Background
Cystic fibrosis (CF) is a monogenic autosomal recessive disorder associated with significant morbidity and mortality that affects approximately 1 in 3000 newborns in the US. Despite significant advances in treatment, the median predicted survival for a CF patient born between 2003 and 2007 is 37 years [
1]. CF arises from mutations in the CF transmembrane conductance regulator (CFTR) that lead to impaired chloride and bicarbonate transport. Impaired CFTR epithelial cell function in patients with CF leads to viscous mucus, impaired mucociliary clearance and airway colonization with pathogenic bacteria, especially
Pseudomonas aeruginosa (P. aeruginosa). These changes in airway biology lead to progressive lung damage and respiratory insufficiency. Recently, the ionocyte was identified as the main CFTR expressing epithelial cell [
2]. However, CFTR expression on both innate and adaptive immune cells has been increasingly appreciated to contribute to immune dysfunction and disease pathogenesis in CF.
The absence of normal CFTR expression on macrophages has been linked to increased inflammatory cytokine production, altered TLR4 trafficking and impaired resolution of infection and inflammation [
3‐
5]. A direct role for CFTR in macrophage function was confirmed by siRNA knockdown of CFTR expression on human alveolar macrophages. Alveolar macrophages with silenced CFTR had increased IL-8 secretion, increased NF-κB phosphorylation and increased caveolin-1 expression [
6]. The T helper 2 (Th2) skewing of CD4 lymphocytes from both patients and
Cftr −/− mice has been well described [
7,
8]. CFTR deficiency has also been linked to diminished regulatory CD4 T cell (Treg) effector function [
9].
B cells are critical for adaptive immune responses and express CFTR mRNA. Human B cells that lack CFTR have impaired chloride conductance as is observed in CFTR-deficient epithelial cells [
10,
11]. B cell-activating factor of tumor necrosis factor family (BAFF) is produced by B cells, T cells and myeloid lineage cells and plays an important role in B cell survival and maturation [
12]. BAFF can bind to three receptors that are constitutively expressed on B cells (BAFF-receptor, transmembrane activator and calcium modulator and cyclophilin ligand interactor and B-cell maturation antigen). BAFF is not produced by B cells at steady state but is induced by antigen-activated helper T cells. BAFF produced by B cells can work in both an autocrine and paracrine manner [
12‐
14]. The importance of BAFF in lung B cell development and immunity was recently reinforced and confirmed to be important in CF. Wild type and
Cftr −/− mice treated with a neutralizing antibody that blocks BAFF resulted in B cell and lung CD4
+ regulatory T cell (Treg) depletion. Blockade of BAFF and resultant B cell depletion increased the lung bacterial burden in both wild type and CFTR deficient mice infected with
P. aeruginosa, although the bacterial load and lung resistance was higher in the CFTR deficient mice [
15]. BAFF has been detected in bronchoalveolar lavage (BAL) fluid from patients with CF and was most elevated in patients infected with
P. aeruginosa. BAFF was not detected in BAL fluid from healthy controls [
16]. BAFF was also shown to be induced in the lungs of wild type mice infected with
P. aeruginosa [
16].
Peribronchial lymphoid follicles (LFs) have been observed in patients with CF and developed in wild type mice in response to bacterial infection. Wild type mice infected with
P. aeruginosa had elevated levels of lung tissue BAFF and B cell chemoattractants including CXCL13 [
16,
17]. Lung B cell BAFF expression has also been shown to correlate with LF development in chronic obstructive pulmonary disease (COPD) [
13]. Lung BAFF and specifically autocrine B cell BAFF production may contribute to the promotion and persistence of airway inflammation as was demonstrated in patients with COPD [
13]. These findings raise questions as to whether LF development may contribute to CF lung pathology. The observation of increased lung BAFF and LFs has been made in lung tissue from patients with CF that have airway colonization with pathogenic bacteria and BAL fluid and lung tissue from wild type mice infected with
P. aeruginosa [
16,
17]. However, a direct role for CFTR in B cell immune function has not been well characterized.
Several murine
Cftr −/− lines deficient in CFTR have been developed and do not develop lung pathology in the absence of direct exposure to pathogenic bacteria [
18]. However, age dependent increases in interstitial macrophages and interstitial thickening have been observed in lung tissue from uninfected
Cftr −/− mice [
19]. A different group examined uninfected
Cftr −/− mice 16 to 20 weeks of age and also observed lung inflammatory cell infiltration that was not present in wild type controls. Interestingly, immunoglobulin chain genes were the genes that were most overexpressed in lung tissue from uninfected
Cftr −/− mice versus wild type controls in this study [
20]. These changes in unchallenged mice suggest that CFTR deficiency may contribute to lung inflammation in the absence of infection over time. Here, we demonstrate a role for CFTR in the promotion of tertiary lung LF development, B cell BAFF and CXCR4 expression and B cell inflammatory cytokine production in the absence of infection.
Discussion
CFTR deficiency on epithelial cells drives alterations in airway biology that promote infection and ineffective airway clearance that lead to the progressive loss of lung function in patients with CF. Patients with identical clinical characteristics and CFTR mutations can have very disparate outcomes. Individual differences in immune responses and maintenance of homeostasis may play a critical role in disease progression and morbidity, independent of CFTR deficiency. However, the findings presented here and prior work in other cell types points to a direct role for CFTR deficiency on immune function in CF that may alter clinical outcomes.
Tertiary LFs can develop in the setting of chronic inflammation and/or infection [
30,
31]. Lung tissue from patients with CF had numerous tertiary LFs, significantly more than the controls. The majority of the LFs were well organized with well-defined germinal centers (Fig.
1a and b). In contrast to secondary lymphoid organs, tertiary LFs are not encapsulated and are directly exposed to local stimuli, antigens and inflammatory cytokines [
30]. The proximity of these LFs to infectious pathogens may facilitate protective immune responses. However, in the case of CF where patients are chronically infected with bacteria that are not eradicated by antibiotics, their proximity may also lead to a feedback loop of inflammatory mediators, leukocyte recruitment and immune activation, which in turn could promote aberrant immune responses.
It is challenging in the setting of chronic infection to tease out what aspects of altered lung immunity are a direct effect of CFTR deficiency and which are driven by chronic infection and immune activation. Prior work has observed lung LF development in wild type mice infected with
P. aeruginosa [
17]. In order to determine the role of CFTR in the development of tertiary LFs and B cell responses, lung tissue from uninfected
Cftr −/− and wild type mice was examined here. Lung tissue from uninfected
Cftr −/− mice had a significantly higher number of LFs compared to wild type controls (Fig.
3a and b). To validate these findings, a second cohort of mice was obtained from a collaborator at a different institution exposed to a different microenvironment and microbiome. We observed a similar increase in the number of pulmonary LFs in uninfected
Cftr −/− mice compared to controls (Fig.
4a and b). Despite the fact that the
Cftr −/− mouse does not recapitulate the phenotype of lung disease observed in humans, there have been reports in the literature of changes in the lungs of uninfected
Cftr −/− mice including lung inflammatory cell infiltration [
19,
20]. Interestingly, the observations were made in mice that were 16 to 24 weeks of age and in the case of Kent, G. et al., were age dependent. It is possible that organized LFs make take time to develop in the absence of infection in
Cftr −/− mice.
The potential important role for age in the development of lymphoid follicle development is also highlighted by the higher number of follicles in the older cohort (cohort 2) compared to cohort 1 (Figs.
3b and
4b). However, differences in environmental or microbiome exposures between the cohorts cannot be excluded. Intestinal isolated lymphoid follicle development has also been shown to significantly increase with age [
32]. As patients with cystic fibrosis achieve longer lifespans, the cumulative effect of immune dysfunction superimposed on the age-related immune senescence may play a role in disease progression.
One of the likely drivers of LF development in the lung is BAFF as has been previously shown in COPD [
13]. We observed a marked increase in BAFF expression in uninfected
Cftr −/− lung LF B cells (Figs.
3a, d and
4a), which may contribute to B cell survival and may drive other autocrine and paracrine effects [
12]. BAFF and tertiary lung lymphoid structures have been shown to have detrimental effects. This was demonstrated in a murine model of COPD where BAFF was blocked by soluble BAFF fusion protein. In addition to reduced inflammation and the prevention of LF development, there was a significant reduction in alveolar wall destruction [
33]. Depletion of BAFF in
Cftr −/− mice prior to infection with
P. aeruginosa depleted B cells and impaired antimicrobial immunity. The mice were euthanized 3 days after infection. LF development was not examined in this study that highlights the importance of effective humoral immunity in CF [
15]. However, aberrant or excessive BAFF expression could amplify inflammation with deleterious effects. The fact that B cell BAFF and Ki67 expression was increased in lung LFs that were also increased in number in uninfected
Cftr −/− mice suggests a direct role for CFTR deficiency in increased BAFF expression, B cell activation and the promotion of LF development.
To further characterize the effects of CFTR deficiency on B cell phenotype and function, lung and splenic B cells from
Cftr −/− mice were analyzed by flow cytometry and stimulated in vitro. There were no differences in costimulatory molecule expression or the percentage of B cells. However, there was a statistically significant increase in MHC class II expression on
Cftr −/− lung B cells compared to wild type (Fig.
5a and b). MHC class II expression on B cells is critical for antigen presentation. In the presence of large amounts of antigen, B cells in tertiary LFs could present antigen to other lymphocytes, which may be further amplified by increased MHC class II expression [
30]. It will remain to be determined if the increase in MHC class II expression is an indirect consequence of an increase in the activation state of B cells in
Cftr −/− mice or if CFTR deficiency plays a direct role in regulating MHC class II expression. In vitro,
Cftr −/− splenic B cells produced more IL-6 when stimulated with LPS than wild type controls (Fig.
5d). IL-6 has been shown to promote lung LF development [
34]. Augmented IL-6 inflammatory responses to chronic airway infection may be an important mechanism that amplifies LF formation in CF and may work synergistically with BAFF that is increased in lung LF B cells from uninfected
Cftr −/− mice. Increased IL-6 mRNA was observed in bronchoalveolar lavage cells stimulated with LPS from mice deficient in myeloid lineage CFTR [
5], which suggests altered cytokine responses and IL-6 production may be present in other cell lineages in the absence of CFTR. These data suggest an important role for CFTR promoting a pro-inflammatory B cell phenotype, which in turn could promote tertiary LF formation.
Uninfected
Cftr −/− lung LF B cells also had significantly increased levels of CXCR4. CXCR4 is dynamically regulated during B cell maturation and plays an important role germinal center organization [
24,
25]. CXCR4 responds to its ligand CXCL12 and its expression helps distinguish centroblasts from centrocytes. BAFF has not been reported to alter CXCR4 expression [
35]. Whether increased CXCR4 expression in CFTR deficiency reflects an altered B cell maturation state and/or if it alters the ability of B cells to respond to germinal center chemokine gradients is unclear. These findings, in addition to increased lung B cell MHC class II expression, increased lung B cell BAFF and Ki67 expression, and increased inflammatory cytokine production upon activation raise concern for dysregulated immune responses to chronic infection and for the loss of immune tolerance to self-antigens that could lead to autoimmunity. Elevated autoantibody levels have been linked to lung disease severity in patients with CF [
36,
37]. In addition, several systemic autoimmune disorders are also associated with CF including CF arthropathy [
38‐
40] and cutaneous vasculitis [
41].
The mechanisms by which CFTR-deficiency alters immune function in B cells have yet to be delineated. B cell chloride conductance is altered in CFTR deficiency [
10,
11]. It has been postulated that altered lymphocyte chloride conductance could alter membrane potential and in turn calcium flux. T lymphocyte intracellular calcium flux has been shown to be increased in CFTR deficient T cells upon stimulation. Increased nuclear factor of activated T-cells (NFAT) nuclear translocation, which is modulated by calcium-associated signaling pathways, was observed in
Cftr −/− T cells upon activation resulting in increased inflammatory cytokine production [
8].
Interactions between CFTR and other molecules important in immune activation may also be important. Activated CFTR deficient macrophages have reduced ezrin protein levels and altered localization [
28]. Ezrin is a member of the ezrin-radixin-moesin family that bridges plasma membrane proteins to the actin cytoskeleton and regulates cellular processes that require membrane remodeling and modulate signaling events. Decreased ezrin levels were linked to reduced PI3K/AKT signaling upon TLR4 activation in CFTR deficient macrophages and promoted a pro-inflammatory phenotype [
28]. Ezrin has also be shown to maintain the topology of signaling molecules in the immunologic synapse and to down regulate Erk1/2 and NFAT signaling [
42] in T lymphocytes. Ezrin plays a critical role in regulating B cell receptor signaling and lipid raft aggregation. Loss of ezrin increases B cell activation and increased MHC class II expression [
42]. Reduced ezrin levels, as have been observed in CFTR deficient macrophages, may alter the function of the immunologic synapse and downstream signaling contributing to pro-inflammatory responses in CFTR deficient B cells as well. The role of these mechanisms in CFTR deficient B cells and their importance in the pathogenesis of CF remains to be defined.
In light of the chronic airway exposure to bacterial pathogens that occurs in cystic fibrosis, the expression of pathogen sensing molecules is likely an important contributor to immune activation. CFTR-dependent aberrant TLR4 trafficking has been observed in CFTR deficient macrophages, which promoted a hyperinflammatory response [
4,
28]. B cell TLR4 expression was examined in CF subjects and controls. In contrast to mice, humans B cells do not constitutively express TLR4. However, human B cells have been shown to upregulate TLR4 in response to inflammatory stimuli [
29]. As expected, based on literature, there was scant expression of TLR4 in lung B cells from control subjects. There was a significantly higher number of TLR4
+ B cells in CF subjects (Fig.
6). Increased B cell TLR4 expression may amplify cytokine production as has been shown in macrophages. TLR4 expression can also contribute to B cell maturation. Interestingly, BAFF and TLR4, both of which had significantly increased expression in lung B cells in CF subjects, have been previously shown to have a synergistic effect on B cell maturation [
27]. Additional studies will be needed to dissect the complex mechanisms that promote B cell dysregulation and lymphoid follicle formation in cystic fibrosis.
The limitations of this study include the fact that with few exceptions human tissue samples studied were from patients with severe lung disease. Lung biopsies are infrequently performed in patients with CF, so correlation of immunohistochemical findings with FEV1 and other clinical parameters over a spectrum of disease was not possible. Given that Cftr −/− mice do not develop the airway pathology observed in patients with cystic fibrosis, the effects of CFTR deficiency on B cell function and activation observed in murine analyses may be more pronounced or altered in human disease. We did not examine the ability of Cftr −/− B cells to promote humoral immunity and host defense as it was beyond the scope of this study. We also did not examine the diverse B cell subsets found in LFs and germinal centers. It is possible that CFTR deficiency may play a more prominent role in specific B cell maturation subsets. In addition, since archival lung tissue from patients with CF was used, correlation of immunohistochemical findings with B cell functional assays using peripheral blood B cells from patients with CF and controls could not be performed.
The data presented here point to an important role for B cells and tertiary LF development in CFTR-mediated immune activation and the pathobiology of lung disease in CF. Although we did not examine other organs typically affected by CF such as the gastrointestinal system, it is possible that altered B cell responses in CF could affect other sites of disease. Immune dysregulation may be an important contributor to the pathobiology of CF where chronic airway infection may augment the effects of CFTR-mediated immune dysfunction. There may be a clinical role for immunomodulation in altering CF outcomes. Finding biomarkers that identify immune dysregulation and identifying precise pharmacologic targets that can alter disease progression in CF merit further investigation.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.