Helicobacter pylori strains containing the preserving liquid were removed from the refrigerator where they were stored at −80°C (standard strains 26695, NSH57, MSD132, and G27 were presented by professor Hongkai Bi from Nanjing Medical University; the clinical strains HPBS001-HPBS016 were isolated by our laboratory), Information on other strains are provided in Supplementary Table 1. Helicobacter pylori strains were cultured on the Columbia blood agar plate (OXOID, UK) or in the brain heart infusion (BHI, OXOID, UK) broth medium containing 10% serum (Pingrui, China). Bacterial species from non-H. pylori group were used as an experimental control and were cultured on nutrient agar (NA) plates or Luria-Bertani (LB) plates.

The active ingredients of F. suspensa purchased from Chengdu Herbpurify Co., Ltd. (CAS: 487-39-8, etc.) were dissolved into 4 mg/ml using absolute ethanol. The first well of a 96-well plate with bacteria was added with 173.6 μl of culture medium and then with 6.4 μl of antibacterial drugs, and every well from the first to the eighth was all diluted in proportion (the drug concentrations of the first to eighth wells were 128, 64, 32, 16, 8, 4, 2, and 1 μg/ml) with negative wells (sterile, only with medium, and drugs) and positive wells (no drugs, only with medium, and bacteria) as controls. Bacteria at the logarithmic growth phase were taken from the solid plate and made into bacterial suspensions with the corresponding mediums: the concentration of H. pylori was adjusted to 0.3 [1.0 × 10⁸ colony forming unit (CFU)/ml] and diluted 10 times to 1.0 × 10⁷ CFU/ml. The concentration of other bacteria at OD600 was adjusted to 0.3 (1 × 10⁸ CFU/ml) and diluted 100 times to 1 × 10⁶ CFU/ml. The concentration of fungi was adjusted to 0.5 (5 × 10⁶ CFU/ml) and diluted 1,000 times to 5 × 10³ CFU/ml. Notably, 10 μl was taken and added from the first to eighth well (the concentration of bacterium per well was about 1.0 × 10⁶ CFU/ml) and the wells with bacteria were cultivated for 24–72 h to judge the results (Huang et al., 2019).

Phillygenin was dissolved to 4 mg/ml by absolute ethanol. The first well of the 96-well plate with bacteria was added with 173.6 μl culture medium (pH 3.0, pH 4.5, pH 6.0, and pH 7.0) with 90 μl culture medium added to the other wells. Thereafter, 6.4 ml of phillygenin was added to the first well, with every well from the first to the fifth well diluted in proportion (the drug concentrations were 128, 64, 32, and 16, 8 μg/ml) and phosphate-buffered saline (PBS) (sangon, China) used as a positive control. The H. pylori G27 strains at the logarithmic growth phase on the solid medium were employed to make a bacterial suspension with a BHI medium; the concentration of the bacteria liquid was adjusted to 1 × 10⁸ CFU/ml (OD600 was 0.3), diluted by ten times, and kept in reserve. The first to fifth wells were added with 10 μl of the bacteria liquid (the concentration of bacteria liquid per well was about 1 × 10⁶ CFU/ml) and cultured in a three-gas incubator. The bacteria liquid after the drug action for a certain period of time (such as 2 h) was diluted (100 times, 1,000 times, etc.), coated on a Columbia agar plate, and cultured in a three-gas incubator for 4–5 days. The number of bacteria growing on the agar plate was calculated, with the drug concentration at which the number of bacteria was suppressed by 99.9% as the MBC.

Helicobacter pylori G27 strain was used to detect the drug resistance of phillygenin. First, the MICs for metronidazole and phillygenin were 2 and 16 μg/ml, respectively. The induction on the strains was performed with one-fourth MIC concentration of metronidazole and phillygenin, with the detection performed every 3 days during a total of 24 days of induction. The induced concentrations were adjusted with the change of MICs. For example, the induced concentration was adjusted to 4 μg/ml when the MIC of metronidazole was 16 μg/ml.

The cell suspensions of gastric epithelial cell (GES)-1 and BGC823 (KeyGEN BioTECH, Nanjing, China) were prepared with their concentrations adjusted to 1 × 10⁵. The suspensions were then inoculated into a 96-well plate, 100 μl per well. Three repeats of the same sample were performed and cultured in the incubator at 37°C for 24 h. Moreover, 10 μl of phillygenin (the working concentrations were 300, 200, 100, 50, and 0 μg/ml) was added to each well and inoculated at 37°C for 24 h. Furthermore, 10 μl of CCK8 (Beyotime, China) was added per well, tapped, and mixed well and then incubated for 4 h. The absorbance at 450 nm was measured, and the survival rate was calculated according to the formula: cell survival rate = [(As-Ab)]/[(Ac-Ab)] × 100%. As refers to the wells that contain the culture medium of cells, drugs, and CCK-8. Ac refers to the wells that contain the culture medium of cells and CCK-8 with no drug. Ab refers to the wells that contain the culture medium of cells and CCK-8, with no cell and drug. A survival curve based on the survival rate was established.

Specific pathogen-free (SPF) C57BL/6 mice aged 6–8 weeks were purchased from Changsha Tianqin Biological Co., Ltd., the number of SPF animal license: SYXK Gui 2017-0004, and animal experiment Ethics Number: No. 2019112501. The evaluation of drug efficacy for animal in vivo as follows was performed in accordance with the same experiment ethics. The mice were raised in an SPF environment and randomly divided into the medicine administration group and negative control group with 10 in each group. Ten times therapeutic dose of phillygenin was administered to the treatment group for 3 consecutive days, once a day; the negative control group was given the PBS buffer solution with the same times of administration and dosage as the administration group. The mice were weighed a day before the administration and weighed after the administration for 7 days. On the third day after the drug withdrawal, the mice in the infection group were weighed, and their average weight was calculated. The blood from their eyeballs was collected. Thereafter, the mice were sacrificed by cervical dislocation, their stomachs, kidneys, livers, and spleens were made into pathological sections, stained with H & E.

Phillygenin, omeprazole (Sigma-Aldrich, Germany), amoxicillin (Sigma-Aldrich, Germany), and clarithromycin (Sigma-Aldrich, Germany) were dissolved and diluted to 10 mg/ml. The models of SPF C57BL/6 mice aged 6 weeks were established (HPBS001). The mice were divided into four groups, namely, the omeprazole + amoxicillin + clarithromycin group (the triple-therapy group), the omeprazole + phillygenin group (28 mg/kg), the omeprazole + phillygenin group (7 mg/kg), and the PBS group, each with 10 mice. In addition, 10 mice that were not infected with H. pylori were treated as the negative control. The treatment group was given an intragastric administration. The groups that contained omeprazole were administered omeprazole first and then other drugs 30 min later. After the administration, the mice were made to fast and deprived of water for 4 h. The dosage was 138.2 mg/kg of omeprazole, 28.5 mg/kg of amoxicillin, and 14.3 mg/kg of clarithromycin, once a day for 3 consecutive days; the control group was given the PBS buffer solution with the same volume and times of administration as above. Two days after the drug withdrawal, the blood was collected from the eyeballs of the mice, which were then sacrificed by cervical dislocation with their stomach tissues taken and broken to acquire H. pylori that was then isolated, cultured, and identified with the amount of colonization calculated. A part of the stomach tissues was made into paraffin sections with H&E staining, TUNEL immunohistochemistry, and fluorescence immunoassay performed thereon.

The OD of the H. pylori G27 bacterial suspension was adjusted to 0.1 and inoculated into a 96-well plate under microaerobic conditions at 37°C for 3 days to form H. pylori biofilms (Hathroubi et al., 2020), which were then added with phillygenin (concentrations were 128, 64, 32, and 16 μg/ml). The antibiofilm effect of phillygenin was evaluated through crystal violet (Macklin, China) staining and detection using the Alamar blue assay (Solarbio, China). Biofilms were stained with LIVE/DEAD BacLight Bacterial Viability Kit (Thermo Fisher, America) in the dark for 15 min and then visualized using the confocal laser scanning microscopy. The wavelength of the SYTO 9-stained live cells that had been activated was 411 nm, while that of the PI-stained dead cells was 568 nm. The protein content of these biofilms was determined by the BCA Protein Quantification Kit (Beyotime, China).

Helicobacter pylori G27 was cultured to the logarithmic phase, and the concentration of the bacteria solution was adjusted to 1 × 10⁷ CFU/ml. Meanwhile, three gradients of the working concentration of phillygenin were set, which were 32, 64, and 128 μg/ml, respectively. Polymyxin B (Macklin, China) was used for the positive control and PBS for the negative control. They were cultured for 2 h and centrifuged to acquire the supernatant and bacteria. Adenosine triphosphate (ATP) was detected by ATP Assay Kit (Beyotime, China) through the multifunctional enzyme microplate reader (BioTek, USA).

In the exploration of the semi-inhibition curve of phillygenin, the concentration of the bacterial suspension was adjusted to 1 × 10⁸ CFU/ml, with drug action lasting (16, 32, 64, and 128 μg/ml) for 0, 2, and 8 h. Thereafter, the OD value at 600 nm (HITACHI, Japan) was measured with the semi-inhibitory drug concentration confirmed when the OD value remained unchanged. The bacteria were extracted for a transcriptome sequencing, which was completed by Nanjing Fengzi Bio-pharm Technology using the second-generation Illumina high-throughput sequencing platform and PE150 sequencing strategy. As a result, a total of 80,448,750 raw read pairs were measured with 79,746,845 clean read pairs obtained after quality control. Bowtie2 and Rockhopper were employed to carry out the comparison and sliced transcript analysis. Thereafter, all genes were quantitatively analyzed, and a total of 1,214 differential genes were identified. A functional enrichment analysis was performed on these differential genes to explore their features.

Helicobacter pylori G27 was cultured to the logarithmic phase with the concentration of the bacterial solution adjusted to 1 × 10⁸ CFU/ml. The bacterial solution was added with phillygenin for a certain time based on the semi-inhibitory growth curve and centrifuged to obtain the precipitate. Zero hour was labeled as F_1 group, 2 h as F_2 group, and 8 h as F_3 group. Three biological repeats were performed in each group, labeled as A, B, and C. The TRIZOL reagent (Takara, China) was used to extract RNA, the transcription of which was reversed into cDNA through a reverse transcription kit (MonPure, China). RT-qPCR was performed applying the LightCycler according to the RT-PCR kit (MonPure, China); the sample was subjected to 95° predenaturation for 30 s, 95° denaturation for 10 s, and 60° annealing and extension for 30 s, about 40 cycles. The primers are displayed in Supplementary Table 2. The 16S rRNA amplicon (Thermo Fisher, America) was used as an internal control for data standardization, and the remaining primers were purchased from Sangon Biotech (Shanghai). The change at the transcriptional level was determined by applying the relative quantification method (2^−ΔΔCT). The dissociation curve was analyzed to verify the homogeneity of the products.

All data were expressed as mean ± standard deviation (SD). Data analyzed by one-way analysis of variance were performed using SPSS 25.0 and p < 0.05 considered to be statistically significant.

A total of 12 components of F. suspensa were screened, among which phillygenin had the best antibacterial effect. The liquid chromatogram of phillygenin is displayed in Figure 1. The MIC of phillygenin against H. pylori was 16–32 μg/ml, and the MICs of other components were all >128 μg/ml, as illustrated in Table 1. The inhibition effects of phillygenin on 18 H. pylori strains were tested, in which it was found to have a good inhibitory effect on sensitive, drug-resistant, and multiple-resistant strains with the MIC being 16–32 μg/ml, as displayed in Table 2. The MBC of phillygenin against H. pylori was 16 times the MIC in the normal medium (PH 7.0), with the antibacterial rate reaching 99.9% after 6 h and 99.999% after 8 h. The antibacterial rate was 90, 99, and 99.9% with phillygenin at a concentration of 8 times the MIC for 4, 6, and 8 h in a dose-and time-dependent manner, as illustrated in Figure 2.

The efficacy of phillygenin on H. pylori in vivo was evaluated by the acute gastritis mice models, which had been confirmed to be infected with H. pylori (HPBS001). According to the counted amount of colonization, the antibacterial effect of phillygenin was better than that of the triple therapy with no significant difference between high and low concentrations, as displayed in Figure 3A. The H&E staining and immunohistochemical images of the phillygenin group showed that the number of the apoptotic cells in the gastric mucosae of the mice and that of the inflammatory factors were significantly decreased (Figure 3B). In addition, the expressions of several inflammatory factors in the samples of gastric tissues were detected, with results revealing that the expression levels of interleukin (IL)-6, tumor necrosis factor (TNF)-α, and IL-1β were the lowest in the phillygenin group, as shown in Figures 3B,C, which suggested that phillygenin had a good antibacterial effect in vivo.

The toxicity test of phillygenin was carried out. The results showed that phillygenin at 100 μg/ml had no obvious cytotoxic effect on GES-1 and gastric cancer cells BGC823, with a survival rate above 90% (Figures 4A,B). After the intragastric administration of 10 times the therapeutic dose of phillygenin, there was no significant change in the weights of mice within 7 days, as shown in Figure 4C. Besides, no obvious pathological damage was found in the stomachs, livers, spleens, and kidneys of mice, as shown in Figure 4D. Phillygenin that was found to have a low toxicity in vitro and in vivo and high safety could be used as the first-line therapy for H. pylori.

To check the selectivity of phillygenin against H. pylori, a total of 20 non-H. pylori strains were used and showed MICs all >128 μg/ml. Phillygenin that was found to have a single effect on H. pylori (Supplementary Table 3) was a narrow-spectrum antibiotic with a specificity and not much impact on other microflora. The inhibitory effects of phillygenin under different pH conditions (i.e., 3.0, 4.5, 6.0, and 7.0) were explored for it exerted effects after entering the stomach under low pH conditions, as shown in Figure 5A. At pH 3.0 and 4.5, the antibacterial rate reached 99.9% 2 h after the administration of 8 times and 16 times MICs of phillygenin, which could not be found at pH 6.0 and 7.0. These results indicated that phillygenin were more effective in an acidic environment with acid resistance. The drug resistance induction of phillygenin on H. pylori G27 strains was detected (Figure 5B). The results showed that during the 24 days drug resistance induction, the MIC of phillygenin only doubled on the 12th day with no change at other times, while the MIC of metronidazole increased by 64 times. Phillygenin that was shown to have difficulty in making H. pylori develop resistance could be used for clinical treatment.

The formation of biofilms is an important cause for the drug resistance of H. pylori (Krzyzek et al., 2021). Phillygenin was shown to inhibit biofilms of H. pylori at 100–150 μg/ml (about 6–8 times MIC) through a detection employing the Alamar blue assay (Figure 6A). Besides, it was found through the crystal violet staining assay that a drug concentration eight times of the MIC could inhibit 50% of the biofilm growth, with an effect better than that of clarithromycin (Figure 6B). Since the main component of biofilm is protein, the detection of the protein content of the biofilms of H. pylori was carried out. The results showed that a drug concentration 8 times of the MIC could inhibit 50% of the protein in biofilms, which was consistent with the results of the crystal violet staining and Alamar blue assay, as shown in Figure 6C. Phillygenin at a concentration 8 times of the MIC could destroy biofilms of H. pylori as shown in Figure 6D photographed by the confocal microscope at a 400 magnification.

ATP serves as the direct energy source for life activities and changes in levels of ATP will affect functions of cell. After the effect of phillygenin on H. pylori for 2 h, the intracellular ATP gradually leaked to the outside of the cells with most thereof leaking at a drug concentration of two times of the MIC. However, almost all the ATP leaked at both the drug concentrations of four and eight times of the MIC. The degree of leakage was equivalent to that of polymyxin B in a dose- and time-dependent manner (Figure 6E).

The detection of the expression of biofilm-related genes was mainly conducted by detecting the differential expressions of phillygenin on suspended bacteria and bacteria biofilms at 0 h and 4 h. The effect of a drug concentration two times of the MIC on suspended bacteria was shown in Figure 6F; the effect of a drug concentration four times of the MIC on suspended bacteria was displayed in Figure 6G. After drug action, the Hp1174 genes that formed through biofilm regulation were significantly downregulated, with the remaining ones upregulated. The effect of a drug concentration two times of the MIC on the bacteria in biofilms was as shown in Figure 6H and that of a drug concentration four times of the MIC on the bacteria in biofilms was displayed in Figure 6I. After the drug action, the spoT genes and Hp1174 genes that regulated the formation of biofilms were significantly downregulated; after the effects of phillygenin at the drug concentrations two times and four times of the MIC, there was no significant change in hefA, HP0939, HP0497, and HP0471. The Hp1174 genes involved in the biofilm formation, whether in the suspension or bacteria in biofilms, were significantly downregulated after drug actions at different concentrations. Therefore, it could be inferred that phillygenin might inhibit biofilm formation mainly through the downregulation of Hp1174 genes. It is noteworthy that in planktonic bacteria, spoT genes were upregulated, while spoT genes in biofilm bacteria were downregulated, which might be attributed to the characteristics of the bifunctional hydrolase in spoT (Bahareh et al., 2017): guanosine tetraphosphate ((p)ppGpp) that was the key to bacterial biofilm formation could be both hydrolyzed and promoted; therefore, it could be inferred that in suspended bacteria, spoT might hydrolyze (p)ppGpp and promote the generation thereof after biofilm formation (Cai et al., 2020).

In the proposed transcriptome sequencing, the drug action of phillygenin at different times was first detected, and the curves of half inhibitory concentrations were drawn as shown in Figure 7A, whereas, there was no change in the OD value at a drug concentration of 2 times of the MIC. The RNA-seq correlation analysis of the results showed a good biological repeatability (Figure 7B). Using the PCA, differences were shown between groups (Figure 7C). Through a pairwise comparison, 929 differential genes were detected between F_1 group and F_2 group, of which 445 were upregulated and 484 were downregulated; 1,015 differential genes were detected between F_1 group and F_3 group, of which 467 were upregulated and 548 were downregulated; 596 differential genes were detected between F_2 group and F_3 group, of which 315 were upregulated and 281 were downregulated. In the Venn diagram of differentially expressed genes, it was shown that there were 299 repetitive differentially expressed genes among three groups (Figure 7D). The differentially expressed genes were drawn into a gene ontology (GO) enrichment analysis histogram and divided into three categories, namely, biological processes, cellular components, and molecular functions. The differential genes between each group were found to mostly concentrate on pathways such as obsolete electron transport, transporter activity, ion binding, etc. (Figure 7E). According to the correlation between groups (log 2-fold change), and the comparison of repeated genes between groups to the exclusion of unknown proteins, six repeated differentially expressed genes that were most relevant were screened (including two upregulated and four downregulated ones), as shown in Table 3. According to the results of Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment and GO analysis, the differential genes mainly concentrated on the pathway for efflux transport, which was further verified by RT-qPCR to compare the differential expression levels of the effect of phillygenin on suspended bacteria with 0 h and 8 h, as shown in Figure 7F. Nhac, caggamma, MATE, and MdoB genes were downregulated, and flagellin A and lptB genes were upregulated, which were consistent with the results of the transcriptome sequencing.