TGFβ induces excessive pulmonary IL-6 secretion in cystic fibrosis via PI3K

Abstract

Cystic Fibrosis (CF) is characterized by impaired mucociliary clearance and pulmonary infections. Accumulating evidence suggests that fundamentally abnormal inflammatory responses also contribute to CF pathology. TGFβ, a pleiotropic cytokine, is a modifier of CF lung disease; its mechanism of action in CF is unclear. Previous studies have shown that TGFβ induces IL-6 secretion from lung epithelium, which may drive worse pulmonary outcomes in CF and other lung diseases. However, the nature of the TGFβ/IL-6 relationship in CF is not fully understood. In this study, we demonstrated that TGFβ and IL-6 concentration were positively associated in bronchoalveolar lavage fluid from children with CF. Furthermore, pulmonary TGFβ exposure in a CF mouse model induced heightened IL-6 secretion when compared with non-CF mice. CF airway epithelial cells had increased IL-6 secretion and PI3K signaling after TGFβ exposure. In wild type airway epithelium, TGFβ exposure and CFTR inhibition synergistically provoked IL-6 secretion. Restoration of CFTR function by a CFTR modulator and inhibition of PI3K signaling both normalized IL-6 secretion from CF airway epithelial cells. These data indicate that TGFβ drives abnormal IL-6 secretion via the PI3K pathway in the CF airway, demonstrating an inherent inflammatory abnormality in CF and suggesting potential therapeutic targets.

Graphical Abstract

New and Noteworthy:

The etiology of IL-6 oversecretion in cystic fibrosis (CF) is unclear, as is the mechanism of CF lung disease modification by TGFβ. We show that TGFβ induces IL-6 oversecretion in human and mouse models of CF. In mechanistic studies, we further demonstrate that loss of CFTR function drives increased IL-6 secretion via the PI3K pathway downstream of TGFβ. Treatment of CF airway epithelial cells with a CFTR modulator rescues this IL-6 oversecretion.

Introduction

Cystic fibrosis (CF) is an autosomal recessive disease affecting an estimated 160,000 people worldwide (1). Dysfunction of the cystic fibrosis transmembrane conductance regulator (CFTR) protein, a chloride and bicarbonate transporter, causes thickened mucus, impaired mucociliary clearance, and recurring infections and inflammation in people with CF. Although inflammation in CF is provoked by mucus stasis and infection, there is also evidence that loss of CFTR function directly causes intrinsic immune dysregulation (2–8). This suggests a fundamental inflammatory defect that may predispose people with CF to early bronchiectasis and progressive lung disease despite aggressive treatment of microbial pathogens. Furthermore, the amplified inflammatory environment may decrease the effectiveness of CFTR modulator medications such as elexacaftor/tezacaftor/ivacaftor (ETI), a therapy that has led to remarkable benefits for many people with CF (9). Understanding abnormal inflammatory responses in CF and other chronic lung diseases is crucial in developing targeted therapies to slow lung disease progression.

Transforming growth factor β, a pleiotropic cytokine, is increased in the CF lung and is a biomarker of lung disease severity and treatment response (10, 11). TGFβ is a potent CF disease modifier, as individuals with higher producing TGFβ polymorphisms have worse lung disease outcomes (12). The mechanism of CF lung disease modification by TGFβ is likely multifactorial, including decreased CFTR expression and function, airway remodeling, and altered inflammation (13–18). TGFβ is a key regulator of lung health and disease beyond CF (19, 20). Detrimental effects of TGFβ are heightened in CF, including mucus secretion and airway hyperreactivity (AHR) (21). Multiple pathways are activated downstream of TGFβ, including the noncanonical phosphoinositide 3-kinase (PI3K) pathway. We have shown that PI3K signaling mediates TGFβ induced AHR and goblet cell hyperplasia in a CF animal model (22). However, the mechanism by which TGFβ drives abnormal inflammation in CF is poorly understood.

IL-6 is another cytokine identified as an important mediator of pulmonary inflammation and disease (23). Numerous studies, including analysis of young children with CF, have described increased IL-6 levels in people with CF beginning in early childhood (24–26). Viral respiratory illnesses and nicotine exposure increase IL-6 in infants with CF, and IL-6 levels are elevated during pulmonary exacerbations (26–28). CF animal models also demonstrate increased pulmonary IL-6 levels, including CF mice, despite their lack of spontaneous significant lung disease (29, 30). Finally, treatment with CFTR modulators reduces IL-6 levels in the lungs of children and adults with CF, although it is unclear if this is a direct effect of improved CFTR function or a secondary consequence of improved mucociliary clearance and decreased microbial burden (31–35).

TGFβ exposure provokes IL-6 expression and secretion in a variety of cells, including A549 cells and lung fibroblasts (36–38). This is further enhanced in CF airway cells. Immortalized CF bronchial epithelial cells have increased TGFβ pathway activation and increased IL-6 production compared to cells expressing wild-type CFTR (39). Together, TGFβ and IL-6 can induce differentiation of Th17 lymphocytes, which are involved in early CF pathologic immune responses (40, 41). However, the link between TGFβ exposure and IL-6 production in CF, and how this may drive a pro-inflammatory phenotype, is unclear.

In this study, we examined the relationship between TGFβ and IL-6 secretion in CF. We hypothesized that increased pulmonary IL-6 secretion in CF is driven, at least in part, by TGFβ. Studies in children with CF and a CF mouse model showed an association between pulmonary TGFβ and IL-6. Mechanistic studies in airway epithelial cells (AECs) demonstrated that AECs secrete IL-6 in response to TGFβ exposure. Furthermore, this TGFβ effect on IL-6 secretion is enhanced in CF cells and mediated via the PI3K signaling pathway. These findings suggest that abnormal TGFβ-induced cell signaling pathways contribute to the pro-inflammatory environment in CF lungs, even in the absence of infection. Some of these data have been previously published in the form of an abstract (42).

Materials and Methods

Human subjects.

BAL specimens were obtained from the CCHMC Pulmonary Biorepository. The study was approved by the Institutional Review Board (IRB) of the Cincinnati Children’s Hospital Medical Center (IRB # 2018–4711 and 2013–3309). All study participants had written informed consent for specimen biobanking prior to the study. Specimens were selected from children with CF under age 6 years who underwent clinically indicated bronchoscopies at the Cincinnati Children’s Hospital from 2003–2015. BAL sample data from 15 individual patients are shown. No patients were on CFTR modulator medications at time of bronchoscopy. Deidentified clinical data, including sex, age, BAL culture results, and indication for bronchoscopy were obtained from the Biorepository.

Mice.

All animal experiments were approved by the Cincinnati Children’s Hospital Institutional Animal Care and Use Committee. Gut-corrected CFTR^tm1kth mice (“CF Mice”) were obtained from the Case Western Reserve University Cystic Fibrosis Mouse Resource Center. These C57BL/6J mice are homozygous for the F508del CFTR mutation (43). Control mice were heterozygous littermates (“Non-CF Mice”). Adult female mice between 9–12 weeks of age were used in this study. Our study exclusively examined female mice, as they develop more pronounced lung disease in other reports (44, 45). It is unknown whether the findings are relevant for male mice. Data shown are from 4 mice per group. As we have previously published (16), mice were intratracheally injected with vehicle (PBS) or Ad-TGFβ (5×10⁷ pfu), a previously described nonreplicating adenoviral vector containing a constitutively active TGFβ1 transgene (46). Previous work with this construct in mice has confirmed no empty vector effects are seen at this dosage (16). Mice were sacrificed 7 days after TGFβ exposure. BAL fluid was collected by flushing 1 mL sterile PBS through a tracheal cannula. Lung tissue was collected for Western blot analysis.

Human airway epithelial cells.

Primary human AECs were obtained from the University of North Carolina-Marisco Lung Institute Tissue Procurement and Cell Culture Core Facility (Chapel Hill, North Carolina; https://www.med.unc.edu/marsicolunginstitute/core-facilities/tissueprocurmentandcellculturecore/). All samples are primary cultured AECs of human bronchial origin. Cells are provided as either wild type CFTR or mutant CFTR (F508del homozygous). Otherwise, cells are provided in a fully de-identified fashion without knowledge of sex, race/ethnicity, or age. Primary cells undergo extensive authentication in the core prior to shipping. Upon receipt, cultures are observed against benchmarked growth parameters (time to confluency, passage success) and morphologically for evidence of ciliation and mucociliary transport. Cultures are routinely fixed and immunostained for critical AEC markers (CFTR, Muc5AC, alpha tubulin, e-cadherin) and morphometric analysis. Mycobacterial testing is performed on all cultures. Data shown are from 3 non-CF donors and 3 CF donors. Cells were grown on permeable inserts at air-liquid interface as previously described (47). Recombinant human TGFβ1 (10 ng/mL; R+D Systems, Minneapolis, MN) or vehicle (4mM HCL/0.1% BSA) was added to Ultroser G-containing media (Crescent Chemical Company, Islandia, NY) for 24 hours. This dosage of TGFβ was selected as physiologically relevant based upon reported TGFβ concentrations in the BAL of children with CF (11) and an estimated BAL to epithelial lining fluid concentration dilution of 1:100 (48). For PI3K inhibition, cells were pretreated with the pan-PI3K inhibitor LY294002 (20 μM; Calbiochem, San Diego, CA) or vehicle (DMSO) for 60 minutes prior to TGFβ. CFTR function was inhibited in WT cells by pretreatment with Inh172 (10 μM; Tocris Bioscience, Minneapolis, MN) for 24 hours prior to TGFβ exposure. CFTR function was rescued in CF cells by pretreatment with elexacaftor/tezacaftor (3 μM/3 μM; Selleck Chemicals, Houston, TX) for 24 hours prior to TGFβ exposure, with the CFTR potentiator ivacaftor (1 μM; Selleck Chemicals) added at time of TGFβ exposure due to prior reports of CFTR instability with prolonged ivacaftor exposure (49).

Cytokine analysis.

TGFβ levels in human BAL were determined via ELISA (R+D Systems). Luminex multiplex assay (Millipore Sigma, Burlington, MA) was used for analysis of all other BAL cytokines. No dilution correction was performed for BAL sample analysis. IL-6 concentrations in cell culture supernatant were determined by ELISA (R+D Systems).

Western blot analysis.

Mouse lung and human AEC lysates were collected and prepared as previously described (16). Protein loading was standardized for all samples: 15 μg for cell lysates and 60 μg for homogenized lung tissue. Primary antibodies used were phosphorylated Stat3 (#9145, Cell Signaling, Danvers, MA; 1:2000 dilution), total Stat3 (#4904, Cell Signaling; 1:2000 dilution), phosphorylated S6 (#4858, Cell Signaling; 1:2000 dilution), and total S6 (#2217, Cell Signaling; 1:1000 dilution). Secondary antibody was anti-rabbit IgG (#7074, Cell Signaling; 1:2000 dilution). Specificity for Stat3 and S6 antibodies has been previously shown in siRNA experiments (50, 51). Protein expression was quantified by chemiluminescent detection with PhosphoImager (Fujifilm, Valhalla, NY) and MultiGauge Software (Fujifilm) was used for analysis.

Ion transport.

I_SC was measured in Ussing chambers as previously described (47). Cell cultures were mounted in chambers under voltage clamp conditions with an apical-low-chloride gradient. Cells were treated sequentially with amiloride (apical 100 μM, Spectrum Chemical Corporation, Gardena, CA), forskolin (apical/basolateral 10 μM, Millipore Sigma)/IBMX (apical/basolateral 100 μM, Acros Organics, Fair Lawn, NJ), ivacaftor (apical 1 μM, Selleck Chemicals), and Inh172 (apical 10 μM, Tocris Biosciences). Change from baseline current to maximum current with CFTR activation is reported as ΔI_SC.

Statistics.

Prism software (GraphPad) was used for analysis. 2-tailed Student’s t test, 1-way ANOVA with Tukey’s multiple comparisons test, or linear regression analysis was used to compare data as indicated in figure legends. Data are displayed as mean ± standard deviation. P<0.05 was considered statistically significant. For primary human AECs, due to the limited availability of biologic samples, each group (control and CF) consists of three samples. Although the small sample size does not allow for formal testing of normality, parametric tests were used for analysis, which is a limitation of this study. To support use of parametric tests, we note that human BAL samples from this study demonstrated a normal distribution of IL-6 concentrations and Western blot data from mice as shown in Figure 2 are also normally distributed, indicating biologic rationale for utilizing parametric tests here. Due to small sample size, these findings should be interpreted carefully and, in future studies, validated with greater numbers.

Figure 2. Pulmonary TGFβ exposure drives greater increases in IL-6 secretion and Stat3 signaling in CF mice compared to non-CF mice.

Intratracheal PBS or Ad-TGFβ was administered to CF (n = 4) and non-CF (n = 4) mice. Concentrations of BAL IL-6 (A), KC (B), IL-1β (C), and TNFα (D) were analyzed. Unlike other cytokines analyzed, IL-6 (A) demonstrated significantly greater elevation in TGFβ exposed CF mice compared to TGFβ exposed non-CF mice. (E) Western blot analysis of whole lung lysates showed increased Stat3 signaling in TGFβ exposed CF mice compared to non-CF mice, as indicated by pStat3/total Stat3. Data were analyzed by one-way ANOVA with Tukey’s multiple comparisons test. NS indicates not significant.

Results

Pulmonary TGFβ correlates with IL-6 in children with CF lung disease.

The relationship between TGFβ elevation and other inflammatory mediators in early CF lung disease is unclear. To investigate associations between pulmonary TGFβ and other CF-relevant cytokines in early disease, BAL fluid was analyzed from 15 children with CF less than age 6 years old undergoing clinically indicated bronchoscopy. None were on CFTR modulator medications. Patient characteristics are described in Table 1. At the time of sample acquisition, the majority of patients were clinically stable, with only 3 (19%) of the patients concurrently diagnosed with a pulmonary exacerbation. Six (40%) of the children grew a microorganism in their BAL sample, most commonly Haemophilus influenzae (27%).

Table 1:

Characteristics for children with CF participating in study.

Patient Characteristics	N = 15
Age (years)	3.6 (1.5)
Sex
Female	8 (53%)
Male	7 (47%)
Ethnicity
Caucasian	14 (93%)
Other	1 (7%)
F508del Mutation
Homozygous	10 (67%)
Heterozygous	3 (20%)
None	2 (13%)
BAL Culture Results
Any organism	6 (40%)
Haemophilus influenzae	4 (27%)
Pseudomonas aeruginosa	0
Staphylococcus aureus	0

Patient characteristics for children with CF undergoing clinically indicated bronchoscopy from whom BAL was analyzed. Data presented as n (%) or mean (sd).

In addition to TGFβ, four other CF relevant cytokines were chosen for analysis: IL-6, IL-8, IL-1β, and TNFα (Figure 1 A–D). Prior studies have demonstrated elevation of these cytokines in the CF lung (27, 52–54). Among these cytokines, only IL-6 demonstrated a significant association with TGFβ levels in the BAL fluid (Figure 1A). An outlier with high IL-6 and TGFβ BAL concentration was noted; this child was not experiencing a pulmonary exacerbation at time of bronchoscopy. With this outlier removed, the association between TGFβ and IL-6 was still strong (R² = 0.32, p = 0.036). These data suggest a link between pulmonary TGFβ and IL-6 in early CF lung disease but do not determine causation. To better understand if TGFβ drives IL-6 production in the CF lung, we evaluated an animal model of TGFβ overexpression.

Figure 1. IL-6 demonstrates a significant association with TGFβ concentration in BAL fluid from children with CF.

BAL fluid from 15 children with CF was analyzed via multiplex immunoassay for (A) IL-6, (B) IL-8, (C) IL-1β, and (D) TNFα. Closed circles indicate BAL from male children, open circles indicate BAL from female children. Unlike other cytokines analyzed, pulmonary IL-6 demonstrated a significant association with TGFβ (P<0.01, linear regression analysis). IL-8, IL-1β, and TNFα concentrations in BAL did not correlate with TGFβ.

Pulmonary TGFβ exposure causes increased BAL fluid IL-6 in a CF mouse model.

CF (F508del/F508del) mice were exposed via intratracheal injection to either PBS control or an adenoviral vector containing a TGFβ transgene (Ad-TGFβ), which induces a sustained, physiologically relevant increase in pulmonary TGFβ (16, 46). Previously, we have shown that pulmonary Ad-TGFβ induces CF-relevant lung disease, including increased mucus secretion, pulmonary inflammation, and abnormal lung mechanics with a more robust phenotype in CF mice (16). TGFβ-induced pulmonary PI3K signaling was higher in CF mice compared to non-CF controls despite similar pulmonary TGFβ exposure, demonstrating a heightened TGFβ response in the CF mouse (16). In the current study, we examined BAL fluid cytokines after pulmonary TGFβ exposure. Mean active TGFβ in BAL of Ad-TGFβ treated mice was 11.2 ± 4.9 ng/mL, with no significant difference between CF and non-CF mice, as previously shown (16). As with the human BAL samples, BAL fluid obtained from these mice was analyzed for IL-6, KC (a murine homologue of IL-8), IL-1β, and TNFα (Figure 2A–D). Only IL-6 demonstrated greater elevation in TGFβ-exposed CF mice compared to TGFβ-exposed non-CF mice (15.6 ± 0.9 μg/mL versus 4.4 ± 1.4 μg/mL, respectively, Figure 2A). TGFβ exposure caused similar increases in BAL fluid TNFα and KC in CF and non-CF mice (Figure 2B, D). BAL fluid IL-1β was variable and did not increase with TGFβ exposure (Figure 2C). Phosphorylated Stat3 protein, an indicator of downstream IL-6 pathway activation, was measured via Western blot analysis of lung homogenates. Both non-CF and CF mice demonstrated significantly increased Stat3 signaling after pulmonary TGFβ exposure; TGFβ-exposed CF mice had 1.8-fold greater Stat3 induction compared to non-CF mice (Figure 2E). Thus, pulmonary TGFβ exposure in CF mice is sufficient to drive heightened pulmonary IL-6 secretion and subsequent IL-6/Stat3 pathway activation.

TGFβ induces increased IL-6 secretion in CF AECs.

To further investigate the cellular mechanism of increased IL-6 secretion in CF, primary human AECs were grown at air-liquid interface. CF (F508del/F508del) and WT AECs were exposed to TGFβ or vehicle, and the supernatant was collected to determine IL-6 secretion after 24 hours. Similar to our mouse studies, TGFβ exposure provoked 2.5-fold greater IL-6 secretion in CF compared to WT AECs (Figure 3A). Downstream IL-6/Stat3 pathway activation was also increased in TGFβ exposed CF cells, but not in WT controls (Supplemental Fig. S1, A-B). Previously, we showed that heightened PI3K signaling in CF mice mediated detrimental effects of pulmonary TGFβ, including AHR and goblet cell hyperplasia (22). To examine whether increased PI3K signaling may also drive enhanced IL-6 secretion in CF AECs, we examined phosphorylation of S6. Phosphorylation of ribosomal protein S6 occurs downstream of multiple pathways, including MAPK/ERK and PI3K/AKT/mTOR (55, 56), making it a nonspecific proxy for PI3K pathway activation; PI3K inhibitor studies were also used to more specifically determine the role of PI3K in IL-6 secretion. TGFβ stimulated phosphorylation of S6 in both WT and CF AECs compared to vehicle, with CF AECs demonstrating a greater increase (12.8-fold increase in CF versus 4.3-fold increase in WT AECs; Figure 3B,C). To determine if CFTR dysfunction was a driver of this altered TGFβ response, WT AECs were treated with a CFTR inhibitor (Inh172, 24-hour pretreatment) and/or TGFβ for 24 hours. While neither TGFβ nor CFTR inhibitor alone caused a significant increase in IL-6 secretion, WT AECs treated with both had significantly increased IL-6 secretion compared to vehicle (Figure 3D). This finding implicates CFTR dysfunction as a critical driver of abnormal TGFβ-induced IL-6 secretion.

Figure 3. TGFβ exposure drives greater increases in IL-6 secretion in CF AECs compared to control AECs; ETI treatment and PI3K inhibition reduces excessive IL-6 secretion.

Primary human airway epithelial cells from CF (n=3) and non-CF (n=3) donors were exposed to TGFβ, and IL-6 secretion and downstream signaling pathways were assessed. (A) ELISA analysis of cell culture supernatant revealed that CF AECs have heightened IL-6 secretion in response to TGFβ. (B,C) TGFβ induced a greater rise in phosphorylation of S6, a proxy for PI3K signaling, in CF AECs (C) compared to WT AECs (B). (D) In WT AECs, TGFβ exposure and CFTR inhibition with Inh172 synergistically increased IL-6 secretion detected by ELISA of cell media. (E) ETI treatment reduced TGFβ-induced IL-6 secretion by CF AECs as detected by ELISA of cell culture supernatant. (F) Phosphorylation of S6/total S6, a proxy for PI3K signaling, was reduced in CF cell lysates after ETI treatment. (G) Treatment with a pan-PI3K inhibitor (LY294002, “LY”) significantly reduced IL-6 secretion from CF AECs. Data were analyzed by two-tailed t-test or one-way ANOVA with Tukey’s multiple comparisons test.

Restoration of CFTR function and inhibition of PI3K signaling reduce IL-6 secretion in CF AECs.

CF AECs were treated with ETI, which improves CFTR function to approximately 28% of WT CFTR function (Supplemental Fig. S2A). Treatment of WT AECs with ETI induced phosphorylation of S6 (Supplemental Fig. S2, B). ETI-treated CF AECs did not demonstrate significant alteration of S6 signaling nor IL-6 secretion (Supplemental Fig. S2, C-D). ETI treatment reduced TGFβ-induced IL-6 secretion to 34% of baseline in CF AECs (Figure 3E). Treatment with ETI also reduced S6 phosphorylation (Figure 3F), indicating that CFTR dysfunction may alter PI3K signaling downstream of TGFβ. To determine if PI3K signaling mediated heightened IL-6 secretion in CF cells, CF AECs were treated with a PI3K inhibitor (LY294002) prior to TGFβ exposure. As expected, this treatment suppressed phosphorylation of S6 (Supplemental Fig. S2E). PI3K inhibition decreased TGFβ-induced IL-6 secretion in CF AECs to less than 33% of that observed in untreated cells (Figure 3G). Taken together, these data suggest that abnormal PI3K signaling in CF cells drives heightened TGFβ-induced IL-6 secretion.

Discussion

IL-6 is a proinflammatory cytokine implicated in the pathogenesis of pulmonary diseases, including asthma, COPD, and pulmonary fibrosis (23, 57). In CF, increased IL-6 has been noted as early as infancy, but the mechanism driving this abnormality was unclear (24–28). In this study, we show that TGFβ induces amplified IL-6 secretion in both animal and cell-based models of CF, and that IL-6 production in response to TGFβ tracks with CFTR activity. IL-6 production is excessive in TGFβ-exposed CF cell and animal models and when CFTR function is pharmacologically inhibited in wild type AECs (Figures 2A, 3A, 3D). ETI treatment of CF AECs corrects this abnormal TGFβ-induced IL-6 secretion, as does PI3K inhibition (Figure 3E, G). These data suggest that TGFβ drives abnormal IL-6 secretion via PI3K signaling in CF AECs, lending support to the notion that CF pulmonary disease is driven in part by inherent pro-inflammatory epithelial dysfunction. Understanding downstream drivers of this inflammation could allow for the use of targeted therapeutics to prevent long term pulmonary damage in CF and related lung diseases.

While the variety of model systems utilized in this study is a strength, more investigation is needed. Individual cellular responses can be impacted by a multitude of factors, including genetic modifiers and the environmental milieu. Furthermore, the CF cell and animal models used in this study carry the F508del CFTR mutation. The applicability of these findings across all individuals with CF, including those with non-F508del mutations, is unclear. Low sample size is an important study limitation, especially in our human AEC experiments. As people with CF live healthier lives in the era of CFTR modulator therapies, fewer individuals are undergoing lung transplantation, limiting availability of these samples. Further studies will be necessary to test these findings in larger populations and with CFTR mutations beyond F508del.

Prior research has established TGFβ as a modifier of CF lung disease (12). Altered inflammatory signaling in response to TGFβ, such as the increased IL-6 secretion demonstrated in this study, may be part of a central mechanism responsible for its contribution to CF lung disease. IL-6 production in CF is implicated in driving heightened Th17 type inflammation, airway smooth muscle abnormalities, and production of other pro-inflammatory cytokines (27, 41, 58). Although treatment with the highly effective modulator combination ETI normalizes TGFβ-dependent IL-6 production in CF AECs (Figure 3E), TGFβ itself can undermine the benefit of ETI treatment by inhibiting CFTR expression (59). This suggests that TGFβ can abrogate the benefits of highly effective modulators on mutant CFTR, priming the CF lung for abnormal TGFβ-driven inflammation.

In this study and in our previous work, we have demonstrated heightened PI3K signaling in CF models (16, 22). Abnormal PI3K signaling may be responsible for driving CF pulmonary pathology, including overproduction of IL-6, and has been suggested as a potential therapeutic target (60–62). A direct linkage between CFTR function and PI3K signaling, however, had not been established. In the current study, we add to the prior observations and show that ETI-based restoration of CFTR function is sufficient to reduce abnormal PI3K function in TGFβ-exposed AECs (Figure 3F). Whether this effect translates to people with CF treated with highly effective modulators is unclear.

Dysregulated inflammation in CF is multifactorial with contributors including mucus stasis, chronic infection, and fundamental abnormalities in inflammatory responses (63). These pro-inflammatory factors are also important in non-CF lung diseases such as asthma, COPD, and non-CF bronchiectasis, where abnormal CFTR function has also been described (64–68). Once inflammation becomes excessive, damage to lung tissue can occur and provoke further mucus obstruction, infection, and inflammation. Our data and those of others suggest that CFTR itself could play a fundamental role in inflammatory homeostasis within and beyond CF lung disease. Understanding drivers of abnormal pulmonary inflammation will be critical as we hone our understanding of CFTR dysfunction and develop new, targeted therapies to treat the hyper-inflammatory pulmonary milieu in chronic lung diseases.

Highly effective modulator therapy, which is available to approximately 90% of people with CF based on their CFTR genotype, has changed the course of CF lung disease by targeting the underlying molecular defect (69). Restoration of CFTR function with these therapies produces marked acute clinical improvements, but abnormal inflammation can persist or rebound; furthermore, the long-term effects of ETI remain unclear (69–72). Our data suggest that ETI attenuates TGFβ-induced IL-6 hypersecretion in CF (Figure 3E), at least acutely. Whether there is persistent inflammation post-ETI, and how this impacts residual CF disease, is a critical question of the post-modulator era.

In summary, our results provide important mechanistic data regarding how TGFβ modifies CF lung disease, elucidating novel relationships between TGFβ, CFTR dysfunction, and dysregulated PI3K and IL-6 signaling. This study also highlights how these signaling relationships may contribute to airway epithelial dysfunction in CF and non-CF lung disease and suggests potential therapeutic targets for further investigation.

Supplementary Material

Supplemental Western Blot Images: https://doi.org/10.6084/m9.figshare.29944301

Supplemental Figure S1: https://doi.org/10.6084/m9.figshare.29944070

Supplemental Figure S2: https://doi.org/10.6084/m9.figshare.29944304

Data Availability:

The raw data supporting the conclusions of this article will be made available by the authors upon request.