Evaluation of the Potential Protective Effects of Lactobacillus Strains against Helicobacter pylori Infection: A Randomized, Double-Blinded, Placebo-Controlled Trial

Background The beneficial effects of probiotic supplementation standard antibiotic therapies for Helicobacter pylori infection have been verified, but the ability of probiotic monotherapy to eradicate H. pylori remains unclear. Aim To evaluate the accuracy and efficacy of specific Lactobacillus strains against H. pylori infection. Methods Seventy-eight patients with H. pylori infection were treated with strain L. crispatus G14-5M (L. crispatus CCFM1118) or L. helveticus M2-09-R02-S146 (L. helveticus CCFM1121) or L. plantarum CCFM8610 at a dose of 2 g twice daily for one month. 14C-urea breath test, the gastrointestinal symptom rating scale, serum pepsinogen concentrations, and serum cytokine concentrations of patients were measured at baseline and end-of-trial to analyze the effect of the Lactobacillus strains in eradicating H. pylori infection and reducing gastrointestinal discomfort in patients. In addition, the composition and abundance of the intestinal microbiota of patients were also measured at end-of-trial. Results The 14C-urea breath test value of the three Lactobacillus treatment groups had decreased significantly, and the eradication rate of H. pylori had increased by the end of the trial. In particular, the eradication rate in the G14-5M treatment group was significantly higher than the placebo group (70.59% vs. 15.38%, P=0.0039), indicating that one-month administration of the G14-5M regimen was sufficient to eradicate H. pylori infection. The ingestion of Lactobacillus strains also ameliorated the gastrointestinal symptom rating scale scores, and the serum interleukin-8 concentrations of H. pylori-infected patients appeared to modulate the gut microbiota of patients. However, none of the Lactobacillus strains had a significant effect on general blood physiological characteristics, serum tumor necrosis factor α concentrations, or serum pepsinogen concentrations in the patients. Conclusion Three Lactobacillus strains significantly alleviate the gastrointestinal discomfort and the gastric inflammatory response of H. pylori-infected patients. The activity of probiotics in eradicating H. pyloriinfection may be species/strain specific.


Introduction
Helicobacter pylori is a spiral Gram-negative bacterium that colonizes human gastric mucosa [1,2]. It is associated with diseases of the upper gastrointestinal tract, such as chronic gastritis, peptic ulcers, atrophy of gastric mucosa, mucosaassociated lymphoid tissue lymphoma, and gastric cancer [3,4]. Standard antibiotic strategies may have adverse consequences, such as causing bacterial antibiotic resistance and gastrointestinal side effects [5,6]. us, several studies have been conducted to develop novel, safe and efficacious therapies to eradicate H. pylori in patients. For instance, probiotics improved the eradication rate and reduced side effects when added to the treatments designed to eradicate H. pylori. Several food factors proved the antimicrobial activity against H. pylori. β-caryophyllene, a volatile bicyclic sesquiterpene compound that can be present in the essential oils of many edible plants such as cloves, oregano, and cinnamon, has been reported to significantly inhibit H. pylori growth via the downregulation of virulence factors in a model using Mongolian gerbils [7]. e flavonoid compounds baicalin and baicalein found in many medicinal plants exhibit an anti-inflammatory effect. Baicalin and baicalein both suppressed the vacA gene expression of H. pylori and interfered with the adhesion and invasion ability of H. pylori to human gastric adenocarcinoma cell line (AGS), as well as decreased H. pylori-induced interleukin (IL)-8 expression [8]. In the mice infection model, high dosages of baicalin and baicalein inhibited H. pylori growth in the mice's stomach [9]. e ability of probiotics to inhibit H. pylori infection has been previously demonstrated. In animal models, Lactobacillus spp. strongly inhibited H. pylori infection by reducing H. pylori colonization [10], alleviating H. pyloriinduced gastric inflammatory responses [11,12], inhibiting urease activity of H. pylori [13], and rebalancing the gastric microbiota [11,13]. Clinical trials have suggested that a combination of Lactobacillus spp. (e.g., L. acidophilus [14,15], L. reuteri [16], L. rhamnosus [17,18], L. plantarum [14], L. bulgaricus [18], L. casei [18], and L. sporogenes [19]) and conventional antibiotic treatment has positive effects on both the eradication rate of H. pylori and/or the incidence of overall side effects. A recent meta-analysis (40 articles, 5792 patients) about the efficacy of probiotic-supplemented therapy on the eradication of H. pylori and incidence of therapy-associated side effects showed that probiotic supplementation improved the eradication rate by approximately 10% relative to the control group, and the side effects of antibiotic treatment (e.g., diarrhea, vomiting and nausea, constipation, epigastric pain, and taste disturbance) also decreased significantly with probiotic supplementation [20]. e mechanisms by which Lactobacillus spp. inhibit H. pylori infection are generally as follows [21]: (1) e production of bactericidal metabolites: Lactobacillus spp. inhibit H. pylori growth by producing short-chain fatty acids (e.g., butyrate, propionate, and acetate) and antibacterial agents (e.g., bulgaricus BB18, L. brevis BK11, lacticins A164, and lacticins BH5) [12,22,23]. For instance, lactacin F, a bacteriocin secreted by L. johnsonii La1, showed a bactericidal effect against pathogens by forming pores in their lipid bilayers, perturbing membrane permeability and membrane potential [24]. (2) Inhibition of H. pylori adherence: Lactobacillus spp. affect the adherence of H. pylori by competing with H. pylori for attachment to the adhesion receptors for Asialo-GM1 and sulfatide [25], inhibiting expression of the adhesin-encoding gene sabA of H. pylori [26] and upregulating the expression of MUC3mRNA in the gastric mucosa (where MUC3 mucin has the ability to inhibit the adherence of pathogens to epithelial cells) [27], all of which further reduce the in vivo colonization of H. pylori. (3) Modulation of the immune response: Lactobacillus spp. decreases the secretion of H. pylori-induced IL-8 or tumor necrosis factor (TNF)-α and increases the secretion of IL-10 in the gastric mucosa [28,29].
Although Lactobacillus strains used in combination with antibiotics have been shown to eradicate H. pylori, few in vivo studies have focused on the use of Lactobacillus monotherapy to treat H. pylori infection. Furthermore, the clinical trial efficacy of single-probiotic strain treatment for H. pylori eradication remains controversial. For instance, it was reported that L. reuteri treatment (2 × 10 10 CFU/day) reduced the load of H. pylori in adults [30], whereas the same dose of L. casei did not [31]. Similarly, Lactobacillus showed strain specificity in the eradication of H. pylori: L. rhamnosus GG significantly increased H. pylori eradication rates in a clinical trial [32], but L. rhamnosus LR06 had no effect [33].
us, there is a clear need for more studies on the effect of treatment with a single-probiotic strain on H. pylori infection. In our preliminary study, we screened 97 strains of Lactobacillus for their ability to inhibit the in vitro growth of H. pylori ( Figure S1), reduce the adherence of H. pylori to IL-8 cells (Figure S2), and stably colonize C57BL/6 mouse gastric mucosa ( Figure S3). We screened out three strains with remarkable bacteriostatic effects, inhibition of H. pylori adherence, and gastric colonization abilities: L. crispatus G14-5M, L. helveticus M2-09-R02-S146, and L. plantarum CCFM8610. We determined that treatment with each of these Lactobacillus strains decreased the concentration of IL-8 secreted by AGS cells cocultured with H. pylori to a value comparable to the control ( Figure S4) and downregulated the expression of the CagA gene of H. pylori ( Figure S5). Furthermore, these three strains exhibited the main properties and safety profile required of a probiotic, as follows: resistance to gastrointestinal juices, biliary salts, NaCl, and low pH; the presence of the CRISPR/Cas system (Table S1); no significant toxin-producing virulence factors (Table S2); and low/no harm of antibiotic resistance genes (Table S3 and Figure S6). erefore, L. crispatus G14-5M, L. helveticus M2-09-R02-S146, and L. plantarum CCFM8610 were selected for a trial in humans.
We aimed to evaluate the accuracy and efficacy of the three Lactobacillus strains in eradicating H. pylori infection in patients, in decreasing their gastrointestinal discomfort, alleviating their gastric inflammatory responses, and regulating their intestinal microbiota.

Patients.
e patients were recruited from adults who visited the hospital and had been diagnosed as positive for H. pylori infection by a 13 C/ 14 C-urea breath test (UBT), a rapid urease test, or a histological examination of biopsy tissue, within three months before the onset of the study. e exclusion criteria were as follows: the presence of a severe disease, such as malignant tumor and severe metabolic disease; the consumption of nonsteroidal anti-inflammatory drugs, corticosteroids, acid-inhibitory drugs (proton-pump inhibitors or H 2 -receptor blockers), or antiflatulent agents; antibiotic treatment one month prior to study start, including H. pylori eradication therapy; a habit of ingesting probiotics, yogurt, or lactic acid bacteria-fermented beverages; a history of previous gastrointestinal surgery; mental illness; and pregnancy or lactation.
Seventy-eight individuals were included in the study, and all patients signed a written informed consent form prior to study entrance. e study was conducted at Tinghu e Lactobacillus strains were cultured, lyophilized, and packaged into small aluminum-foil sachets by a probiotic-strain manufacturer (Jiangsu Wecare Biotechnology Co., Ltd., Suzhou, Jiangsu, China). e number of viable bacteria in Lactobacillus products during the experimental period was 5 × 10 9 CFU/g, measured once a week. e placebo products contained soy protein and maltodextrin, provided by the same manufacturer.
All of the products (2 g/sachet) were in a powder form and had the same appearance, packaging, and color. ey were stored in a refrigerator at 4°C.

Study Design.
e human trial followed a randomized, double-blind, placebo-controlled design. Sample sizes were determined based on similar previous studies [30,31,34,35]. A table of random numbers generated by computer was used to allocate patients to one of four groups, namely, a L. crispatus G14-5M treatment group (n � 19), a L. helveticus M2-09-R02-S146 treatment group (n � 20), a L. plantarum CCFM8610 treatment group (n � 20) and a placebo group (n � 19). Patients were asked to ingest two sachets of probiotic products or placebo products daily (once in the morning and once in the evening) for a month. Both the researchers and the patients were blind to the contents of the products during the study. e patients were followed up weekly by a researcher via phone, who was also unaware of the patient's allocation. e primary endpoint was a decrease in H. pylori load evaluated by 14 C-UBT. e secondary endpoints were a decrease in gastrointestinal discomfort (assessed by a gastrointestinal symptom rating scale (GSRS)), an alleviation of gastric mucosal inflammation (assessed by the ratio of serum pepsinogens [PGs] I and II, and the serum concentrations of inflammatory factors), and changes in the gut microbiota of the patients.

Evaluation Parameters
2.4.1. 14 C-Urea Breath Test. We used the 14 C-UBT to confirm the status of H. pylori infection one day before the treatment and one day after the month-long treatment. Begins with the oral administration of 14 C labeled urea. H. pylori produce the urea splitting enzyme Urease, which ultimately cleaves the labeled urea to ammonia and bicarbonate. Bicarbonate is the precursor of CO 2 that is incorporated into breath. After an overnight fast, all patients swallowed a capsule containing 14 C-urea with 20 mL of water. Fifteen minutes after capsule intake, each patient blew into a dry cartridge until the breath-card indicator turned from orange to yellow. 14 CO 2 collected by the breath card was measured with the H. pylori analyzer, and disintegrations per minute (DPM) > 100 were judged as positive for H. pylori infection.

e Gastrointestinal Symptom Rating Scale.
e GSRS is a questionnaire recommended by Japanese guidelines for evaluating gastrointestinal symptoms in functional dyspepsia [36].
Each of 15 gastrointestinal symptom items, such as abdominal pain, heartburn, and acid regurgitation, was scored from 0 to 3 according to severity during the past week. A higher score indicated more severe symptoms. e questionnaire was filled in one day before the treatment and one day after the month-long treatment, i.e., a total of two times.

Serum Pepsinogen Concentrations.
e blood samples of patients were collected one day before the treatment and one day after the month-long treatment, and serum was obtained by centrifugation. Serum PG (PG I and PG II) concentrations were detected using an enzyme-linked immunosorbent assay (ELISA) kit (Fcmacs Biotech Co., Ltd.), following the protocol recommended by the manufacturer.

Cytokine Analysis.
Serum cytokine concentrations were detected using an ELISA kit (Fcmacs Biotech Co., Ltd.), following the protocol recommended by the manufacturer.

Composition and Abundance of the Intestinal
Microbiota. Patients provided one stool sample after the completion of the study (within three days). Stool samples were collected in sterile plastic containers and stored at 4°C until they reached the laboratory. Upon arrival, stool samples were immediately stored at −80°C until DNA extraction. DNA was extracted from the stool samples using the FastDNA ™ SPIN Kit for Feces (MP Biomedicals, USA), following the manufacturer's protocol. e polymerase chain reaction methods and primers for amplifying the V3-V4 region and the groEL gene of the 16S rDNA were based on the previously published protocols [37,38]. Lactobacillusspecific primer sets were developed for the hypervariable region of the groEL gene, a single-copy gene that undergoes rapid mutation and evolution.
is methodology could accurately perform taxonomic identification of Lactobacillus down to the species level. e accuracy of the method has been demonstrated in fermented yak milk samples and human, rat, and mouse fecal samples.
Library preparation and sequencing were based on the method proposed by Yang et al. [39]. e composition and abundance of the intestinal microbiota of patients were analyzed with the Quantitative Insights Into Microbial Ecology software package (Flagstaff, AZ).

Statistical
Analysis. All data were expressed as means ± standard errors of the mean. Fisher's exact tests, one-way Canadian Journal of Infectious Diseases and Medical Microbiology analyses of variance (ANOVA), and t-tests were performed (using SPSS version 22.0 software) for the comparison of results, such as H. pylori eradication rate, serum PG concentration, serum cytokine concentration, Shannon index, observed species index and taxa abundance count in different groups. e differences between groups were judged by ANOVA, and the differences between the two groups were judged by a t-test or chi-square test. P < 0.05 was considered as significant.

Results
Seventy-eight patients who were positive for H. pylori infection participated in the trial. Six patients in the placebo group, two patients in the G14-5M treatment group, and one patient in the M2-09-R02-S146 treatment group withdrew from the trial, which meant that 69 patients [placebo group (n � 13), G14-5M treatment group (n � 17), M2-09-R02-S146 treatment group (n � 19), and CCFM8610 treatment group (n � 20)] completed the study (Figure 1).

General Characteristics of Patients.
No statistically significant differences were observed in the mean age, male to female ratio, number of smokers, or number of alcoholic drinkers between the groups of patients who completed the study (Table 1).
Compared with the placebo treatment, the Lactobacillus strain treatments did not significantly affect the general blood physiological characteristics of patients (Table 2).

3.2.
e Eradication Rate of Helicobacter Pylori. Compared with the placebo group, the H. pylori eradication rate ( 14 C-UBT results) was increased in the three Lactobacillus treatment groups at the end of the trial, and the eradication rate in the G14-5M treatment group was significantly higher than those of the other groups (Table 3). Specifically, the 14 C-UBT value of the placebo group showed no significant change before and after the trial, but the 14 C-UBT values of each of the Lactobacillus treatment groups exhibited a significant (70-120 dpm/mmol) decrease ( Figure 2). e letters a and b above the bars indicate significant differences (P < 0.05) between the groups.

Effect of Consumption of Lactobacillus Strains on Gastrointestinal Symptom Rating Scale Scores.
e average GSRS scores of H. pylori-infected patients in the four groups were all greater than 6.00 at baseline (Figure 3), indicating that they had functional dyspepsia. After one month of treatment with Lactobacillus strains, the scores of the three treatment groups were less than 2.50, indicating that their gastrointestinal symptoms were significantly improved compared to baseline (P < 0.001).
"ns" indicates no significant differences (P > 0.05) between the baseline and end-of-trial. " * * * " indicates significant differences (P < 0.001) between the baseline and end-of-trial.

Effects of Consumption of Lactobacillus Strains on Serum Concentrations of Pepsinogens and Inflammatory Cytokines.
Compared with the placebo treatment, the Lactobacillus strain treatments did not significantly affect the concentrations of PG I, PG II, or the PG I/PG II ratio in patients' serum (Table 4).
One month after Lactobacillus treatment, the mean serum IL-8 concentration in the G14-5M treatment group and the M2-09-R02-S146 treatment group had decreased to 6.16 pg/mL (P < 0.05) and 7.09 pg/mL (P < 0.01), respectively, which was much lower than the mean serum IL-8 concentration in the placebo treatment group (Table 4). In contrast, treatment with any of the three Lactobacillus strains did not cause striking changes in serum TNF-α concentrations (P > 0.05).  Figure 4(c) shows that the distribution of samples in each treatment group was similar and that there was no obvious clustering, indicating that treatment with Lactobacillus strains had little effect on the composition and structure of intestinal microbial communities. e letter a above the bars indicates no significant differences (P > 0.05) between the groups.

Gut Microbiome Composition in Helicobacter Pylori-Infected Patients after Lactobacillus Strain Treatment.
Compared with the placebo treatment, the administration of the three Lactobacillus strains did not significantly affect the structure of the gut microbiota at the phylum level ( Figure 5(a)). Further analysis of the composition at the genus level ( Figure 5(b)) showed that all of the treatment groups exhibited an increase in the relative abundances of Lactobacillus and Ruminococcus, and a decrease in the relative abundances of Parasutterella and Dialister after one month of Lactobacillus strain treatment, relative to placebo. Moreover, compared with the placebo treatment, the relative abundance of Prevotella was reduced in the M2-09-R02-S146 treatment group, and the relative abundances of Escherichia-Shigella and Blautia were reduced in the CCFM8610 treatment group.
ere were some differences in the composition of Lactobacillus communities at the species level between the four groups ( Figure 5(c)). e relative abundances of L. crispatus, L. helveticus, and L. plantarum were increased in the G14-5M, M2-09-R02-S146, and CCFM8610 treatment groups, respectively, consistent with the species of Lactobacillus with which each of these groups was treated.

Discussion
In this double-blind randomized controlled trial, we evaluated the efficacy of Lactobacillus strains in eliminating H. pylori infection. Compared with the placebo treatment, the 14 C-UBT value had decreased significantly in the three Lactobacillus treatment groups, and the eradication rate of H. pylori had increased significantly in the L. crispatus G14-5M treatment group at the end of the trial (Figure 2). However, the eradication rates of H. pylori in the three Lactobacillus-treated groups were different, indicating that the ability of probiotics to inhibit H. pylori infection was species-specific, which is consistent with the findings of previous studies [23,40,41]. In addition, the types and amounts of short-chain fatty acids and bacteriocins secreted by different Lactobacillus species can affect their abilities to inhibit H. pylori in the stomach [42,43]. To date, it does not appear clear whether probiotics may be more effective in particular subgroups, and if predictive factors for treatment success can be identified. e complex physiological environment of the human body may affect the ability of probiotics to antagonize H. pylori. In addition, clinical outcomes may be related to the timing of probiotics intake. Sakamoto et al. [44] reported the efficacy of yogurt containing L. gasseri OLL2716 (LG21) in suppressing H. pylori.
ere was no significant difference in the UBT levels at weeks 0 and 9. However, consumption of the yogurt for 18 weeks reduced gastric mucosal inflammation indicating that long-term administration is necessary. It is also of concern that there are essential factors such as H. pylori infection strain, the host genetic background, and the host microbiome, that may influence the efficacy of probiotics. Studies indicated that the susceptibility to H. pylori infection and the outcome of the infection vary according to both H. pylori and/or host genetic background [45,46]. In conclusion, further research into the mechanisms underlying the direct and indirect effects of probiotics on H. pylori could help not only to better refine treatment types but also contribute to a better understanding of some aspects of H. pylori pathogenesis. e patients in each group had symptoms of gastrointestinal discomfort before treatment.
e Lactobacillus treatment groups had significantly lower GSRS scores by the end of the trial, indicating the ability of Lactobacillus to relieve gastrointestinal discomfort in patients (Figure 3). Gastrointestinal inflammation and H. pylori infection may play a role in functional dyspepsia [47]. Several clinical trials have demonstrated that a diet enriched in Lactobacillus spp. may alleviate dyspeptic symptoms [34,41,48]. e lower incidence of gastrointestinal discomfort in the treatment groups may be due to the suppression of H. pylori colonization by competition from Lactobacillus strains in the gastrointestinal tract. Furthermore, Lactobacillus strains may reduce the occurrence of adverse gastrointestinal symptoms by maintaining intestinal homeostasis via creating a lower colonic pH that favors the growth of nonpathogenic species, by stimulating immunity, or by producing antimicrobial substances [49].
IL-8, produced by gastric epithelial cells, is a key cytokine in H. pylori-associated gastritis [50]. In this study, we   Figure S4) have also shown that Lactobacillus treatment decreased the concentration of IL-8 secreted by AGS cells cocultured with H. pylori, to a value comparable to the control. Nuclear transcription factor kappa B (NF-κB) is a master regulator of proinflammatory cytokines and antiapoptotic signaling molecules, which can be activated by H. pylori through several different bacterial components and host signaling  " * * " (P � 0.0039) and " * " (P � 0.0176) indicate significant differences between the G14-5M treatment group and the placebo group. PP: per-protocol analysis; ITT: intention-to-treat population. e data of the placebo group have been previously published in Food and Fermentation Industries (DOI: 10.13995/ j.cnki.11-1802/ts.024742).
In addition, Ryan et al. [55] have proposed that the suppression of IL-8 secretion is a result of Lactobacillus spp. downregulating the expression of CagA pathogenicity island genes of H. pylori. e expression of other proinflammatory cytokines, such as TNF-α, increases in H. pylori-infected mucosa [51]. Serum PG concentrations are associated with the functional activity of the gastric mucosa, and a PGI/PGII ratio < 3 is a marker of atrophic gastritis [56]. In this study, we found that Lactobacillus treatment did not affect the serum concentrations of TNF-α or PG, which echoes the findings of previous studies [41,49].
H. pylori infection elicits significantly different population structures in the gastric, oral and intestinal microbiota, which affects microbiota homeostasis and weakens the body's defense against microorganisms with pathogenic potential [57][58][59]. Frost et al. [60] identified differences in the relative abundances of 13 intestinal microbiota genera, such as Bacteroides, Prevotella, and Parasutterella, between H. pylori-infected cases and controls.
ey also demonstrated that a high abundance of Prevotella was positively associated with H. pylori infection. In this study, we found that compared with placebo, the Lactobacillus strain treatments decreased the relative abundances of Parasutterella and Prevotella in the intestinal microbiota of patients. e treatments also decreased the abundance of specific gut microbes that have been reported to be associated with oral diseases such as periodontitis (Dialister) [61], enteric diseases such as diarrhea (Escherichia-Shigella) [62], and metabolic syndromes such as hypertriglyceridemia, fatty liver disease, and insulin resistance (Blautia) [63].
Notably, Lactobacillus strain treatments also increased the relative abundance of Ruminococcaceae, which is an important butyrate-producing family of microbes. Butyrate plays a central role in maintaining gut homeostasis [64,65]. Furthermore, the colonization of applied Lactobacillus strains not only increased the relative abundance of Lactobacillus at the genus level but also led to changes in the proportion of various intra-genus species. is may have been due to synergetic or antagonistic interactions between treatment Lactobacillus strains and those Lactobacillus species that were already present in patients.
Lactobacillus strains intervention did not affect the richness and diversity of the intestinal microbiota. Diversity is an important indicator of the productivity, function, and stability of gut microecosystems; however, the diversity in gut microbiota will not be as simple as "more diversity is better" [66]. It is reasonable to conclude that the diversity of the fecal microbiota was not significantly affected by probiotics administration [67]. Probiotics intervention usually significantly altered the proportion of fecal microbiota at the genus level and species level, with the overall community complexity and richness unaffected. is may be due to the influence of intestinal microbiota balance in adults. It may also be attributed to the relatively larger size and the number of overall intestinal microbiota, compared with probiotics administered. G14-5M M2-09-R02-S146 CCFM8610   " * " indicates significant differences (P < 0.05) between baseline and end-of-trial. " * * " indicates significant differences (P < 0.01) between baseline and end-oftrial.

Conclusion
Overall, the findings demonstrated that the 14 C-UBT value of the three Lactobacillus treatment groups had decreased significantly by the end of the trial. e eradication rate of H. pylori was significantly elevated by a one-month treatment with a L. crispatus G14-5M regimen. Treatment with Lactobacillus strains also reduced the GSRS score, serum IL-8 concentrations, and the abundance of specific gut microbes that have been linked to H. pylori infection. e three Lactobacillus strains had no significant effect on the physiological indicators of patients. Taken together, these data suggest that the role of probiotics in patients with H. pylori infection may be species/strain specific.

Data Availability
e data used to support the findings of this study are included within the article and the supplementary materials.

Disclosure
Shumin Wang and Meiyi Zhang contributed to the work equally and are the co-first authors.  Table S1. distribution of the CRISPR-Cas regions among Lactobacillus strains. Table S2. potential virulence factor of Lactobacillus strains. Table S3.distribution of the intact prophage regions among Lactobacillus strains. Table S4. antibiotic resistance of Lactobacillus strains. Figure S1. the inhibition of Lactobacillus strains on the growth of Helicobacter pylori. Figure S2. inhibitory effect of Lactobacillus strains on the adhesion of H. pylori to the human gastric adenocarcinoma cell line (AGS). Figure S3. colonization of Lactobacillus strains in the stomach of mice. Figure S4. changes in IL-8 production of AGS after Lactobacillus strains intervention. Figure S5. changes in virulence factor expression after Lactobacillus strains intervention. Figure S6. distribution of resistance genes in Lactobacillus strains. (Supplementary Materials)