All Bordetella spp. were cultured on 100 x 15mm Petri dishes containing Difco Bordet-Gengou (BG) agar (BD, cat. 248200) supplemented with 10% sheep defibrinated blood and 20 μg/mL streptomycin, herein referred to as BGS agar (Dewan et al., 2017; Gestal et al., 2018; Gestal et al., 2019). For liquid cultures, Luria-Bertani or Stainer-Scholte media were used (Gestal et al., 2018). The Bordetella strains used include the B. pertussis strain Bp536, B. parapertussis strain Bpp12822, and B. bronchiseptica strain BbRB50 (Gestal et al., 2018). The RB50ΔbtrS knockout (Gestal et al., 2019) and the T3SS knockout (RB50ΔbscN) mutants were generated in a previous study (Yuk et al., 1998). See Table 1 for the bacterial strains used for the following experiments and their associated references.

To evaluate the antimicrobial activity of VIP (Sigma Aldrich, cat number V6130-250UG) the experiments were performed in a 96-well plate (Corning, Coastar, Cat number 3591). Inoculums of 0.1 OD₆₀₀ of the 3 classical Bordetella spp. were inoculated in phenol red-free DMEM (Gibco, Cat #3105-028) containing 10% FBS (Gibco, Cat #10-082-147) supplemented with different concentrations of VIP or in artificial sputum matrix media (BioChemazone, Cat #BZ274), to mimic the host environment. The initial concentration was 25nM, considering that this will be a physiologically relevant concentration of VIP (Henning and Sawmiller, 2001). Concentration was doubled up to 200nM, which has been reported to be maximum concentration possible in the brain. To further investigate the ability of VIP to kill bacteria, we then tested up to 1000 nM of VIP in DMEN. Additionally we also evaluated the effects on VIP on synthetic sputum media. The OD₆₀₀ measurements were performed overnight (≥12 hours) using the Biotek Synergy H1^® 96-well plate reader. To calculate the replication time, we followed previously published protocols (Pope et al., 2010) combined with the built-in analytic calculations available in GraphPad Prism v9.5.0.

Wildtype C57BL/6J mice were purchased from Jackson Laboratories, Bar Harbor, ME. VPAC2^-/- mice were gifted to us from Dr. Xiao-Hong in the department of Pharmacology, Toxicology, and Neuroscience at LSU Health Shreveport, who originally purchased the mice from Jackson Laboratories (B6.129P2-Vipr2^tm1Ajh/J). Breeding colonies were maintained under the care of the employees and the veterinarians at the animal facility at Louisiana State University health science center at Shreveport, LA. All animal experiments were carried out in accordance with all institutional guidelines (AUP:20-038; AUP:22-031) and performed in at least two or more independent experiments, with the total number of mice used for each experiment indicated in the figure legends. Mice were euthanized using 5% CO₂ followed by cervical dislocation. Following euthanasia, the nasal cavity, trachea, and lungs were collected in 1 mL of cold PBS in 2 mL tissue homogenization tubes containing a mixture of 0.5 mm and 1.4 mm ceramic beads.

To evaluate the effects of preincubation with VIP on virulence, bacteria were cultured for 24h in phenol-free DMEM (Gibco) containing 10% FBS (Gibco) supplemented with a 200nM concentration of VIP. These overnight cultures were used to prepare the inoculums as previously described (Gestal et al., 2018). We euthanized mice at day 3 post-infection (Sukumar et al., 2009; Boehm et al., 2019; Wolf et al., 2021) to accurately analyze the effects of VIP on early events of Bordetella spp: colonization, as measuring bacterial burden at later timepoints such as day 7 or day 14 post-infection would be more representative of the infection peak or the initiation of infection clearance, respectively.

To evaluate the effects of the VIP/VPAC2 function on Bordetella spp. colonization in the lower respiratory tract, serial dilutions of the bacterial strains indicated in Table 1 (10² - 10⁶ CFU/mL) were prepared then plated to confirm proper dilution factors were used. C57BL/6J and VPAC2^-/- mice were then inoculated with the increasing bacterial concentrations, and bacterial colonies from nasal cavity, trachea, and lungs of the mice (n = 4-8 mice per condition) were enumerated at 7 days post-infection. Measuring bacterial burden at day 7 post-infection is defined as the peak of Bordetella spp. infection, and it also supported by previously published methods used to assess the minimal required dose of bacteria for successful infection (Belhart et al., 2019). Thus, from these values, 50% of the bacterial inoculum required to efficiently infect the mice used for these experiments (also referred to as ID₅₀) was calculated.

C57BL/6J and VPAC2^-/- mice were intranasally challenged with 30-50μL of PBS containing 1x10⁵ CFU/mL of B. pertussis, B. parapertussis, and B. bronchiseptica (bacterial strains listed in Table 1 ). Bacterial colonies were enumerated from the nasal cavity, trachea, and lungs by plating serial dilutions in BG agar with 20 μg/ml of Streptomycin once every 7 days, starting from day 7 and ending on day 28 post-infection.

For the evaluation of the effects of the administration of an antagonist of the VPAC2 receptor, the compound PG 99-465 was purchased from Bachem Americas, Inc. (cat number H-7292.1000BA). C57BL/6J mice were intranasally challenged with the RB50 strain of B. bronchiseptica. The administration of the treatment consisted of daily 50μL doses of an 8.5μM dilution of PG 99-465. This was to ensure proper administration of a 0.1mg/1kg dose in a volume high enough to flood the murine lungs and generate a localized effect. The preparation of the 8.5 μM dilution of PG 99-465 started from the 0.5mg of compound in the commercial format and adding 1x PBS until reaching the desired concentration. The mice were treated until day 14 post-infection, then the bacterial burden was enumerated. Another group of mice were sacrificed at day 7 post-infection for hematoxylin and eosin (H&E) staining to assess inflammation in the lungs.

All results were graphed in GraphPrism v9.5.0 and statistical significance was calculated using the Sidaks two-way ANOVA multiple comparison test for all animal experiments. All experiments were performed in at least two or more independent biological replicates. The exact number of mice and technical replicates is indicated in each figure legend. A p-value <0.05 was considered statistically significant. In the figures the asterisks correspond with *p ≤ 0.05, ** p ≤ 0.01, ***p ≤ 0.001, and ****p ≤ 0.0001. Statistical significance is shown only in the relevant figures.

Previous literature has shown that the Vasoactive Intestinal Peptide (VIP) has an antimicrobial effect on some bacterial species such as Escherichia coli, Pseudomonas aeruginosa and Streptococcus mutans (El Karim et al., 2008; Carion et al., 2015). Therefore, we first evaluated the antimicrobial activity of VIP against all three classical Bordetella species.

To assess the bactericidal or bacteriostatic effect of VIP we performed our studies using a physiologically relevant media, DMEM with 10% FBS or artificial sputum media, supplemented with increasing concentrations of VIP. Our lowest concentration corresponds with the reported concentration of VIP in serum, 25 nM (Ghaferi et al., 2008), and our highest is greater than the reported concentration of VIP in brain, 100nM (Goadsby and Edvinsson, 1994) or previously tested for antimicrobial activity of VIP (Carion et al., 2015). The results show no significant differences between the duplication times of B. pertussis, B. parapertussis or B. bronchiseptica, for any of the concentrations evaluated in our studies ( Figure 1 ). We were not able to identify defect in growth or bacterial killing in DMEM or sputum. Thus, we concluded that VIP does not possess bactericidal nor bacteriostatic activity against the three classical Bordetella spp. at physiologically relevant conditions.

We wanted to investigate if prior incubation with VIP would prime Bordetella spp. to better respond to host cues and enhance the bacteria’s ability to colonize the respiratory tract, similar to what has been shown with other host components. To test this hypothesis, overnight cultures of the three classical Bordetella spp. in DMEM supplemented with 10% FBS with or without 200 nM of VIP (as indicated in the figure legend), were used to intranasally inoculate mice. At day 3 post-infection, mice were euthanized to enumerate colonies in the respiratory tract to assess for the ability of the bacteria to colonize the lungs.

The results show no significant differences in bacterial ability to colonize the respiratory tract. It is worth noting that there are limitations to infecting mice with B. pertussis and B. parapertussis. Because the bacteria were inoculated intranasally, it is possible that the inoculums were not fully inhaled due to bubbles or leaks from the nose, leading to inconsistent lung colonization. In addition, colony growth can be unpredictable since both bacteria are not natural pathogens of mice. The number of B. pertussis ( Figure 2A ), B. parapertussis ( Figure 2B ), and B. bronchiseptica ( Figure 2C ) that was isolated from the lungs of mice at day 3 post-infection was not affected by previous incubation with VIP. Similar results were previously observed with pre-incubation with blood and serum where despite increased in vitro cytotoxicity was observed after incubation with serum, no differences in the bacterial ability to colonize mice were detected (Gestal et al., 2018). Thus, we concluded that VIP does not possess any kind of activity that enhances the colonization properties of any of the three classical Bordetella species.

Previous literature has shown that VPAC2 and VPAC1 vary levels of expression during infection with different pathogens (Ipp et al., 2014; Askar et al., 2020; Yu et al., 2021). Moreover, pathogens such as Salmonella spp. utilize VIP/VPAC axis to promote intracellular survival in macrophages (Askar et al., 2020), suggesting that VIP/VPAC axis might influence infection dynamics and persistence. C57BL/6J and VPAC2^-/- mice were intranasally inoculated with increasing bacterial concentrations of B. pertussis, B. parapertussis, and B. bronchiseptica, and euthanized at day 7 post-infection to enumerate colonies from the respiratory tract (Belhart et al., 2019).

Our results show that the ability of B. pertussis to colonize in the lungs of the VPAC2^-/- mice was substantially depleted ( Figure 3A ), as a 10-fold increase of bacterial inoculum was required to establish an efficient lung infection. Similarly, B. parapertussis colonized C57BL/6J (Weyrich et al., 2012) at a lower infectious dose than the VPAC2^-/- mice ( Figure 3B ). Finally, very low doses of B. bronchiseptica RB50 were required to colonize the C57BL/6J lungs efficiently (Weyrich et al., 2012; Belhart et al., 2019; Gestal et al., 2019) ( Figure 3C ). However, at least 1 x 10⁴ CFUs of RB50 were required to colonize the lungs of VPAC2^-/- mice. Overall, this reduced colonization can be the product of a stronger pro-inflammatory response of VPAC2^-/- mice, or it can be due to the role of VIP in mucus secretion, which can affect colonization as previously reported (Goetzl et al., 2001; Kim et al., 2006; Samarasinghe et al., 2010; Wu et al., 2011). Nevertheless, our results altogether suggest that the absence of a functional VIP/VPAC2 signaling impairs classical Bordetella spp. colonization.

One critical aspect of Bordetella spp. infection is the pneumonic phase that in patients lead to serius illness or even death (Rath et al., 2008; Barger-Kamate et al., 2016; Jiang et al., 2021). Identifying immunotherapies that prevent long term lung damage is of great importance. As VIP is commonly found in lungs, and we saw a defect in VPAC2^-/- mice colonization, we wanted to investigate if VIP/VPAC2 signaling plays a role during Bordetella clearance from the lungs. To test this hypothesis, VPAC2^-/- or C57BL/6J mice were challenged with B. pertussis, B. parapertussis, or B. bronchiseptica then sacrificed at different time points to evaluate the bacterial burden in the respiratory tract.

Following infection with the B. pertussis strain Bp536, we first analyzed the colonization in the nasal cavity ( Figure 4C ). Differences were not detectable up to day 28 post-infection. Similarly, the trachea did not reveal any major differences in bacterial load ( Figure 4B ). In the lungs, however, bacterial enumeration in the VPAC2^-/- lungs dropped slightly at day 14 post-infection, with differences in the bacterial burden between VPAC2^-/- and C57BL/6J becoming more distinct as early as day 20 ( Figure 4A ). Thus, our results suggest that disruption of VIP/VPAC2 signaling axis promotes clearance the infection from the lungs.

During Bpp infection, no differences were detected in the colonization of the nasal cavity ( Figure 4F ). However, in the trachea and lungs ( Figures 4D–F , respectively) we were able to detect a more rapid decrease in bacterial burden at day 28 post-infection, suggesting that, just like B. pertussis, B. parapertussis infection is more rapidly cleared from the lower respiratory tract of VPAC2^-/- mice.

Our results demonstrate that the absence of a functional VIP/VPAC2 axis promotes rapid clearance of B. pertussis and B. parapertussis from the lower respiratory tract. However, these two bacteria are not natural pathogens of mice and cause only an acute disease that is efficiently cleared by the immune system rather than the characteristic long-term and persistent infection of the human pertussis disease (Soumana et al., 2021). To investigate the effects of disrupting VIP/VPAC2 signaling in bacterial clearance in a natural setting of long-term chronic disease, we used the well-established B. bronchiseptica murine model (Gestal et al., 2019) following the same experimental setting as previously indicated.

Our results did not reveal significant differences in the bacterial burden detected in the nasal cavity ( Figure 4I ) and trachea ( Figure 4H ) between the C57BL/6J and VPAC2^-/- mice. In the lungs, we observed a clear decrease in the bacterial burden by day 14 post-infection ( Figure 4G ). Overall, these results indicate that the VIP/VPAC2 signaling pathway significantly impacts the clearance of B. pertussis, B. parapertussis, and B. bronchiseptica from the lungs. Because of these observed similarities, the B. bronchiseptica mouse model can be used for a better understanding of host-pathogen interactions at the molecular level, as the differences found became clearer when using a natural pathogen of mice.

Our previous results revealed that using B. bronchiseptica in this study could provide a more mechanistic understanding of the role of VIP/VPAC2 signaling in the infection dynamics. We previously investigated the mechanisms by which Bordetella spp. suppress host immune responses, and we described a Bordetella sigma factor, btrS, a regulator of immunosuppressive pathways in all three of the classical Bordetella species (Gestal et al., 2019). Absence of btrS results in rapid clearance and robust immune responses, with the generation of long-lasting protective immunity (Gestal et al., 2022). We wanted to investigate if the bacterial btrS-signaling pathway is involved in the modulation of VIP/VPAC2 signaling. To determine the effects of btrS on the manipulation of VIP/VPAC2 signaling, we evaluated the differences between murine bacterial infection by using the calculated ID₅₀ values following inoculation with the wildtype BbRB50 strain, a mutant RB50 strain lacking the sigma factor btrS (RB50ΔbtrS). The ID₅₀ values were assessed in these experiments because we predicted that if VIP/VPAC2 axis is modulated by btrS, there would be no observable difference in the ID₅₀ of RB50ΔbtrS in either C57BL6/J or VPAC2^-/- mice. However, if btrS is not involved in the modulation of VIP/VPAC2, then a reduction in the capability of colonizing VPAC2^-/- mice would be observed, as was previously with the three Bordetella spp. strains. While differences in the ability to colonize VPAC2^-/- were observed when investigating the RB50 strain ( Figure 5A ). Our results did not reveal differences in the ID₅₀ of the RB50ΔbtrS, suggesting that in the absence of btrS there are no alterations in the bacterial ability to colonize neither C57BL6/J nor VPAC2^-/- mice ( Figure 5B ), contrary to the wildtype B. bronchiseptica strain which fails to colonize in VPAC2^-/- mice.

One of the main virulence factors regulated by the btrS-pathway is the type 3 secretion system (T3SS). Using the same experimental setting as previously described, we wanted to evaluate the role of the T3SS in the modulation of VIP/VPAC2 signaling, we again performed an ID₅₀ study. We expected that if manipulation of VIP/VPAC2 signaling is mediated by the T3SS, we will not be able to observe differences in the ID₅₀ between C57BL/6J and VPAC2^-/- mice ( Figure 5C ). Our results did not show difference between the ID₅₀ in C57BL/6J mice and the VPAC2^-/- mice with the T3SS mutant of B. bronchiseptica, suggesting that, the modulation of the VIP/VPAC2 axis by Bordetella spp. might be mediated by T3SS. Although VIP/VPAC2 axis is also involved in the modulation between Th1/Th2 responses, the fact that no differences were found in the ID₅₀ of the RB50ΔbscN while differences were found with all the wildtype bacteria, led us to hypothesized that T3SS might mediate Bordetella spp. manipulation of VIP/VPAC2 axis to promote colonization.

Our previous results show that blocking VIP/VPAC2 axis promotes rapid clearance of the lower respiratory tract, suggesting that, the use of an antagonist of the VPAC2 receptor should have a similar effect in clearance to that observed in the VPAC2^-/- mutant mice. To evaluate if antagonists of VPAC2 promote clearance, we used the substance PG 99-465 as a treatment for C57BL/6J mice infected with B. bronchiseptica. The mice were treated every 24 hours following inoculation with a 50μL dose of PG 99-465 (8.5μM). At 7 and 14 days post-infection, the mice were euthanized to evaluate lung pathology and colonization respectively.

Although no differences in lung colonization were observed ( Figure 6A ), this could be the consequence of the short half-life of the antagonist, which is a very small and unstable peptide. However, the pathology results show a clear difference in inflammation of the bronchia and consolidation of the respiratory parenchyma between the untreated and treated mice ( Figure 6B ). The pathology in mice treated with the antagonist resembled the pathology observed in VPAC2^-/- mice, which successfully recover from infection. Overall, this suggest that VPAC2 receptor could be targeted for novel immune therapies against Bordetella spp. infections.