Bacteria strains and culture conditions
H. pylori NCTC 11637, Sydney strain 1 (SS1) and clinical strains used in this study have been identified using 16S rDNA sequencing. H. pylori strains were cultured under microaerophilic condition (5% O2, 10% CO2 and 85% N2) on solid culture media which composed of Karmali Agar Base (CM0935, Oxoid, UK) and 5% sterile defibrinated sheep blood (XLF Medical Sales Co., Beijing, China).
Bacteria growth at different concentrations of zinc compounds
Culture mediums containing different concentrations of zinc acetate or polaprezinc were prepared. H. pylori NCTC 11637 was resuspended to 105 cfu/ml and 100 μl of bacteria suspension was evenly coated on culture medium and colonies were counted after incubating at 37 ℃ for 3 days.
Minimal inhibitory concentrations (MICs) of zinc compounds to H. pylori
H. pylori strains - two standard strains (NCTC 11637 and SS1) and eight clinical strains (19004, 20050, 27054, 28038, 827, L1, L2, L3) - were resuspended in sterile PBS to 2 McFarland. 100 μl bacteria suspension was evenly coated on culture medium with 0.0005 mmol/ml, 0.001 mmol/ml, 0.0025 mmol/ml, 0.005 mmol/ml, 0.01 mmol/ml of zinc compounds. After incubating at 37 ℃ for 3 days, concentration of the zinc compounds in the plate with no bacterial grow was regarded as MIC.
Zinc compound effects on MICs of antibiotics to H. pylori
H. pylori NCTC 11637 strain was used. Culture mediums containing half MIC of zinc compounds (0.0005 mmol/ml for polaprezinc and 0.00125 mmol/ml for zinc acetate) were prepared. In vitro MICs of three antibiotics: levofloxacin, clarithromycin and tetracycline were tested using the Epsilometer test (E-test, Liofilchem, Italy). The MIC value was read at the point where the bacteriostatic ring intersects the E-test strip. The clinical breakpoints for levofloxacin, clarithromycin and tetracycline were defined as: >1 μg/ml, >0.5 μg/ml, and >1 μg/ml, respectively as per European Committee on Antimicrobial Susceptibility Testing Breakpoints .
Genome DNA extraction and analyses
H. pylori NCTC 11637 was divided into two groups. The experimental group was incubated on medium contains 0.00125 mmol/ml zinc acetate for 3 days and the control group was incubated on blank medium for 3 days. DNA was extracted using sodium dodecyl sulphate (SDS) method . Qubit RNA Assay Kit in Qubit 2.0 Fluorometer (Life Technologies, USA) was used to quantify the extracted DNA, and its purity and integrity was determined by agarose gel electrophoresis. Eligible DNA samples were broken into fragments of about 350 bp using Covaris ultrasonic crushing instrument and NEBNext Ultra DNA Library Prep Kit for Illumina Kit (NEB, USA) was used for sequencing library preparation. Library quality control was performed using Agilent 2100 Bioanalyzer (Agilent Technologies, Beijing, China) and the sequencing procedure was completed by Illumina NovaSeq PE150. Library construction, sequencing and comparative analysis between control and experimental groups were performed by Novogene Tech (Beijing) Co., Ltd.
RNA preparation and transcriptomic analyses
H. pylori NCTC 11637 strain grouping was the same as Genome DNA analysis and each group were run using three replicates. Total RNA was extracted with RNAprep Pure Cell/Bacteria Kit (DP430, TIANGEN, China) and quality control was assessed using agar gel electrophoresis, NanoPhotometer spectrophotometer and Agilent 2100 bioanalyzer for integrity and purity. Eligible RNA sample was treated with Ribo-Zero™ Magnetic Kit (Bacteria) (Epicentre, USA) to remove rRNA. The cDNA library was constructed according to the Strand-specific approach . For the quality of the cDNA library, Qubit2.0 Fluorometer was used for preliminary quantification and then qRT-PCR was used for accurate quantification to ensure its effective concentration was above 2 nM. Library construction and sequencing were conducted by Novogene Tech (Beijing) Co., Ltd. Raw reads were filtered and mapped to the H. pylori NCTC 11637 genome using Bowtie2. Gene expression counts were quantified using HTSeq. Differentially expressed genes (DEGs) were analysed using DESeq2. Genes with absolute log2 fold changes > 0.5 and multiple p-value < 0.05 were considered as DEGs.
Gene expression measurement by qRT-PCR
Quantitative real-time polymerase chain reaction (qRT-PCR) assays were applied to detect the targeted gene expression. The same samples as in the RNA-sequencing transcriptomic analyses were used, including 3 controls incubated on blank medium and 3 incubated on medium contains 0.00125 mmol/ml zinc acetate. Based on transcriptomic analyses, 6 differentially expressed genes were selected for qRT-PCR validation. Gene-specific primers were designed using Primer Premier 5.0 and first strand cDNA was synthesized using HiFiScript cDNA Synthesis Kit (Cowin Bio, China). qRT-PCR reactions were performed in technical triplicates using an ABI StepOne Real-Time PCR System (Perkin-Elmer Applied Biosystems, Foster City, CA). Reaction component and process are shown in Supplementary Table 6. Gene expression and log2 fold changes were analyzed using the 2-ΔΔCT algorithm.
GO and KEGG pathway analyses
First we mapped genes to Entrez Gene symbols and then gene sets were annotated by GO (Gene Ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) databases [20, 21]. GO enrichment analysis (cellular component (CC), molecular function (MF), and biological process (BP)) and KEGG pathway enrichment analysis of DEGs were performed using clusterProfiler R package (v3.4.4), while the H. pylori 11637 reference strain annotated by Pfam (protein families database) was used as background, and Benjamini–Hochberg adjusted p-value < 0.05 was considered significantly enriched.
Construction of ssrA gene knocking-out strain
ssrA gene knock-out in H. pylori NCTC 11637 was constructed according to the published protocol  and summarised in Supplementary Table 7. The plasmid PILL570, which contains a kanamycin resistance gene, was kindly provided by Professor Yundong Sun from Shandong University, China. The recombinant plasmid for knocking out oipA gene (used as positive control in this study) was kindly provided by Professor Yong Xie, Nanchang University, China. Briefly. We amplified the upstream and downstream homologous arms of ssrA and then inserted them into PILL570 and transformed the reconstructed plasmid into the bacteria. The targeted genes (ssrA and oipA) would be replaced by the kanamycin resistance gene.
Knock-down of ssrA by siRNA
Levofloxacin-resistant H. pylori NCTC 11637 strain with a MIC of 4-6 μg/ml induced by gradient levofloxacin exposure in vitro was used for analysis. SiRNA against ssrA was designed and synthesized by Saisofi Biotech (Jiangsu) Co., Ltd, China. - the sequence is shown in Supplementary Table 8. SiRNA transfection was performed following the published protocols . H. pylori strains were harvested from plates and washed three times with sterile 10% ice glycerol solution, 100 μl mixture was blended with 10 μg of siRNA and 109 CFU bacteria was transferred to 0.1 cm electroporation cups (Biorad, USA). Transfection was performed by electroporation using 2500V for 5 ms. At 30 min, 1 h and 2 h after electro-transformation, bacterial RNA was extracted and reverse transcribed according to manufacturer’s protocols (Vazyme, China). The qRT-PCR reaction components and processes are detailed in Supplementary Table 9. MICs of levofloxacin to H. pylori after ssrA was knocked down at different time points were tested using the E-strip method.