Dehydrated media and supplements were procured from HiMedia laboratories, India. Taq DNA polymerase, dNTPs, IPTG and adjuvants were from Sigma–Aldrich, India. Pfu DNA polymerase was from Fermentas, United States. pRSET vector series, Escherichia coli DH5α and BL21DE3 strains are from Invitrogen, United States. K. pneumoniae ATCC 13883 and ATCC 10031 strains were procured from MicroBiologics Inc., United States. All the cell culture reagents were from Sigma–Aldrich, India. Specific pathogen-free BALB/c mice used in the study were received from Central Animal Housing Facility of Defence Food Research Laboratory (DFRL), Mysore city, India.

Genomic DNA of K. pneumoniae ATCC 13883 strain was used for amplification of ompK36 and ompA genes for chimeric gene construction. Additionally, K. pneumoniae E-1716 a clinical isolate belonging to K1 serotype isolated from sputum sample collected from a patient admitted at Sri Dharmasthala Manjunatheshwara (SDM) College of medical science and hospital, Dharwad, India was used for challenge experiments based on its higher virulence (lower LD₅₀ values) potential. For animal experiments, K. pneumoniae was cultured from glycerol stocks in BHI broth and incubated overnight at 37°C with vigorous shaking. Following centrifugation, bacterial cells were washed two times and suspended in 1x PBS solution. LD₅₀values were determined by intraperitoneal administration of different serial dilutions of three K. pneumoniae strains.

Primers were designed for amplification of ompA and ompK36 genes based on gene sequences available at GenBank (Table 1). The individual genes were fused using PCR by ‘gene splicing by overlap extension’ (SOE-PCR) method employing the primers shown in Table 1. Restriction sites (BamHI and Hind III) were added to 5′ end of primers so as to clone the resulting fusion gene in-frame with the coding sequence of pRSETA vector. The recombinant plasmid was transferred to E. coli DH5α and BL21(DE3) strains. A single transformed BL21(DE3) clone was subjected to induction with IPTG followed by testing for chimeric protein expression by SDS-PAGE analysis and Western blotting with anti-histidine antibodies. For bulk purification, recombinant clone was cultured in 500 ml Luria-Bertani broth followed by purification of histidine-tagged (6xHis) protein using Ni-NTA resin (Qiagen) by gravity flow protocol as per manufacturer’s instructions.

Six to seven-week-old female BALB/c mice were grouped into three batches of 10 mice each. Blood was drawn from all the mice 2 days prior to the start of immunization schedule and the pooled serum served as negative control in all the experiments. Two groups of mice received 50 μg of recombinant protein subcutaneously in emulsion with Freund’s complete adjuvant followed by three booster doses at 10 days interval with the same quantity of protein in Freund’s incomplete adjuvant. Third group served as a control receiving only adjuvants in 1x PBS by intraperitoneal route. Blood was drawn 2 days prior to first immunization, 1 day prior to the second booster and 7 days past the last immunization and the pooled sera were used for evaluation of antibody titers and isotypes. Of the 20 immunized mice, two sets of 6 mice each were used in active challenge and memory groups. Three mice were used for splenocyte stimulation experiments. Remaining mice served as a backup in case of accidental deaths. Of the 10 sham-immunized mice, 6 were used in active challenge and 3 were used for splenocyte proliferation assay. For passive immunization and challenge naïve mice were used.

Antibody titers of the pooled serum were measured after the second and the third booster immunizations by indirect ELISA employing r-AK36 protein as antigen. Maximum dilution at which the mean absorbance (OD₆₀₀) was at least two times more or equals the mean value of negative control was considered as endpoint titers. Serum samples were diluted 1:500 in PBS followed by antibody isotype determination employing the mouse monoclonal antibody isotyping reagents (Sigma, United States) as per manufacturer’s recommendations. Sham-immunized serum served as controls during isotype determination.

Western blot analysis was performed to test the reactivity of r-AK36 hyperimmune sera with native OmpA and OmpK36 proteins from various K. pneumoniae strains along with other Enterobacteriaceae members. Whole-cell lysates of various bacterial strains were separated on 12% SDS-PAGE gels followed by electrotransfer on to nitrocellulose membrane. Membranes were blocked in 5% skim milk solution followed by washing in PBS-Tween 20 (0.05%) and probing with 1:1000 dilutions of anti-r-AK36 sera. After washing, membranes were treated with 1:3000 dilutions of goat anti-mouse IgG conjugated with horseradish peroxidase (HRP). Blots were developed either with diaminobenzidine tetrahydrochloride (Sigma–Aldrich, India) and 30% H₂O₂ or ready to use tetramethylbenzidine (Bangalore Genei, India) solution.

Seven days after the last immunization, aseptically isolated spleens from immunized and sham-immunized mice were gently ruptured on a wire mesh using sterile glass rod with round edges to break splenic capsule and separating connective tissue followed by flushing with sterile DMEM. The resultant splenocytes were treated with ACK buffer (150 mM NH₄Cl, 10 mM KHCO₃ and 0.1 mM EDTA) to remove RBCs. The splenocytes were resuspended in Dulbecco’s Modified Eagle’s medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum (Hyclone, Thermo scientific, United States), 5 mM glutamine, 50 U/ml penicillin, 50 μg/ml of streptomycin and 0.2% NaHCO₃. This experiment was performed three times with three mice from active and sham-immunized groups. The number of splenocytes used for the stimulation studies was 10⁶ cells/well. The antigen (recombinant protein) was used at a concentration of 2, 4, 6, 8, and 10 μg/well. Concanavalin A (Sigma–Aldrich, United States) was included as a mitogen for positive control at a concentration of 10 μg/ml. The experiment was performed in triplicates. The plates were incubated at 37°C in 5% CO₂ for 72 h. The proliferation was measured using MTT reagent (Sigma, India) as described elsewhere (Reddy et al., 2015). Lymphocyte proliferation was calculated as stimulation index employing the following formula: Stimulation index = OD₅₇₀ with antigen/OD₅₇₀ without antigen. The supernatants from cultured splenocytes were evaluated for cytokine estimation (IL-2, IL-4, IL-10, IFN-γ, and TNF-α) employing commercial kits (MabTech, Nacka, Sweden) based on sandwich ELISA.

To determine the antibacterial activity of r-AK36 antibodies against K. pneumoniae, microtiter plate-based in vitro antimicrobial susceptibility test was carried out as per CLSI recommendations. In brief, 1 mg/ml of antibody stock was serially diluted in PBS to 100, 75, 50, 25, and 10 μg/ml concentrations and 100 μl volumes of each dilution were tested for antimicrobial activity in triplicates. Approximately 100 μl of 10⁶/ml CFU of K. pneumoniae ATCC 13883, K. pneumoniae ATCC 10031 and K. pneumoniae E-1716 strain diluted in double-strength Mueller-Hinton broth were added to each antibody dilution in three replicates. Only the PBS with no antibody and K. pneumoniae in MH broth served as negative control. Only PBS along with MH broth served as a blank. Rifampicin antibiotic served as positive control during antimicrobial testing. Absorbance at OD₆₀₀ was recorded before the start of experiment followed by incubation of plates at 37°C overnight. The lowest antibody dilution at which there was no increase in turbidity after overnight incubation was considered as MIC of the polyclonal antibodies.

Effect of sub-inhibitory concentrations of r-AK36 antibodies on biofilm formation was assessed by microtiter plate method as described by Sun et al. (2005) with minor modifications. K. pneumoniae standard and clinical isolates were grown overnight in BHI broth at 37°C. Bacterial cells were harvested and the cell count was adjusted to 10⁶ CFU/ml using sterile double-strength BHI broth (with 1% glucose) and 100 μl was mixed with 100 μl of 50 μg/ml polyclonal antibodies and incubated at 37°C overnight with shaking. Post-incubation, plate was washed two times in PBS to remove non-adherent cells. Adherent biofilm was fixed by adding 200 μl of methanol for 10 min. After drying the plates, 250 μl of 1% crystal violet was added to each well and incubated for 5 min. Plates were washed under running water, dried and the adherent biofilm was solubilised by adding 250 μl of 33% glacial acetic acid. OD was measured at 570 nm in Tecan multimode plate reader. A control was included for each strain with only BHI broth and glucose without antibodies. Each sample was tested in triplicates. The percent inhibition of biofilm formation of each K. pneumoniae strain was determined using the formula: (OD₅₇₀ in control – OD₅₇₀ in treatment)/(OD₅₇₀ in control) × 100.

For studying the protective efficacy of the recombinant protein, the mice were challenged with a hypervirulent strain (K1 serotype) of K. pneumoniae whose lethal dose (LD₅₀) had been determined to be 3 × 10⁵ CFU. Two groups of six mice each of both immunized and sham-immunized were challenged intraperitoneally 10 days after the last immunization with 10⁶ CFU in 100 μL of 0.9% saline. The protective efficacy of the recombinant protein was checked by evaluating the survival rates after 15 days of challenge. Every day the animals were observed for signs of illness, such as weight loss, change in coat appearance, and lack of movement. On day 15 the remaining mice were terminated to collect blood for antibacterial assay and organs were collected for testing bacterial load. Similar challenge experiments were performed to test the persistence of immunization and memory response 120 days after the last immunization and observed for survival rates.

After the 15 days observation period following bacterial challenge, two mice were humanely euthanized by CO₂ sedation followed by cervical dislocation. Before excising the organs, the surface of the viscera was swiped with sterile cotton swabs to check the bacterial translocation and cultured in BHI broth. Further kidney, spleen, liver, and lungs were excised and transferred to sterile preweighed Eppendorf tubes. The tissues from organs were macerated by plastic tissue grinder and the resultant homogenate and blood samples were plated on plate count agar. The colony count was measured following incubation at 37°C for 24 h.

The in vivo prophylactic and neutralization efficacy of the hyperimmunepolysera was determined by performing passive antibody administration where in 6-week-old naïve female BALB/c mice (n = 6) were injected intraperitoneally with 200 μL of polysera and left for 24 h followed by challenge with 2 × LD₁₀₀ (4 × 10⁵ CFU) of K. pneumoniae E-1716 strain. A control group (n = 6) was injected with same quantity of PBS alone followed by challenge with same quantity of lethal dosage. Mice were observed for a period of 14 days for any possible deaths. Survival of mice post-challenge during the observation period is represented by Kaplan–Meier graphs using GraphPad Prism 5.0 version.

The data were presented either as mean ± SD or individual values+grand median. Mantel-Cox (log-rank) test was used to compare the survival curves and Student’s t-test was used for other statistical comparisons. All graphical illustrations were constructed by GraphPad Prism 5 software. Statistical difference between different groups were analyzed by univariate ANOVA using Minitab 17. Significance (P)-value summary: ^∗P ≤ 0.05, ^∗∗P ≤ 0.01, ^∗∗∗P ≤ 0.001; no significance #.

All the animal experiments, biosafety procedures and genetic manipulations undertaken in this study have been performed with permission and following the guidelines laid by the Institutional Animal Ethics Committee (IAEC) and the Institutional Bio-Safety Committee (IBSC) of Defence Food Research Laboratory, Mysore. Permission for using the experimental animals was approved by Institutional Animal Ethical Committee during the meeting held on 4th December, 2015 at DFRL, Mysore. Animals were received from DFRL animal housing facility (28/1999/CPCSEA dt. 11 March, 1999) and maintained under hygienic conditions in animal house of Microbiology Department of DFRL, Mysore. The clinical isolate used for challenge experiments in this study was received as pure culture from the culture collection of SDM hospital after taking mutual consents of Head of microbiology departments of SDM hospital and DFRL. Direct sampling of clinical specimens was not carried out in this study. The IAEC of DFRL approved the protocols employed.

Fusion gene comprising immunodominant regions of K. pneumoniae ompA and ompK36 genes was generated by SOE-PCR (Figures 1A,B) and cloned into pRSETA vector. The chimeric r-AK36 gene was sequenced and was found not to possess any deletions or point mutations when compared to the expected sequence. Transformation of recombinant plasmid in E. coli host was followed by induction with IPTG which resulted in the expression of a protein at 61 kDa region as observed in SDS-PAGE (Figure 1C) and Western blot analysis (Figure 1D). This was in accordance with the predicted size of the protein deduced from the fusion gene sequence in vector. Histidine-tagged (6xhis) recombinant protein in inclusion bodies was extracted and purified from 500 ml induced culture by immobilized metal affinity chromatography under denaturing conditions (Figure 1E). Purified protein was subjected to step dialysis with decreasing concentrations of urea under reducing environment. Final protein concentration was quantified and adjusted to 2.5 mg/ml by Lowry’s calorimetric assay against BSA standard.

In immunized mice, a progressive and significant increase of anti-r-AK36 antibodies was observed. After three booster doses (Figure 2A), the end-point antibody titers reached 1:64000, while the control sera from sham-immunized mice showed no reactivity with r-AK36 in indirect ELISA (Figure 2B). Pooled sera from r-AK36-immunized mice, when subjected to isotype analysis, a Th2-type immune response was observed indicated by the predominance of IgG1 immunoglobulin. Surprisingly, IgG2b levels also predominated the immune response indicating Th1 polarization. IgM antibodies also persisted in serum even after last immunization and Ig class switching (Figure 2C). Despite significant IgG2b levels, the relative degree of contribution by Th2- and Th1-type humoral immune responses determined as IgG1:IgG2a ratio is >1, indicating r-AK36 in CFA/IFA adjuvant system induced primarily Th2-type antibody response in BALB/c mice.

The specific reactivity of anti-r-AK36 antibodies with the native OMPs of various K. pneumoniae and other bacterial species was assessed by Western blot analysis. Anti-r-AK36 antibodies showed strong reactivity with native OmpA (37 kDa) and OmpK36 (39.6 kDa) proteins from different K. pneumoniae standard and isolated strains (Figure 3). Other Gram-negative rods of Enterobacteriaceae family showed varying reactivities at different molecular sizes. This is not surprising considering the conserved nature of epitopes on OmpA and OmpK36 proteins among Enterobacteriaceae members. Reactivities of anti-r-AK36 antibodies with different bacterial species are provided in Table 2.

The r-AK36 antigen induced a significant proliferation of splenocytes (whole lymphocytes) from r-AK36-immunized mice upon reinduction by the same antigen at a concentration of 4 μg per million cells. When splenocytes from sham-immunized mice were induced with r-AK36, no significant proliferation was observed (Figure 4A). As a positive control, 10 μg/well of Concanavalin A was used and a significant increase in proliferation was observed in the lymphocyte population, confirming the efficiency of the cell purification process.

The Th2polarization of immune response suggested a role of IL-4 lymphokine in mediating IgG1 production. Also, significant proliferative index observed in r-AK36-immunized mice splenocytes prompted us to assess the levels of pro-inflammatory and anti-inflammatory cytokines from conditioned media after lymphocyte proliferation. No increase in the production of IL-4 cytokine was observed from r-AK36-immunized mice splenocytes but a significant rise in the same was observed when sham-immunized mice splenocytes were induced with r-AK36 protein (Figures 4B,C). On the other hand, IL-2, a cytokine that promotes differentiation of T-cells to effector and memory T-cells was significantly produced from r-AK36-immunized mice. IL-2 plays a role in the development of T cell immunologic memory by expanding the number and function of antigen-selected T cell clones. Also, IFN-γ was found to be produced significantly from the r-AK36-immunized mice splenocytes (Figure 4B). IFN-γ was found to be crucial for controlling pulmonary pneumoniae caused by K. pneumoniae. IFN-γ is involved in host immunological defense against K. pneumoniae infection (Yoshida et al., 2001). Splenocytes from sham-immunized mice upon induction by r-AK36 produced significant levels of IL-2 cytokine (Figure 4C).

To challenge the immunized mice with K. pneumoniae, firstly we ventured into determining LDs of various strains. The LD (LD₁₀₀ and LD₅₀) of different K. pneumoniae strains varied significantly (Table 3). In general, these values were lower for strains recovered from clinical sources rather than the standard cultures indicating their higher virulence potential. Groups of 4 mice were used for each dilution during LD determination. K. pneumoniae E-1716 strain isolated from sputum sample from an infected patient had lower LD₅₀ values than K. pneumoniae ATCC 13883 and ATCC 10031 strains. The clinical strain was also different in terms of pathology induced post-challenge and the rapidity of mortality. The clinical strain induced acute respiratory distress, lethargy and rapid death between 24 and 48 h of pathogen administration. On the other hand, mice challenged with standard strains showed reduced activity and dullness in movement and delayed death beyond 48 h of administration.

Groups of six r-AK36-immunized and sham-immunized mice were challenged with ∼10⁶ CFU (3 × LD₁₀₀) of virulent K. pneumoniae E-1716 strain by intraperitoneal route 7 days after the last booster dose immunization. At least 83% of mice survived beyond the observation period with only one death on day 11 (Figure 5). None of the sham-immunized mice survived beyond 72 h indicating the protective role of r-AK36 subunit vaccine. Gross pathology of lungs of challenged sham-immunized mice showed enlarged, swollen and dull gray regions affecting entire lobes with edema.

To assess the antibacterial activity of anti-r-AK36 antibodies, three assays were utilized, viz., MIC, biofilm inhibition and antibacterial protection conferred by passive transfer of antibodies. The MIC values were moderately uniform over a different range of K. pneumoniae strains. The MIC value of r-AK36 antibodies collected 7 days past the last booster dose immunization was determined to be in the range of 60–100 μg/ml. The inhibition of bacteria is also time and concentration dependant as seen from the Figure 6A.

Antibodies against surface proteins can inhibit biofilm formation by coating the pathogens surface and targeting for destruction. Anti-r-AK36 antibodies at subinhibitory concentrations reduced the ability of different K. pneumoniae strains belonging to various serotypes to form biofilm in vitro. Since the ability of K. pneumoniae to form biofilm varies between strains, a preliminary test was carried out to ascertain the biofilm forming potential. The percent reduction of biofilm varied significantly between different strains. The amount of biofilm formation quantified by crystal violet staining with and without addition of antibody and the percent inhibition of biofilm formation is provided in Table 4.

Passive administration of anti-r-AK36 antibodies protected the mice significantly from mortality upon challenge by 4 × 10⁵ CFU of K. pneumoniae E-1716. Approximately 67% of mice survived beyond the observation period post-challenge (Figure 6B).

Persistence of immunity was evaluated by subjecting r-AK36-immunized mice to lethal challenge with K. pneumoniae E-1716 strain 120 days past final booster. Approximately 50% of the challenged mice survived beyond the observation period in memory group (Figure 7).

Viable bacterial counts from various organs such as lungs, spleen, kidney, and liver were evaluated from challenged mice sacrificed after the observation period. Results indicated that lungs of sham-immunized mice showed higher infiltration of K. pneumoniae while lungs of r-AK36-immunized mice showed significantly reduced bacterial counts. The same trend in CFU counts was seen with other organs namely kidney, liver, and spleen. Bacterial infiltration was highest in lungs followed by liver, kidney, and colon (Table 5).