Intrinsic neutrophil dysregulation in programmed cell death promotes neutrophilic inflammation in cystic fibrosis

Abstract

Neutrophilic inflammation constitutes a major pathology in cystic fibrosis (CF), a life-shortening genetic disorder. However, the underlying mechanisms remain incompletely understood. Here we report our investigation into CF neutrophil recruitment, apoptosis and clearance to determine if any neutrophil intrinsic defect is involved in the pathogenesis. Congenic fluorescent wild-type (WT) and CF mice were generated and used as bone marrow (BM) donors. Non-fluorescent WT and CF recipients were transplanted with a mixture of EGFP-WT and DsRed-CF BM cells and interrogated for neutrophil recruitment and apoptosis in the same lung after zymosan challenge. Additionally, ex vivo neutrophil migration and apoptosis were also examined to confirm the in vivo findings. Moreover, DsRed-WT or DsRed-CF BM cells were transfused into non-fluorescent CF recipients that had been intratracheally challenged with zymosan. Percentages of apoptotic cell-laden macrophages in the lungs were compared. The data indicate that CF and WT neutrophils exhibited a similar migratory capacity and were equally recruited to the same lung, regardless of the recipient phenotype. However, CF neutrophils displayed a significantly delayed apoptosis. Such an anomaly was linked to reduced hypochlorous acid production. Moreover, CF macrophages exhibited a significantly higher expression of efferocytosis marker MerTK, and a greater capacity of efferocytosis, implying that macrophage phagocytosis of apoptotic cells is not impaired in this experimental setting. These findings suggest that intrinsically dysregulated programmed cell death affects CF neutrophil fate decision and fuels neutrophilic inflammation in CF lungs.

1. Introduction

Cystic fibrosis (CF), an autosomal recessive genetic disorder, primarily affects epithelia-lined organs, including the respiratory, gastrointestinal and reproductive systems [1]. In adult CF patients, lung complications claim the most morbidity and mortality, presenting as chronic bacterial infection, persistent neutrophilic inflammation, and mucopurulent small airway obstruction [2,3]. Due to these outstanding epithelial manifestations, CF is generally considered a disease of epithelial dysfunction, from which neutrophilic inflammation is believed to arise [4,5]. However, previous studies from us and others demonstrated that CFTR is expressed in normal neutrophils [6–8]. Mutant CFTR fails to target the phagosomes [9], leading to insufficient chloride supply to the organelle and resulting in a deficit in hypochlorous acid (HOCl) production, one of the key microbicides for neutrophil-mediated bacterial killing [6,10–12]. Moreover, CF neutrophils show other irregularities, including suboptimal activation [13], cleavage of CXCR1 [14], hyper-sensitivity to LPS stimulation [15], alteration in inflammatory signaling [16], hyper-production of IL-8 [17, 18], abnormal extracellular trap formation [19], hyper-oxidation of glutathione [20], dysregulation of protease secretion and degranulation, delayed apoptosis [21,22], and abnormal granule release [23]. CF neutrophils also express higher levels of CD64, a marker of neutrophil activation, and lower levels of Toll-like receptor-2 (TLR2) compared with healthy blood neutrophils. Further, TLR2, TLR4 and IL-8 expressions in CF airway neutrophils were higher than in CF blood neutrophils [24]. Interrogation of the transcriptional program via single-cell RNA sequencing revealed a prevalence of immature pro-inflammatory neutrophils in CF sputum [25,26]. Despite these extensive studies, it remains unknown whether neutrophils per se are responsible for the neutrophilic inflammation in CF lungs.

In this report, we seek to determine if CF neutrophils are ill-programmed for recruitment, apoptosis and clearance, thus facilitating neutrophilic inflammation. Bone marrow transplantation (BMT) with a mixture of fluorescent CF and WT donor bone marrow (BM) cells was performed to investigate neutrophil recruitment and apoptosis within the same lung. Furthermore, fluorescent CF or WT BM cells were transfused into non-fluorescent CF mice. Macrophage phagocytosis of apoptotic cells in CF lungs was interrogated. The data obtained provide the first evidence suggesting that dysregulated apoptosis of CF neutrophils is a factor contributing to pulmonary neutrophilic inflammation, and this defect is associated with deficient HOCl production within the cells.

2. Materials and methods

2.1. Animals and animal use ethics

The non-fluorescent CF mice used in this study were whole-body CFTR-exon-11 deletion mice, originally provided by Dr. Mitchell Drumm at Case Western University. Heterozygous CF (Cftr^+/−) mice served as breeders to produce homozygous CF mice (Cftr^−/−) and their sibling WT (Cftr^+/+) mice. DsRed mice (B6.Cg-Tg(CAG-DsRed*MST) 1Nagy/J), EGFP mice (C57BL/6-Tg(CAG-EGFP)131Osb/LeySopJ), and MPO-knockout (Mpo^−/−) mice (B6.129 ×1-Mpotm1Lus/J) were purchased from The Jackson Laboratory. Fluorescent CF (Cftr^−/−DsRed^+/+ or Cftr^−/−Egfp^+/+) mice and their sibling WT controls (Cftr^+/+DsRed^+/+ or Cftr^+/+Egfp^+/+) were established by breeding non-fluorescent heterozygous CF (Cftr^+/−DsRed^−/−or Cftr^+/−Egfp^−/−) mice with fluorescent (Cftr^+/+DsRed^+/+ or Cftr^+/+Egfp^+/+) mice. After multiple rounds of breeding and screening, the fluorescent CF mice were identified under UV light, flow cytometry, and PCR genotyping. All CF mice, unless otherwise indicated, were maintained on a drinking water containing Polyethylene glycol 3350 (60 g/L). All animals used in the study were age and sex-matched, ranging from 6 to 12 weeks old, and were handled according to the ethical standards set by the Guide for Care and Use of Laboratory Animals of the National Institutes of Health. The experimental protocols were approved by the Louisiana State University Health Sciences Center Animal Care and Use Committee.

2.2. Bone marrow transplantation

Non-fluorescent WT or CF recipients were, respectively, irradiated with a total dose of 900 or 700 Rad, delivered by a split administration 4 h apart (RS-2000 X-ray irradiator, Rad Source). DsRed-CF mice and EGFP-WT mice served as donors. Bone marrow (BM) cells from both femurs and tibias of donor mice were collected under sterile conditions and filtered through a 70-μm cell strainer. After red blood cell lysis, the nucleated cells were washed and resuspended in RPMI 1640 medium supplemented with 1x Antibiotic-Antimycotic (Gibco). Then, the CF and WT BM cells were mixed at a ratio of 1:1, and injected retro-orbitally into the irradiated mice (~ 2 ×10⁶ BM nucleated cells per recipient). After BMT, all mice were on laxative water supplemented with 2 % SulfaTrim pediatric suspension (800 mg/L sulfamethoxazole and 160 mg/L trimethoptim) for the first 2 weeks, then on laxative water alone. Bone marrow cell engraftment was confirmed 4 weeks post BMT by measuring blood cell fluorescence by flow cytometry (BD FACSymphony^™ A3).

2.3. Zymosan preparation and induction of lung inflammation

Zymosan A powder (Cat. No. 21175, Cayman Chemical, Ann Arbor, MI) was suspended and boiled in distilled water for 30 min. The suspension was spun down, washed with PBS twice, and sonicated before use. To induce lung inflammation, mice were intratracheally challenged with 10 mg/kg zymosan by laryngopharyngeal installation. At different time points (Days 1, 3, and 5 post-challenge), blood samples were harvested. After euthanasia, their lungs were lavaged and the bronchoalveolar lavage fluid (BALF) harvested for analyses. For pharmacological inhibition of CFTR or MPO, mice were injected i.p. with 10 mg/kg CFTR Inhibitor (CFTR_inh-172; Cayman Chemical) or 80 mg/kg MPO inhibitor (4-ABAH; Millipore Sigma) every 12 h starting from 3 h before zymosan intratracheal challenge. Three days after challenge, assigned animals were sacrificed, their lungs lavaged and the BALF analyzed.

2.4. Assessment of neutrophil recruitment and apoptosis in lung

BALF cells were subjected to cell differential counting and apoptosis assessment. For non-BMT studies, the staining reagent panel included Fc blocker (1:100), Fixable Viability Stain 780 (FVD, 1:500), Ly6G-BV421 (1:200), CD11b-BUV395 (1:200), CD3ε-PE (1:200), MerTK-BV711 (1:200), and Annexin V-FITC (1:20). For BMT studies, the staining panel included Fc blocker (1:100), Fixable Viability Stain 780 (FVD, 1:500), Ly6G-BV421 (1:200), CD11b-BUV395 (1:200), MerTK-BV711 (1:200), and Annexin V-APC (1:20). Neutrophils were defined as CD11b⁺Ly6G⁺, Macrophages were defined as CD11b⁺Ly6G⁻, and T cells were defined as CD11b⁻CD3⁺. The gating strategies for neutrophil recruitment and apoptosis are shown in Supplementary Figure S1 & S2.

2.5. Assessment of macrophage efferocytosis of neutrophils in lung

BM nucleated cells from DsRed-WT and DsRed-CF mice were obtained as described above, resuspended in RPMI 1640 medium supplemented with 1x Antibiotic-Antimycotic (Gibco). The cells (1 ×10⁷ per recipient) were injected retro-orbitally into each non-fluorescent CF recipient mouse that had been pre-challenged with 10 mg/kg zymosan intratracheally. Day 3 post-challenge, mice were euthanized for lung lavage and assessment of neutrophil-laden macrophages in the BALF by Amnis ImageStream MK II Flow Cytometer. The staining panel included Fc blocker (1:100), Fixable Viability Stain 780 (FVD, 1:500), Ly6G-BV421 (1:200), and CD11b-FITC (1:200). The gating strategy is shown in Supplementary Figure S3.

2.6. Ex vivo neutrophil migration and apoptosis assays

Mouse neutrophils were isolated from bone marrow by Percoll gradient (52 %/64 %/72 %) centrifugation, as previously published [27]. Mature neutrophils were harvested from the interface between 64 % and 72 %, washed twice with cold RPMI 1640 medium, and resuspended in RPMI 1640 supplement with 2 % FBS and 1x Antibiotic-Antimycotic at 2 × 10⁶ cells/ ml. Neutrophil migration assay was performed using a HTS Transwell-96 Permeable Support with 3.0 μm pore polycarbonate membrane (Corning). MIP-2 (Recombinant Mouse CXCL2/MIP-2) (R&D System) and Leukotriene B4 (Cayman Chemical) were diluted in PRMI 1640 supplied with 2 % FBS to a final concentration of 1 ng/ml and 10 nM, respectively. To the bottom of each well 235 μl of chemoattractant-containing cell culture medium or plain cell culture medium was placed, while into the insert 75 μl of WT or CF cell suspension was added. Then, the transwell plates were placed in an incubator at 5 % CO₂, 37°C for 2 h. Cells migrated to the bottom well were counted. For ex vivo neutrophil apoptosis assessment, bone marrow cells were co-cultured with serum-opsonized zymosan particles at an MOI of 1:5 in regular chloride Ringer’s buffer supplemented with 10 % FBS and 1x Antibiotic-Antimycotic up to 12 h. At each indicated time point, the cells were stained with Fc blocker (1:100), Fixable Viability Stain 780 (FVD, 1:500), Ly6G-BV421 (1:200), and Annexin V-FITC (1:20), followed by flow cytometry analysis (BD FACSymphony^™ A3).

2.7. Statistics

Data in bar graphs were presented as mean ± SD. Data presented in box plots showed the minimum, lower quartile Q1, median, upper quartile Q3, and maximum. Student’s t-test was used for one-factor group comparisons, if the data passed the normality test. A two-way ANOVA method with Tukey test was used for 2-factor group comparisons. Fisher’s exact test was used for survival assays. Statistical significance was considered at P<0.05. All statistical analyses and graphs were made by GraphPad Prism 10.

3. Results

3.1. Generation and characterization of congenic fluorescent CF and WT mice

Fluorescent CF mice were generated by breeding non-fluorescent CF mice with EGFP^+/+ or DsRed^+/+ WT mice. After multiple rounds of breeding and screening, fluorescent CF mice were obtained and identifiable under UV light (Fig. 1A). To confirm Cftr knockout, PCR genotyping was performed to amplify a Cftr fragment encompassing Exon-11, giving rise to a large WT band and/or a small Exon-11-deleted band, depending on the genotype (Fig. 1B). To evaluate fluorescent protein expression in leukocytes, mean fluorescence intensity (MFI) of blood nucleated cells was measured by flow cytometry. The MFIs of the cells from homozygous DsRed or EGFP animals were approximately double those of their corresponding heterozygous siblings (Fig. 1C & D). To validate neutrophil expression of the fluorescent proteins, neutrophils (CD45⁺Ly6G⁺) from EGFP-CF or DsRed-CF mice were interrogated (Fig. 1E). Fluorescence of EGFP or DsRed in the neutrophils was clearly distinguishable (Fig. 1F & G). Therefore, these fluorescent CF mice robustly express EGFP or DsRed in their mature neutrophils, which is critical to our subsequent investigations.

Fig. 1.

Fluorescent CF mice. (A) Fluorescence was detected in the hairless areas (feet, ears and tail) of the fluorescent CF mice under an epi-illumination of UV. Left: EGFP-CF mouse; Middle: non-fluorescent CF mouse; Right: DsRed CF mouse. (B) Genotyping PCR showing the gel electrophoresis bands specific to Cftr^+/+, Cftr^+/− and Cftr^−/−. (C) Scatter plot and histogram data of EGFP expression in EGFP^−/−, EGFP^+/−, and EGFP^+/+ mice. (D) Scatter plot and histogram data of DsRed expression in DsRed^−/−, DsRed^+/−, and DsRed^+/+ mice. (E) Flow cytometry gating of neutrophils from the blood. (F & G) Histograms showing fluorescence intensity of neutrophils in fluorescent CF mice and non-fluorescent CF mice. (F) EGFP fluorescence intensity in the neutrophils of EGFP^+/+ mice and EGFP^−/− mice. (G) DsRed fluorescence intensity in the neutrophils of DsRed^+/+ mice and DsRed^−/− mice. (H & I) Survival rate at Day 35 of fluorescent CF mice and their sibling WT controls without Colyte water. (H) EGFP mice. (I) DsRed mice. Fisher’s exact test was performed to compare the survival rate of CF mice at Day 35 with WT controls. (J) Comparison of survival patterns of DsRed-CF, EGFP-CF and non-fluorescent CF mouse colonies under laxative application.

CF mice typically develop severe intestinal disease and die of intestinal obstruction before or during weaning [28,29]. However, laxative supplementation into their drinking water largely overcomes the intestinal problem and sustains the mice to adulthood [30]. To assess whether our fluorescent CF mice develop a similar intestinal disease as their parental non-fluorescent CF mice, 11 EGFP-CF mice, 6 EGFP-WT mice, 12 DsRed-CF mice, and 11 DsRed-WT mice were maintained under regular mouse drinking water without laxative from birth. By Day 35, all (100 %) EGFP CF mice, and 11 (91.7 %) DsRed CF mice died, while WT mice had no mortality (Fig. 1H & I). Moreover, maintaining the fluorescent CF mice on laxative water led to a significant increase in survival rate. Across all litters, survived CF mice accounted for roughly 15 % of the offspring, despite being lower than the theoretical 25 % frequency. The percentage of WT mice was ~25 % and that of heterozygotes ~50 % (Fig. 1J), close to the theoretical numbers. This pattern of survivability of the EGFP-CF or DsRed-CF colony was comparable to that of the published non-fluorescent CF mice [29]. Therefore, the genotype and phenotype of the fluorescent CF mice are consistent with those of the non-fluorescent CF mice.

3.2. CFTR loss of function in neutrophils does not affect their recruitment to inflamed lungs

To compare the first-wave neutrophil recruitment to an inflamed lung, non-fluorescent WT and CF mice were intratracheally challenged with 10 mg/kg zymosan (Fig. 2A). The rationale for selecting zymosan is to ensure identical single-time activation of both WT and CF lungs, and zymosan is a CF-related immunostimulant, as there is a strong association between persistent lung colonization with Candida albicans and disease progression in CF patients [31–33]. At Day 1 post challenge, neutrophils mobilized to both WT and CF lungs accounted for ~80 % of the BAL cells, while other immune cells, such as macrophages and T cells, only occupied a small portion (Fig. 2B), suggesting that the first wave of immune cell recruitment and the early inflammation is largely a neutrophil-dominant event. No significant difference in the first-wave neutrophil recruitment was found between the two genotypes. To determine which cell type, neutrophils or epithelial cells, in the lung dictated the neutrophilic inflammation, we conducted a competitive BMT experiment to compare the recruitment of WT and CF neutrophils to the same lung. Briefly, non-fluorescent WT or CF recipient mice were conditioned by lethal irradiation, and a mixture of DsRed-CF and EGFP-WT bone marrow cells at an equal ratio was infused into each recipient (Fig. 2C). Before transplantation, ~80 % of the donor BM cells were found expressing the respective fluorescence (Fig. 2D & E). After 1:1 mixing, the ratio between EGFP-positive cells and DsRed-positive cells was close to 1:1 (Fig. 2F). Four weeks post-BMT, engraftment was confirmed by measuring fluorescent leukocytes in the blood via flow cytometry. There were ~43 % of EGFP and 45 % DsRed leukocytes in the blood of each BMT mouse, maintaining a roughly equal ratio (Fig. 2G). Then, the established chimeric mice were intratracheally challenged with zymosan. At Day 1 post challenge, the dominant BAL cells were neutrophils (CD45⁺Ly6G⁺) in both WT and CF recipient mice, and the lungs were infiltrated by an equal number of EGFP-WT and DsRed-CF neutrophils, after normalizing to the blood neutrophil percentage (Fig. 2H–J). This result suggests that migration and recruitment of the CF and WT neutrophils to the lungs are comparable, which are independent of the lung phenotypes. To confirm the above in vivo results, we performed an ex vivo neutrophil transmigration assay. WT and CF BM neutrophils were placed into the upper chamber of a permeable support insert, and MIP-2 or LTB4 was added to the bottom media. Compared with non-chemoattractant control group, MIP-2 and LTB4 significantly increased both CF and WT neutrophil transmigration, and no difference was detected between the genotypes (Supplementary Figure S4). Taken all together, CFTR loss of function does not affect neutrophil sensing of chemoattractants and migrating to the lung.

Fig. 2.

Lung neutrophil recruitment with sterile zymosan challenge. (A) Workflow for comparing the first-wave neutrophil recruitment to the lungs of WT and CF mice. (B) Differentials of BAL cells at Day 1 post challenge. Neutrophils were defined as CD11b⁺Ly6G⁺, Macrophages were defined as CD11b⁺Ly6G⁻, and T cells were defined as CD11b⁻CD3⁺. Data are presented as mean ± SD. Statistical differences were determined by two-way ANOVA with Tukey’s post hoc test. (C) Workflow for comparing the first-wave neutrophil recruitment to the lungs of WT and CF recipient mice after bone marrow transplantation (BMT). (D-G) Confirmation of the ratio of WT and CF bone marrow cells before and after BMT. (H) WT and CF neutrophils in the BALF were distinguished by EGFP and DsRed in CD11b⁺Ly6G⁺ cells. (I) WT and CF neutrophils from the BALF of WT recipients. (J) WT and CF neutrophils from the BALF of CF recipients. Data are presented as mean ± SD. Statistical differences were determined by a two-tailed Student’s t-test. All experiments were independently repeated at least three times with a sample size of n = 4–6 per group.

3.3. Delayed neutrophil clearance exacerbates pulmonary neutrophilic inflammation in CF lungs

Time-dependent profiling of BAL cells following zymosan intratracheal challenge was conducted (Fig. 3A). While the total cell counts of lung inflammatory cells were comparable between WT and CF mice at Day 1 post-challenge, a more rapid decrease in the number was observed in WT lungs (Fig. 3B), and the CF lungs had a much slower pace of such a decrease (Fig. 3B). Specifically, CF lungs exhibited significantly more neutrophils on day 3 post-challenge (Fig. 3C). Other immune cell types, such as macrophages and T cells, gradually increased but without any genotype difference (Fig. 3D & E). Thus, zymosan challenge triggers neutrophilic inflammation in the CF lungs, to which the slow removal of neutrophils is a key contributing factor.

Fig. 3.

Neutrophilic inflammation in the CF lungs after zymosan challenge. (A) Workflow for profiling BALF cells. (B) BAL cell concentrations in BALF from WT and CF mice at different time points. (C-E) BAL cell differentials of WT and CF mice in a time-dependent manner. (C) Concentrations of BAL neutrophils. (D) Concentrations of BAL macrophages. (E) Concentrations of BAL T cells. Neutrophils were defined as CD11b⁺Ly6G⁺, Macrophages were defined as CD11b⁺Ly6G⁻, and T cells were defined as CD11b⁻CD3⁺. Data are presented as mean ± SD. Statistical differences were determined by two-way ANOVA with Tukey’s post hoc test. **p < 0.01. All experiments were independently repeated at least three times with a sample size of n = 4–6 per group.

3.4. CF neutrophils instead of pulmonary epithelial cells are responsible for delayed neutrophil apoptosis

To investigate the fate of those lung-homed neutrophils, WT and CF mice were similarly challenged, and the BAL cells obtained at different time points were subjected to Annexin V staining and flow cytometric analysis (Fig. 4A). A significantly lower percentage of apoptotic neutrophils was detected in the BAL cells from CF mice, as compared to that from WT mice at Day 3 post challenge. Such a difference diminished at Day 5 post challenge (Fig. 4B), suggesting that CF neutrophils have a delayed apoptosis, which retards the lung resolution of inflammation and promotes neutrophilic inflammation. This phenotype was recapitulated in WT mice treated with CFTR_inh-172 (Fig. 4C & D). To determine that this delayed apoptosis is intrinsic to neutrophils, we isolated the neutrophils from WT and CF mice and cultured them ex vivo. To the cultures, serum-opsonized zymosan was added to activate the cells. At different times, the cells were sampled and analyzed for apoptosis. CF neutrophils consistently showed reduced apoptosis from 30 min to 12 h post-stimulation (Fig. 4E & F), as compared to WT counterparts.

Fig. 4.

Delayed apoptosis of CF neutrophils. (A) Workflow for assaying neutrophil apoptosis in non-BMT mice in a time-dependent manner. (B) Percentages of apoptotic neutrophils in the BALF from WT and CF mice at Day 1, 3 and 5 post challenge. Statistical differences were determined by two-way ANOVA with Tukey’s post hoc test. *p < 0.05. (C) Workflow for assaying neutrophil apoptosis in WT mice with CFTR_inh-172 treatment. (D) Percentages of apoptotic neutrophils in the BALF from WT mice treated with or without CFTR_inh-172 at Day 3 post zymosan challenge. Statistical differences were determined by a two-tailed Student’s t-test. *p < 0.05. (E) Workflow for assaying neutrophil apoptosis ex vivo. (F) Percentages of apoptotic neutrophils at 0.5, 6 and 12 h-post challenge of opsonized zymosan. Statistical differences were determined by two-way ANOVA with Tukey’s post hoc test. *p < 0.05, ****p < 0.0001. (G) WT and CF apoptotic neutrophils in the same BALF were distinguished by EGFP and DsRed in CD11b⁺Ly6G⁺Annexin V⁺FVD⁻ cells. (H) WT and CF apoptotic neutrophils in the same BALF of WT recipients. (I) WT and CF apoptotic neutrophils in the same BALF of CF recipients. Data are presented as mean ± SD. Statistical differences were determined by a two-tailed, unpaired Student’s t-test. ***p < 0.001, ****p < 0.0001. All experiments were independently repeated three times with a sample size of n = 4–6 per group.

To further validate the finding, WT or CF recipient mice that had received a mixture of an equal ratio of DsRed-CF and EGFP-WT BM cells were intratracheally challenged by zymosan as before (Fig. 4G). At Day 3 post challenge, CD45⁺Ly6G⁺ BAL cells were gated, and Annexin V/FVD staining distinguished apoptotic cells and live cells (Fig. 4H). Data show that in all lungs, regardless of their genotypes, CF neutrophils had a significantly lower percentage of apoptosis as compared to WT neutrophils (Fig. 4I & J). This finding clearly suggests that the delayed apoptosis observed in CF neutrophils is intrinsic to neutrophils, rather than being influenced by lung epithelial cells or microenvironments.

3.5. CF neutrophil-delayed apoptosis is linked to reduced HOCl production

Our previous reports have documented that defective CFTR in CF neutrophils fail to transport chloride to phagosomes [10,27], and compromise HOCl production in the organelle [6]. As HOCl is a potent oxidant that not only arms neutrophils to fight against pathogens, but also make the cells prone to collateral damage, we hypothesized that the insufficient HOCl production by CF neutrophils underlies the delayed apoptosis. According to the bio-synthesis reaction $({\text{H}}_2{\text{O}}_2+{\text{H}}^{+}+{\text{Cl}}^{-}\stackrel{\text{MPO}}{\to }\text{HOCl}+{\text{H}}_2\text{O})$, HOCl production is limited by the availability of the substrate Cl⁻, and the activity of myeloperoxidase (MPO). If our hypothesis is correct, interception of HOCl synthesis should replicate the CF phenotype. To this end, we challenged WT and Mpo^−/− mice intratracheally with zymosan, and three days later the lungs were lavaged and the cells were subjected to neutrophil and Annexin-V staining. Flow cytometry data acquired indicate that Mpo^−/− lung neutrophils had a significantly lower apoptosis rate (Fig. 5A). This phenomenon was also recapitulated in WT mice treated with MPO-specific inhibitor 4-ABAH (Fig. 5B). Furthermore, we isolated neutrophils from WT and Mpo^−/− mice, which were then cultured and challenged with serum-opsonized zymosan for 4 h. After Annexin-V staining, Mpo^−/− neutrophils showed significantly reduced apoptosis (Fig. 5C). To directly implicate HOCl in the delayed apoptosis event, we treated WT neutrophils with 50 mM methionine, a potent HOCl scavenger, over zymosan challenge. As compared to the non-treated cells, the methionine-treated neutrophils had a significantly lower apoptosis rate (Fig. 5D), indicating that diminishing HOCl production in neutrophils delays their apoptosis.

Fig. 5.

Delayed neutrophil apoptosis by HClO blocking. (A) Percentages of apoptotic neutrophils in the BALF from WT and Mpo^−/− mice at Day 3 post zymosan challenge. Statistical differences were determined by a two-tailed Student’s t-test. ****p < 0.0001. (B) Percentages of apoptotic neutrophils in the BALF from WT mice treated with or without 4-ABAH MPO inhibitor at Day 3 post zymosan challenge. Statistical differences were determined by a two-tailed Student’s t-test. **p < 0.01. (C) Percentage of apoptotic neutrophils isolated from WT or MPO-deficient (Mpo^−/−) mice, 4 h-post challenge of opsonized zymosan. Statistical significance was determined using a two-tailed Student’s t-test. *p < 0.05. (D) Percentage of apoptotic neutrophils in the presence or absence of the HOCl scavenger methionine (50 mM), 4 h-post challenge of opsonized zymosan. Statistical significance was determined using a two-tailed Student’s t-test (**p < 0.01). All experiments were independently repeated three times with a sample size of n = 4–6 per group.

3.6. CF macrophages have enhanced efferocytosis

Another potential contributing factor to CF neutrophilic inflammation is the rate of neutrophil disposal or clearance. Macrophages play a pivotal role in this process, referred to as efferocytosis characterized by upregulation of cell surface marker MerTK and phagocytosis of apoptotic cells [34]. To compare macrophage MerTK expression, we similarly challenged WT and CF mice intratracheally with zymosan (Fig. 6A). The obtained BAL cells were immunostained for MerTK expression. Flow cytometry data demonstrate an increased population of MerTK⁺ cells with high granularity in a time-dependent manner in both WT and CF BAL cells (Fig. 6B). This heightened granularity suggests phagocytosis of either zymosan particles or apoptotic cells. The CD11b⁺Ly6G⁻ macrophage population was focused (Fig. 6C). By Day 3 after zymosan challenge, the CF lungs exhibited a significantly higher number of the CD11b⁺Ly6G⁻MerTK⁺ macrophages, suggesting more macrophages in CF lungs are primed for and/or participating in apoptotic cell clearance (Fig. 6D).

Fig. 6.

Clearance of apoptotic cells by phagocytic macrophages. (A) Workflow for quantifying phagocytic macrophages in non-BMT mice. (B) Increased granularity of MerTK⁺ cells in the BALF post zymosan challenge. (C) Determination of MerTK⁺ macrophages in the BALF. (D) Percentages of MerTK⁺ macrophages in the BALF from WT and CF mice at Day 3 and Day 5 post-challenge. Statistical differences were determined by two-way ANOVA with Tukey’s post hoc test. **p < 0.01. (E) Workflow for quantification of apoptotic cell-laden macrophages. (F) Co-localization of DsRed-positive cells with macrophages (CD11b⁺Ly6G⁻). (G) Representative pictures for DsRed-positive cells-laden macrophages. (H) Percentages of WT apoptotic cell-laden macrophages and CF apoptotic cell-laden macrophages. Data are presented as mean ± SD. Statistical differences were determined by a two-tailed, unpaired Student’s t-test. *p < 0.05. All experiments were independently repeated three times with a sample size of n = 4–6 per group.

To compare the capability of WT and CF macrophages to clear apoptotic cells, BM nucleated cells from DsRed-WT or DsRed-CF mice were harvested and infused into non-fluorescent CF recipient mice that had been pre-challenged with zymosan (Fig. 6E). At Day 3 post-challenge, the recipient macrophages from BALF were immunostained for macrophages (CD11b⁺Ly6G⁻). The stained macrophages with DsRed-positivity were captured by Imaging flow cytometry (Fig. 6F & G). The percentage of DsRed-positive macrophages from the mice transfused with CF BM cells was notably higher than that from those animals that had received WT BM cells (Fig. 6H). These findings align well with the increased presence of MerTK⁺ macrophages in CF lungs, indicating an increased capacity of macrophages to clear apoptotic cells, largely neutrophils, in CF in our experimental setting.

4. Discussion

CF lung disease is characterized by chronic airway infection, neutrophil-dominant inflammation and mucopurulent obstruction. Epithelial CFTR dysfunction, leading to airway surface liquid depletion and airway mucus overproduction, has been extensively studied and is widely believed to be the driver of CF pathogenesis [35]. Based on this concept, current therapies, including the highly effective modulator therapy, have been developed. Individuals with CF possessing at least one allele of the F508del mutation are highly responsive to ETI, a triple combination of Elexacaftor, Tezacaftor and Ivacaftor, which leads to a near correction of CF-related epithelial dysfunction [36,37]. Evidence includes 1) normalization of sweat chloride levels, 2) diminished mucus production, and 3) significant improvements in disease severity and quality of life of the treated patients. Despite these significant impacts, CF lung infection and inflammation persist [38], suggesting that other systems implicated in the pathogenesis may be less amenable to the modulation. In CF lungs, neutrophils predominate all inflammatory cells. This can be a sequela of 1) increased recruitment, 2) prolonged survival, 3) reduced efferocytosis or clearance, or 4) a combination of any of these. In this study, we observed that upon lung stimulation with zymosan, a non-proliferative agent, CF neutrophils exhibited a delayed apoptosis, which led to excessive neutrophil accumulation. More importantly, such an anomaly was independent of lung phenotype, suggesting that it is the neutrophils, not the epithelial cells, that are responsible for neutrophilic inflammation in CF lungs. This finding has important implications, predicting that CF therapies must include targeting the defective neutrophils for an effective treatment of the neutrophilic inflammation.

In this report, we generated DsRed-CF and EGFP-CF mice and their congenic WT control mice. Using these animals, we conducted competitive BMT, allowing an unbiased comparison of recruitment, cell death, and clearance between WT and CF neutrophils within the same lung environment. This experimental design eliminated systematic variations arising from individual differences and environmental influences. Thus, our data provide a more precise insight into the underlying cause of neutrophilic inflammation in CF lungs.

Neutrophils are the first responder and the major type of immune cells fighting extracellular pathogen infection. To execute this function, they preferably phagocytose and contain microbes in phagosomes where the microbicidal toxins, either pre- or de-novo synthesized, are released [39,40]. Phagocyte NADPH oxidase (NOX2) shuttles electrons from cytoplasmic NADPH to molecular oxygen in phagosomes or the extracellular space to produce superoxide $({}_{}{}^{•}\text{O}_2^{-})$, which is converted to hydrogen peroxide (H₂O₂) [41], an oxidant with moderate antimicrobial strength. Subsequently, MPO facilitates a two-electron oxidation of chloride using the H₂O₂, yielding HOCl that possesses a significantly higher oxidation potential [42]. Because HOCl is neutral in charge and small in size, it can easily penetrate bacterial cell walls and membranes, a key feature that ensures its effectiveness as an antimicrobial agent. Once inside a bacterium, HOCl can disrupt cell membranes, oxidize key cellular components, and denature proteins. It can also damage DNA and RNA, and inhibit protein and DNA synthesis [43]. In a similar fashion, neutrophils themselves can be damaged and die with the “enemies”. There are at least two distinct regulated cell death pathways in neutrophils that interplay with each other to decide the way of the cell death. Initially, neutrophils initiate apoptotic signaling pathways in response to intrinsic stressors such as oxidative damage or via extrinsic cues like pathogen-associated molecular patterns (PAMPs), ultimately committing to programmed cell death. These apoptotic neutrophils will be removed by macrophages. However, if an apoptotic neutrophil is not cleared promptly, then it has time to open the pores formed by gasdermin E (GSDME) through the mitochondrial- and death receptor-mediated pathway [44]. This leads to the activation of peptidylarginine deiminase 4 (PAD4), which results in NETosis as a second death process. Our data indicate that CF neutrophils have a delayed apoptosis, which is consistent with previous findings [21,22,45,46]. Importantly, we provide an original evidence suggesting that reduced HOCl production in CF neutrophils is primarily responsible for the observed dysregulated apoptosis. Reactive oxygen species have long been known to prompt programmed neutrophil death [47]. Direct evidence comes from the neutrophils with oxidant product deficits. Neutrophils from patients with chronic granulomatous disease (CGD) produce no oxidants at all, including ${}_{}{}^{•}\text{O}_2^{-}$, H₂O₂ and HOCl, and have shown a prolonged apoptosis [48]. Moreover, MPO-deficient neutrophils which lack HOCl production revealed a significantly decreased rate of apoptosis induced by three different stimuli – phorbol 12-myristate 13-acetate (PMA), opsonized streptococcus (OST), and N-for-myl-met-leu-phe (fMLP) [49]. In the current study, we investigated zymosan induction of neutrophil apoptosis in genetic knockout (Cftr^−/− and Mpo^−/−) mice or WT mice treated with CFTR inhibitor or MPO inhibitor. The data clearly demonstrated an association of HOCl production with CF neutrophil apoptosis.

Our data also show that WT and CF neutrophils have no difference in mobilization to an inflamed lung. Neutrophil recruitment is a multistep process that includes sensing and navigating chemoattractant gradients, exiting the bloodstream via extravasation, migrating through the interstitial tissue, and traversing the lung epithelial barrier to reach the target site in the airway lumen. Our data indicated that CF neutrophils are comparable with WT neutrophils in their responsiveness to lung recruitment. In the current study, zymosan was used to induce lung inflammation, which mostly activates Toll-like receptor-2 (TLR-2) and Dectin-1 [50]. Previous publications reported that when challenged with P. aeruginosa or its lipopolysaccharides (LPS), WT and CF lungs have a comparable neutrophil recruitment [51,52]. It is noteworthy that these stimulations are through TLR-4. Intriguingly, CF macrophages have shown an enhanced efferocytosis of neutrophils, suggesting that CFTR dysfunction does not impair their capacity for phagocytosis and clearance of apoptotic cells in our experimental setting. This conclusion does not preclude the possibility of impairments in other regulated cell death pathways. Indeed, it was reported that CF neutrophils display a higher activity in producing neutrophil extracellular traps (NETs), a process termed NETosis [22]. Future studies are warranted to understand how the way of CF neutrophil demise is decided.

It is important to recognize the close interplay between infection and inflammation. In CF lungs, impaired bacterial clearance allows persistent bacterial colonization, which continuously stimulates the lung and exacerbates inflammation. Employing a reductionist approach, we utilized zymosan to study sterile inflammation. This approach effectively eliminates the confounding influence of infection. Since zymosan does not propagate or replicate, its stimulatory effect is one-time and transient. Our data showed that zymosan-induced neutrophilic inflammation was resolved by Day 5 post-challenge. Of note, neutrophil accumulation should be understood as a dynamic process. There is still a possibility that CF lungs might be less effective than WT ones in inactivating zymosan, potentially prolonging its stimulatory effect and resulting in a stronger second-waves of neutrophil recruitment following the initial response, which might also contribute to the observed neutrophilic inflammation. Future studies are needed to further dissect this process.

In conclusion, this study indicates that CF neutrophil inherent defect in HOCl production results in their delayed apoptosis, thereby promoting neutrophilic inflammation in CF. Therefore, targeting the neutrophil defect may be essential for resolving CF-related neutrophilic inflammation in the modulator era.

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.biopha.2026.119092.

Data availability

Data will be made available on request.