The B Lymphocyte Differentiation Factor (BAFF) Is Expressed in the Airways of Children with CF and in Lungs of Mice Infected with Pseudomonas aeruginosa

Background Chronic lung infection with Pseudomonas aeruginosa remains a major cause of mortality and morbidity among individuals with CF. Expression of mediators promoting recruitment and differentiation of B cells, or supporting antibody production is poorly understood yet could be key to controlling infection. Methods BAFF was measured in BAL from children with CF, both with and without P. aeruginosa, and controls. Mice were intra-nasally infected with P. aeruginosa strain LESB65 for up to 7 days. Cellular infiltration and expression of B cell chemoattractants and B cell differentiation factor, BAFF were measured in lung tissue. Results BAFF expression was elevated in both P. aeruginosa negative and positive CF patients and in P. aeruginosa infected mice post infection. Expression of the B cell chemoattractants CXCL13, CCL19 and CCL21 increased progressively post infection. Conclusions In a mouse model, infection with P. aeruginosa was associated with elevated expression of BAFF and other B cell chemoattractants suggesting a role for airway B cell recruitment and differentiation in the local adaptive immune response to P. aeruginosa. The paediatric CF airway, irrespective of pseudomonal infection, was found to be associated with an elevated level of BAFF implying that BAFF expression is not specific to pseudomonas infection and may be a feature of the CF airway. Despite the observed presence of a potent B cell activator, chronic colonisation is common suggesting that this response is ineffective.


Introduction
Chronic lung disease is the major cause of morbidity and mortality among individuals with Cystic Fibrosis [CF]. Prominent among the causes of such lung disease is chronic infection with the opportunistic pathogen Pseudomonas aeruginosa [1]. Immune defence against respiratory tract pathogens is dependent upon a complex interplay of both innate and adaptive components that may prevent or clear infection. However CF is a leading example of an illness in which this complex interplay between innate and adaptive immunity is not protective as evidenced by long term P. aeruginosa infection.
Production of pathogen-specific antibody by B lymphocytes is normally an important mechanism of airway defence against P. aeruginosa infection [2,3]. Studies, in mice and humans, have characterised the specificity and magnitude of antibody responses to P. aeruginosa infection [3] and notably antibody production to vaccine candidates has been correlated with protection upon subsequent P. aeruginosa challenge [4] but antibody production does not correlate with protection in the CF airway.
To improve strategies for airway vaccine design and new therapeutics against P. aeruginosa, a better understanding of antibody induction in the airways is needed. However, the mediators that promote recruitment and differentiation of B cells to the lung or support airway antibody production during infection are poorly understood, both with respect to P. aeruginosa and airway pathogens in general [5]. We have previously reported production of the B cell differentiation factor, B cell activating factor of the TNF family [BAFF, also known as BLys or TNFS13B] by human lung epithelia in response to viral infection [6]. Studies using influenza infected mice show that homeostatic chemokines including CXCL13, CCL19 and CCL21 are essential for effective pulmonary defence, regulating B and T cell recruitment to the airway, stimulating dendritic cells and influencing formation of inducible bronchial associated lymphoid tissues [iBALT] [7]. In mice infected with P. aeruginosa, Eppert et al showed a possible role for CCL19 and CCL21 in regulating lymphocyte or dendritic cell recruitment but only examined one day post infection [8]. The BAFF response to ongoing airway P. aeruginosa infection has not been studied, but recombinant BAFF, used as an adjuvant, has been shown to enhance the airway response to heat-killed P. aeruginosa in mice [9].
We firstly sought to establish if expression of BAFF is a feature of the airway response to chronic bacterial infection, including P. aeruginosa, analysing lower airway samples from children with CF, with and without P. aeruginosa infection, compared to healthy control patients. To validate our findings further we used a 7-day murine respiratory inhalation model to establish stable infection of the airways and study the inflammatory response to P. aeruginosa, specifically with regards to the recruitment and potential stimulation of B cells. In our murine respiratory inhalation model, mice are challenged with the clinically relevant P. aeruginosa Liverpool epidemic strain [LES] isolate LESB65, which establishes a stable 7 day infection that is restricted to the airway [10]. In contrast to models using non-colonising P. aeruginosa strains that are rapidly cleared by components of the innate immune system or more aggressive strains that cause sepsis and host death, this model allows the lung response to in situ infection to develop over a longer time period, such as occurs in CF [10].
Our results demonstrate for the first time the kinetics of B cell chemoattractant and differentiation factor responses both following P. aeruginosa infection and in relation to lung B cell recruitment in a murine model. An elevated level of BAFF was found to be associated with the paediatric CF airway, irrespective of the presence or absence of pseudomonal infection, implying that the expression is not specific to pseudomonas infection and may be a feature of the CF airway. The importance of the B cell response in the CF lung remains unclear and its effectiveness is complicated by other factors predisposing to poor bacteria eradication such as impaired mucocilary clearance.

Patient samples
Bronchoalveolar lavage [BAL] was obtained bronchoscopically from CF patients who were being investigated electively for persistent respiratory symptoms. Lower respiratory lavage fluid was obtained non-bronchoscopically from the remainder of CF patients and non-CF, healthy control patients, when anaesthetised for routine elective surgery as described previously [11]. Microbiological analysis was conducted on BAL from all CF patients in the microbiology department of this specialist hospital [BAL from non-CF healthy control children was not sent for analysis]. Lavage fluid was aliquoted and stored at 270uC until used for analysis.

Ethics Statement
The paediatric ethics committee in Liverpool approved the study and all samples were collected following written informed consent from parents [REC:]06/Q1502/142].

Mice
Female BALB/cOlaHsd mice were purchased from Harlan Laboratories [Bicester, UK] and acclimatised for one week prior to use. Mice used for infection experiments were between 9 and 12 weeks of age and were age matched for each experiment. The U.K. Home Office approved the experimental protocols.

Infection of Mice
Animals were anaesthetised with a mixture of oxygen and isofluorane and infected intranasally with 1610 6 CFU P. aeruginosa LESB65 in 50 ml phosphate buffered saline [10].Mice did not develop visible signs of disease and were culled at pre-determined times post-infection by cervical dislocation. Lungs were removed and either homogenised for CFU counts and ELISA or prepared for flow cytometry by incubation for 30 min at 37uC with 10 mg/ ml collagenase D [Amersham], followed by homogenization and lysis of erythrocytes [12].

Bacterial Counts
Serial dilutions to 10 27 made using 20 ml samples of homogenised lung were analysed in sterile 96-well microtiter plates [GIBCO, Paisley, United Kingdom] containing 180 ml of sterile nanopure water per well. Pseudomonas selective agar plates were marked into six sectors, and three 20 ml aliquots of each dilution were plated into each sector; this procedure was repeated in duplicate. Once dry, plates were incubated at 37uC overnight. Colonies were counted in sectors containing a measurable number [approximately ,300].

ELISA assays
Lung homogenates were prepared in ice cold PBS containing a protease inhibitor cocktail before being clarified by centrifugation for five minutes at 4uC. Total protein concentrations were determined by using a DC protein assay kit [Biorad] according to the manufacturer's instructions and expression per milligram of lung homogenate calculated. Chemokines CXCL13, CCL19 and CCL21 were measured using DuoSet ELISA assays as described by the manufacturer [R&D systems] and both murine and human BAFF using quantikine ELISA kits [R&D systems].

Statistical Analysis
Data were analyzed using independent samples t tests or Mann-Whitney U tests depending on data distribution determined by the 1-sample Kolmogorov-Smirnov test. Data were analyzed using SPSS 20 software [SPSS Inc].

BAFF Expression in Cystic Fibrosis Patient Bronchoalveolar Lavage
To determine if BAFF expression is a characteristic of human lung bacterial infection, BAFF was measured in BAL from 20 school age children with CF [mean [SD] age, 8.0 [4.5] years]. Samples from six of these children were obtained bronchoscopically when they were being investigated for persistent respiratory symptoms. Samples from the remaining 14 CF patients and 7 healthy control children [6.8 [3.5] years] were obtained nonbronchoscopically when these children attended hospital for routine elective surgery. BAL from six of the CF patients cultured P. aeruginosa, with the remainder being culture positive for other organisms including Haemophilus influenzae, Stenotrophomonas maltophilia and Aspergillus species [n = 6] or BAL fluid culture negative [n = 8] [a full list of all organisms is available in Table S1]. BAFF expression was significantly elevated in both P. aeruginosa positive and negative patient groups [p,0.01] compared to the control group [ Figure 1].

Expression of mBAFF and Lymphoid Cell Chemoattractants CXCL13, CCL19 and CCL21
To determine if increased expression of BAFF is also a feature of the response to airway P. aeruginosa infection in this murine model system, mean [SD] protein levels were measured in lung homogenate samples prepared from mice at day 0 and at 1, 2, 3 and 7 days post infection [all n = 5]. BAFF was found to be significantly elevated at day 1 [892 [199] Figure 5].
Expression of the chemoattractants CXCL13, CCL19 and CCL21 were measured to assess how B lymphocytes could be recruited into the lung [ Figure 5].

Discussion
Although many studies have focused on the mechanism and importance of inflammatory cell recruitment into the lung following P. aeruginosa infection, few studies have examined the B cell response within the airway. Here we demonstrate, to our knowledge for the first time, that BAFF expression is increased in BAL from the lungs of patients with CF. Furthermore, using a respiratory inhalation model of stable long-term lung infection with P. aeruginosa LESB65 in mice, we show, to our knowledge for the first time, recruitment of both B and T lymphocytes, expression of the B cell chemo-attractant cytokines CXCL13, CCL19 and CCL21 and up-regulation of the B cell differentiation factor BAFF.
We have previously shown that the B cell differentiation factor BAFF, a major B cell growth factor that promotes B cell maturation and antibody production is expressed in the human airway during a range of viral infections including respiratory syncytial virus [6]. To determine if BAFF expression is a component of the human airway response to P. aeruginosa infection, we measured BAFF expression in BAL from children with CF both with and without P. aeruginosa infection. Both groups showed significant elevation in BAFF expression in comparison to the control group. The P. aeruginosa negative CF patient group, some of whom were culture positive for other bacterial infections, also showed elevated BAFF expression, indicating that this response is not P. aeruginosa-specific. The recent observation that BAFF is expressed during Mycobacterial infection is consistent with this suggestion [13]. Furthermore, our previous finding that BAFF expression occurs in response to a range of viral infections [6], supports this being a general mechanism in airway defence. The findings presented here cannot discount the possibility that elevated BAFF expression may be an innate characteristic of the CF airway associated with the underlying CFTR defect or of immune or inflammatory processes within the CF airway.
To further determine if expression of this B cell differentiation factor correlated with increased expression of B cell recruitment factors and increased B cell numbers in the lung, we used an animal model of chronic pseudomonas infection. Neutrophils were the major inflammatory cell recruited to the lung, with numbers peaking at day 1 post-challenge before reducing at days 2 and 3. The 24-48 hour peak of neutrophil influx coincides with a significant drop in bacterial load in the lung and the subsequent decline in lung neutrophil numbers is matched by a corresponding steady decrease in bacterial CFUs. This pattern of rapid neutrophil influx and gradual decline is in keeping with other respiratory pathogen infection models of the lungs e.g. streptococcus pneumonia, where early and rapid neutrophil influx is ineffectual in clearance and containment of bacterial loads [14]. In contrast, B cell and T cell [both CD4 and CD8] influx into the lung progressively increased over the study period of 7 days postinfection. To our knowledge no previous studies have documented B cell recruitment into the lung following P aeruginosa infection. In contrast, T cell responses have been studied with CD4 cell Th2 and Th17 type responses being described in human P. aeruginosa infection [15][16][17], while CD4 Th1 cells have been reported to be protective in mice [18].
In order to understand better how the observed recruitment of B and T cells into the lungs is mediated, we looked at expression of the homeostatic chemokines, CXCL13, CCL19 and CCL21, known to influence lymphocyte migration and to be expressed by epithelial, endothelial and dendritic cells [7,19]. Increased CXCL13 expression from day 1 post infection preceded increased CCL19 and CCL21. Elevation of CXCL13, primarily a B cell chemoattractant with highest expression on days 1 and 2 coincided with the increase in B cell numbers, suggesting that this chemoattractant may be important in the local B cell response to P. aeruginosa, possibly supporting the recruitment of both B1 and B2 B cell types [19].
In contrast, expression of the B cell, T cell and DC chemoattractants CCL19 and CCL21 increased progressively to 7 days post infection with CCL19 elevated at day 1 and rising before CCL21. A previous study, using the PAO1 strain, reported increased expression of CCL19 but not CCL21 on day 1 following P. aeruginosa infection but did not study later time points [8]. The data presented here show that both are upregulated post infection but with differential kinetics, CCL19 levels initially being increased before CCL21, although CCL21 is expressed at higher levels by day 7. Our data supports a role for these two chemokines in later recruitment of B and T lymphocytes and potentially DCs whilst CCL19 could also have a role in the early recruitment of B cells. CXCL13, CCL19 and CCL21 have previously been shown to be essential components of the local airway immune response to viral infection [7] and to be involved in iBALT formation, a recognised feature of long term pathogen exposure [20]. It is possible that recruitment of both B and T cells seen in this study represents the early stages of iBALT formation and the induction of a local adaptive immune response.
Our results demonstrate that increased BAFF expression is a characteristic of the airway immune response at similar time points to increased expression of B cell chemoattractant and elevated B cell numbers. This suggests that not only are B cells recruited to the airway but that the local environment is capable of supporting B cell survival, differentiation and antibody production. That airway BAFF can act to promote antibody secretion is supported by previous work showing BAFF, when used as an adjuvant in mucosal immunisation with heat killed Pseudomonas, resulted in increased antibody production [9]. The mechanism leading to expression of BAFF in the mouse airway in response to P. aeruginosa has not been defined here. In viral infection it is expressed by infected epithelial cells and dependent on interferon beta production [6,21]. Airway epithelial cells from CF patients have been shown to produce less interferon in response to viral infection [22,23]. If a similar mechanism occurs in response to bacterial challenge, reduced levels of BAFF and consequent lower local antibody production may be a further factor compromising airway defence in this patient group.
Collectively our results demonstrate differential expression of both B and T lymphocyte chemoattractants, recruitment of B cells and significant expression in both mouse and humans of the key differentiation factor BAFF, potentially capable of supporting B cell maturation in the lung.