Efficacy of Sucralfate-Combined Quadruple Therapy on Gastric Mucosal Injury Induced by Helicobacter pylori and Its Effect on Gastrointestinal Flora

Background This study explored the therapeutic efficacy of standard triple therapy combined with sucralfate suspension gel as well as the mechanisms of action in mouse models of H. pylori infection. Materials and Methods C57BL/6J mice were randomly divided into 5 groups: NC (natural control), HP (H. pylori infection), RAC (rabeprazole, amoxicillin, and clarithromycin), RACS (RAC and sucralfate suspension gel), and RACB (RAC and bismuth potassium citrate). HE staining and electron microscopy were performed to estimate histological and ultrastructural damages. The IL-8, IL-10, and TNF-α of gastric antrum tissues were measured by immunohistochemistry and qRT-PCR. ZO-1 and Occludin were also detected with immunohistochemistry. The genomes of gastric and fecal microbiota were sequenced. Results The eradication rate of H. pylori in the RACS group was higher than the RAC group. RACS therapy had protective effects on H. pylori-induced histological and ultrastructural damages, which were superior to the RAC group. RACS therapy reduced the protein and mRNA levels of IL-8 compared with the RAC group. The expression of Occludin in the RACS group was significantly higher than that of the RAC group. The composition of gastric and fecal microbiota for RACS was similar to the RACB group according to PCA. Conclusions The RACS regimen eradicated H. pylori infection effectively and showed RACS had protective effects against H. pylori-induced histological and ultrastructural damage. The mechanisms of RACS effects included decreasing IL-8, enhancing Occludin, and transforming gastric microbiota. Moreover, RACS and RACB have a similar effect on gastrointestinal flora.


Introduction
Helicobacter pylori (H. pylori), a Gram-negative, microaerophilic bacterium, is closely associated with chronic gastritis, peptic ulcers, and gastric adenocarcinoma [1]. This bacterium is able to colonize and survive in the gastric environment by several mechanisms, including the adherence to the epithelium and breakdown of urea with production of ammonium which neutralizes the gastric acidity [1]. Improving the eradication rate of H. pylori is particularly important due to the high rates of H. pylori infection and high morbidity of gastric cancer in China [2,3]. The eradication rates of standard triple therapy have been declining due to increased antibiotic resistance [4,5]. In China, bismuth-containing quadruple therapy is currently the recommended first-line treatment [6], but its administration is limited due to adverse effects of bismuth.
Recently, the efficacy of gastric mucosal protective agents in H. pylori eradication has been widely estimated. Several mucosal protective agents combined with PPI+antibiotic therapy have been validated to increase eradication rates and reduce side effects [7][8][9][10][11][12]. Sucralfate suspension gel (SC) is a sucrose sulfate compound [13]. Its clinical efficacy for ulcer and chronic gastritis has been observed, but the role of sucralfate suspension gel for eradicating H. pylori has not been determined. The mechanisms of nonantibiotic drugs to eliminate H. pylori can be summarized as decreasing inflammatory factors, enhancing the mucosal barrier, transforming gastric microbiota, and so on [14].
In this study, we tried to explore the effect of standard triple therapy+sucralfate suspension gel (rabeprazole, amoxicillin, and clarithromycin and sucralfate suspension gel (RACS)) on H. pylori-induced histological and ultrastructural damages, inflammatory factors, tight junction protein, and gastric and fecal microbiota in H. pylori infection mouse models.

Materials and Methods
2.1. Experimental Animals, Medicine, and Strains. Male C57BL/6J mice at an age of 6-8 weeks and weight of 18~22 g were purchased from SPF Biotechnology Company (Beijing, China). H. pylori Sydney strain 1 (SS1) was cultured for H. pylori infection mouse models. SC was provided by Kunming Jida Pharmaceutical Co. Ltd. The experiment was approved by the Animal Ethical Committee of the First Hospital of Peking University (No. J201819).

Animal Model of H. pylori Infection and Treatments.
After adaption to their environment for 1 week, mice were randomly divided into 5 groups: natural control group (NC, n = 6), H. pylori infection group (HP, n = 12), standard triple therapy group (rabeprazole, amoxicillin, and clarithromycin (RAC), n = 12), standard triple therapy and sucralfate group (RAC and SC (RACS), n = 12), and bismuthcontaining quadruple therapy group (RAC and bismuth potassium citrate (RACB), n = 12). Except for the NC group, mice were given doses of 1 × 10 9 CFU SS1 in 0.2 ml brucella broth by oral gavage (every other day, 5 times total). After postinfection, one mouse in each group was killed randomly and H. pylori colonization was confirmed by immunohistochemical staining. One week after H. pylori infection, the RAC group was fed with 4 mg/kg omeprazole, 206 mg/kg amoxicillin, and 103 mg/kg clarithromycin. Based on standard triple therapy for the RAC group, 206 mg/kg SC was added in the RACS group and 123 mg/kg bismuth was used simultaneously in the RACB group (twice daily for 14 days). Animals in the NC and HP groups were given the same volume of normal saline. All groups of mice were sacrificed on six days after the last administration of eradication therapies.
H. pylori colonization was tested by immunohistochemical staining and PCR (the sequences are shown in Supporting Information Table 1). Gastric tissues were cut lengthwise in order to observe both the antrum and the body simultaneously and to confirm H. pylori on 30 slides unless positive result was observed.

HE Staining and Electron
Microscopy. Gastric tissues were fixed by 10% neutral formalin, embedded in paraffin, and cut into 4 μm thick sections. The sections were stained with standard hematoxylin and eosin (HE) staining by Liu Y and independently scored by two blinded investigators (Teng GG and Wu T). The score for epithelial damage (EDS) was based on the following criteria: 1 (normal mucosa), 2 (mucosal surface cell damage), 3 (glandular cell damage), and 4 (erosion, bleeding, or ulcers). The ultrastructure of the gastric antrum was observed with transmission electron microscopy.

Statistical
Analysis. The data were analyzed with SPSS 21.0 and R software. Continuous variables were displayed as means and standard deviations. To investigate whether differences among different groups are statistically significant, data were analyzed by one-way analysis of variance (ANOVA) followed by the Tukey test or Kruskal-Wallis test and followed by the Nemenyi test for multiple groups. The eradication rate was calculated using Fisher's exact test. The sequencing analyses for gastric microbiota or fecal microbiota were conducted with R software. P values < 0.05 were defined as statistically significant.

Results
3.1. RACS Therapy May Be Superior to RAC Therapy for H. pylori Infection in Mice. The colonization rate of H. pylori in the HP group was 91.67% (11/12), while it was 0.00% (0/6) in the NC group. The eradication rates were 66.67% (8/12) in the RAC group, 83.33% (10/12) in the RACS group, and 91.67% (11/12) in the RACB group. These results suggest that eradication efficacy of RACS may be effective, which still requires further clinical trials.

Protective Effects of RACS on H. pylori-Induced
Histological and Ultrastructural Damages. As shown in Figure 1(a), there were intact structures of gastric mucosa without inflammatory cell infiltration in the NC group. The gastric mucosal epithelium was unclear, and erosion or ulcers were observed in the HP group. The histological damage for three treatment groups was attenuated with mucosal surface cell damage along with mild inflammatory cell infiltration. The EDS is shown in Supporting Information Table 2. The EDS in the RACS group was significantly decreased compared with that in the RAC group (P = 0:019) and was lower than that in the RACB group without significance (P = 0:382).
The ultrastructure of the gastric antrum is shown in Figure 1(b). The normal cell structures and abundant secretory granules were seen in the NC group. The mitochondria and endoplasmic reticulum were swollen with sparse microvilli and decreased secretory granules in the HP group. The ultrastructural damage of three treatment groups was reduced, which was similar to the NC group, but swollen mitochondria were still seen in the RAC group. These data showed that RACS therapy has protective effects against H. pylori-induced histological and ultrastructural damages.

RACS Inhibited the Overexpression of IL-8 Induced by H.
pylori. The mRNA levels of IL-8 (also known as chemokine ligand 15 (Cxcl15)), IL-10, and TNF-α expression were upregulated in the HP group compared with the NC group (P = 0:017; P = 0:247; and P = 0:038). Additionally, the IL-8, IL-10, and TNF-α mRNA levels significantly decreased after different eradication therapies. The IL-8 mRNA level of the RACS group was significantly lower than that of the RAC group (P = 0:041) but similar to that of the RACB group (P = 0:988). No significant differences were found in IL-10 or TNF-α mRNA levels between the RACS and RAC therapies (P = 0:136; P = 0:975). The protein levels of IL-8, IL-10, and TNF-α followed the same trend (Table 1, Supporting Information Figure 1). These data indicate that RACS therapy induces an anti-inflammatory response, especially by reducing IL-8.

RACS Enhanced Expression of the Tight Junction Protein
Occludin. As shown in Table 2 and Supporting Information Figure 2, ZO-1 and Occludin IHC scores were downregulated in the HP group compared with the NC group (P = 0:009; P < 0:001). The expression of ZO-1 and Occludin in the three treatment groups was higher than that in the HP group. No significant difference was noted in ZO-1 protein levels between the RACS and RAC therapies (P = 0:961). The Occludin expression for the RACS group and RACB group was elevated significantly compared with that for the RAC group (P < 0:001). The data suggest that RACS significantly enhanced expression of Occludin.

Alteration of Gastric Microbiota Composition during
Eradication Therapy in H. pylori-Infected Mice 3.5.1. Composition of the Gastric Microbiota. A total of 871,470 clean reads with an average of 29,049 reads per sample were generated from 30 gastric tissues, and 819 OTUs at a 97% similarity level were generated afterwards. The most abundant phyla in the gastric tissues were Bacteroidetes, Firmicutes, Proteobacteria, and Verrucomicrobia with average relative abundances of 48.69%, 41.57%, 5.65%, and 2.95%, respectively (Figure 2(a)). At the genus level, the microbiota of both the NC and HP groups were dominated by Lactobacillus with relative abundances of 64.64% and 90.47%, respectively, while the microbiota of the treatment groups were dominated by Bacteroides, Parabacteroides, and Barnesiella, with average relative abundances of 25.97%, 11.49%, and 8.57%, respectively (Figure 2(b)).

Alpha and Beta Diversities of the Gastric Microbiota.
The diversity of gastric microbiota was evaluated through alpha diversity (Shannon and Simpson indexes) and beta diversity (PCoA and Anosim). As shown in Figures 2(c) and 2(d), the Shannon and Simpson indexes decreased in the HP group compared with the NC group (P = 0:026; P = 0:041), and no significant difference in alpha diversity indexes was observed between the HP group and the three treatment groups or among the three treated groups. The results of a PCoA based on unweighted UniFrac metrics are displayed in Figure 2(e); the correlated Anosim demonstrated a significant difference between the HP and NC groups (R = 0:235, P = 0:012), among the three treatment groups (RAC, RACS, and RACB) and HP group (R = 0:356, P = 0:009; R = 0:591, P = 0:003; and R = 0:376, P = 0:013), and between the RAC and RACS groups (R = 0:304, P = 0:011). No significant difference was found between the RACS and RACB groups (R = 0:135, P = 0:066). These findings showed that H. pylori infection significantly decreased the alpha diversity of gastric microbiota and changed gastric microbial composition. Moreover, the three treatments also had significant impact on gastric microbial structure, while alpha diversity of these three groups did not differ from the other groups.

Composition Alteration in the Taxa of the Gastric
Microbiota. To explore the distinct species among different groups, we used the Wilcoxon rank sum test and the Kruskal-Wallis rank sum test to conduct differential abundance analyses at all levels (phylum, family, class, order, and genus). The 102 differential abundant taxa of gastric microbiota were found among all groups, and 48 taxa were detected at the genus level. The top 20 different abundances of microbes among five groups are shown in Figure 3(a), and phylum level abundance was high for Bacteroidetes and Proteobacteria, while abundance for Firmicutes was lower in the three eradication groups. We found that the phylum Bacteroidetes, class Bacteroidia, order Bacteroidetes, family Bacteroidaceae, and genus Bacteroides were all more abundant in the therapy groups. As shown in Figure 3(b), the most affected specific genera in gastric microbiota were Lactobacillus, Bacteroides, Parabacteroides, Barnesiella, Blautia, Clostridium XlVa, and Alistipes. The PCA based on the relative abundance of all differential taxa or genera are shown in Figures 3(c) and 3(d), and we could not separate RACS from the RACB group, respectively, while the other groups were obviously distinct.     (Figure 4(a)). At the genus level, the fecal microbiota of both the NC and HP groups were dominated by Alistipes, Lactobacillus, and Barnesiella, while therapeutic groups were dominated by Bacteroides, Parabacteroides, Akkermansia, and Barnesiella (Figure 4(b)).

Alpha and Beta
Diversities of the Fecal Microbiota. The diversity of gut microbiota was evaluated as mentioned above. As shown in Figures 4(c) and 4(d), the Shannon and Simpson indexes showed no significant difference between the HP and NC groups (P = 0:180; P = 0:065). Those indexes decreased in the three treatment groups (RAC, RACS, and RACB) compared with the HP group. Furthermore, the Shannon and Simpson indexes of the RACS group were lower than the indexes of the RAC group (P = 0:009; P = 0:041), which were similar to the RACB group (P = 1:000; P = 0:699).

Composition Alteration in the Taxa of the Fecal
Microbiota. We also evaluated differential abundance analyses at all levels as mentioned above. The 94 differential abundant taxa of fecal microbiota were found among all groups, and 40 taxa were found at the genus level. The top 20 different abundances of microbes among five groups in fecal microbiota are displayed in Figure 5(a), and phylum level abundance of Bacteroidetes was higher while Firmicutes was lower after H. pylori infection or eradication treatments. We found that all levels of Bacteroidetes were affected among all groups. The most affected specific genera in fecal microbiota were Bacteroides, Parabacteroides, Akkermansia, Clostridium XlVa, Blautia, Escherichia/Shigella, Oscillibacter, and Clostridium XlVb ( Figure 5(b)). The PCA based on relative abundance of all differential taxa or genera are shown in Figures 5(c) and 5(d), and we also could not separate RACS from the RACB group, while the other groups were obviously distinct.

Discussion
In this study, we confirmed eradication efficacy of RACS in mice. RACS therapy also had a therapeutic effect in H. pylori-induced histological and ultrastructural damages, which was better than the RAC group and similar to the RACB group. The preliminary results indicated that the RACS regimen eradicated H. pylori infection effectively, which needs to be confirmed through further clinical studies.
Except for a direct bactericidal effect, the mechanisms of nonantibiotic drugs for eliminating H. pylori can be summarized as decreasing inflammatory factors, enhancing the mucosal barrier, transforming gastric microbiota, and so on 14 . After H. pylori infection, the local inflammation of gastric mucosa was caused by neutrophil granulocytes. A series of cytokines, such as IL-4, IL-6, IL-8, IL-10, and IL-12, were upregulated in gastric mucosal tissues [15]. These inflammatory factors formed a complicated network of immune inflammation to induce gastric mucosal damages [16]. The results of this study showed that IL-8 expression of the RACS group was significantly lower than that of the RAC group, while IL-10 and TNF-α of the RACS group was similar to those of the RAC group. These data showed that RACS therapy suppressed the inflammatory response by decreasing cytokines, especially by reducing IL-8 to ameliorate H. pylori-induced injury.
Tight junction proteins play an important role in the gastric epithelial barrier [17], including ZO-1 and Occludin. ZO-1 is a cytoskeletal protein of tight junction proteins [18], and Occludin is a transmembrane protein located on tight junction proteins [19]. Fan et al. [20] concluded that H. pylori infection dysregulated gastric epithelial barrier function by reducing ZO-1 and Occludin. Our results indicated that Occludin expression of the RACS group was elevated significantly compared with that of the RAC group, while no significant difference was noted in ZO-1 protein levels. These data suggested that RACS enhanced expression of the tight junction protein Occludin.
The main phyla of the gastric microbiota are Proteobacteria, Firmicutes, Bacteroidetes, and Actinobacteria in healthy individuals [21]. Additionally, the diversity of the human gastric microbiota decreased after H. pylori infection [22]. Our data revealed that the most abundant phyla of the mouse gastric microbiota were Bacteroidetes, Firmicutes, Proteobacteria, Verrucomicrobia, and Actinobacteria. H. pylori infection decreased alpha diversity and changed beta diversity, which was similar to the previous clinical study [22]. The treatment regimens markedly affected beta diversity while alpha diversity decreased insignificantly in mice. Li et al. concluded that alterations in gastric microbiota and reduction in bacterial diversity induced by H. pylori could be restored through antibiotic treatment in human beings [23]. However, our results showed that the gastric flora of the treatment groups was still significantly different compared to that of the normal mice, which indicated that the eradication drugs may affect the gastric flora, or the gastric flora needs a longer time to be restored after H. pylori eradication.

BioMed Research International
Antibiotic treatments can alter richness, diversity, and composition of gut microbiota in mice with a controlled environment [24]. This study showed that most abundant phyla of mouse fecal microbiota were Bacteroidetes, Firmicutes, and Proteobacteria, which were similar with human gut flora [25]. Additionally, the alpha diversity of gut microbiota in the HP group decreased compared with that in the NC group (P > 0:05) in mice, whereas an increase was found in a previous human research [25]. Three eradication therapies significantly altered diversity in mouse fecal microbiota. We observed disorders of Bacteroidetes and Firmicutes after H. pylori infection or eradication treatments in mice, which was a change associated with type 2 diabetes and Crohn's disease [26,27]. Bacteroidetes has been reported to be associated with immunity and metabolism in primary biliary cirrhosis patients [28]. To our knowledge, the effects of different eradication regimens on gut microbiota composition have not been compared directly in patients or mice. We found the composition of mouse fecal microbiota after RACS was similar to the RACB group in PCA. It is noteworthy that genus Akkermansia of RACB mice was more prominent than that of RACS mice, although the difference was not significant, whereas Akkermansia decreased after bismuth-containing eradication in previous clinical studies [29,30]. Akkermansia is a mucin-degrading beneficial bacterium, and it has been shown to reduce gut barrier disruption and insulin resistance [31,32].
The limitation of this study is that we only discussed mouse gastrointestinal microbiota compositions without human results. Mice are used to easily control the diet or other environmental factors on microbial diversity of the intestinal tract and to relate this back to intervention measures. Although many common genera are shared in the human and murine intestines, these differ in abundance, which could weaken the application value of the mouse results [33]. Additionally, humans take different tablets before or after meals to eradicate H. pylori, while mice are given combined medicines simultaneously by oral gavage.
Intragastric administration may make it easier for drugs to enter the gastrointestinal with a high dose and have obvious effects on microbial dysbiosis.
In conclusion, our results indicate that the RACS regimen might eradicate H. pylori effectively. RACS therapy has protective effects against H. pylori-induced histological and ultrastructural damages. The mechanisms of RACS for eliminating H. pylori included decreasing IL-8, enhancing Occludin, and transforming gastric microbiota. Moreover, RACS and RACB have similar effects on gastrointestinal flora.

Data Availability
The data used to support the findings of this study are included within the article.