Background
Chronic bronchitis and chronic obstructive pulmonary disease (COPD) are common adverse respiratory health effects among persons exposed to agriculture organic dust work environments [
1,
2]. Agriculture organic dusts are complex, comprised of particulates and enriched with a diversity and abundance of gram negative and gram positive microbial components [
3‐
5]. These exposures elicit innate immune responses through activation of Toll-like receptors (TLR; i.e. TLR2, TLR4, TLR9), scavenger receptors, and myeloid differentiation factor 88 (MyD88) signaling pathways [
6‐
9]. Human and animal studies have also described roles for activated lung macrophages, CD4
+ T cells, and a Th1/Th17 lung microenvironment as related to agriculture organic dust exposures [
10‐
14]. Earlier animal studies demonstrated that repetitive organic dust extract (ODE) exposures induce discrete peribronchiolar and lung parenchymal lymphoid aggregates comprised of both T and B cells [
15]. However, the precise role of B cells in mediating agriculture organic dust-induced lung inflammation has not been thoroughly investigated. This is increasingly relevant given the emerging role that B cells are understood to play in COPD and COPD-associated autoimmunity [
16], pathways that could potentially be targeted therapeutically.
Increased numbers of large and small airway associated B cells correlate with COPD severity [
17], but the role of these B cells is controversial. B cell responses can be protective against microbial infection through production of an antibody response to promote pathogen clearance, but B cells might also be harmful by directing an autoantibody response against lung tissue antigens [
18]. It has long been recognized that agricultural workers exposed to a variety of environmental agents mount antigen-driven, humoral responses. This has been most evident in hypersensitivity pneumonitis whereby the inciting antigen is due to microorganisms found in hay or grain [
19]. More recently, cigarette smoke exposure resulting in lung inflammation and oxidative stress has been shown to induce autoantibodies such as anti-citrullinated protein antibody (ACPA) that are implicated in rheumatoid arthritis (RA) as well as autoantibodies to heat shock protein and oxidized lipoproteins [
20,
21]. Moreover, malondialdehyde-acetaldehyde (MAA) protein adducts are a general product of oxidative stress that are expressed in a variety of inflammatory disease states including atherosclerosis [
22,
23], alcoholic liver disease [
24], rheumatoid arthritis [
25], and alcohol/smoking-associated lung inflammation [
26]. This is important because MAA adducts are highly immunogenic and lead to robust adaptive immune responses to both the MAA epitope and carrier proteins. It is not known whether agriculture organic dust-induced lung inflammation induces autoreactive antibodies to endogenous proteins including those characterized by the presence of MAA adducts or citrullinated residues. The relationship of agricultural dust exposures with B cell mediated lung inflammation and autoimmunity development may have added relevance as farming occupations as well as other pulmonary inhalant have been associated with autoimmune diseases including RA [
27‐
31].
In this study we hypothesize that B cells are important in mediating organic dust-induced lung disease and immunoglobulin responses, including the expression of autoantibodies. To test this hypothesis, organic dust-induced lung inflammatory consequences, lung B cell subset responses, and systemic IgM, IgA, IgG, and IgE response in wild-type (WT) and B-cell receptor (BCR) knock out (KO) mice were investigated. Additionally, we sought to determine whether autoantibodies to endogenous products associated with autoimmune disease pathogenesis including ACPA and anti-MAA antibodies were induced. Our studies show that repetitive agriculture organic dust-induced lung inflammation is dependent upon B cells, and that this dust exposure induces autoantibody responses.
Methods
Aqueous organic dust extract (ODE) was prepared as previously described [
15]. Briefly, settled dust was collected from horizontal surfaces (~1 m above floor level) of swine confinement feeding facilities (~400-600 animals) located in Colfax County, Nebraska (population density approximates 25 people per square mile). Dust (1 g) was incubated in sterile Hank’s Balanced Salt Solution (10 mL; Mediatech, Manassas, VA) at room temperature for 1 h and centrifuged for 30 min at 2850 ×
g. The supernatant was removed and then spun again for 30 min at 2850 g. The final supernate was filter-sterilized (0.22 μm) to remove microorganisms and coarse particles. Endotoxin concentrations in 100% ODE ranged from 1200 to 1400 EU/mL as determined using the limulus amebocyte lysate assay (Lonza, Walkersville, MD). Muramic acid levels (bacterial cell wall peptidoglycans) were previously determined by mass spectrometry to be approximately 70 ng/mg [
32]. Stock ODE was batched prepared, stored at −20 °C, and aliquots were diluted for each experiment to a final concentration (vol/vol) of 12.5% for animal studies in sterile phosphate buffered saline (PBS; pH 7.4; diluent). ODE 12.5% has been previously shown to elicit optimal experimental outcomes in mice and is well tolerated [
15].
Animals
All animal procedures were approved by the University of Nebraska Medical Center Institutional Animal Care and Use Committee and were in accordance with the NIH guidelines for the use of rodents. WT C57BL/6 mice and B cell receptor KO (Strain B6.129S2-Igh-6tm1Cgn/J) on C57BL/6 background were purchased from The Jackson Laboratory (Bar Harbor, ME). Male mice, between 6 and 10 weeks of age, were used for all studies because the occupational animal confinement facility industry is predominately male. All mice had ad libitum access to standard rodent chow and filtered water through the course of the studies.
Animal exposure model
An established intranasal inhalation repetitive exposure animal model was utilized whereby mice were lightly sedated under isoflurane and received treatment with either 50 μL of sterile saline (PBS) or 12.5% ODE daily for 3 weeks [
15]. Animals were euthanized 5 h following the final exposure for experimental endpoint quantification. No respiratory distress, signs of stress, or weight loss throughout the treatment period was observed.
Bronchoalveolar lavage fluid and lung tissue homogenate cell analysis
Immediately after the animals were euthanized, bronchoalveolar lavage fluid (BALF) was accumulated using 3 × 1 mL PBS. Total cell numbers from the combined recovered lavage were enumerated and differential cell counts determined from cytospin-prepared slides (cytopro cytocentrifuge, ELITech Group, Logan, UT) stained with DiffQuick (Siemens, Newark, DE). From cell-free supernate of the first lavage fraction, tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6) and murine neutrophil chemokines, CXCL1 and CXLC4, were quantitated by ELISA (R&D Systems, Minneapolis, MN) because they have been implicated in agriculture exposure-induced lung inflammation [
9,
10,
15]. IL-17 and total hyaluronan, B cell chemoattractants [
33‐
36], were also characterized in total lung homogenates. Briefly, lung homogenates were prepared by homogenizing ½ lung samples in 500 μl of sterile PBS as previously described [
12]. Levels were determined according to manufacturer’s instruction using an enzyme-linked immunosorbent assay kit for IL-17A (Abcam, Cambridge, MA) and hyaluronic acid (Echelon Biosciences, Inc., Salt Lake City, UT) with sensitivities of 6.25 pg/ml and 12.5 ng/ml, respectively.
Histopathology
Following lung lavage, whole lungs were excised and slowly inflated (15 cm H
2O pressure) with 10% formalin (Sigma, St. Louis, MO) for 24 h to preserve pulmonary architecture as previously described [
15]. Fixed lungs were processed, embedded in paraffin, and entire lung sections were cut (4-5 μm) and stained with hematoxylin and eosin (H&E). Each slide was entirely reviewed at scanning magnifications (2X, 4X, and 10X objectives; Nikon Eclipse Model E600 microscope, Nikon, Tokyo, Japan) and semi-quantitatively assessed for the degree and distribution of lung inflammation by a pathologist (W.W.W.), blinded to the treatment conditions, utilizing a previously published scoring system [
15]. This scoring system evaluates the spectrum of inflammatory changes for: 1) alveolar compartment inflammation, 2) bronchiolar compartment inflammation, and 3) intrapulmonary lymphoid aggregates and is compared to a standard set of photomicrographs. The bronchiolar compartment consists of inflammatory infiltrates in both the peri-bronchiolar and per-vascular space. Each parameter was independently assigned a value from 0 to 3, with a higher score indicating greater inflammatory changes in the lung.
Cell staining and flow cytometry
In separate studies following collection of lung lavage fluid cells, the chest cavity was opened and the right ventricle was infused with 10 mL of sterile PBS with heparin to remove blood from the pulmonary vasculature. Half of the lung was harvested and subjected to an automated dissociation procedure using a gentleMACS Dissociator instrument (Miltenyi Biotech, Auburn, CA) in a solution containing collagenase type I (324 U/mL; Fisher, Pittsburgh, PA), bovine DNAse (75 U/mL), porcine heparin (25 U/mL), and PBS with Ca2+ and Mg2+. Cell solution was passed through nylon mesh (40 μM; Fisher) to remove any large fragments and red blood cells were lysed using a 0.84% (w/v) ammonium chloride treatment (5 min at 4 °C). All reagents not specifically stated were obtained from Sigma. The cells were then resuspended in PBS, and lung cells were isolated by density gradient centrifugation over Ficoll-Paque PLUS (GE Healthcare, Uppsala, Sweden), enumerated, and reported as “total lung cells.” Viability of the final cell preparation was assessed by trypan blue exclusion and a LIVE/DEAD fixable Violet Dead Cell Stain Kit (Life Technologies, Carlsbad, CA). Ultimately, <1% of gated cells were not viable, with no difference noted between the saline and ODE-treated groups (data not shown).
BALF and lung cells from each animal were incubated with CD16/32 (Fc Block, BD Biosciences, San Jose, CA) to minimize non-specific antibody staining, and then stained with monoclonal antibodies (mAb) directed against rat anti-mouse CD45 (clone: 30-F11), CD11b (clone: M1/70), Ly6G (clone RB6-8C5), CD4 (clone RM 4-5), CD8 (clone: 53-6.7), and hamster anti-mouse CD3 (clone: 145-2C11) from BD Biosciences (San Jose, California), rat anti-mouse CD19 (clone: eBio1D3) and CD11c (clone: N418) from eBiosciences (Waltham, MA), and rat anti-mouse CD5 (clone: 53-7.3), CD273 (clone: TY25), CD73 (clone: TY/11.8) from Biolegend (San Diego, CA). Parallel cell preparations were treated with appropriate isotype control antibodies, and compensation was performed with antibody capture beads (BD Biosciences) stained separately with each individual mAb used in test samples.
The gating strategy for CD11c
+CD11b
lo alveolar macrophages, CD11c
+CD11
hi exudate macrophages, Ly6G
+ neutrophils, CD3
+CD4
+ and CD3
+CD8
+ T cells were previously reported [
12,
13]. B1, B2 B and memory B cell subpopulations were identified according to published reports by others: B1 B cells (CD19
+CD11b
+), B1a B cells (CD19
+CD11b
+CD5
+), B2 B cells (CD19
+CD11b
−) [
37], and the spectrum of murine memory B cell subsets identified by CD273 and CD73 expression of CD19
+ B cells [
38]. The percentage of all respective cell populations were determined from CD45
+ lung leukocytes after excluding debris and doublets. This percentage was multiplied by the respective total BALF or lung cell numbers to determine specific cell population numbers for each animal.
Preparation of modified proteins
For these experiments, aqueous human albumin (Alb) purchased from Talecris Biotherapeutics, Inc., Research Triangle Park, NC was incubated with 1 mM Acetaldehyde (AA) obtained from Aldrich Chemical Co. (Milwaukee, WI, USA) and 2 mM Malondialdehyde (MDA) obtained as the sodium salt (MDA–Na) by treatment of tetramethoxypropane (Aldrich Chemical Co.) with NaOH, according to the method of Iwata and Kikugawa [
39] to form the MAA adduct. Alternatively, a quantitative ELISA for MAA adducted protein was performed on mouse lung homogenates as previously described [
26].
Serum
Whole blood was collected from mice at the time of euthanasia from the axillary artery. Blood (400 μL) was placed in BD Microtainer Tubes (Becton, Dickinson and Company, Franklin Lakes, NJ), centrifuged for 2 min at 6000 × g and supernatant collected. Serum IgM, IgA, IgG and IgE were quantified according to manufacturer’s instruction using a Quantikine enzyme-linked immunosorbent assay kit (Affymetrix eBioscience, Santa Clara, CA).
Serum anti-MAA antibodies (IgG) were quantified as previously described [
25]. Aqueous human albumin was adducted with malondialdehyde and acetaldehyde (2:1). ELISA plates were coated with 2 μg/well of MAA-albumin or albumin. Additional wells were coated with known concentrations of murine IgG isotype standards from which relative antibody concentration were extrapolated. Plates were incubated overnight at 4 °C, washed and blocked with 3% bovine serum albumin, and incubated with each mouse serum at a 1:1000 dilution. Following incubation at 37 °C for 1 h, a secondary horseradish peroxidase goat anti-mouse antibody specific for IgG Fcγ fragment specific (Jackson ImmunoResearch, West Grove, PA) was added. Plates were developed using TMB substrate, absorbance determined at 450 nm using an Epoch Plate reader (BioTek, Winooski, VT) and analyzed using Gene 5 Software (BioTek). Data are presented as arbitrary units (AU) of the antibody detected in the assay as this reflects the amount of antibody present in a sample relative to a standard curve.
Serum anti-citrullinated peptide antigen (ACPA) was determined using a modification of the second-generation anti-CCP antibody ELISA (Diastat; Axis-Shield Diagnostics). As previously described [
40], mouse serum was incubated on the anti-CCP plate, washed, and then a HRP goat anti-mouse antibody specific for IgG Fcγ fragment specific (Jackson ImmunoResearch) detection antibody replaced the anti-human IgG antibody from the kit.
Immunohistochemistry
Formalin-fixed, paraffin-embedded lung tissue sections of 4- to5-μm-thickness lung tissue were deparaffinized and antigen unmasking was performed using antigen retrieval techniques as per Bethyl Laboratories protocol and buffers (Montgomery, TX). The rabbit anti-MAA antibody and rabbit IgG were first labeled with a Zenon™ Alexa Fluor™ 405 rabbit IgG labeling kit prior to incubation with the sample (Life Technologies Corporation, Eugene, Oregon). Slides were blocked with 2% goat serum and sections stained for the presence of MAA adducted proteins using an affinity purified MAA-specific rabbit polyclonal antibody, anti-peptidyl-citrulline, clone F95 antibody (EMD Millipore Corporation, Temecula, CA) and macrophages using a ALEXA FLUOR™ 594 Conjugated Rabbit Anti-CD68 Polyclonal Antibody (Bioss Antibodies, Woburn, MA). The anti-citrulline antibody was detected using a Cy™3-conjugated AffiniPure F(ab’)2 Fragment Goat Anti-Mouse IgM, μ Chain Specific (minimal cross-reaction to Human, Bovine, and Horse Serum Proteins) (Jackson ImmunoResearch). Isotype controls included a Zenon™ labeled rabbit IgG, mouse IgM + the Cy3 goat anti-mouse secondary, and ALEXA FLUOR® 594 Conjugated Rabbit IgG (Bioss Antibodies). Slides containing the primary antibodies were incubated overnight at 4 °C in a humidified chamber. The next day slides were washed in PBS and incubated with the secondary antibody for 1 h, washed, mounted, and imaged using a Zeiss 510 Meta Confocal Laser Scanning Microscope. All images were analyzed using ZEN 2012 software (Zeiss). Image quantification was done using Image J software (National Institutes of Health) and represented as mean (± standard error of mean [SEM] pixel density).
Statistical methods
Data are presented as the mean ± standard error of mean (SEM). To detect significant changes between groups, a one-way analysis of variance (ANOVA) was utilized and a post hoc test (Tukey/LSD) or nonparametric Mann-Whitney test was performed to account for multiple comparisons if the p value was <0.05. All statistical analysis were performed using GraphPad Prism software (La Jolla, CA) and statistical significance accepted at p < 0.05.
Discussion
In this study, we demonstrated that B cells are central to the development of lung lymphoid aggregates following organic dust exposures and that these exposures induce B2, B1, and memory B cell subpopulations with a systemic immunoglobulin response. Moreover, we are the first to describe that organic dust-induced lung inflammation induces an autoantibody response of ACPA and anti-MAA antibody positivity with detection of both citrullinated and MAA modified proteins in lung tissue, which was abrogated in BCR KO animals. Taken together, these data suggest that B cells comprising ODE-induced lymphoid aggregates could be playing an important role in perpetuating agriculture-related lung disease, and that part of the inflammatory response is associated with the production of endogenous proteins resulting in self-reactivity. It is also possible that these autoantibodies could potentially serve as biomarkers of disease and/or subsequent generation of systemic autoimmune responses in exposed workers.
Despite reductions in hypersensitivity pneumonitis and allergic disorders in farmers, studies continue to show that respiratory complaints remain highly prevalent and airflow obstruction/COPD is common, even amongst non-smokers [
1,
41,
42]. Of the various farming environments, livestock workers demonstrate the highest prevalence of respiratory symptoms, chronic bronchitis, and COPD [
1,
2,
42]. With limitations in current therapeutic approaches for this population, a better understanding of the pathogenesis of disease could lead to alternative approaches. With the emerging interest in B cells and COPD pathogenesis and autoimmunity [
16], we investigated the role of B cells and ODE exposure-induced lung disease. As in human COPD where there is an expansion of B cells, memory B cells, and B cell-rich lymphoid follicles [
16], we found that B cells were important in explaining repetitive ODE exposure-induced lung pathogenesis. In the absence of B cells, there were decreased TNF-α, IL-6, and CXCL2 concentrations in lavage fluid, a failure to develop lymphoid aggregates with reductions being most prominent in the bronchiolar as opposed to alveolar compartment, and decreased lung exudative (CD11c
+CD11b
+) macrophages following ODE exposure. However, there was no reduction in ODE-induced lung neutrophil or T lymphocyte influx. As prior studies demonstrated that macrophages play a role in regulating airway inflammation following ODE exposure [
13] and that lymphoid aggregates are comprised of T cell, B cell and macrophages [
15], it is possible that the failure to form lymphoid aggregates explains the reduced inflammatory consequences. Furthermore, B cells as well as macrophages are sources of TNF-α, IL-6, and CXCL2 [
13,
34]. This current study might also suggest that targeting B cells (e.g. anti-CD20 therapy) could reduce disease in affected agriculture workers. Moreover, our studies demonstrated an expansion of the conventional B2 B cells as well as increases in B1 and memory B cells.
B1- and B2- B cells secrete immunoglobulin, but murine B1 subsets (CD11b
+CD5
+ and CD11b
+CD5
−) are implicated in innate defense against mucosal pathogens and autoimmunity development [
34]. ODE exposure increased B1 B cells, albeit the relative numbers of these cells were low. We speculate that these cells might be important in the autoreactive response to MAA and citrullinated modified proteins demonstrated following lung inflammation induced by ODE. Despite the increasing evidence of B1 B cell importance in mice, an equivalent subset in humans has not been definitively characterized due to difference between murine and human CD5 expression on B cells [
34]. However, there are elevated fractions of human B1 B cells that express CD11b in patients with lupus [
43]. To the best of our knowledge, there are no studies investigating the role of B1 B cells and COPD. Memory B cells (CD20
+CD27
+) are increased in the serum of patients with COPD [
44] and plasma cells are found in COPD lung [
45]. Memory B cells are a heterogeneous population [
38], and our studies found expansion of these cells following ODE exposures. These current studies would support phenotyping B cell populations in agriculture workers to determine associations with disease.
There was an increase in total serum IgG levels in animals repetitively exposed to ODE, and to a much lesser degree, there was also an increase in IgE levels. The significance of the small degree of serum IgE increase is not known as there were no corresponding findings of an influx of eosinophils into the BALF or lung tissues following ODE exposure. We did find that repetitive ODE exposure increased specific IgG responses to MAA and citrullinated modified proteins. ACPAs serve as informative biomarkers of RA, and although they appear unable to induce arthritis alone, they can enhance the development of arthritis in mice when mild synovitis is present. To date, pathogenic consequences of anti-MAA antibodies have not been described. We suggest that these autoantibodies might serve as biomarkers of inflammatory disease in agriculture exposed persons, and future studies in humans are warranted. These antigens were detected in ODE-exposed lung tissue of WT mice. Historically, specific serum immunoglobulin responses to microorganisms in hay or grain attributable to agriculture work have been well described as in hypersensitivity pneumonitis [
19]. Recently, it was demonstrated that occupational exposure to livestock (i.e. swine, poultry, and cattle) was associated with a trend (
p = 0.1) toward increased levels of anti-ganglioside autoantibodies [
46]. These authors had investigated anti-ganglioside autoantibody association because there was a higher prevalence of self-reported symptoms of peripheral neuropathy in the Agricultural Health Study farmers who worked with animals [
47]. Interestingly, the farming occupation is associated with an increased prevalence of RA morbidity and mortality [
28,
29].
There is an emergence of autoantibodies that recognize post-translational protein modifications that form under conditions of oxidative stress and chronic inflammation [
48]. Serum ACPA is a prominent example whereby antibodies are directed against a wide array of citrullinated proteins and commonly associated with RA [
48] or B-cell chemotactic factors. ACPA can present years prior to the development of autoimmune disease; and moreover, airborne stimuli including tobacco smoking and air pollution (industrial PM2.5 and SO
2 emissions) have been linked to ACPA positivity [
21,
49]. Less extensively studied are self-protein modifications through malondialdehyde (MDA) or MAA adduction. A large number of self-proteins are modified by MDA under inflammatory conditions and acetaldehyde can further react with MDA to form immunogenic MAA protein adducts [
50]. It has been proposed that in the settings of chronic inflammation, citrullination and MDA/MAA-modifications could simultaneously arise [
51]. We focused on the formation of MAA protein modifications because it has been shown that an increase of anti-MAA reactivity was associated with ACPA positivity without any direct cross-reactivity between the MAA and citrulline distinct epitope binding in RA [
25]. Furthermore, unlike the high lung levels of MAA adducted protein observed in alcohol and cigarette smoke co-exposure [
26], the lower concentration of ODE-induced MAA adducted protein would suggest that MAA represents an outcome, but not a cause of lung inflammation. Mechanisms underpinning the reduction in citrullinated and malondialdehyde-modified proteins in the lung of basal and ODE treated BCR KO mice are not understood and while this might be partly explained by the overall reduction in inflammatory mediators, lack of B cells, and reduction in exudative macrophages, further studies are necessary. It has been shown that MAA and citrulline co-localize in inflamed synovial tissues of RA patients with CD27 B cells [
52]. Nonetheless, our study supports that repetitive ODE exposure-induced lung inflammation is associated with increased citrullinated and MAA modified proteins in the lung tissue with a corresponding serum ACPA and anti-MAA response.
There are limitations in this study. We do not know which lung inflammation-induced proteins underwent citrullination and/or MAA modification; it is reasonable to suspect that there are many potential candidates, one of which may be surfactant protein [
26]. It is also possible that other autoantibodies could be induced by ODE-mediated lung jury including proteins modified by carbamylation and lipid oxidation. We have not identified specific anti-CIT or anti-MAA B cell clones, but others have described that B1 and memory B cells can produce these autoreactive antibodies [
51]. B cell recruitment, maturation and activation can involve interaction with macrophages, T cells, resident B cells, and numerous mediators including, but not limited to IL-17A, IL-6, CXCL13, CXCL12, B cell activating factor (BAFF), a proliferating inducing ligand (APRIL) [
34], and extracellular matrix proteins such as hyaluronan [
33,
35]. Prior work has demonstrated key roles for IL-17A, Th1/Th17 lymphocytes, IL-6, and macrophages in ODE-mediated inflammation [
12,
13,
53], and in this study we demonstrated again that ODE induces IL-17, but also found increased total hyaluronan as potential mechanisms to explain B cell recruitment. However, future studies are warranted to further understand this and other processes in ODE-mediated B cell regulation. It is also not known whether inhalant ODE exposure is directly linked to arthritis manifestations, and this line of work could be investigated in the context of collagen-induced arthritis or other relevant animal models of RA. These future studies could provide insight into the lung-autoimmune disease burden associated with agriculture work.