IL-22 regulates inflammatory responses to agricultural dust-induced airway inflammation

Abstract

IL-22 is a unique cytokine that is upregulated in many chronic inflammatory diseases, including asthma, and modulates tissue responses during inflammation. However, the role of IL-22 in the resolution of inflammation and how this contributes to lung repair processes are largely unknown. Here, we tested the hypothesis that IL-22 signaling is critical in inflammation resolution after repetitive exposure to agricultural dust. Using an established mouse model of organic dust extract-induced lung inflammation, we found that IL-22 knockout mice have an enhanced response to agricultural dust as evidenced by an exacerbated increase in infiltrating immune cells and lung pathology as compared to wild-type controls. We further identified that, in response to dust, IL-22 is expressed in airway epithelium and in Ym1+ macrophages found within the parenchyma in response to dust. The increase in IL-22 expression was accompanied by increases in IL-22 receptor IL-22R1 within the lung epithelium. In addition, we found that alveolar macrophages in vivo as well as THP-1 cells in vitro express IL-22, and this expression is modulated by dust exposure. Furthermore, subcellular localization of IL-22 appears to be in the Golgi of resting THP1 human monocytes, and treatment with dust extracts is associated with IL-22 release into the cytosolic compartment from the Golgi reservoirs during dust extract exposure. Taken together, we have identified a significant role for macrophage-mediated IL-22 signaling that is activated in dust-induced lung inflammation in mice.

1. Introduction

Half of agricultural workers report lower respiratory symptoms [1, 2], and occupational exposure is thought to be involved in 30% of asthma (>25 million) and 15% of chronic obstructive pulmonary disease (COPD) cases (9 million, including chronic bronchitis) in the US [3–5]. Per the Centers for Disease Control, COPD (including chronic bronchitis and emphysema) affects 15 million Americans with >140,000 people dying of COPD annually. National costs of asthma were estimated at $81.9 billion (medical and missing days at work or school) between 2008–2013, and the projected cost of COPD, the third leading cause of death worldwide, was $49 billion in 2020 alone [6]. In the agricultural industry, a major occupational hazard is hog farming, since hog farmers are at increased risk for developing chronic bronchitis, asthma, and COPD, due to exposure to dust in enclosed confinement facilities [2, 7, 8]. Hog confinement facility dusts lead to inflammation and modulate the immune response; however, the mechanisms leading to susceptibility versus resilience to chronic lung disease in exposed workers are incompletely understood, and bridging this gap is important in the prevention of chronic pulmonary disease driven by such exposures.

An IL-10 family cytokine, IL-22 has key immunomodulatory roles in inflammation, injury and host defense [9]. Among its family members, it is a unique cytokine in its effects on non-hematopoietic cells and exhibiting a broad range of actions in regeneration and host protection [10], which is mainly mediated by the heterodimer IL-22 receptor (IL-22R1/IL-10R2). IL-22 is produced by both innate and adaptive immune cells, and production has been primarily identified in NK cells, innate lymphoid cells, and T cells [9]. Upon its release by these cells, IL-22 has been shown to regulate epithelial repair and maintain mucosal homeostasis following lung injury induced by influenza virus as well as cystic fibrosis pathogen Pseudomonas aeruginosa [11–13]. While IL-22 ameliorates tissue inflammatory responses, tipping the scale toward repair, paradoxically, IL-22 is upregulated in many chronic inflammatory diseases, including asthma [14]. An increase in IL-22 in bronchial mucosa of patients with COPD [15] and serum of patients with asthma has been reported [16]. IL-22 functions at the mucosal barriers by influencing changes in the expression of genes involved in epithelial barrier integrity, mucus layer modifications, tight junction maintenance and production of antimicrobial compounds. These biological functions involve several mechanisms, including activation of STAT3 signaling [17], modulation of TGF-β-mediated epithelial to mesenchymal transition in primary bronchial epithelial cells, and proliferation of human airway smooth muscle cells via the MAPK and NFκB pathways [18, 19].

Unlike its IL-10 family members, IL-22 has been shown to be both anti-inflammatory and pro-inflammatory depending on the local microenvironment (i.e., pro-inflammatory in the presence of IL-17 and anti-inflammatory in the presence of IL-23) [20]. A dual role for IL-22 has been reported in allergic airway inflammation in OVA-sensitized/challenged mice in which IL-22 is a requisite player in the initial allergic airway response but may negatively regulate allergic inflammation following sensitization [16]. To better understand the role of IL-22, a whole-body IL-22 knockout (KO) mouse model has been developed [21, 22]. IL-22 KO mice fail to clear bacterial infection and epithelial damage caused by Citrobacter rodentium and display reduced pSTAT3 activity and mucosal wound healing in the intestinal epithelium following DSS treatment [23]. However, the role of IL-22 during chronic inflammation following environmental stimuli have not been investigated before.

Given the previously observed associations of IL-22 in the lung and disease pathology of asthma, COPD and allergic airway disease, here we sought to identify the role of IL-22 in inflammation in response to agricultural dust. To do so, we employed IL-22 KO mice to elucidate the role of IL-22 after repetitive exposure to agricultural dust, investigating its impacts on the lung inflammatory response to dust exposure as well as impacts on macrophage function.

2. Materials and Methods

2.1. In vivo repetitive hog dust extract (DE) exposure model

Extracts of organic dusts from swine confinement facilities (DE) were prepared as described before, and an intranasal dust exposure model was used in all in vivo studies [24]. Settled dust was collected from surfaces at least one-meter above the ground in a large swine confinement animal facility in the Midwest that house approximately 500 animals. Dusts have been collected from different sites and in different seasons and are assayed to ensure comparability in inflammatory outcomes based on induction of inflammatory mediator release when applied to human bronchial epithelial cells in vitro and when given as a single intranasal challenge in mice in vivo. Only after assuring comparability in these responses are dusts used for our investigations. Aliquots were frozen in 50 mL conical tubes and kept at −20°C until an aqueous dust extract was prepared. We now include a step-by-step preparation of the dust extract as Supplementary Material, as previously described [25].

All mouse studies were conducted at the University of California Riverside (UCR) and approved by the UCR Institutional Animal Care and Use committee. Briefly, male and female C57BL/6J mice and IL-22 KO mice of 6–8 weeks age (Jackson Laboratories, Bar Harbor, ME, USA) were housed in pathogen-free conditions and given free access to standard mouse chow and water. Intranasal (IN) exposure to 12.5% DE was administered in 50 μl volume of sterile saline under light isoflurane anesthesia [26]. A group of animals were instilled with saline only as controls using the same 3-week regimen. Repetitive exposure included five times weekly IN administration of the DE for three weeks. At the end of each experiment, bronchoalveolar lavage fluid (BALF) was collected as three separate 1-mL saline washes. Washes were centrifuged and the first BALF aliquot supernatant fraction was reserved for cytokine analyses prior to combining the cell pellets of all three washes for use in total and differential cell counting. Differential cell counts were performed by creating cytospins from 100,000 cells and staining cells with a Diff Quik staining kit (Hema3 Stat Pack, Fisherbrand, Kalamazoo, MI, USA). Following the BAL washes, left lung lobes were tied off and reserved for use in RNA isolation, and the right lung lobes were inflated with formalin at 20 cm pressure prior to being paraffin-embedded (FFPE) and sectioned by the University of California Irvine Pathology Research Services Core Facility. Some FFPE sections were stained with hematoxylin and eosin by the UCI core for use in histopathology scoring.

2.2. Tissue Culture

THP-1 human monocyte cell lines (ATCC-TIB 202) were employed. THP-1 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with high glucose and glutamine (HyClone), supplemented with 10% FBS and Penicillin/ Streptomycin (10,000 units/mL, Thermo Fisher Scientific). Alveolar macrophages were obtained from the BALF of five C57BL/6 mice and IL-22 KO mice. Peritoneal macrophages were obtained by injecting 3 mL of saline solution into the peritoneal cavity and collecting cells as described before [27].

2.3. Immunofluorescent staining of cells and tissue sections

To determine IL-22 and IL-22 receptor expression, 5 μm FFPE lung tissue sections were deparaffinized and rehydrated with serial ethanol washes and PBS. Antigen retrieval was performed with Diva DeCloaker per the manufacturer’s protocol (Biocare), and tissues were incubated with antibodies against IL-22 (Abcam ab18498), rat IL-22R (Millipore MA524205), GSL (Griffonia Simplicifolia Lectin-1, B-1205-5) or Ym1 (Stem Cell Technologies, #60130), then incubated overnight at 4°C.

Cytospins prepared from THP-1 cells grown on collagen-coated coverslips were fixed in 10% formalin after 24-hour treatment with DE. Cells were then permeabilized with 0.2% Triton X-100 in PBS for 5 min at room temperature and rinsed with PBS with 0.1% Tween-20 in PBS (PBST). Cells were incubated overnight at 4°C with antibodies against goat-IL22 (Abcam, catalog no: ab18498), rat IL-22R (Millipore MA524205), and rabbit GM-130 (D6B1) (Cell Signaling, catalog no:12480S). The next day, cells were washed with PBST for 5 min 3 times and incubated with secondary antibodies; rabbit anti-goat IgG (Dylight 550, SA5-10079, 1:500 dilution), donkey anti-rat IgG FITC (catalog no: AP189F, 1:250 dilution) and Cy5 Streptavidin (SA-1500-1, 1:500 dilution) for 2 hours at room temperature. After 3 washes in PBST for 5 min each, cells were mounted using Prolong Gold antifade reagent with DAPI. A ‘no primary antibody’ control was also included (Suppl. Figure 1 and 2).

Cells and tissue sections were visualized with an Echo Revolve (San Diego, CA, USA) or Confocal microscopy (Zeiss 880 Airyscan, Ontario, USA) using ZEN software. Image analysis was performed using Fiji open-source software (Version 1.52p, imagej.nih.gov) as described previously [28]. Briefly, quantification of Ym1, IL-22 and IL-22R1 immunofluorescence signal was performed in 7–9 images obtained for each mouse. At least 25 individual cells with positive staining in each image were quantified as integrated density/area. Data points were graphed as the average of each image (n =7–9 images/mouse) as well as average of each group (i.e. n=3 mice/group).

2.4. Scoring of lung pathology

Lung sections were either stained with hematoxylin and eosin (H&E) or Masson’s trichrome for collagen deposition and images were obtained with 10X and 20X objectives on an Echo Revolve microscope (San Diego, CA, USA). The observer was blinded to the treatment group and any experimental identifiers on each slide. Lung inflammation was evaluated by quantification and scoring of lymphoid aggregates, alveolar inflammation, and epithelial hyperplasia in H&E-stained slides, whereas lung pro-fibrotic changes were evaluated in Masson’s trichrome-stained slides. Peri-bronchial regions were evaluated by scanning the entire lung section with 20X and 40X objectives, then enumerating the number of cells in the alveolar sacs in a total of 5 images at 40X objective, which resulted in 5 values for each mouse. An average of these values represented one mouse in statistical analysis. Lymphoid aggregates, which we defined as 20 lymphocytes aggregated very closely with each other (no pink spaces between the cells), were counted manually by scanning the whole lung section with 4X and 10X objectives. To quantify epithelial hyperplasia, epithelial cell nuclei around the airways were counted and divided by the total area of that airway as μm². A total of 10 values were obtained for each mouse by assessing 10 bronchioles. Percentiles were obtained in GraphPad Prism (version 7) using descriptive statistics and scores 0–4 was assigned based on each percentile for a given pathology parameter.

Lung pro-fibrotic changes were evaluated on Masson’s trichrome-stained lung sections using a 0-to-4-point scale, representing different percentiles. Deposition of collagen was evaluated by determining the area of blue staining in 10 images of each slide (at 10X and 20X objectives). Three male and three female lung sections were used for the scoring of fibrosis for each group. QuPath imaging software (Version 0.3.0) was used to determine the area of collagen deposition.

2.5. Cytokine measurements in BALF

Levels of IL-22, IL-6, CXCL-1 and TNF-α in BALF were detected using R&D Systems DuoSet enzyme linked immunoassay (ELISA) kits; mouse IL-22 (Duoset mouse IL-22 ELISA kit, DY582), IL-6 (Duoset mouse IL-6 ELISA kit, DY406), CXCL-1 (Duoset mouse CXCL-1/KC ELISA kit, DY453) and TNF-α (Duoset mouse TNF-α ELISA kit, DY410) (R&D systems, Minneapolis, MN) and were performed according to manufacturer’s instructions. ELISA plates were read using the VarioSkan Luxe Microplate reader (Thermo Scientific, VL0000D0) at the wavelengths recommended in the ELISA kit protocols.

2.6. qRT-PCR and NanoString gene expression assay

THP-1 cells were cultured in a 12-well plate overnight and exposed to 1% dust for 24 hours. At the end of each experiment cells were harvested and RNA was extracted using an RNA mini-isolation kit (Ambion, PureLink RNA Mini kit, Thermo Fisher Scientific, Kalamazoo, MI, USA). RNA samples were converted to cDNA using a cDNA conversion kit (High-Capacity cDNA Reverse Transcription Kit, Fisher Scientific, catalog no: 43-688-14), and qRT-PCR was performed using the following primers. Human IL-22 forward: 5’-GCAGGCTTGACAAGTCCAACT-3’ and reverse 5’-GCCTCCTTAGCCAGCATGAA-3’; hIL-22R forward: 5′-CTCCACAGCGGCATAGCCT-3′and reverse: 5′-ACATGCAGCTTCCAGCTGG-3;hIL10R2 forward: 5′-GGCTGAATTTGCAGATGAGCA-3′and reverse 5′-GAAGACCGAGGCCATGAGG-3′; IL-22BP forward: 5′-AGGGTACAATTTCAGTCCCGA-3′and reverse: 5′-CGGCGTCATGCTCCATTCTGA-3′[29].

For gene expression assays, we employed NanoString gene expression technology. At the end of each repetitive exposure experiment, left lungs from each mouse were put into 1 mL RNA Later, incubated at 4°C overnight then stored at −80°C until analysis. These samples were randomly selected for gene expression analysis to obtain at least three mice per treatment group per sex (total of 6 lungs per treatment group). Lung samples were thawed on ice, rinsed with PBS to remove RNA Later, homogenized in 1 ml Trizol and RNA was extracted as per manufacturer’s instructions using a PureLink RNA Mini Kit (Invitrogen, Carlsbad, California, USA). RNA integrity and sample quality were determined using a NanoDrop ND-100 (NanoDrop Technologies, Inc, Wilmington, DE, USA) and an Agilent Bioanalyzer 2100 (UC Riverside Institute for Integrative Genome Biology Core Facility, Agilent Technologies, Santa Clara, CA, USA). To evaluate changes in gene expression among the treatment groups, we employed nCounter gene expression assays, namely NanoString technology, an amplification-free technology that directly quantifies RNA transcript levels. Specifically, we used NanoString mouse immunology panels (NanoString Technologies, Seattle, WA, USA) that include 561 inflammation and immunology-related genes. Briefly, 50 ng total RNA was hybridized with the panel codeset probes for 18 hours at 65°C following NanoString instructions. Thirty microliters of the diluted hybridization complex were then loaded onto nCounter Sprint Cartridges and transcripts were quantified. Gene expression data analysis was performed using nSolver 4.0 Software and NanoString Advanced Analysis. All samples passed the nSolver QC test. Then, the raw data were normalized using ten housekeeping genes (Rpl19, Ppia, Oaz1, Sdha, Alas1, Eef1g, Gusb, Gapdh, Hprt, Tbp). Heat maps were generated from the normalized data in nSolver using agglomerative clustering function (nSolver 4.0 User Manual). While nSolver uses geometrical means, Advanced Analysis uses a GNORM algorithm for normalization (nSolver 4.0 User Manual). Advanced analysis was performed on the raw data and differential expression analysis was performed to identify the most upregulated or downregulated genes among the treatments. Advanced analysis provides gene set analysis which consists of predefined genes for each signaling pathway, as well as a pathway analysis that help interpret the changes found in gene set analyses. Before performing advanced analysis, we examined the correlation of each housekeeping gene with each other and defined a cutoff of 11 as the sum of correlations. With these criteria, we determined that among the housekeeping genes Alas1, Eef1g, G6pdx, Gapdh, Gusb, Hprt, Oaz1, Ppia, Rpl19, Sdha and Tbp show a sum of correlations equal to or greater than 11, and thus were used in the advanced analysis for housekeeping gene normalization (NanoString, MAN-C0011-04 Gene Expression Data Analysis Guidelines). We also defined a transcript ‘count threshold’, which we defined as two fold of the highest background-to-noise ratio (average of negative controls/ sample + 2*standard deviation of negative controls/sample).To further dissect differences between the wild-type (WT) and IL-22 KO animals subjected to repetitive DE exposure, we also took advantage of the STRING database (https://string-db.org/).

2.7. Statistical Analyses

To determine changes between treatment groups, we used two-way and three-way ANOVA followed by Tukey post-hoc multiple comparison tests. In vitro experiments were analyzed using two-way ANOVA GraphPad Prism software (version 9) was used to generate the graphs and to perform three-way or two-way ANOVA analyses to determine statistical significance of results among all the groups, and a P value less than 0.05 was considered statistically significant.

3. RESULTS

3.1. Effects of repetitive HDE exposure on IL-22/ IL-22R1 expression

We have previously shown that exposure to DE leads to airway inflammation characterized by infiltrating neutrophils and macrophages in a mouse model of repetitive dust exposure [24] that models hog dust exposure in people [26]. Since IL-22 is reported to be involved in both epithelial repair and inflammation in airway disease, we assessed whether IL-22 expression is altered during DE-induced lung inflammation. To evaluate the role of IL-22 in hog dust-driven inflammation, we used two different approaches. First we assessed the protein expression levels of IL-22 in mouse lungs in DE-exposed and saline controls, and then measured the IL-22 levels in the BALF. We found that the immunofluorescence signal of IL-22 increased in mouse lungs exposed to DE as compared to saline controls (Figure 1A). Of note the bright red autofluorescence appears to come from red blood cells as seen in the IL-22 KO images. We used IL-22 KO mouse lungs as negative controls which showed minimal background staining. The distribution of IL-22 expression was evident both in the lung parenchyma and around the airways. Similarly, when we measured the BALF IL-22 levels we found that mice exposed to DE had at least 2-fold more IL-22 in BALF as compared to their saline counterparts (Figure 1B, P = 0.017, F = 34.2). We also considered sex differences in response to DE to increase the translation of our results to individuals exposed to hog farm dusts. Sex differences in pulmonary inflammation have been observed among agricultural workers before [30]. When separated by sex, the difference in IL-22 levels between the control and DE-exposed mice reached statistical significance in males (P < 0.0001, F = 88.7) and not in females (P = 0.17, F = 42.35), indicating the overall difference stems from male sex.

Figure 1. Effects of repetitive dust exposure on IL-22 release in bronchoalveolar lavage fluid (BALF) and tissue IL-22 expression in mouse lungs.

Wild-type C67BL/6 mice were subjected to repetitive dust exposure for 3 weeks (15 total intranasal instillations). At the end of the exposure, lungs were harvested and inflated with 10% buffered formalin, and fixed tissues were paraffin embedded then stained for IL-22 as indicated in the Methods. (A) Immunofluorescence staining of IL-22 in wild-type mice exposed to either saline or 12.5% dust extract (DE). (B) IL-22 levels in BALF collected at the end of the 3-week exposure, and IL-22 concentration was determined using a mouse IL-22 ELISA kit. Error bars are standard error of mean (* P ≤ 0.05). WT saline group includes 18 saline (12 males, 6 females) and 16 DE-exposed mice (10 males, 6 females).

IL-22 binds to a heterodimeric receptor composed of IL-10R2 and IL-22R1 subunits. This receptor has been shown to be on non-hematopoietic cells, and it is believed not to be present on immune cells [31]. Here, we evaluated the expression of the IL22 receptor IL22R1 in the lungs and found that IL22R1 expression increased with repetitive DE exposure (Figure 2A). Consistent with previous reports [31], IL-22R1 expression was nearly exclusively in the airway epithelium of large, small and terminal airways. Quantification of the immunofluorescence signal for IL-22R1 was performed using Image J and data were presented as fluorescence intensity/area of individual cells as described in Methods. We found that repetitive DE exposure significantly increased the expression of the IL22R1 receptor in the airway epithelium (Figure 2B, P < 0.0001, F = 195.4 for average of images on the left, and neared significance P = 0.08, F = 4483 for average of each group on the right).

Figure 2. Expression of the IL22 receptor in the airway epithelium of mice exposed to either saline or 12.5% DE.

Wild-type C67BL/6 mice were subjected to repetitive dust exposure for 3 weeks (15 total intranasal instillations). At the end of the exposure, lungs were harvested and inflated with 10% buffered formalin, and fixed tissues were paraffin embedded then stained for IL-22R1 as indicated in the Methods. (A) Immunofluorescence staining of IL-22R1 in wild-type mice exposed to either saline or 12.5% dust extract (DE). (B) Quantification of the IF data in (A), data on left panel includes 7–9 image average values obtained from 3 mice/group. Data on the right panel averages all the images and the data represent the average of 3 mice/group. Error bars are standard error of mean (**** P ≤ 0.0001).

3.2. Effects of genetic deletion of IL-22 in DE-induced airway inflammation

Because we found that DE stimulates IL-22 production in the lung, and IL-22 has been reported to have a dual role that is anti- or pro-inflammatory depending on the disease setting [20], we next assessed whether increased expression and BALF levels of IL-22 in DE-exposed mice contributes to tissue repair or to worsening inflammation. To address this question, we employed IL-22 KO mice which have been used to define the immunomodulatory roles of IL-22 in the lung during allergen- or infection-induced inflammation [32, 33]. Wild-type and IL-22 KO mice were exposed to DE or saline for 3 weeks (a total of 15 instillations, 5 days/week), and then BALF was collected at the end of each repetitive exposure study. As expected, we found significantly increased total cell counts (P < 0.01, Figure 3A) and neutrophils (P < 0.05, Figure 3B) in wild-type mice exposed to DE, in both DE-exposed males and females as compared to their saline controls. The 3-way ANOVA analysis of the total cell counts showed significant main effects of both genotype (P = 0.0001, F = 7.37) and exposure (P < 0.0001, F = 61.08) and a genotype × exposure interaction (P = 0.0026, F = 4.26). Similar results were obtained for PMNs, with significant main effects of both genotype (P = 0.0006, F = 5.45) and exposure (P < 0.0001, F = 63.75) and a genotype × exposure interaction (P = 0.0024, F = 4.15). Macrophages displayed a similar pattern as well, with significant main effects of both genotype (P = 0.0051, F = 8.23) and exposure (P < 0.0001, F = 34.17) and a genotype × exposure interaction (P = 0.07, F = 3.13, Figure 3C). Multiple comparisons showed a similar pattern in total BALF infiltrating leukocytes for the IL-22 KO mice subjected to repetitive DE exposure as compared their saline counterparts (P < 0.001); however, in this case the increase reached 3–4-fold in the IL-22 KO mice as compared to a 2–3-fold increase in WT animals (Figure 3A). This difference between the WT and IL-22 KO animals exposed to DE was statistically significant only in females (WT + DE vs. KO + DE, P < 0.01). We also examined lymphocytes and eosinophils in the BALF among all the groups, but no statistically significant differences were observed (Figure 3D and 3E). Pictures of cell differential staining are shown in Supplementary Information. Supporting these results and previous reports, we also found an increase in BALF IL-6 levels with DE exposure regardless of genotype (P = 0.002, F = 15.06, Figure 3F), where it reached significance between the IL-22 KO saline controls vs. IL-22 KO mice receiving the repetitive dust exposure (P = 0.03); the wild-type saline controls vs. IL-22 KO mice receiving the repetitive dust exposure trended towards significant also (P = 0.06). We also investigated CXCL-1 and TNF-α, where we saw a significant main effect of dust on CXCL-1 levels (P = 0.05) and no significant changes in TNF-α (Suppl. Figure 3). Other cytokines were determined to have low levels in BALF (IL-17A and IL-10) as well as low transcript levels in the lung (IL-17A, IL-17B, IL-17E, IL-23a, IL-23R, IL-10, IL-10ra, IL-10rb) (Suppl. Figure 4A and 4B).

Figure 3. Changes in infiltrating immune cells to the lung after repetitive dust exposure in wild-type versus IL-22 KO mice.

Mice were exposed to either saline or 12.5% DE for 3 weeks (15 total intranasal instillations). Following 5 hours after the last exposure, BALF was collected and cytospins were prepared. Infiltrating immune cells were enumerated after diff-quik staining of the cytospin slides. (A) Total infiltrating cells, (B) Neutrophils, (C) Macrophages, (D) Lymphocytes, and (E) Eosinophils. Main effects of dust exposure, genotype and interaction were determined by 3-way ANOVA and indicated on each figure. Error bars are standard error of mean (**** P≤ 0.0001). (F) Secretion of IL-6 was measured in BALF as a biomarker of inflammation. WT saline (6 females, 8 males), wt DE (5 females, 9 males), KO saline (6 females, 8 males) and KO DE (7 females, 9 males).

3.3. IL-22 is expressed in YM1 positive tissue macrophages following repetitive DE-exposure.

In assessing IL-22 expression, we identified immunofluorescence staining of IL-22 in epithelial cells as well as cells found in the lung parenchyma. We have previously identified that the number of macrophages increases at the end of a 3-week repetitive DE exposure [24, 34]. Together with the previously reported retention of macrophages during the recovery phase, we hypothesized that the observed IL-22-expressing cells in the lung parenchyma were mostly macrophages. To test this hypothesis, we stained lung tissue sections for a general macrophage marker GSL (Griffonia simplicifolia lectin II) and for the alternatively activated macrophage polarization marker YM1. We found that the IL-22-expressing cells in the lung tissues exhibited dual-positivity for both YM1 (Figure 4A) and GSL (Figure 4B). Airway epithelium also showed some IL-22 staining in DE-exposed lungs. Overall, IL-22 expression in the lung was reflected as a trend towards an increase in the lungs following 3-week exposure to DE (P=0.09; Figure 4C).

Figure 4. IL-22 expression in the lung after repetitive dust exposure.

Immunostaining of paraffin embedded lung sections with (A) IL-22 and YM1 and (B) YM1 and GSL (macrophage marker). (C) Quantification of tissue IL-22 expression. 7–9 images were taken from saline controls and 12.5% DE exposed mouse lungs. For each image, 25 cells were quantified as integrated intensity/area of each cell and these 25 values were averages to obtain one value/ image resulting in 7–9 values for each mouse. Error bars are standard error of mean, n= 3 wt saline and 5 WT DE-exposed mice.

3.4. Effect of DE on IL-22 and IL-22R1 expression in mouse alveolar macrophages and THP-1 human monocytes

The observation that IL-22 is expressed in lung tissue macrophages prompted us to further explore the expression of IL-22 and its receptor IL22R1 in alveolar macrophages, particularly given their role in inflammation resolution and tissue homeostasis [35]. We collected alveolar macrophages from five wild-type mice and treated the cells with either PBS or 1% DE for 24 hours ex vivo. This amount of DE is well-tolerated in cells as it was shown that 1% DE has no effect on cell viability or proliferation after 48 hours of treatment [36]. At the end of this experiment, cells were stained for IL-22, IL-22R1 and YM1. We found both IL-22 and IL-22R1 to be expressed in these cells regardless of the DE treatment (Figure 5A and 5B). In addition, the alveolar macrophages were all YM1+. We tested the same hypothesis in THP-1 human monocytes to assess the translatability of our findings among murine and human cells. Different from alveolar macrophages, these cells showed statistically significant increases in both IL-22 and IL-22R1 expression following DE treatment (Figure 6A). Quantification of the IF data also supported this observation (Figure 6B and 6C, IL-22 P = 0.0002, F = 1.006, and IL-22R1 P = 0.015, F = 1.068, control vs. DE, average of each image). The expression of IL-22 increased by 1% DE treatment showed a distinct pattern in IF staining. While PBS treated-control cells exhibited a punctate staining for IL-22, cells receiving 1% DE showed cytosolic staining of IL-22. We hypothesized that this distinct pattern might be related to a mechanism where these cells package IL-22 in the Golgi in steady state and only release in the presence of an inflammatory insult like DE. Thus, we used a cis-Golgi protein marker, GM130 and found that there was an overlap in the IL-22/Golgi staining. We next quantified IL-22 expression in the Golgi and cytosol and evaluated Golgi-to-cytosolic ratio between the control and DE-treated cells. Consistent with our observation, this ratio decreased in DE-exposed cells (Figure 6D).

Figure 5. Expression of IL-22 and its receptor in alveolar macrophages.

Alveolar macrophages were obtained from five C57BL/6 wild-type mice and plated onto glass coverslips. Cells were treated either with saline or 1% DE for 24 hours. (A) At the end of exposures, coverslips were fixed in 10% formalin and stained for IL-22, IL-22R1 and YM1. (B) Quantification of data in A. Data are from three independent experiments. A total of 100–137 cells were quantified from all three experiments, and results are shown as integrated intensity/ area. Error bars are standard error of mean.

Figure 6. Changes in IL-22 and IL-22 receptor expression in THP1 human monocytes after dust exposure.

THP1 cells were subjected to either saline or 1% DE-exposure for 24 hours. At the end of the exposures cytospin slides were fixed in 10% formalin, and immunofluorescence staining was performed for IL-22, cis-Golgi marker GM130 and IL22R1. (A) Representative IF images from 3 independent experiments are shown. Quantification of IF data was performed as integrated intensity/ area of each cell, 20 cells/image from 5–7 images per experiment (B) IL-22 expression and (C) IL22R1 expression. (D) Change in IL-22 expression in the Golgi and cytosolic compartments after dust exposure. A smaller Golgi-to-cytosolic ratio indicates that IL-22 is mostly in the cytosolic compartment. (E) Time course of IL-22 gene expression in THP-1 cells 2, 4 and 6 hours following 1% DE exposure. Error bars are standard error of mean (* P ≤ 0.05, *** P ≤ 0.001; **** P ≤ 0.0001).

To further explore the role of macrophage-derived IL-22, we performed in vitro experiments to assess the control of IL-22 expression and its transcriptional regulation. Human monocytes (THP-1) were exposed to 1% DE and temporal gene expression determined by qPCR using the primers indicated in the Methods section. At 2 hours, we observed a significant upregulation of IL-22 gene expression that returned to baseline levels by 4 hours of treatment (Figure 6E). We also assessed the expression of IL-22, IL-22 Binding Protein (IL-22 BP), IL-22R1, and IL-10R2 at 24 hours post-treatment. In these studies, THP-1 cells expressed both IL-22 and IL-22R1 transcripts in response to DE treatment and IL-22 BP but none reached statistical significance at this timepoint (Suppl. Figure 5). Peritoneal macrophages were collected for stimulation experiments with DE; however, we were unable to detect IL-22 by ELISA (data not shown).

3.5. Changes in inflammatory and immunological gene expression following repetitive DE exposure in WT and IL-22 KO mice.

To assess the lung mRNA transcript levels of genes following DE challenge in mice, we employed a NanoString Mouse Immunology gene expression panel. Gene expression was evaluated in a total of 24 mouse lungs with six mice (3 female and 3 male) per group (wt saline, wt DE, KO saline, KO DE). The data generated in nSolver were analyzed in different ways to dissect the differences of genotype, DE and sex, separately. In these analyses, we compared changes in lung gene expression among the groups by taking WT saline as the reference control as well as by taking WT DE as reference control to directly compare WT + DE vs. IL-22 KO + DE. After housekeeping gene and reference sample normalization, we performed principal component analysis (PCA) of all the samples. Only those that showed a clear clustering of groups were evaluated further. We first evaluated male mice lung gene expression. Assessment of differentially expressed genes (DEG) showed a total of 64 genes upregulated between the WT saline controls and WT + DE exposed animals, whereas 85 genes were upregulated between the WT saline controls and IL-22 KO + DE exposed mice. The number of downregulated genes were relatively low as compared to upregulated genes (Table 1). A volcano plot of the most DEG is shown on Figure 7A presented separately by sex. These genes were similar with our previous results in WT mice after a single dust exposure [37]. Gene set analysis revealed that pathways related to inflammatory (PC1 = 0.82) and immune response (PC1 = 0.75), activation of MAPK (PC1 = 0.83), chemokine/ cytokine activity (PC1 = 0.88), positive regulation of T cell activation (PC1 = 0.84), JAK/STAT signaling (PC1 = 0.81), collagen (PC1 = 0.86), extracellular matrix (PC1 = 0.75), integrin complex (PC1 = 0.76), calcium-mediated signaling/cell-signaling (PC1 = 0.88), cell proliferation (PC1 = 0.8) and anti-apoptosis (PC1 = 0.79) were altered following repetitive DE exposure (Suppl. Figure 6). In females, upregulated genes were similar to males with the exception that females had a smaller number of upregulated genes (Figure 7A, right panel) with a total of 35 genes upregulated between the WT saline controls and WT + DE exposed animals, and 40 genes upregulated between the WT saline controls and IL-22 KO + DE exposed mice.

Table 1.

Differentially expressed and statistically significant genes were summarized in males and females separately.

Exposure groups	Number of differentially expressed genes as compared to wild-type saline controls (unadjusted p-values)
Males	Upregulated genes	Downregulated genes	Total number of genes differentially expressed
WT saline vs. WT + DE	163	19	182
WT saline vs. KO saline	10	3	13
WT saline vs. KO + DE	175	23	198
WT + DE vs. KO + DE	4	5	9
Females
WT saline vs. WT + DE	124	24	148
WT saline vs. KO saline	53	7	60
WT saline vs. KO + DE	123	35	158
WT + DE vs. KO + DE	3	11	14

Figure 7. Sex-dependent changes in inflammatory and immunological gene expression following repetitive exposure to DE in wild-type and IL-22 KO mice.

RNA was isolated from mouse lungs, and a nCounter Immunology NanoString gene expression panel was run. Data were analyzed using ‘wt saline controls’ as reference in nSolver software and advanced analysis was performed. (A) Volcano plot of the overall changes in male (3 mice/ group) and female mice (3 mice/ group) following repetitive DE exposure. (B) Overall clustering of genes with PCA analysis are shown in each experimental group (24 samples total, 6 mice / group (3 female and 3 male) in the NanoString Immunology panel. (C) A representation of changes in gene expression following repetitive dust exposure, the most differentially regulated genes are shown in a volcano plot for the IL-22 KO female lungs (WT saline males were used as reference for NanoString analysis). n=6 mice (3 females, 3 males) / group. (D) Volcano plot of changes in gene expression between the IL-22 KO males vs. IL-22 KO females, n=6 mice (3 females, 3 males) / group. (E) The top 20 most significantly regulated genes between the IL-22 KO males vs. IL-22 KO females were entered into STRING database to identify potential protein-protein interactions. These 20 DEG had a statistically significant PPI enrichment p-value < 1.0E-16 and were significantly associated with B cell homeostasis, T-cell differentiation, wound healing involved in inflammatory response and regulation of Th17 cell differentiation.

Next, we combined all 24 samples and reanalyzed to get a more complete picture of the changes in gene expression. We found that male and females clustered separately regardless of their genotype with a PC1 of 0.43 (Figure 7B). The most differentially regulated genes are shown in the volcano plot (Figure 7C), which were in parallel to those results when male and females were analyzed separately as expected. To follow up on the differences observed between the two sexes, we investigated the differences in the top 20 DEG specifically in the KO DE male vs. KO DE females. STRING database revealed a statistically significant interaction among these 20 DEG (PPI enrichment p-value < 1.0E-16), which mainly correlated with B cell homeostasis (false discovery rate 0.0002), T-cell differentiation (false discovery rate 0.0002), wound healing involved in inflammatory response (false discovery rate 0.0004), and regulation of Th17 cell differentiation (false discovery rate 0.0011) (Figure 7D and 7E).

Given this finding and the role of IL-22 in tissue repair [38], we next investigated differences related to tissue repair in WT and KO DE-exposed animals. In these analyses, we focused on male mice because of the higher number of upregulated genes in males as compared to female counterparts. NanoString data analysis uses two types of normalization, one with a housekeeping gene and another with a reference control. To meet these criteria, we used WT + DE animals as the reference for advanced data analysis to allow for a direct comparison between the WT + DE and KO + DE animals for gene expression. In this analysis, unadjusted p values identified that 12 genes—Itga2b, Itg2b, H2-eb1, Psmb7, Batf3, Tyk2, C6, Ltb4r1, Zap70, Klrc1, CD209g and Fn1 were different between the two groups in male sex. When we entered these genes in the STRING database, we found a significant interaction among these genes with a PPI enrichment P value of 0.0027 (Figure 8A). Among this set of proteins, gene ontology identified a significant difference for ‘integrin complex’ with a false discovery rate of 0.007. To further evaluate these 12 genes, we performed a two-way ANOVA of the log2 counts among the groups. We identified that Itga2b (P=0.003), Itg2b (P=0.013), H2-Eb1(P=0.0013), Psmb7(P=0.043), Batf3(P=0.016) and Tyk2 (P=0.0078) had a significant main effect of genotype. However, among these genes, only Itga2B gene expression came close to statistical significance between the WT + DE and IL-22 KO + DE animals (P = 0.06). In females, we identified only four genes to be different between WT + DE and IL-22 KO + DE animals, which included Marco, IL6st, B2m and H2-K1, indicating a greater compensation of these pathways by upregulation of genes related to extracellular matrix remodeling and tissue repair in the absence of IL-22 in male sex.

Figure 8. Effect of genetic deletion of IL-22 on gene expression of those genes involved in cellular signaling and inflammatory response.

Twelve genes were significantly different in expression between the WT vs. IL-22 KO male mice exposed to repetitive DE. (A) When entered in the STRING database, we found significant interactions between these genes. Among these genes Itga2B (integrin submit α−2) showed a significant main effect on genotype with trending difference in IL-22 KO DE vs. WT DE groups (P = 0.06, n = 3 male mice/ group, data are mean ± SEM). Upon identifying Itga2B as an important difference between WT and IL-22 KO animals in males, we also examined integrin complex related gene expression in all samples that included all 24 samples (3 male + 3 female/ group). We found (B) Itgb2 (integrin submit β−2) as well as other relevant genesets in the NanoString Immunology panel showing significant differences between WT and IL-22 KO mice following repetitive DE exposure: (C) integrin complex, (D) cytoplasm and (E) inflammatory response. Gene expression changes were represented as log2 counts of the integrin alpha IIB and integrin beta 2. Z-scores were plotted for genesets to show differences among all the groups. Main effects of genotype, exposure and sex were determined by 3-way ANOVA. Error bars are standard error of mean (* P ≤ 0.05, *** P ≤ 0.001; **** P ≤ 0.0001), n=3 male mice randomly selected from different experiments.

Because we identified differences in genes involved in integrin complex between the DE-exposed WT and IL-22 KO males, we further explored whether there were any sex-dependent differences within the IL-22 KO animals exposed to DE. This analysis identified differences in the ‘integrin complex’ geneset, with statistically significant upregulation of Itgax, Itgb2, Itga1, Itga6 and Itgb1 genes in KO + DE males as compared to KO + DE females. Among all the 24 analyzed samples, Itgb2 gene expression and genes included in the ‘integrin complex’ geneset trended towards an increase in males as compared to females (Figure 8 B and 8C). In this geneset, we found Itgb2 (log2 fold change = 1.66), Itgax (log2 fold change = 1.93), Itgal (log2 fold change = 1.2) in IL-22 KO males as compared to WT saline males. In IL-22 KO females, we also found changes in Itgb2 (log2 fold change = 1.14), Itgax (log2 fold change = 1.18), Itgal (log2 fold change = 0.7), Itga6 (log2 fold change = − 0.981) and Itgb1 (log2 fold change = −0.556) as compared to reference controls (WT saline males). Other genesets that showed sex-dependent differences included ‘cytoplasm’ and ‘inflammatory response’ genesets (Figure 8D and 8E). Overall, our data are consistent with previous studies reporting sex differences in immune response, for example sex steroids affecting Th1/Th2 production and interferon gamma production [39]. Also, our data suggest that the main difference between the WT and IL-22 KO mice is related to integrin signaling, as well as differences in T-cell differentiation and phagosome as per the correlations from the STRING database protein interaction analysis.

3.6. Histopathological changes following repetitive DE exposure in WT and IL-22 KO mice.

Considering that the gene expression results suggested integrin signaling might be a key factor for observed differences in infiltrating immune cells to the lung between WT and IL-22 KO mice, we next evaluated histopathology in the lung following repetitive DE exposure. In our previous studies, we identified alveolar inflammation, lymphoid aggregate (LA) formation, and epithelial hyperplasia as key changes in histopathology after repetitive DE exposure [28]. In these current studies, we found significant differences in both LA formation and epithelial hyperplasia after dust exposure as compared to saline controls (Figure 9A and 9B). We found a significant main effect of DE for lymphoid aggregate formation (F = 23.68, P = 0.0067) and multiple comparison analysis showed a trend for an increase in lymphoid aggregate formation in WT mice exposed to DE. We also evaluated epithelial hyperplasia given that one of the tissue repair effects of IL-22 includes epithelial cell proliferation. We observed a significant main effect of genotype for epithelial hyperplasia (F = 53.99, P =0.001); in saline mice, IL-22 KO mice exhibited a significantly increased epithelial hyperplasia score compared to WT saline mice; scores were not altered by DE treatment. As an outcome of lung fibrosis, we measured collagen deposition around the airways and vasculature (Figure 9C); 3-way ANOVA analysis of collagen deposition area revealed no differences among the groups.

Figure 9. Histopathological changes following repetitive exposure to HDE in wild-type and IL-22 KO mice.

H & E - stained paraffin embedded lung sections were evaluated using Echo Revolve brightfield microscopy as described in the Methods. (A) Lymphoid aggregate scores (1–5). (B) Epithelial hyperplasia scores are shown. N = 4 (2 females and 2 males) for WT saline, n = 5 for WT + DE mice (3 females, 2 males), and n = 4 for IL-22 KO mice (3 males, 1–2 females) and n = 4. Mice were randomly selected from different experiments performed on different dates. Orange arrows show lymphoid aggregates, green arrows show epithelial hyperplasia. (C) Collagen deposition (blue staining) as a measure of fibrosis in the lung. Error bars are standard error of mean (* P ≤ 0.05, *** P ≤ 0.001; **** P ≤ 0.0001).

4. Discussion

IL-22 is a unique cytokine and increased understanding of its function has great potential in managing lung inflammation, which contributes significantly to the pathogenesis of respiratory diseases like chronic bronchitis, asthma and chronic obstructive pulmonary disease [10, 14, 40, 41]. Here, we identified that IL-22 is expressed in the airway epithelium and macrophages in response to a 3-week repetitive DE exposure. The increase in IL-22 expression was accompanied by increases in the IL-22 receptor, IL-22R1, within the lung epithelium. In addition, we found that alveolar macrophages in vivo as well as THP-1 cells in vitro are unexpected sources of IL-22 production in the lung and express IL-22R1.

Accumulating evidence indicates that agricultural workers have a higher risk for developing respiratory diseases compared to the general population [8, 42]. Consistent with this observation, agriculture is considered one of the most dangerous occupations due to physical injury, chronic exposure to toxic substances, gases, fumes and dust generated during agricultural production [43]. Exposure to these substances leads to lung inflammation, and if left untreated, it progresses into chronic inflammatory respiratory diseases. Due to the consistent exposure of the airway to pathogens and immunogens, these respiratory diseases are heterogeneous and impacted by both genetic and environmental factors [7, 44, 45]. Common treatment options such as bronchodilators and inhaled corticosteroids treat the symptoms rather than curing these diseases, which suggests a deeper understanding of the underlying mechanisms of lung disease is needed to develop novel therapeutic targets. Additionally, many individuals have diseases that are poorly managed by the standard pharmaceutics. In our previous studies, we have characterized in vitro and in vivo models of organic dust-induced inflammation for inflammatory outcomes of interest, whereby we have identified DE-induced release of proinflammatory mediators TNF-α, IL-6, IL-8, and matrix metalloproteases that are classic mediators of inflammation. Meanwhile, we have also identified a compensatory response, whereby DE exposure also induces the release of pro-repair/pro-resolution mediators including amphiregulin, FGF-10, and resolvin D1 [24, 46, 47]. These previous findings prompted us to identify likely mediators involved at the interface of inflammation, resolution, and tissue repair following environmental dust exposure.

One such mediator involved in inflammation and tissue repair is IL-22, a pleiotropic cytokine with dual inflammatory and repair functions depending on the setting. IL-22 has been implicated in psoriasis, colitis, infections, and respiratory diseases (asthma and COPD) [14, 40, 41, 48, 49]. It has been previously reported that NKT cells, innate lymphoid cells (ILC3), γδ T cells, Th1 and Th22 cells are cellular sources of IL-22 [50]. A key report that utilized IL-22 KO reporter mice to assess IL-22-expressing populations suggested that IL-22 expression is mostly expressed in lymphoid lineages [22]. However, authors of this study first gated on the lymphoid population, thus excluding any potential IL-22 expression from myeloid or other cell lineages. In our model of DE-induced lung inflammation, we identified IL-22 expression in both large, small and terminal airway epithelium, which is consistent with the fact that IL-22 acts on non-hematopoietic cells[20]. However, our finding that IL-22 is expressed in Ym1+ macrophages following dust exposure is a novel and important observation (Figure 1 and Figure 4). Like IL-22 itself, the IL-22R1 receptor expression was also significantly induced upon DE exposure (Figure 2). To elucidate whether the increase in IL-22 expression after DE exposure is an adaptation to DE exposure or an attempt for inflammation resolution, we tested the impacts of DE on the lung inflammatory response in IL-22 KO mice after repetitive DE exposure. Through these investigations, we found that IL-22 KO animals have higher infiltrating immune cells as compared to their wild-type counterparts (Figure 3) in addition to alterations in lung histopathology (Figure 8). To our surprise, we did not observe a significant increase in TNF-alpha levels following DE exposure in BALF. The methods used to measure cytokine levels as well as sampling time after the last dust exposure might have affected these cytokine levels as the source, collection and preparation of the aqueous dust extract are the same as previously published [51]. Together these data suggest a protective role of IL-22 in our repetitive dust exposure model.

Despite the prevailing dogma that IL-22 expression is lymphocyte-restricted, a limited number of reports have indeed identified expression of IL-22 in macrophages in BALF and lung tissue [52, 53]. These reports also identified by RT-PCR that IL-22 is expressed in THP-1 human monocytes, alveolar macrophages and alveolar epithelial cells in the lung [52, 53]. Furthermore, in one study IL-22R1 receptor was found to be expressed robustly in alveolar macrophages and slightly in THP-1 monocytes [52]. Another study investigated IL-22 release in mice challenged with LPS or peptidoglycan, where CD3+ T cell-depleted lung cells were isolated from these mice and cultured in vitro for 24 hours in the presence of LPS, peptidoglycan and IL-23 [53]. They found that IL-22 was released in the medium after stimulation with both peptidoglycan and IL-23. While we focused our studies to the complex agricultural dust and not any single component of the dust such as peptidoglycan, this release of IL-22 after peptidoglycan stimulation is consistent with our results of BALF IL-22 levels after dust exposure, because peptidoglycan has been shown to be a component of dust that induces certain of the effects of DE exposure in the mouse lung [26, 36]. Furthermore, this is the only study that showed a double positive staining of IL-22 with a macrophage marker F4/80 in BALF of Balb/c mice exposed to peptidoglycan [53]. In our study, both alveolar macrophages and THP-1 human monocytes expressed IL-22 and its receptor IL-22R1 (Figure 5 and 6). The lack of an investigation of other immune cell types is a limitation of our study, and it is an interesting area of research that needs to be studied in future studies. While our results show a protective role of IL-22 in our model, we are limited by the cytokines measured and we cannot exclude the possibility that other IL-10 family cytokines are involved in the phenotype that we observe in the IL-22 KO mice. However, IL-10 family cytokines are considered anti-inflammatory, and to the best of our knowledge there are no reports indicating that IL-22 KO mice lacks any of the other IL-10 family cytokines, therefore we believe that any other IL-10 family cytokines would have minimal impact on the exacerbation of inflammation seen in the IL-22 KO animals.

In human monocytes, we identified a unique immunostaining of IL-22, where IL-22 had a punctate staining in resting cells while it diffused into the cytosolic compartment in the presence of DE (Figure 6). This suggests a unique mechanism that regulates IL-22 function through its subcellular localization in the way that IL-22 is released into the cytosol, potentially to allow for its binding to its receptor for autoregulatory functions. Another interesting mechanism of IL-22 production was reported in psoriatic mast cells. Mast cells release exosomes after induction with IFN-γ, which transferred cytoplasmic PLA₂ activity to neighboring CD1α-expressing cells, stimulating CD1α-reactive T cells to produce IL-22 and IL-17A [54]. Also, IL-22 has been shown to induce the transcription factor STAT3; IL-22 KO mice have been shown to be more susceptible to certain infections with increased mortality [55].

The pathways we found to be altered after repetitive DE exposure included inflammatory and immune response genes, MAPK, chemokine/ cytokine activity, positive regulation of T cell activation, JAK/STAT signaling, collagen, extracellular matrix, integrin complex, calcium-mediated signaling/cell-signaling, cell proliferation and anti-apoptosis (Figure 7 and Suppl. Figure 6). These pathway alterations are consistent with the role of IL-22 in tissue repair and epithelial cell regeneration [56]. We observed similar results with gene set analysis after a single dust exposure as well as with repetitive dust exposure [28, 37]. Additional analyses revealed that 12 genes related to integrin signaling were different between the WT and IL-22 KO animals exposed to DE. Integrins are cell surface receptors that connect cells to their extracellular matrix, and thus they play a key role in cell invasion, adhesion, cell motility, and epithelial-to-mesenchymal transition, all of which are important processes in wound repair and fibrosis. In its simplest terms, fibrosis is defined as scar formation in tissue mainly due to excessive formation of extracellular matrix proteins and disorganization of the tissue architecture. Not only fibroblasts, but also epithelial cells and macrophages contribute to fibrosis [57]. Interestingly, IL-22 has been reported to be regulated by integrins in the gut [58], and IL-22-mediated signaling in fibroblasts regulates extracellular matrix production and myofibroblast differentiation during proper wound healing[59]. Also, IL-22 has been found to be upregulated as part of the inflammatory response after tissue injury. Yet, IL-22 KO mice have severe wounds after acute skin wounding [38]. Despite the transcriptional changes we observed in the immunology panel (Figure 8), histopathological analysis of fibrotic changes were not apparent in the IL-22 KO animals, at least at the time point we assessed (Figure 9). This finding suggests that the tissue repair effects of IL-22 are consistent with epithelial cell proliferation at the tissue level as well as modulation of intracellular signaling mediated by integrins. Studies investigating detailed mechanisms are warranted.

In conclusion, we have identified a novel role for IL-22 signaling in the lung that is activated by repetitive DE exposure. There has been great interest in targeting IL-22 in clinical trials to treat atopic dermatitis, psoriasis, alcoholic hepatitis, and infectious COPD. Our data indicate that IL-22 may be important for repair and more work is warranted to elucidate its pleiotropic effects in response to different types of lung inflammation/injury. This work is needed in order to safely translate IL-22 signaling-related therapeutics to the clinic for the potential benefit of patients, including those with occupational lung diseases.

Supplementary Material

Supp Figure 4A

Supplementary Figure 4. (A) BALF cytokine levels of IL-17A and IL-10 determinex in a mouse Th17 cytokine panel 6-plex kit, IL-17A: n=3 per group except for wt saline (n=2). Data are mean +/− SEM, and IL-10: N=2 per group except for KO DE (n=3). Data are mean+/−SEM. and (B) transcript levels of cytokines and their receptors, IL-17A, IL-17B, IL-17E, IL-23A, IL-23R, TNF, IL-10, IL-10RA, IL-10RB as determined in the NanoString Immunology panel. 3 male + 3 female/ group. Data were obtained in nSolver Analysis Software. Box plots show median value with a horizontal line defined by the first and third quartiles, and error bars represent 1.5x the interquartile range.

Supp Figure 2

Supplementary Figure 2. ‘No primary antibody control’ for IL-22, IL-22R and YM1 immunofluorescence staining in a positive control (dust exposed) sample.

Supp Figure 3

Supplementary Figure 3. Changes in BALF cytokines CXCL-1 and TNF-α at the end of the 3-week repetitive dust exposure.

Supp Figure 5

Supplementary Figure 5. Gene expression of IL-22, IL-22R1 and IL-22 binding protein in THP-1 cells.

Supp Figure 6

Supplementary Figure 6. Differences in male vs. female clustering of genes in different genesets in the NanoString Immunology panel.

Supp Information

Supplementary Information. Preparation of hog dust extract and pictures of cell differentials in BALF following a 3-week saline vs. DE instillation in mice.

Supp Figure 1

Supplementary Figure 1. ‘No primary antibody control’ for IL-22 expression in mouse lung exposed to DE.

Data Availability Statement

The NanoString gene expression data can be found in the NCBI Gene Expression Omnibus (GEO) repository (to be uploaded on acceptance).