The host genetic background defines diverse immune-reactivity and susceptibility to chronic Pseudomonas aeruginosa respiratory infection

Patients with P. aeruginosa airways infection show markedly variable clinical phenotypes likely influenced by genetic backgrounds. Here, we investigated the cellular events involved in resistance and susceptibility to P. aeruginosa chronic infection using genetically distinct inbred mouse strains. As for patients, different murine genotypes revealed variable susceptibility to infection. When directly compared, resistant C3H/HeOuJ and susceptible A/J strains revealed distinct immune responsiveness to the pathogen. In C3H/HeOuJ resistant mice, IL17-producing cells rapidly and transiently infiltrated the infected lung, and this was paralleled by the acute accumulation of alveolar macrophages, bacterial clearance and resolution of infection. In contrast, A/J susceptible mice revealed a more delayed and prolonged lung infiltration by IL17+ and IFNγ+ cells, persistence of innate inflammatory cells and establishment of chronic infection. We conclude that the host genetic background confers diverse immunoreactivity to P. aeruginosa and IL17-producing cells might contribute to the progress of chronic lung infection.

. Survival, body weight, percentage of chronicity and pulmonary bacterial burden after P. aeruginosa infection in inbred mouse strains. DBA/2 J (n = 10), 129S2/SvPasCRL (n = 12), A/J (n = 24), C3H/HeOuJ (n = 26) and C57BL/6NCrl (n = 11) mice were inoculated with 2 × 10 6 CFU of P. aeruginosa mucoid clinical isolate AA43 embedded in agar beads by intratracheal injection, and monitored for survival (A) and body weight change (B) for a period of 14 days after challenge. In addition, the clearance (white) and rate of chronic colonization (colored) (C) and the bacterial burden in the lungs (D) after 14 days from P. aeruginosa challenge were determined in surviving mice. (B) The error bars represent the standard error of the mean (SEM). (D) Dots represent lung CFU in individual mice and horizontal lines represent median values reported in log scale. The data are pooled from two to three independent experiments. Statistical significance by Mantel-Cox test (A), Fisher's test followed by Bonferroni correction for multiple comparisons (C), and Mann-Whitney U test and nonparametric Kruskal-Wallis test followed by post-hoc Dunn test to correct for multiple comparisons (D) is indicated: *p < 0.05. Two-way ANOVA with Bonferroni's post-hoc test (B) is reported in Table S1. the early stage of the P. aeruginosa infection when compared to C57BL/6NCrl and C3H/HeOuJ mice, that instead revealed better survival rates and lower morbidity (Fig. 1A,B and Supplementary Tables 2 and 3).
Next, we evaluated the predisposition to the establishment of long-term chronic infection. At 14 days, susceptible A/J, DBA/2 J and 129S2/SvPasCrl mice showed significantly higher incidence of chronic infection when compared to the resistant C3H/HeOuJ strain, while the C57BL/6NCrl strain showed an intermediate phenotype ( Fig. 1C and Supplementary Table 3). In addition, although not statistically different, in average pulmonary bacterial loads were 1 Log higher in susceptible A/J, DBA/2 J, and 129S2/SvPasCrl than in C3H/HeOuJ and C57BL/6NCrl ( Fig. 1D and Supplementary Table 3).

Figure 2. Percentage of infected mice, pulmonary bacterial burden and level of leukocyte recruitment (CD45 + cells) during chronic P. aeruginosa lung infection in resistant and susceptible mice. Susceptible
A/J and resistant C3H/HeOuJ mice were inoculated with 2 × 10 6 CFU of P. aeruginosa mucoid clinical isolate AA43 embedded in agar beads by intratracheal injection and sacrificed at day 2, 7 and 14 post challenge. At each time point, the clearance (white) or rate of chronic colonization (colored) (A) and the bacterial burden in the lungs (B) were evaluated. Dots represent CFU per lung in individual mice and horizontal lines represent median values reported in log scale (B). Leukocytes (CD45 + cells) were measured in lung cell suspension by flow cytometric analysis (C). The error bars represent the SEM (C). The data are pooled from two to three independent experiments. Statistical significance by Fisher's test followed by Bonferroni correction for multiple comparisons (A), and Mann-Whitney U test and nonparametric Kruskal-Wallis test followed by post-hoc Dunn test to correct for multiple comparisons (B,C) is indicated: *p < 0.05, **p < 0.01, ***p < 0.001. Colored stars indicate the statistical significance of the difference between each mouse strain at the specific time point compared with its own naïve counterpart.
Scientific RepoRts | 6:36924 | DOI: 10.1038/srep36924 The most susceptible and resistant mice (i.e. A/J and C3H/HeOuJ, respectively) were further characterized in time course analyses. At day 2, both A/J and C3H/HeOuJ mice held P. aeruginosa in the lungs ( Fig. 2A). Nevertheless, by this time, resistant C3H/HeOuJ mice displayed a lower bacterial burden (4.2 × 10 5 CFUs) compared to susceptible A/J mice (2.1 × 10 7 CFUs) (Fig. 2B). Although by day 7 and 14, the number of infected mice and the pulmonary bacterial load decreased in both strains, the frequency of chronically infected mice was significantly higher in A/J mice than in C3H/HeOuJ (day 7: 94% versus 47%; day 14: 56% versus 17%) ( Fig. 2A). Of note, although CD45 + inflammatory cells equally infiltrated lungs of infected mice by day 2, they persisted to higher numbers in chronically infected A/J susceptible mice when compared to chronically infected C3H/HeOuJ    When measuring pulmonary cytokines/chemokines, we found that several of them were expressed at higher levels in the lungs of A/J susceptible mice compared to resistant C3H/HeOuJ early after infection (day 2) ( Table 1), likely reflecting higher bacterial loads (Fig. 2B). By day 7, cytokines/chemokines expression decreased in both strains, although at slower rate in susceptible mice compared to resistant, to reach similar levels at day 14 (Table 1).
Thus, as reported in humans, also inbred mouse strains reveal different susceptibility to infection, with C3H/ HeOuJ and A/J being the most resistant and susceptible ones, respectively, and diverse pulmonary pathology.

Pulmonary innate immune response in P. aeruginosa-susceptible and resistant mice.
To understand whether susceptibility/resistance to infection could be linked to host immune reactivity, we performed FACS-assisted immunophenotyping of lung CD45 + infiltrates during P. aeruginosa infection. By day 2 neutrophils, monocytes, and macrophages were all represented within infected lungs. Neutrophils (CD45 + Gr1 high CD11b high ) and to a lower extent monocytes/small macrophages (CD45 + CD11b + CD11c − ) numbers decreased thereafter, and yet persisted to higher numbers in A/J susceptible mice (Fig. 4A,B, Supplementary Fig. 1A,B). Of note, C3H/HeOuJ and A/J-derived bone marrow neutrophils and peritoneal macrophages displayed similar in vitro phagocytic activity (Fig. S2), indicating that different bacterial burden are unlike to be attributable to defective phagocytic capacity of susceptible mice. Alveolar macrophages (CD45 + CD11b − CD11c + ; Fig. 4C, Supplementary Fig. 1C) and myeloid dendritic cells (CD45 + CD11b + CD11c + ; Fig. 4D, Supplementary Fig. 1D) accumulated in the infected lung by day 7. Of note, at 2 and 7 days lungs of resistant C3H/HeOuJ mice contained a significantly higher fraction of alveolar macrophages when compared to those derived from susceptible A/J mice, supporting their direct involvement in host defense and acute resolution of the airways inflammation (Fig. 4C, Supplementary Fig. 1C).
Together these data indicate that shortly after infection the lungs of both susceptible and resistant strains are infiltrated by component of the innate immune response, evoking an inflammatory response. Data indicate that alveolar macrophages best accumulate in the C3H/HeOuJ resistant strains at early time points, and this is associated to a better bacterial clearance. This event might be defective in A/J mice, where infection persists contributing to establishment of chronicity. Lymphocytes-mediated pulmonary immune response in P. aeruginosa-susceptible and resistant mice. T cells orchestrate the activity of neutrophils and alveolar macrophages over the course of P. aeruginosa infection 23 . We found that at day 2, CD3 + and CD4 + T cell infiltrated the infected lungs to higher extents in C3H/ HeOuJ resistant mice than in A/J susceptible ones (Fig. 5A,B, Supplementary Fig. 3A,B). By day 7 T cell infiltration increased in C3H/HeOuJ mice and decreased thereafter. In contrast CD3 + and CD4 + T cell subsets progressively accumulated in A/J mice both at day 7 and day 14, being significantly enriched for when compared to C3H/HeOuJ. Similar trends were observed for CD8 + T cells and B cells, accumulating to higher numbers in A/J mice at day 7 and day 14 compared to C3H/HeOuJ ones (Fig. 5C,D, Supplementary Fig. 3C,D). Thus, these data indicate that a superior CD4 + T cell responses shortly after infection is recruited in inbred strain with a better survival and resistance to chronic infection; on the other side, a delayed response is associated with susceptibility to P. aeruginosa chronic infection.
Thus, the rapid infiltration of the lung by IL17-producing cells appears to promote the accumulation of alveolar macrophages and CD8 + IL17 + , favouring bacterial clearance, resistance to chronic infection and improved survival. At difference, suboptimal and delayed CD4 + IL17 + responses suggest to favour neutrophils persistence in infected lungs and chronic inflammation, likely contributing to the establishment of chronic infection.

Discussions
In this work, we have exploited a mouse model of chronic P. aeruginosa infection, and demonstrated that the host genetic background confers different immunoreactivity to the pathogen, contributing to predisposition to chronic infection.
To mimic the progressive bronchopulmonary infection typical of CF and COPD patients and study the impact of the genetic background on the establishment of chronic infections, we challenged five different inbred murine strains with the AA43 P. aeruginosa adapted clinical isolate embedded in agar beads 21,24 . By evaluating mortality, changes in body weight, the capacity to efficiently clear the pathogen, and the frequency of persistent infection, we defined a susceptibility range and ranked the mice accordingly. We found the A/J, 129S2/SvPasCrl and DBA/2 J strains to be most susceptible to P. aeruginosa chronic infection, while the C57BL/6NCrl and C3H/HeOuJ strains to be most resistant. Of note, these results are in line with previous studies 5, 14 and with results obtained in acute infection models 9 , the only exception being the C57BL/6NCrl strain, which revealed an intermediate susceptibility to acute infection and instead proved resistant to chronic disease. Thus, in general terms, it appears that host genetic backgrounds predisposing to acute P. aeruginosa infections also appear more prone to accommodate long-term chronic bronchopulmonary disease. Translating this observation to CF patients, it is plausible to suggest that the first colonization, successive episodes of re-colonization and later chronic infection are all events influenced by the genetic profile of the host.
To define possible causes for predisposition, we characterized immune-related events possibly linked to diverse P. aeruginosa susceptibility and resistance. Early responses to the pathogen (day 2) included lung infiltration by inflammatory phagocytes, neutrophils and macrophages, which occurred to a similar extent in susceptible (A/J) and resistant (C3H/HeOuJ) hosts. Also the in vitro phagocytic capacity of lung inflammatory neutrophils and macrophages derived from resistant C3H/HeOuJ and susceptible A/J mice was found to be similar, as also previously reported by Morissette and co-authors in the case of resistant (BALB/c) and susceptible (DBA2) mice 6 . Thus, the early phases of the response to the pathogen by components of the innate immune response appear to be similar between susceptible and resistant host, and could not account for predisposition to development of chronic infection. This observation differs from data generated when studying the AA2 acute infection model, which showed inter-strain differences in the early recruitment of the innate immune response 9 . Differences in the early phases of acute versus chronic infection models might be due to the P. aeruginosa clinical isolates adopted in the two studies (the AA2 vs AA43), in the modality of infection (free bacteria vs agar-beads embedded ones) and to the time of analysis (6-18 h vs day 2). Further studies will be needed to better define the molecular events subtending innate cell responsiveness in AA2 and AA43 models.
Nevertheless, resistant C3H/HeOuJ mice showed a prevalence of alveolar macrophages in infected lungs at day 2 and day 7. This might favor resolution of inflammation via phagocytosis of dying neutrophils and initiate mechanisms of repair 25 , which were evident in lung histopathological analyses. Accordingly, neutrophils numbers were lower by day 7 and 14 in resistant mice when compared to susceptible A/J, which revealed sustained neutrophil lung infiltration over the course of chronic infection.
When analyzing the contribution of adaptive immune responses, P. aeruginosa-resistant and susceptible hosts appeared also clearly different. FACS-assisted immunophenotyping revealed superior lung infiltration by T cells in resistant C3H/HeOuJ mice compared to A/J susceptible ones by day 2 (including CD4 + , CD8 + and TCRγ δ + T cell subsets), which were highly enriched for IL17-producing cells. Given the fast kinetic of infiltration, we believe that CD4 + IL17 + cells, that predominate the early phases of P. aeruginosa infection, belong to the "natural" Th17 (nTh17) cell subset, described in the gut, liver, oral mucosa and lungs predominate the early phases of P. aeruginosa infection. These cells, differently from conventional Th17 (cTh17) cells, which require antigen-driven priming, can be mobilized in hours or days 26 . Our data are in line with other studies, reporting IL17 producing Th17 cells and TCRγ δ + T cells as critical in the fast host immune defense against P. aeruginosa 27 . Whether other cellular sources (e.g. pulmonary group 3 innate lymphoid cells and natural killer cells), previously described to be critical in other models of P. aeruginosa infection 28 , also contribute to the early phases of chronic infection, remains to be determined.
Previous studies had defined Th1-like cells as critical against P. aeruginosa infection. Indeed, Th1-prone C3H/HeN mice were reported to have a better disease outcome when compared to the Th2-prone BALB/c mice 14,15 . In our work, while IL4-producing CD4 + T cells were negligible in both resistant and susceptible mice, we detected equal levels of IFNγ -producing CD4 + T cells in response to the pathogen. Instead, we found the early IL17-dominated response best observed in resistant mice is better associated with the efficient clearance of P. aeruginosa and the progressive resolution of inflammation, along with the protection against chronic infection, and an overall milder clinical outcome. Compared to resistant C3H/HeOuJ mice, susceptible A/J mice showed a delayed infiltration of the lung by CD4 + IL17 + cells and a more prolonged CD4 + IFNγ + cell response. Results obtained in susceptible mice are highly reminiscent of those we recently reported in the C57BL/6NCrl background 17 , and support a model by which IL17 might provide a protective role in early phases of infection, and instead might be detrimental afterwards and sustain immunopathological manifestations when chronically produced in the context of inflammation. In a recent study, we found that IL-17 is active at different phases of chronic airways infection and indeed plays a double-edged activity 17,28 . On one side we found IL-17 to contribute to the control of P. aeruginosa burden, while on the other, IL-17 appeared to worsen pulmonary neutrophilia and tissue damage. We believe that the current study supports a previous report and a model whereby IL-17 produced by neutrophils and CD4 + T cells plays a critical role in the early phase of infection by contributing to pathogen clearance. In the case of delayed recruitment of such cells, bacterial persistence causes abnormal accumulation of these cells, later on responsible for chronic inflammation and tissue damage.
Thus, overall our results indicate that the host genetic background defines distinct immune-reactivity to P. aeruginosa infection, and consequently a diverse susceptibility to chronic infections. Our results suggest a key role for IL17-producing cells in the early phases of infections. Mice with pre-existing IL17-producing cells are most likely to eradicate the pathogen at the time of first encounter, and to resolve pathogen-induced inflammation, enabling limited pathological manifestations. Thus, strategies capable to increase the frequency of IL17 + cells might be envisaged to protect patients against P. aeruginosa-induced chronic infection, especially in CF. Conversely, prolonged IL17 expression correlated with chronicity of infection and likely exacerbated immunopathology. This knowledge can be further explored in patients to assess the risk of P. aeruginosa chronic infection based on their genetic profile or treatments inhibiting prolonged IL17 expression.

Materials and Methods
Bacterial strains. The AA43 P. aeruginosa clinical strain was originally isolated at a late stage of chronic infection from a CF patient 22,29 , cultured in trypticase soy broth (TSB) and plated on trypticase soy agar (TSA).
Mouse model of P. aeruginosa chronic infection. Ten-twelve weeks old inbred mice were used: A/J, C3H/HeOuJ and DBA/2 J were purchased from Jackson Laboratories, C57BL/6NCrl and 129S2/SvPasCRL were obtained from Charles River Laboratories. Mice were infected with 2 × 10 6 CFU of P. aeruginosa embedded in agar beads, as previously described 22,24 . Animals were sacrificed after 2, 7 and 14 days by CO 2 administration and lungs were recovered.
Animal studies were conducted according to protocols approved by San Raffaele Scientific Institute (Milan, Italy) Institutional Animal Care and Use Committee (IACUC) (Permit number: 502).
Lung single-cell suspension and flow cytometric analysis. Lung single-cell suspensions were obtained by mashing the lungs through a 70-μ M cell strainer in RPMI + 5% FBS and serially diluted and plated on TSA for CFU counts. All the procedures were previously described 17 . Briefly, lung cell suspensions (1-3 × 10 6 cells) were incubated with blocking buffer (5% rat serum and 95% culture supernatant of 24G2 anti-FcR mAb-producing hybridoma cells) for 10 min at 4 °C. Then, cells were stained for 20 min at 4 °C in the darkness with different combinations of antibodies (BD Biosciences; listed in Supplementary Table 1). For intracellular cytokine staining, 1-3 × 10 6 cells were stimulated with PMA (50 ng/ml) and Ionomycin (1 μ g/ml) for 4 h at 37 °C (the last 2 h with Brefeldin A (5 μ g/ml)). Cells were surface stained, fixed, permeabilized and then stained for intracellular cytokines for 30 min at RT in darkness. Cells were washed in permeabilization buffer and collected using a FACSCANTO flow cytometry (BD Biosciences) and then analyzed using FlowJo software.
Histologic and immunofluorescence analysis. Lungs were embedded in paraffin and 2-mm sections were stained by Haematoxylin-Eosin for histological analysis. De-paraffinized lung sections were stained with Scientific RepoRts | 6:36924 | DOI: 10.1038/srep36924 rabbit antiserum specific for P. aeruginosa and Texas Red-labelled goat anti-rabbit IgG as described previously for bacterial localization 22 . The slides were examined using an Axioplan fluorescence microscope (Zeiss).
Cytokine analysis. Part of the harvested lung single-cell suspensions, obtained by mechanical dissociation, were centrifuged at 14000 rpm for 30 min at 4 °C and the supernatants (SN) were stored at − 80 °C for quantification of total protein content with Bradford's assay (Bio-RAD). Protein content was quantified at the final concentration of 500 μ g/ml. A panel of murine chemokines and cytokines were measured using Bio-Plex pro TM Mouse Cytokine Standard 23-Plex, Group I and Bio-Plex pro TM Mouse Cytokine Custom TH17 Mouse Group III.
Phagocytosis and intracellular killing assay. Bone marrow cells were isolated from naïve mice and neutrophils were negatively selected with STEMCELL mouse enrichment Kit (catalog #19762) 30 . Peritoneal cells were harvested 3 days after i.p. injection of 1 ml thioglycolate broth (BD) by two lavages with 5 ml PBS. The purity of cells determined by Diff-Quick staining was 85-90%.
For neutrophils infection, P. aeruginosa cells where opsonized with 10% mouse serum for 30 min at room temperature in agitation. Both neutrophils and macrophages were infected with P. aeruginosa AA43 at mid exponential phase, with a MOI of 100:1 (bacteria:neutrophils). After 30 min of incubation at 37 °C cells were treated with Polymvxins B (100 μ g/ml) (Sigma), washed, lysed with H 2 0 and plated on TSA. To evaluate intracellular killing capacity, treatment with Polymvxins B was extended for 30 min, 60 min and 90 min for macrophages. Statistical analysis. Data analysis was performed using a nonparametric two-taled Mann-Whitney U test for single comparisons, such as when comparing data between two murine strains at a specific time-point (e.g. CFU levels). To compare data for a specific murine strain across multiple time-points a nonparametric Kruskal-Wallis test was used followed by post-hoc Dunn test to correct for multiple comparisons. Rates of mortality and infection were compared using Fisher's test followed by Bonferroni correction for multiple comparisons. Mantel-Cox test was used to compare survival between pairs. The changes in body weight were compared using Two-way ANOVA with Bonferroni's post-hoc test. P < 0.05 was considered significant.