The NCBI GenBank was used to retrieve the nucleotide sequence of A. baumannii ompA (GenBank: AY485227.1). Two pairs of specific PCR primers (P1A/P1B and P2A/P2B) were designed: ompA-P1A, 5′-GGAAAGTCTATCAAGTGTTTGTAGGATC CAAA-3′ (BamHI); ompA-P1B, 5′-AACTTCTACTACAGGAG CAGCAGGCTCTCGAGAAA-3′ (XhoI); ompA-P2A, 5′-ATCA AGCCGTACGTATTATTAGGTGCTCGAGAAA-3′ (XhoI); and ompA-P2B, 5′-TTACTGTTCAAACT-3′ (XhoI). Using P1A/P1B and P2A/P2B as primers and the A. baumannii standard strain ATCC17978 [American Type Culture Collection (ATCC), United States]as a template, the left and right fragments of the ompA gene were amplified as the homology arms of homologous recombination. The pMD18-T (TaKaRa Bio, Japan) vector and the target fragment were ligated with DNA ligase (TaKaRa Bio, Japan) to transform DH5α (TaKaRa Bio, Japan) competent cells. Then, the positive plasmids were named p18T-ompA-P1 and p18T-ompA-P2 and verified by sequencing.

The recombinant plasmids p18T-ompA-P1 and p18T-ompA-P2 were extracted. BamHI and XhoI was used for double digestion for the p18T-ompA-P1 recombinant plasmid. In contrast, XhoI and SphI was used for double digestion for the p18T-ompA-P2 recombinant plasmid. The pMD18-T vector plasmid was subjected to double digestion with the BamHI and SphI. The target fragments were purified by a gel extraction kit.

Double digestion at 37°C for 3 h with SphI and BamHI was performed for the p18T-ompA recombinant plasmid and pDS132 (ATCC, United States) plasmid. The 6,000-bp band of pDS132 double digestion product and the 1,500-bp band of p18T-ompA double digestion product were extracted. The products extracted from the above two gel extraction kits were ligated overnight at 16°C. The ligation products were transformed into E. coli competent cells and the required pDS132-ompA recombinant plasmid were screened using LB (Sangon Biotech, China) plates containing Amp + (100 μg/mL). Then, a single colony was selected into the LB liquid medium containing the same concentration of Amp + and underwent shaking culture at 37°C. This was until the optical density (OD) value was about 1.0, which was measured at 600 nm. The positive recombinant plasmid after PCR and digestion identification was named pDS13-ompA.

For the screening of the ΔompA of A. baumannii, it was constructed by conjugative transduction of a double-parent filter membrane (Griffiths et al., 2000). E. coli (pDS132-ompA) was used as the donor strain. The A. baumannii standard strain ATCC17978 was used as the recipient strain for conjugative transduction. The identification of the ΔompA strain of A. baumannii was performed through a successful growth in Kanamycin (0.02 mg/mL) resistance selective medium and confirmed using western blot analysis and PCR identification.

The primer sequences of the full-length ompA gene were designed and amplified (ompA-F1, 5′-GGAAAGTCTATCAA GTGTTTGTATG-3′; and ompA-R1: 5′-CGAGTCGCTTTTT TACTGTTCA-3′). Using the A. baumannii standard strain ATCC17978 as a template, the fragment size was about 1,235 bp. The PCR product of the amplified full-length ompA gene was subjected to gel electrophoresis. The DNA fragment was extracted and ligated to the pMD18-T vector and transformed into DH5α competent cells. The positive clone pT-ompA was selected and submitted to Genewiz Suzhou Biotechnology for sequencing. Double digestion at 37°C for 3 h with AccI and BglII was performed for the pT-ompA recombinant plasmid and pWH1266 (ATCC, United States) vector plasmid.

The 7,000-bp band of the pWH1266 double digestion product and the 1,200-bp band of the pT-ompA double digestion product were extracted. The products extracted from the above two gel extraction kits were ligated overnight at 16°C, and the ligation products were transformed into ΔompA competent cells. An LB plate containing Amp + (100 μg/mL) was used to screen the required pWH1266-ompA recombinant plasmid. A single colony was selected into the LB liquid medium containing the same concentration of Amp + and underwent shaking culture at 37°C until the OD value was about 1.0. After western blot and PCR confirmation, the ΔompA + pWH1266-ompA strain (+ ompA) was obtained.

The A. baumannii wild-type strain (A. baumannii standard strain ATCC17978), ΔompA strain, and + ompA strain after activation overnight were transferred to the LB culture medium with 2mM FeSO₄ added at 1:100. Then, transferred to the LB culture medium added with 350 μM (350 μmol/L) 2,2′-bipyridine solution for culture in the shaker at 37°C for 8–9 h and 3–4 h, respectively; then, the bacteria were collected. Western blot was used to detect the OmpA contents in the three groups.

The concentration of RAW246.7 [National Collection of Authenticated Cell Cultures (NCTC), United Kingdom]macrophages was adjusted to 2 × 10³/mL with complete culture medium [10% fetal bovine serum (FBS) and 2% double-antibody were added to Dulbecco’s modified eagle medium (DMEM) purchased from Thermo Fisher Scientific, United States] and seeded in 96-well plates with 100 μL per well. The RAW246.7 macrophages were added with 8logCFU/mL of each A. baumannii strain. Normal saline was used as the control group. After 24 h of treatment, an MTT kit (cell proliferation kit I, Sigma-Aldrich, Germany)was used to detect cell viability. After adding MTT for 2 h, the Vario Skan Flash multi-functional microplate reader (Thermo Fisher Scientific, United States) was used to measure the optical density (OD) value of each well at 450 nm.

Female nude mice (4–6-week-old) were obtained from Vital River Laboratory Animal Technology (Beijing, China) and kept in a specific pathogen-free environment at 25°C, 55% humidity, and 12-h light cycles with free access sterilized food and water. The experiments were approved by the animal ethics committee of the Ethics Committee of Longyan First Affiliated Hospital of Fujian Medical University, Longyan, Fujian, China (2017026 and 2020067). The mice were checked for their health status, and animal welfare supervision was provided by a certified veterinarian. All animal experiments were carried out in accordance with the Chinese governing laws on the use of medical laboratory animals (authorization no. 551998, 2013, by the Ministry of Health). Two A. baumannii wild-type strains, ΔompA strain, and + ompA strain, were inoculated into LB culture medium. They were placed in a 37°C incubator for 20–24 h and then mixed with 10% (w/v) porcine mucoprotein at 1:1 to prepare a bacterial suspension. The bacterial content was determined by the spread plate method. The minimum lethal concentration and the maximum non-lethal concentration of the isolates for mice were first measured through preliminary testing. Then formal testing was carried out on this basis. First, 20 mice were randomly assigned to five groups (the experimental groups) and then administrated with 1 mL of normal saline and 1 mL of a bacterial suspension at an intraperitoneal injection dose of 5.7 × 10⁸ CFU/mL, 1.14 × 10⁸ CFU/mL, 2.28 × 10⁷ CFU/mL, 4.56 × 10⁶ CFU/mL, and 9.12 × 10⁵ CFU/mL. The blank control group was injected with 1 mL of normal saline. After 7 days of observation, the disease onset and death of mice and pathological necropsy changes were recorded. The modified Karber method (Rath et al., 2011) was used to calculate the LD50. The mice were sacrificed by cervical dislocation, and the spleens were immediately harvested and ground. Smears were prepared and stained with the Gram solution to observe the bacteria.

Spleen tissue homogenates (10%, wt/vol) were prepared with cold 1 × PBS (pH 7.2). TNF-α, IL-1β, IL-6, and IL-8 levels were measured in spleen tissues using mice enzyme-linked immunosorbent assay kits (Thermo Fisher Scientific, Waltham, MA, United States) according to the instructions.

The strain of A. baumannii was lysed and collected with Bacterial Lysis Buffer (Cat# C500003, Sangon Biotech, China) supplemented with protease inhibitor cocktail (Cat# P8340, Roche). The lysis products were electrophoresed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to polyvinylidene fluoride (PVDF) membranes. Incubation of the suggested primary antibodies OmpA (Abcam) and secondary antibodies (Abcam) was implemented in a 5% milk blocking reagent. Autoradiograms were quantified through densitometry controlled by GAPDH. The assay was repeated three times.

Complete RNA was extracted from the strain of A. baumannii with TRIzol reagent (Cat# 15596018, Life Technologies) in keeping with the manufacturer’s instructions. Total RNA of 2 μg was reverse transcribed according to the protocol for the PrimeScript^® RT Master Mix Perfect Real-Time (Cat# RR047A, TaKaRa Bio, Japan). The qPCR reactions were implemented with SYBR Green Master Mix (Cat# RR820A, TaKaRa Bio, Japan) on a quantitative PCR instrument (Applied Biosystems 7900HT, Thermo Fisher Scientific).

All statistical analyses were performed using SPSS 22.0 (IBM, Armonk, NY, United States) and GraphPad Prism 5 (GraphPad Software Inc., San Diego, CA, United States). All continuous data are presented as means ± standard deviations and were analyzed using Student’s t-test or ANOVA with the LSD post hoc test. P-values < 0.05 were considered statistically significant.

The A. baumannii wild-type strain was 5 × 10⁷ CFU/mL used. The ompA expression of A. baumannii cultured in the LB culture medium added with 2mM FeSO₄ was higher than that cultured after addition with 350 μM (350 μmol/L) of the iron-chelating agent 2,2′-bipyridine (Figure 1).

The results of the MTT test showed that the A. baumannii wild type and + ompA bacterial suspensions had a remarkable toxic effect on the RAW246.7 macrophages (P < 0.05). On the other hand, the ΔompA bacterial suspension had a significantly reduced toxic effect on the RAW246.7 macrophages (P < 0.05) (Figure 2A).

The LD50 of the A. baumannii wild-type strain was 5 × 10⁷ CFU/mL (Table 1). The LD50 of the ΔompA strain was > 5 × 10⁸ CFU/mL (Table 1). The LD50 of the + ompA strain was 5 × 10⁷–1 × 10⁸ CFU/mL (Table 1).

The levels of the inflammatory factors IL-1β, IL-6, IL-8, and TNFα and the number of viable bacteria in the mice spleen were significantly increased in the + ompA strain treatment group compared with the ΔompA strain group (all P < 0.05) (Figures 2B–F). In addition, the levels were higher in the presence of iron ions than in the presence of the chelating agent. The invasion of the spleen by the + ompA strain was higher than by the ΔompA strain and higher in the presence of iron than in the presence of the chelating agent. The details of strain construction are provided in Supplementary Data and Supplementary Figure 1.

The p18T-ompA recombinant plasmid was identified by double digestion with BamH I/XhoI and Xho I/Sph I, respectively. After digestion, an ompA-P1 fragment with a size of 753 bp was present and a size of 737 bp. The p18T-ompA recombinant plasmid and pDS132 plasmid were subjected to double digestion with Sph I and BamH I. After double digestion of pDS132, a 6000-bp band of the product was extracted. Similarly, after double digestion of p18T-ompA, a 1500-bp band of the product was extracted (see Figure 3).