The LAC-4 (33) and AB5075 (38) strains of A. baumannii have previously been reported as highly virulent in mouse models and reproducing the main features of the pathology of a human pulmonary infection by A. baumannii. Interestingly, these strains are not closely related. They differ substantially in allelic variants of housekeeping genes and do not qualify as single- or double-locus variants of each other under either the Pasteur or Oxford schemes. LAC-4 is classified as Sequence Type (ST)10 and has not been assigned to any Global Clone (31). LAC-4 is sometimes referred to as Clonal Complex (CC)10, whereas AB5075 is classified as ST1 and belongs to Clonal Group I (39).

In our C57BL/6 intranasal infection model, we found that AB5075 exhibits lower virulence than LAC-4. Bacterial loads in the lungs at 24 hours post-infection were higher for LAC-4 ( Supplementary Figure S1A ), and this strain induced significant mortality at a dose of 2×10⁷ CFU, whereas all mice survived the same dose of AB5075 ( Supplementary Figure S2B ). Therefore, we choose to use strain LAC-4 to test the acquisition of protective memory against A. baumannii infection.

A vaccine composed of 5x10⁵ or 5x10⁷ CFU of formalin-killed A. baumannii LAC-4 administered intranasally has been shown (35) to fully protect wild-type C57BL/6 mice against a challenge with 7x10⁷ CFU of live LAC-4 that killed 80% of control unvaccinated mice. We choose to use a similar experimental model with a vaccine composed of heat-killed (HK) A. baumannii and tested whether intranasal or intraperitoneal administration of HK A. baumannii LAC-4 or AB5075 strains could induce long-lasting immunity against a semi lethal challenge with the live LAC-4 strain.

Wild-type C57BL/6 mice were inoculated intranasally or intraperitoneally with a solution of PBS (control unvaccinated group) or 10⁸ CFU of HK AB5075 or HK LAC-4 in PBS. Six weeks later, 2x10⁷ CFU of live LAC-4 was administered intranasally to all groups. Survival was evaluated in each group of mice up to 5 days post-challenge ( Figure 1A ). The bacteria count in the lungs ( Figure 1B ) and spleen ( Figure 1C ) of the mice was measured by counting the CFUs at 24 hours post-challenge.

We observed that all groups immunized with HK LAC-4 or HK AB5075 showed significantly improved survival compared to the unvaccinated controls ( Figure 1A ). As expected, immunization with LAC-4 provided stronger protection against LAC-4 infection than AB5075 immunization. Nevertheless, AB5075 immunization still conferred partial but statistically significant protection against LAC-4, indicating that vaccination with one strain of A. baumannii can induce cross-protection against genetically distant strains. Survival of HK LAC-4 groups correlated with bacterial control in both the lungs ( Figure 1B ) and spleen ( Figure 1C ). Interestingly, this is not the case with the protection conferred by HK AB5075, suggesting that survival is driven more by the nature of the inflammatory response than by the absolute bacterial burden in the organs.

Since protection against A. baumannii in mice vaccinated with formalin-killed LAC-4 was attributed to the presence of specific antibodies (35), we measured the presence of antibodies specific for LAC-4 and AB5075 in the serum of mice vaccinated intraperitoneally or intranasally with LAC-4 or AB5075. Our results show that immunization via the intraperitoneal route induces higher levels of specific IgM, IgG1 and IgG3 antibodies than immunization via the intranasal route ( Figure 2 ). As expected, immunization induces a higher level of antibodies against the vaccination strain compared to the other strain. Nevertheless, immunization with LAC-4 induces a low but statistically significant level of antibodies able to recognize AB5075 and vice versa, which may explain the moderate cross protection we observed following a challenge with LAC-4.

To better understand the nature of the protection conferred by the HK LAC-4 vaccine, we compared the production of proinflammatory cytokines induced by an intranasal infection with 2x10⁶ CFU of live LAC-4 and the intranasal administration of 10⁸ CFU of HK LAC-4. We administered a lower dose of live LAC-4 than dead LAC-4 because mice could not survive 24 hours after injection of 10⁸ CFU of live LAC-4. The levels of various cytokines were measured at 4 and 24 hours post-intranasal inoculation in the lungs and the spleen by q-PCR reaction ( Figure 3 ). We observed that HK LAC-4 induces innate proinflammatory cytokines e.g. IL-1, IL-6, TNF-α, as well as cytokines associated with both Th1 (IFN-γ) and Th17 (IL-17A, IL-17F and IL-22) responses. However, compared with live LAC-4, HK LAC-4 induces less IFN-γ but more IL-17F and IL-22, suggesting that HK LAC-4 would be particularly effective in inducing Th17 immunity.

To determine which elements of the adaptive immune response are essential for control of an intranasal infection, we studied the survival of different strains of mice deficient for key elements of the immune system following infection with LAC-4.

Wild-type mice as well as mice rendered genetically deficient for the Th1 response (IL-12p35^-/-, IFN-γ-/-), for the Th17 signaling (IL-17RA^-/-) and for selected lymphocyte populations (δTCR^-/-, TAP-1^-/-, MHCII^-/-, CD3^-/-, MuMT^-/-) were infected intranasally with 2x10⁶ CFU of live LAC-4. Survival was measured in each group for 120 hours post-infection ( Figure 4A ). We observed that the absence of IL-17RA and δTCR has dramatic consequences. Almost all mice died within 24 hours in these groups (0 and 6% survival in the 17RA- and δTCR-KO groups, respectively). The levels of IL-17A, IL-17F, IL-22 and TNF-α were measured at 4 and 24 hours post-intranasal inoculation in the lungs of wild-type and δTCR^-/- mice by q-PCR reaction ( Supplementary Figure S2 ). We observed that mice deficient for δTCR do not produce fewer mRNA for these cytokines than wild-type mice, which suggests that the absence of γδ+T cells does not impair the production of these cytokines. A deficiency in TAP1 also induces significant early mortality with only 22% survival. CD3 deficiency (which affects all T cells) is associated with 46% survival of mice, while MHCII deficiency [which affects CD4⁺ T cells (40)] and MuMT [which affects B cells (41)] allows 76% and 80% survival, respectively. A statistical analysis of these results shows that the absence of CD3 significantly reduces the survival of the animals, whereas the deficiency of MHC-II or MuMT has no significant effects on it.

Surprisingly, the CFU counts observed in the lungs and spleen of infected mice weakly correlated with survival of the mouse ( Supplementary Figure S3 ). All groups of deficient mice had a higher level of bacteria in the lungs and spleen than wild-type mice and the differences were quite small, particularly in spleen, between the deficient mice. These results suggest that it is the nature of the inflammatory immune response rather than the number of bacteria in organs that determines the animal’s survival.

Taken together, these results suggest that γδ T cells, TAP-1 dependent lymphocyte populations and IL-17RA-dependent signaling play a very important role in the control of primary intranasal LAC-4 infection while the role of IL-12 and the IFN-γ dependent Th1 response as well as CD4⁺ T cells and B cells is more moderate.

After identifying the genetic deficiencies leading to increased susceptibility to primary infection, we tested the capacity of intranasal immunization with HK LAC-4 to increase the survival of genetically immunodeficient mice against pulmonary LAC-4 infection.

Wild-type mice were inoculated intranasally with PBS (control, unvaccinated group) or 10⁸ CFU of HK LAC-4 in PBS. The survival of these mice was compared with that of various genetically immunodeficient mice vaccinated with HK LAC-4 and subjected to an intranasal challenge with a dose of 2x10⁶ CFU of live LAC-4 ( Figure 4B ). We observed that the survival rates in all groups of vaccinated immunodeficient mice were high (88-100%). This suggests that vaccination compensates for the absence of the Th1 or Th17 response as well as the TAP-1 and CD3 deficiencies.

To identify the key elements of protective memory immunity induced by HK LAC-4, we challenged vaccinated mice with a higher dose of live LAC-4, 2x10⁷ CFU. At this dose, most unvaccinated mice, whether wild-type or genetically immunodeficient, succumb to infection within 48 hours ( Figure 5A ). Interestingly, the vaccine does not provide equivalent protection against a challenge with 2x10⁷ CFU of live LAC-4 in all groups of immunodeficient mice ( Figure 5B ). MuMT^-/- mice showed the lowest survival rate, at 5%, thus confirming the key role of antibodies in protective memory against A. baumannii. Vaccinated TAP1^-/-, δTCR^-/- and IL-17A^-/-mice also show statistically significant reduction of survival, at 59%, 57% and 43% respectively, compared to vaccinated wild-type mice (94%). These findings demonstrate that vaccination cannot completely compensate for the absence of cell populations that depend on TAP1 and δTCR or of the IL-17RA-dependent Th17 response. In contrast, vaccine protection can be induced in the absence of CD3 or MHCII-dependent T cells as well as the IL-12_p35 or IFN-γ dependent Th1 response.

Since the humoral response is a key element of protection against A. baumannii infection induced by HK LAC-4 in our experimental model, we attempted to identify the main antigenic targets of the protective antibodies. To this end, we produced a LAC-4 strain deficient for the itrA gene (as described in the Materials and Methods section), which is essential for bacterial capsule synthesis (42), and compared the vaccine capacity of this ΔitrA strain to that of the wild-type strain.

Wild-type C57BL/6 mice were vaccinated intranasally with 10⁸ CFU of HK wild-type or ΔitrA LAC-4 strains. At 6 weeks post-vaccination, mice were infected with 2x10⁷ CFU of live wild-type LAC-4 strain. As expected, most mice vaccinated with wild-type HK LAC-4 survived (89%) ( Figure 6A ). In striking contrast, mice vaccinated with HK ΔitrA LAC-4 had a very low survival rate (20%), close to that of unvaccinated mice (6%). Measurement by ELISA of the levels of antibodies specific to wild-type LAC-4 in the blood of vaccinated animals showed that mice vaccinated with HK ΔitrA did not have IgM, IgG1 and IgG3 antibodies recognizing wild-type LAC-4 ( Figure 6B ).

To better identify the targets of protective antibodies against LAC-4 infection, we conducted Western blot analyses ( Figure 6C ) comparing antigens recognized by sera from mice vaccinated with PBS (control), wild-type, or capsule-deficient ΔitrA LAC-4 strains. Extracts from wild-type LAC-4 and ΔitrA LAC-4 were tested. Sera from mice immunized with wild-type LAC-4 recognized a large count of high–molecular-weight antigens in extract from wild-type LAC-4, whereas no corresponding signal was detected in capsule-deficient ΔitrA LAC-4 extract. This intense signal is not revealed by Coomassie blue staining. We also detected with serum from mice vaccinated with wild-type LAC-4 strains the presence of low molecular weight antigens in both extracts that are visible in Coomassie blue. Overall, these results demonstrate that protective antibodies induced by vaccination with HK wild-type LAC-4 are directed against non-proteinaceous capsule antigens and against some low molecular weight protein antigens.

Since hospitalized patients at risk for A. baumannii infection are often immunocompromised (9), we tested the ability of intranasal vaccination with HK LAC-4 to protect cyclophosphamide-treated mice against intranasal infection with LAC-4. Wild-type C57BL/6 mice were inoculated intranasally with PBS (control) or 10⁸ CFU of HK LAC-4. Five weeks later, half of the mice in each group were inoculated intraperitoneally with PBS or underwent cyclophosphamide treatment, as described in the Materials and Methods. One week post- cyclophosphamide treatment (6 weeks post-vaccination), all groups of mice received 2×10⁶ CFU of live LAC-4 intranasally. Survival of the mice was measured up to 120 hours post-challenge.

As shown in Figure 7 , in agreement with a previous study (37), cyclophosphamide treatment greatly reduced the survival of infected wild-type C57BL/6 mice. We observed that intraperitoneal immunization with HK LAC-4 significantly reduced mortality in mice treated with cyclophosphamide. Vaccine protection, however, was not complete in the cyclophosphamide-treated mice and the protection provided did not exceed 65%.

To evaluate the ability of HK LAC-4 to induce an adaptive immune response in humans, we measured in vitro the activation of dendritic cells and T lymphocytes from Peripheral Blood Mononuclear Cells (PBMC) of healthy donors in response to HK LAC-4.

Monocytes isolated from human PBMC were differentiated for 5 days to dendritic cells in the presence of IL-4+GM-CSF and then stimulated for 48 hours with 2x10⁶ CFU/ml of HK A. baumannii LAC-4, HK Brucella melitensis, HK Escherichia coli or 1 µg/ml of LPS from E. coli. B. melitensis is a stealthy bacterium that expresses an atypical and poorly activating LPS (43). HK B. melitensis was used as a negative control in our experiments. We used HK E. coli and commercial LPS as positive controls. The expression of dendritic cell surface maturation markers was analyzed by flow cytometry. We observed that HK LAC-4 significantly up-regulated CD40, CD80, CD83 and CD86 maturation markers and down-regulated CD209 (a lectin receptor mostly expressed by immature DCs) ( Figure 8A ). Moreover, HK LAC-4 induced significant secretion of IL-23 and IL-12_p70 in our experimental setting ( Figure 8B ).

To evaluate the capacity of HK LAC-4 to activate human T lymphocytes, total PBMC were stimulated for 7 days with 2x10⁶ CFU/ml of HK A. baumannii LAC-4, HK B. melitensis or HK E. coli. We used 2 µg/ml CEF+CEFTA (pool of peptides-MHC class I and II-restricted T-cell epitopes from human Cytomegalovirus, Epstein Barr virus, Influenza virus, Tetanus toxin and Adenovirus 5) and 1 µg/ml of SEB (Staphylococcal enterotoxin B) as the positive control for CD4⁺ and CD8⁺ T cell proliferation and 10 µg/ml HMBPP (4-hydroxy-3-methyl-but-2-enyl pyrophosphate) as the positive control for γδ⁺ T cell proliferation. We observed that HK LAC-4 induced significant proliferation of CD4⁺ T cells, CD8⁺ T cells and γδ+ T cells ( Figure 9A ) as well as easily detectable secretion of IFN-γ, IL-17A and TNF-α ( Figure 9B ). To determine the cell source of these cytokines, we analyzed stimulated total PBMC by intracellular flow cytometry. Our results showed ( Figure 10 ) that HK LAC-4 induced significant production of IFN-γ by CD8⁺ and γδ+ T cells, of IL-17A by CD4⁺, CD8⁺ and γδ ⁺ T cells and of TNF-α by CD4⁺ and γδ+ T cells.

Overall, these results show that HK LAC-4 can induce in vitro dendritic cell maturation as well as activation of all T cell subsets, suggesting that it might be able to activate lymphocytes in patients.

The procedures used in this study and the handling of the mice complied with current European legislation (Directive 86/609/EEC). The Animal Welfare Committee of the Université de Namur (UNamur, Belgium) reviewed and approved the complete protocol for Acinetobacter baumannii infection (Permit Number: UN-LE-20/343, UN-LE-22/378). Whole blood for the isolation of human PBMC was collected from healthy donors as described in the ethical protocol/amendment IXP-001_V3 (Belgium; Reg. No. B6702014215858), protocol IXP-003_V1 (Belgium; Reg. No. B707201627607) or protocol IXP-004_V1 (The Netherlands; Reg. No. NL57912.075.16).

Wild-type C57BL/6 mice were obtained from Harlan (Bicester, UK). IFN-γ^−/−, δTCR^−/−, CD3ϵ^−/−, MuMT^−/− C57BL/6 mice were purchased from The Jackson Laboratory (Bar Harbor, ME). IL-12_p35 ^−/− C57BL/6 mice were obtained from Dr. B. Ryffel (University of Orleans, France). IL-17RA^−/− C57BL/6 mice were obtained from Dr. K. Huygen (Belgian Scientific Institute for Public Health, Brussels, Belgium). TAP1^−/− C57BL/6 mice and MHCII^−/− C57BL/6 mice were acquired from Jörg Reimann (University of Ulm, Ulm, Germany). All wild-type and deficient mice used in this study were bred in the animal facility of the Gosselies campus of the Université Libre de Bruxelles (ULB, Belgium).

Cryopreserved primary PBMC were isolated from whole blood donated by healthy volunteers. Written informed consent was obtained from all patients enrolled in the study. All blood samples were tested and found negative for hepatitis B virus (HBV), hepatitis C virus (HCV), and human immunodeficiency virus (HIV). PBMC were separated from the blood by density gradient centrifugation and subsequently cryopreserved in fetal bovine serum (FBS), supplemented with 10% dimethyl sulfoxide, by controlled rate freezing. The PBMC were kept in cryogenic storage (−180 °C) until use.

A. baumannii LAC-4 (31, 32) (ASM78673v1 genome, https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_000786735.1/) was obtained from Wangxue Chen (Human Health Therapeutics Research Center, National Research Council of Canada, Ottawa, ON, Canada). A. baumannii AB5075 was isolated in 2008 from a combatant wound infection (38) and was obtained from Dr. C. Van der Henst (Structural Biology Brussels, Vrije Universiteit Brussel (VUB) and Microbial Resistance and Drug Discovery, VIB-VUB Center for Structural Biology, VIB, Flanders Institute for Biotechnology, Brussels, Belgium). A. baumannii was always handled in BSL-2 or BSL-3 containment at the Université Libre de Namur (UNamur, Namur).

To generate the non-capsulated A. baumannii LAC-4 strain, the itrA3 gene (ABLAC_36890 from the LAC-4 reference genome CP007712) was obtained using the integrative plasmid pK18-apraR [pUC derivative (75)] generated using the following primers: itrA3_up_for_short: GCAAGTATCGAAGCAATATCAG; itrA3_up_rev_short:GGCCCAATTCGCCCTATAGTGAGTCGGGATTAATGCGGTTAAGGAAATTACG; itrA3_down_rev_short: CCTAATTTTGGATGTTCTAAGCC; itrA3_down_for_short:CCGACTCACTATAGGGCGAATTGGGCCGCTCCAGATCAGTTAGTAGG. Briefly, the pK18-ΔitrA3 was integrated into the parental wild-type LAC-4 strain and excised using the sucrose counter-selection marker as previously described (42). The lack of capsule of the LAC-4ΔitrA3 strain was confirmed by density gradients as previously described (76).

Bacteria stored at -80°C in LB-glycerol 30% were inoculated in LB medium and incubated overnight at 37°C under agitation until the stationary phase. Then the culture was diluted to an OD600 (Optical Density measured at a wavelength of 600 nm) of 0.2 for AB5075 and 0.4 for LAC-4 and then was grown until an OD600 of 0.7-0.8. This log-phase culture was then centrifugated for 10 minutes at 5000 g and the pellet was washed twice with Phosphate-Buffered Saline (PBS). Then it was diluted to the target concentration prior to infection. The infectious dose was assessed by plating serial dilutions.

Mice were anesthetized with a cocktail of Xylazine (9 mg/kg) and Ketamine (36 mg/kg) in PBS before being inoculated by intranasal injection with the indicated dose of A. baumannii in 30 µl of PBS. Mice infected intraperitoneally were injected with the indicated dose in 500 µL. Control animals were inoculated with the same volume of PBS. The infectious doses were validated by plating serial dilutions of the inoculums. At the selected time after infection, mice were sacrificed by cervical dislocation. Immediately after sacrifice, the lungs and spleen were collected for bacterial count and qRT-PCR analyses.

To carry out vaccinations as well as in vitro tests, different species and strains of bacteria were killed by heat. We used the previously described A. baumannii strains as well as Escherichia coli K12 MG1655 and B. melitensis strain 16M (Biotype1; ATCC 23456; American Type Culture Collection). To prepare Heat-Killed (HK) bacteria, overnight cultures of bacteria were grown under shaking at 37°C in 2YT media (Luria-Bertani (LB) broth 32 g/L with yeast extract 5 g/L and peptone 10 g/L) for B. melitensis and in LB media for A. baumannii and E. coli. B. melitensis and E. coli from an overnight liquid culture were washed twice in PBS (10 minutes at 3500 g for Brucella and 5000 g for E. coli). A. baumannii from an overnight liquid culture were washed twice in PBS (10 minutes at 5000 g) and diluted to an optical density of 0.4 in fresh LM media and growth for 90 minutes at 37°C under constant agitation. and then washed twice in PBS. After washing, the bacteria were resuspended to a concentration of 2x10⁹ CFU/ml in PBS. This solution was heated at 80°C for 1 hour in a water bath. Following this treatment, the bacterial solution was stored at -20°C before use. Each batch of killed bacteria was verified to confirm the killing by growing it in an incubator at 37°C and counting the CFU after 24 hours for A. baumannii and E. coli and after 72 hours for B. melitensis.

C57BL/6 mice were unvaccinated or vaccinated by intranasal or intraperitoneal injection with 10⁸ CFU of strains of the A. baumannii of interest, as indicated. Samples of blood for antibody detection were collected 4 weeks post-vaccination for analysis of the humoral response against A. baumannii. Six weeks post-vaccination, mice were challenged intranasally with live LAC-4 bacteria (2x10⁶ or 2x10⁷ CFU, as indicated). In some experiments, monohydrate cyclophosphamide (Sigma) diluted in PBS was administrated intraperitoneally at 4 days (150 mg/kg) and 1 day (100 mg/kg) before the challenge. Lungs and spleen were harvested 24 and 120 hours post-challenge and the CFU levels in each organ were counted.

Organs were homogenized in PBS/0.1% X-100 Triton (Sigma-Aldrich). We performed successive serial dilutions in PBS to obtain the most accurate bacterial count and plated them on LB medium. The CFU were counted after 1 day of incubation at 37°C.

RNA from the whole lungs and spleen was extracted by the TRIzol™/Chloroform method. Briefly, the lungs and spleen were harvested from mice and ground in 1 mL of TRIzol™ (TRIzol™ Reagent - Invitrogen™) with a Polytron Homogenizer. After 5 minutes of incubation at room temperature, 200 µL of chloroform were added and the tube was mixed vigorously, then incubated for 2–3 minutes at room temperature. After centrifugation for 15 minutes (12,000 g at 4°C), the aqueous phase was recovered and RNA was extracted using the “NucleoSpin RNA Plus (Macherey-Nagel)” kit according to the manufacturer’s protocol.

The RNA was then reverse transcribed with Superscript II reverse transcriptase (Invitrogen) using hexamer random primers as described by the manufacturer. A condition without reverse transcriptase was also conducted in parallel as a negative control. cDNA was then mixed with SybrGreen mix (FastStart Universal SYBR Green Master (Rox) - Roche) and the appropriate primer sets and subjected to qRT-PCR in a LightCycler 480 (Roche). The forward and reverse primers used were 5’-GGCAACTGTTCCTGAACTCA-3’ and 5’-GGGTCCGTCAACTTCAAAGA-3’ for IL-1β, 5’-CCTGTCTATACCACTTCACA-3’ and 5’-CTCTTTTCTCATTTCCACGATTTCC-3’ for IL-6, 5’-GCCTCCCTCTCATCAGTTCTA-3’ and 5’-GCTACGACGTGGGCTACAG-3’ for TNF-α, 5’-TGCCAAGTTTGAGGTCAACA-3’ and 5’-GAATCAGCAGCGACTCCTTT-3’ for IFN-γ, 5’-ATCCCTCAAAGCTCAGCGTGTC-3’ and 5’-GGGTCTTCATTGCGGTGGAGAG-3’ for IL-17A, 5’-CCCTGGAGGATAACACTGTGA-3’ and 5’-AATTCACGTGGGACAGAAATG-3’ for IL-17F and 5’-CAGCAGCCATACATCGTCAA-3’ and 5’-GCCGGACATCTGTGTTGTTA-3’ for IL-22. Ribosomal Protein L32 (RPL32, forward 5’-ACATCGGTTATGGGAGCAAC-3’, reverse 5’-TCCAGCTCCTTGACATTGTG-3’) mRNA was used as the reference housekeeping gene for normalization. A total of 40 two-step cycles were performed as follows: 95°C for 15 seconds and 60°C for 1 minute preceded by 10 minutes at 95°C for activation. Melting curves were then performed to assess the primer specificity. The target mRNA fold change was calculated based on the 2^-ΔΔCt formula, using the RPL32 gene as the reference gene and RNA from naïve mice as the standard condition. Four biological replicates were performed for each gene tested.

The presence of Acinetobacter baumannii-specific murine IgM, IgG1 and IgG3 was determined by ELISA. Polystyrene plates (269620; Nunc) were coated with Heat-Killed (HK) A. baumannii LAC-4 or AB5075 at a dose of 10⁷ CFU/mL, as indicated, and incubated overnight at 4°C. The plates were blocked for 2 hours at room temperature with 200 μl/well of PBS-1% Bovine Serum Albumin (BSA). Then, plates were incubated with 50 μl/well of plasma in serial dilutions in PBS-0.1% BSA. The plasma of uninfected mice and PBS were used as negative controls. After four washes with PBS, isotype-specific goat anti-mouse HRP conjugated antibodies were added (50 μl/well) at appropriate dilutions (anti-IgM from Sigma-Aldrich; anti IgG1 LO-MG1–13 HRPO and LO-anti IgG3 MG3–13 HRPO from LOIMEX). After 1 hour of incubation at room temperature, plates were washed four times in PBS and 100 μl/well of TMB substrate solution (BD OptEiA Kit) was added. After 15 minutes of incubation at room temperature in the dark, the enzyme reaction was stopped by adding 25 μl/well of 2 N H₂SO₄. The absorbance was measured at 450 nm.

Bacterial cultures of A. baumannii were normalized at 2 x 10⁹ bacteria/mL, then heat killed as previously described. Two milliliters of culture were concentrated in 300 µL of PBS containing the loading buffer (2% SDS, 5% β-mercaptoethanol, 10% glycerol, 0.03 M Tris HCl pH6.8) and samples were heated at 90 °C for 10 minutes before being loaded on 4-20% mini-PROTEAN acrylamide gels (BioRad), with 20 µL of samples per well. Migration was performed for 1 hour at 150V. After migration, one gel was kept for Coomassie Blue staining, while the others had their proteins transferred onto a nitrocellulose membrane which was blocked overnight in PBS supplemented with 0.05% Tween (PBS-T) and 5% milk. For primary antibodies, sera from 8 different vaccinated mice were pooled together and diluted 4 times in PBS-T supplemented with 1% milk. Incubation of the membranes with their respective primary antibodies were performed statically for 1 hour on clean petri dishes with drops of 300 µL for each membrane. Membranes were then washed in PBS-T 4 times for 5 min, before being incubated with secondary antibodies (polyclonal Rabbit anti-Mouse Ig/HRP, Agilent) diluted 5000 times in PBS-T supplemented with 1% milk. After 4 washings, proteins were revealed using Clarity Western ECL Substrate (Biorad) with Image Quant LAS 4000 (General Electric).

Peripheral Blood Mononuclear Cells (PBMC) from healthy human donors were thawed from the ImmunXperts biobank. Monocytes were isolated from PBMC using a MACS magnetic separation column (Miltenyi) following the manufacturer’s instructions and purity was evaluated by CD14 staining using a BD Lsrfortessa X-20 flow cytometer. Cells were then resuspended to a cellular density of 10⁶cells/ml and plated on a tissue culture treated 24-well plate (1ml per well) in CellGenix DC medium (CellGenix) with added Gentamicin antibiotic (Sigma-Aldrich), IL-4 and GM-CSF for 5 days. At day 5, cells were stained for FACS analysis with several DC activation markers to assess their immature (iDC) state (CD14 Miltenyi, CD40 BD, CD80 BD, CD83 Miltenyi, CD86 Miltenyi, CD209 Miltenyi and HLA-DR BD). On the same day, the appropriate stimuli (various heat-killed bacteria and LPS from E. coli from Sigma-Aldrich) were added to the cell culture for 48 hours. At day 7, cells were stained for FACS analysis with the same markers used at day 5 to assess their mature state (mDC). Supernatant at day 7 was collected for further analysis.

Supernatant from the cell culture was collected and stored at -80°C. After thawing, the presence of the cytokines IL-12p70, IL-23 and IL-17A was detected using LEGEND MAX™ Human ELISA kits (BioLegend) with pre-coated plates following the manufacturer’s instructions.

Supernatant from cell culture was collected and stored at -80°C. After thawing, the presence of the cytokines IFN-γ and TNF-α was detected using the Luminex technology following the manufacturer’s instructions (MILLIPLEX MAP Human Cytokine/Chemokine kit, Merck).

Peripheral Blood Mononuclear Cells (PBMC) from healthy human donors were thawed from the ImmunXperts biobank and incubated for 6 days at 37°C with the appropriate stimuli: CEF+CEFTA (pool of 35 peptides-MHC class I and II-restricted T-cell epitopes from human Cytomegalovirus, Epstein Barr virus, Influenza virus, Tetanus toxin and Adenovirus 5 from Mabtech), HMBPP (4-Hydroxy-3-methylbut-2-enyl diphosphate lithium salt, Sigma-Aldrich) and various heat-killed bacteria. T cell proliferation was assessed by measuring the incorporation of 5-Ethynyl-2´-deoxyuridine (EdU). EdU is a thymidine analogue which is incorporated into the DNA of dividing cells during the S phase. On day 6, cell cultures were pulsed with EdU at 1 µM for approximately 16 hours. The next day, cell culture supernatants were collected and stored at -80°C for cytokine analysis. The remaining cells were fluorescently stained for viability and T cell surface markers (CD3 BD, CD4 BD, CD8 BD and γδTCR-Miltenyi), fixed, permeabilized and the incorporated EdU was stained with a fluorescent azide (Click-iT™ EdU Flow Cytometry Assay Kit, Invitrogen, Thermofisher Scientific). Cells were then acquired on a BD Lsrfortessa X-20.

Peripheral Blood Mononuclear Cells (PBMC) from healthy human donors were thawed from the ImmunXperts biobank and incubated for 7 days at 37°C with the appropriate stimuli. At day 7, part of the supernatant was collected, and the remaining cells were treated with Brefeldin A/Monensin (Sigma) for 4–5 hours at 37°C following the manufacturer’s instructions. For some conditions, cells were only stimulated for the last 4–5 hours with Brefeldin A/Monensin. Cells were then stained for viability and extracellular markers (CD3-BD, CD4-BD, CD8-BD and γδTCR-Miltenyi), fixed and permeabilized (BD Cytofix/Cytoperm™ Fixation/Permeabilization Kit) and intracellularly stained for the cytokines IFN-γ-BD, TNF-α-BD and IL-17A-eBiosciences. Cells were then acquired on a BD Fortessa X-20.

For comparisons of CFU counts between groups, statistical analyses were performed using the non-parametric one-way ANOVA followed by multiple comparison tests in GraphPad Prism. Survival curves were compared using the log-rank (Mantel–Cox) test in GraphPad Prism. To analyze ELISA data, we use two-way Anova with Dunnett’s multiple comparisons test in GraphPad Prism. A p-value < 0.05 was considered statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).