan markets in Qingdao. The sea cucumbers were boiled for 10 min and processed in three different ways. To produce dried A. japonicus, the boiled sea cucumbers were put in a cold-air oven at 20˚C for 6 days. They were soaked in distilled water for 2 days and boiled for 30 - 40 minutes. After cooling, the sea cucumbers were soaked in distilled water for 2 - 3 days at 4˚C. To prepare instant A. japonicus, the boiled sea cucumbers were kept in the condiments for 1 hour and sterilized at 121˚C for 15 minutes). Finally, A. japonicus was treated with papain at 50˚C for 12 hours and the papain inactivated at 100˚C for 10 min. The resulting mixture was centrifuged and the supernatant retained. The processed A. japonicus were stored at −40˚C until use. The sea cucumbers were crushed and dissolved in saline before being administered to the mice.

All reagents used in the animal experiments were of analytical purity.

Male balb/c mice (19 ± 1 g, 4 weeks old) were obtained from Vital River (Beijing, China). During the experimental period, the mice were housed in a room maintained under a 12 h light/dark cycle at 24˚C. The mice had free access to standard laboratory pellet chow (Kangda, Jinan, China) and fresh water.

The mice were randomly assigned to 5 groups (10 mice/group): normal control group, model group, and A. japonicus groups with three different processing methods. The A. japonicus groups were given one of three sea cucumber regimens followed by 512 mg∙kg⁻¹ by gavage for 28 days, while the normal and model groups were given normal saline as control. Except for the normal group, the others were treated with cyclophosphamide (50 mg∙kg⁻¹) by intraperitoneal injection on the 26^th and 27^th feeding day. The normal mice were injected with saline as control. The mice body weights were recorded every two days. The feces were collected on the 28^th day. Mice were sacrificed by cervical dislocation on the 29^th day and the small intestine was collected for analysis. All experimental procedures were conducted according to the guidelines provided by the ethical committee of experimental animal care at Ocean University of China (OUC, Qingdao, China) and the standards for laboratory animals of China (GB 14922-94, GB 14923-94 and GB/T 14925-94).

Bacterial genomic DNA was extracted from fecal samples with the QIAamp DNA stool kit (Qiagen, Germantown, MD), according to the manufacturer’s instruction.

Bacterial universal primers GC-338F (5’-CGCCCGGGGCGCGCCCCGGGGCGGGGC GGGGGCGC-GGGGGGCCTACGGGAGGCAGCAG-3’) and 518R (5’-ATTACCGC GGCTGCTGG-3’) were used to amplify the 16S rDNA gene V3 region by PCR. PCR amplification was carried out in a MJ Mini Personal Thermal Cycler (BIO-RAD, America). Each reaction was performed in a total volume of 50 μL containing DNA template 100 ng, 10 × PCR buffer 5 μL, dNTP mix (10 mM) 3.2 μL, MgCl₂ (25 mM) 3 μL, each primer (20 mM) 0.5 μL, and 5 U/μL of Taq DNA polymerase (Takara, Japan) 1 μL. The thermal cycler profile consisted of an initial denaturation step at 94˚C for 5 min, followed by 30 cycles at 94˚C for 30 s, 55˚C for 30 s, 72˚C for 30 s, and a final extension at 72˚C for 10 min. DGGE was carried out with a Dcode System apparatus (Bio-Rad, Herts, UK) and a gradient from 35% - 55% (7 M urea and 40% deionized formamide were considered to be the 100% denaturant). The gel images were converted into digital data using Canoco. Principal component analysis (PCA) and cluster analysis were employed to compare the gut microbiota composition among groups. Important bands were retrieved, sequenced (Sangon Biotech Co. Ltd., Shanghai, China), and identified in Genbank, as described [15] .

Bifidobacteria and Lactobacillus were assayed by quantitative real-time PCR. 25 μL of the reaction volume was used for the quantitative real-time PCR assay that consisted of 12.5 μL of Maxima SYBR Green qPCR Master Mix (Roche, Switzerland), 10 μM of primers (0.75 μL each of forward and reverse primers), 8.5 μL ddH₂O, and 2.5 μL of template. The thermal conditions consisted of an initial denaturation at 95˚C for 10 min followed by 45 cycles of denaturation at 95˚C for 15 s, annealing at 60˚C for 20 s, and extension at 72˚C for 30 s. The sequences of the primers used in this study are described in supplementary Table S.

The expression levels of lysozyme, immunoglobulin A (IgA) and polymeric immunoglobulin receptor (pIgR) in the small intestine were investigated by western blot analysis. Briefly, the tissues were homogenized in RIPA buffer (50 mM Tris, 150 mM NaCl, 1 mM PMSF, 1 mM EDTA, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, pH 7.4) and put on ice for 20 minutes and then centrifuged at 12,000 × g for 5 min at 4˚C. The protein concentration of the supernatant was measured using the BCA protein assay kit (Beyotime Institute of Biotechnology, Shanghai, China) with bovine serum albumin as the standard sample. Proteins were separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinyl difluoride (PVDF) membranes (GE, Fairfield, CT, USA). The membrane was blocked with TBST (10 mM Tris, 150 mM NaCl, and 0.1% Tween 20, pH 7.6) containing 5% skim milk (BD, Franklin Lakes, NJ, USA) and incubated with the indicated primary antibody in TBST overnight at 4˚C. Subsequently, the membrane was incubated with the secondary antibody and visualized by enhanced chemiluminescence (ECL) kit (GE, Fairfield, CT, USA). Equal protein loading was assessed by blotting for GAPDH.

Bacteria-specific primers were designed by Primer Premier 5.0 software.

A one-way analysis of variance between different groups was carried out using SPSS V 18.0 software. Data were presented as mean ± standard error and statistical significance was defined to be p < 0.05.

According to the 16S rRNA V3 region PCR-DGGE patterns, the overall composition after the Cy treatment showed a significant difference from that in normal mice (Fig- ure 1(A)). After treated with Cy, the number of bands in model group reduced sharply, and the intensities of bands in some migration locations decreased compared with the normal group. The results indicated that the diversity and structure of the gut microbiota in the model group changed after Cy treatment. The gut microbiota differences in mice treated with A. japonicus were reduced notably. The structures of the gut microbiota in the A. japonicus groups were similar with that in the normal group. The results were also shown clearly in the virtual figure (Figure 2(B)) of PCR-DGGE fingerprints. The diversity indexes of the microbiota in every group showed that the diversity of intestinal flora in the model group decreased, and A. japonicus helped to reverse the changes (Table 1), especially in the dried A. japonicus group.

The principal components analysis (PCA) of the fingerprints showed that mice were classified into three clusters, with the first principal component (PC1) accounting for

Each group includes 10 mice. We selected 5 fecal samples randomly into one detecting sample and the other five ones was another sample. Every data was mean of two detecting samples.

32.3% of the variation (Figure 1(A)). The normal and dried A. japonicus group belonged to one cluster, the instant and enzymatic A. japonicus group belonged to another cluster, and the model group was in a third cluster. It confirmed that Cy caused changes in gut microbiota, and A. japonicus helped reverse the changes. The cluster dendrogram showed similar results (Figure 1(B)). These results suggest that A. japonicus could ameliorate the gut microbiota imbalance induced by Cy in mice.

In model mice, the intensities of band 6 (b6), b9, b12, b24, b25 and b28 were significantly reduced. By sequence analysis, these bands were related to Prevotella dentalis DSM 3688 (96% homology), Moryella indoligenes (99%), Barnesiella intestinihominis (99%), Ruminococcaceae bacterium LM158 (98%), Clostridiales bacterium CIEAF 020 (97%), Johnsonella sp. oral taxon 166 (97%) (Figure 2(A) and Table 2). These bacteria belong to Bacteroidetes and Clostridia groups. Bacteroidetes is an intestinally adapted bacterium involved in carbohydrate fermentation, polysaccharide and steroid metabolism, and maintaining intestinal physiological functions, amongst other functions [16] . Most Clostridial bacteria can hydrolyze polysaccharide and protein. The reduction of Bacteroidetes and Clostridia may have an effect on digestion and absorption of polysaccharide and protein. The intensities of bands in A. japonicus groups increased to near normal levels. We speculate that administration of A. japonicus helped to restore normal intestinal physiological function.

A number of bacteria are friendly and helpful to hosts, but in low abundance, such as probiotic Bifidobacterium and Lactobacillus. Bifidobacterium can reduce intestinal pH to inhibit the growth of harmful bacteria, produce vitamins and amino acids, stimulate immune responses, protect the intestinal barrier, etc. [17] [18] [19] . Lactobacillus can ferment carbohydrates to produce lactic acid and help digestion and absorption, stimulate immunoglobulin production to enhance the host immune function, etc [20] [21] . Hence, the amount of probiotic Bifidobacterium and Lactobacillus was quantified by real-time PCR in this study. The results show that Cy caused a significant reduction of Bifidobacterium (Figure 3(A)). A. japonicus administration increased the quantity of Bifidobacterium and dried A. japonicus showed a significant effect (p < 0.05). However, the amount of Lactobacillus showed no obvious change in this model (Figure 3(B)). These data imply that A. japonicus intake protects the intestine by increasing the quantity of Bifidobacterium.

Ten important bands were cut from the gel, sequenced and then identified in Genbank.

Lysozyme is an important bactericidal substance in body fluids, which works as a mucosal chemical barrier and represents the innate mucosal immunity effector. SIgA is composed of two immunoglobulin A (IgA) molecules, joined by a J-chain and a secretory component. pIgR transporting dimeric IgA (dIgA) across the epithelial cell function is a necessary condition for the secretion of sIgA to exocrine fluid [22] . Hence, the expression of lysozyme, IgA and pIgR was detected to reflect the mucosal immune function.

The results showed that expression of lysozyme increased with A. japonicus intake, while that in the model group reduced significantly (p < 0.05) (Figure 4(A)). Compared with the normal group, the IgA and pIgR expression of mice in model group decreased significantly (Figure 4(B) and Figure 4(C)). Administration of dried and instant A. japonicas significantly improved the expression of IgA (p < 0.01). The three types of A. japonicus could all increase pIgR expression; however, they showed no statistical significance. All these data indicated that A. japonicus intake could ameliorate intestinal mucosal injury induced by Cy.